Chemical Processes with Participation of Biological and Related Compounds

Chemical Processes with Participation of Biological and Related Compounds Chemical Processes with Participation of Bi...

Author: T.N. Lomova | G.E. Zaikov

120 downloads 1430 Views 7MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Chemical Processes with Participation of Biological and Related Compounds

Chemical Processes with Participation of Biological and Related Compounds Biophysical and -Chemical Aspects of Porphyrins, Pigments, Drugs, Biodegradable Polymers and Nanofibers

Edited by

T.N. Lomova G.E. Zaikov

LEIDEN • BOSTON 2008

This book is printed on acid-free paper.

ISBN 978 90 04 16210 5 Copyright 2008 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Hotei Publishing, IDC Publishers, Martinus Nijhoff Publishers and VSP. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change. printed in the netherlands

Contents

Preface

xi

Foreword

1

Chapter 1 Synthetic and Natural Bacteriochlorins: Synthesis, Properties and Applications M.A. Grin and A.F. Mironov Introduction 1 Synthetic bacteriochlorins 2 Natural bacteriochlorins and their chemical modifications 3 Amphiphilic and water-soluble derivatives of bacteriochlorins 4 Conclusion References Chapter 2 meso-Phenylporphyrins as Synthetic Models of Natural Porphyrins: Synthesis and Modification A.S. Semeykin, S.A. Syrbu and O.I. Koifman

5 5 6 14 30 41 41

45

Introduction 45 1 Synthesis of meso(5,10,15,20)-tetrasubstituted porphyrins 46 55 1.1 Synthesis of meso-tetra-β-octasubstituted porphyrins 1.2 Synthesis of nonsymmetric meso-tetrasubstituted porphyrins 58 2 meso(5,15)-Disubstituted porphyrins 66 3 Synthesis of meso(5)-Phenyl-b-octaalkylporphyrins 74 4 Reactions of introduction and modification of substituents in phenyl rings of meso-phenyl-substituted porphyrins 77 4.1 Introduction of substituents into phenyl rings of meso-phenylporphyrins 77 4.2 Modification of functional groups in phenyl rings of meso-phenylporphyrins 79 4.2.1 meso-Oxyphenylporphyrins and their modification 80 4.2.2 meso-Aminoporphyrins and their modification 82 References 84

vi

CONTENTS

Chapter 3 The Mechanism of Catalytic Action of the Coordination Centres of Catalase Synthetic Models T.N. Lomova, M.E. Klyueva and M.V. Klyuev Introduction 1 Substituted copper(II) porphyrins as catalysts of the hydrogen peroxide disproportionation reaction 2 Kinetic regularities and mechanisms of peroxide decomposition reactions in the presence of acido complexes of highly substituted manganese porphyrins 2.1 Kinetics and reaction mechanism of oxidation of manganese(III) porphyrins by hydrogen peroxide 2.2 Kinetics of peroxide disproportionation in the presence of manganese(III) porphyrins 3 Conclusion References Chapter 4 Complexation of Porphyrins with Ions and Organic Molecules N.Zh. Mamardashvili, V.V. Borovkov, G.M. Mamardashvili, Y. Inoue and O.I. Koifman Introduction 1 Complexation of porphyrins with ions 2 Complexation of porphyrins with organic molecules: the thermodynamic aspect 3 Complexation of porphyrins with organic molecules: the chirality aspect 3.1 Host–guest systems based on monomeric porphyrins 3.2 Host–guest systems based on dimeric and oligomeric porphyrins 4 Conclusion References Chapter 5 Chemical Activation of Porphyrins in Coordination Core Reactions D.B. Berezin and B.D. Berezin Introduction 1 Porphyrin ligands with localized and delocalized NH bonds 1.1 Factors causing the delocalization of NH bonds in porphyrin molecules 1.1.1 Spatial structure and polarization of the porphyrin molecule 1.1.2 NH activation in the course of porphyrin–solution component interaction and porphyrin–solid phase interaction 1.2 Interaction of organic solvents and porphyrins with delocalized-type bonds 1.2.1 Acid-base interactions 1.2.2 Tautomeric processes 1.3 Quantitative assessment of the state of NH bonds in porphyrin molecules

93 93 95

96 98 108 114 115 117

117 118 125 144 144 151 162 163 169 170 171 172 172 181 185 185 190 199

CONTENTS

1.3.1 Spectral criterion 1.3.2 Kinetic criterion 1.3.3 Quantum chemical criterion 1.3.4 Insufficiency of absorption spectrum analysis of porphyrins 1.4 Nonplanar structure of the macrocycle and chemical NH activity in its coordination core 2 Reactivity of the coordination core in molecules of porphyrin ligands 2.1 Complexation reactions 2.2 Proton ionization of NH bonds in H2P molecules 2.3 Nucleophilic substitution reactions in the coordination core 3 Biosignificance of the phenomenon of NH activation References Chapter 6 Synthesis, Structure Peculiarities and Biological Properties of Macroheterocyclic Compounds M.K. Islyaikin, E.A. Danilova, Yu.V. Romanenko, O.G. Khelevina and T.N. Lomova Introduction 1 Synthesis, structure and properties of the initial compounds 2 ABBB-type macroheterocyclic compounds 3 Azolophthalocyanines 4 State of triazoleporphyrazines in proton-donor media 5 ABAB-type macroheterocyclic compounds 6 ABABAB-type macroheterocyclic compounds 7 Coordination properties 8 Studies of the practically valuable properties of macroheterocyclic compounds and their metal complexes 8.1 Biological properties References Chapter 7 The Photochemical Aspect of Reactions of Flavonols with Molecular Oxygen E.A. Venedictov Introduction 1 Structure of flavonols 2 Spectral luminescent properties of flavonols 3 Photoproduction of 1O2 (1∆g) 4 1O2 Reactions 5 Photochemical properties of coordination compounds of quercetine 6 Conclusion References Chapter 8 Solvation of Drugs as a Key for Understanding Partitioning and Passive Transport Exemplified by NSAIDs German L. Perlovich and Annette Bauer-Brandl Introduction

vii

199 200 203 203 204 208 208 210 211 211 212

219

219 221 228 230 236 244 249 253 262 262 266

271 271 272 276 278 279 284 288 288

291 292

CONTENTS

viii

1

Partitioning and diffusion 1.1 The partitioning (distribution) process 1.2 Influence of the solution pH on the partition/distribution coefficients 1.3 Diffusion of drugs 2 Solvation of drugs: relevance and theoretical approaches 2.1 The main definitions 2.2 Models describing the solvation of molecules 3 Experimental methods to measure solvation characteristics and choice of subjects 3.1 Sublimation experiment 3.2 Method of isothermal saturated solubility 3.3 Isothermal calorimetry 3.4 Choice of drugs 4 Crystal structures of NSAIDs 4.1 Description of hydrogen bond networks topology by graph set assignment 4.2 Analysis of packing architectures of NSAIDs crystal lattices 5 Thermochemical and thermodynamic properties of NSAIDs 5.1 Thermodynamic characteristics of sublimation of NSAIDs 5.1.1 Differences of racemate and enantiomer ibuprofen crystal lattices 5.2 Thermochemical characteristics of NSAIDs 6 The difference between partitioning and distribution of NSAIDs from the thermodynamic point of view 6.1 Solvation characteristics of dissociated and non-dissociated (+)- and (±)-IBP 6.2 Solvation characteristics of dissociated and non-dissociated forms of the other NSAIDs 6.3 Solvation characteristics of transfer process of dissociated and non-dissociated molecules from buffer to n-octanol 7 Correlation between biopharmaceutically relevant parameters and solvation characteristics References Chapter 9 Biodamage of Materials: Adhesion of Microorganisms on the Surface of Materials K.Z. Gumargalieva, I.G. Kalinina, S.A. Semenov and G.E. Zaikov Introduction 1 Results and discussion References Chapter 10 Controlled Release of Aseptic Drug from Poly(3-hydroxybutyrate) Films: A Combination of Diffusion and Zero-order Kinetics R.Y. Kosenko, Y.N. Pankova, A.L. Iordanskii, A.P. Bonartsev, and G.E. Zaikov Introduction

292 293 294 294 296 296 297 299 299 301 301 301 302 302 303 307 307 309 312 314 316 317 318 322 324

327

328 328 339

341

342

CONTENTS

1 Experimental 2 Results and discussion 3 Conclusion 4 Acknowledgement References Chapter 11 Transport of Water as a Structurally Sensitive Process Characterizing the Morphology of Biodegradable Polymer Systems A.L. Iordanskii, Yu.N. Pankova, R.Yu. Kosenko, A.A. Ol’khov and G.E. Zaikov Introduction 1 Experimental 2 Hydrophobization of poly-(3-hydroxybutyrate) 3 Hydrophilization of poly-(3-hydroxybutyrate) References Chapter 12 A Novel Technique for Measurement of Electrospun Nanofiber M. Ziabari, V. Mottaghitalab and A.K. Haghi Introduction 1 Methodology 1.1 Simulation of electrospun web 1.2 Fiber diameter measurement 1.2.1 Manual method 1.2.2 Distance transform 1.2.3 Direct tracking 1.3 Real webs treatment 2 Experimental 3 Results and discussion 4 Conclusion References Chapter 13 Image Analysis of Pore Size Distribution in Electrospun Nanofiber Webs: New Trends and Developments M. Ziabari, V. Mottaghitalab and A.K. Haghi Introduction 1 Methodology 1.1 Sieving methods 1.2 Mercury porosimetry 1.3 Flow porosimetry (bubble point method) 1.4 Image analysis 1.4.1 Real webs 1.4.2 Simulated webs 2 Experimental 3 Results and discussion 4 Conclusion References

ix

342 343 347 347 347

349

350 350 351 354 359 361 361 363 363 363 364 364 365 366 367 367 373 374

375 375 376 378 378 379 379 380 381 382 383 388 389

CONTENTS

x

Chapter 14 Electrospun Biodegradable and Biocompatible Natural Nanofibers: A Detailed Review A.K. Haghi and R.K. Haghi 1

Introduction 1.1 Electrospinning setup 2 Effect of systematic parameters on electrospun nanofibers 2.1 Solution properties 2.1.1 Viscosity 2.1.2 Solution concentration 2.1.3 Molecular weight 2.1.4 Surface tension 2.1.5 Number of entanglements 2.1.6 Solution conductivity 2.1.7 Effect of salt addition 2.1.8 Solvent 2.2 Processing condition 2.2.1 Applied voltage 2.2.2 Feed rate 2.2.3 Distance of needle tip to collector 3 Theory and modeling 4 Natural fibers 5 Electrospinning of silk fibers 5.1 Introduction 5.2 Crystal structure of silk (fibroin) at various stages of electrospinning 5.3 Spinning dope preparation for electrospinning 5.3.1 Degumming 5.3.2 Dissolving of fibroin 5.4 Electrospinning of silk fibroin 5.4.1 Effect of silk polymer concentration on fiber diameter 5.4.2 Effect of voltage and spinning distance on morphology and diameter 5.5 Characterization 6 Electrospinning of cellulose and cellulose acetate 6.1 Electrospinning of CTA solution 7 Concluding remarks References Nomenclature

391 391 392 393 393 393 393 395 396 396 396 397 399 399 399 402 403 405 407 408 408 408 411 411 411 411 412 414 416 417 418 419 420 422

To learn anything without thinking is absolutely useless, Thinking about something without analysing and Studying the subject of thinking is dangerous Confucius, 551–479 BC Ancient China Chemistry is a miracle, interest, delight, The future and basis of well-being of people Yury M. Luzhkov Mayor of Moscow November 12, 2003

Preface Everything around us has been created by chemistry (glass, gas, medicines, food, metals, polymers, etc.). There is no pathos of “chemical” enthusiasm in this definition. Chemistry is the basis of everything, which life produces (microbes, plants, animals, human beings). Any living body is a giant chemical reactor with millions of coordinated chemical reactions proceeding in it. Molecular biology, molecular genetics and gene engineering, biotechnology and intellect – all this is chemistry. Nobelist Prof. N.N. Semenov and his pupils (Profs. M.N. Emanuel, V.N. Kondratyev, V.I. Goldansky, A.L. Buchachenko, K.I. Zamoraev, Yu.B. Monakov) insisted on that. And we fully agree with them. There is a certain shift in the world of chemistry from studies of more or less simple reactions to get, for example, sulfuric acid, ammonia, phenol and acetone to research into complicated biochemical and biological processes. This is required for solving medical problems, because medicine is our health and our good life. It is common knowledge that it is better to be healthy and rich than to be poor and sick. Chemistry is more than a science. It is a festivity. This volume presents the reviews of chemists working in the field of biochemistry, biology and, in the final analysis, for medicine and health. We look forward to readers’ comments, will be grateful for them and will definitely use them in our future research. Prof. Gennady E. Zaikov Moscow, Russia Prof. Tatyana N. Lomova Ivanovo, Russia

x

INTRODUCTION

To propagate education is to extend prosperity. I mean the general prosperity but not one’s private wealth. With extension of prosperity, part of evil disappears. Alfred Nobel, Sweden

Foreword Research into the kinetics and thermodynamics of processes involving biologically active substances and multistage mechanisms of their interactions in living systems is of primary interest for modern science. Many problems arising in the course of the studies may be solved by investigating the reactivity of BAS in model systems and synthetic analogs of naturally occurring biomolecules with a complex and often unknown structure. The authors of this book have been for many years engaged in studies of the reactivity of BAS in various aggregate states and in relation to the molecular structures and supramolecular forms. The book covers many aspects of the effects of relatively simple biomolecules as model enzymes, molecular receptors, photosensitizers, pharmacophores, and biopharmaceutical agents. The quantitative characteristics of the transitions of cations, anions and small organic molecules, enzymic catalysis, and diffusion of molecules through biological membranes are presented. The mechanisms of the processes are discussed. The biological activity of the compounds studied is assessed. Hydrated porphyrin forms and their unique properties are of great interest. Chapter 1 deals with naturally occurring bacteriochlorins, their properties, isolation and chemical modifications. A separate part is concerned with synthetic bacteriochlorins; here, general and particular methods of production of these compounds, their spectral and physicochemical characteristics, are considered. Special consideration is given to the use of naturally occurring and synthetic bacteriochlorins in the production of new-generation photosensitizers for photomedicine. Chapter 2 describes methods of synthesis of meso-mono-, di-, tri-, and tetraphenylporphyrins along with schemes of addition and modification of substituents in them. Approaches are shown to fine setting of the physicochemical properties of porphyrins obtained. Chapter 3 represents the results of research into the catalytic activity of manganese(III)- and copper(II)-porphyrins alkyl- and phenyl-substituted in β- and meso-positions in the reaction of decomposition of hydrogen peroxide in the DMFA–KOH–H 2O system. The ion-molecular mechanism of the decomposition with kinetically significant stages of

2

Foreword

two-electron oxidation and subsequent partial reduction of metalporphyrin was determined as well as acid–basic equilibria of peroxide. It is shown that the efficient catalysis of decomposition of hydrogen peroxide is determined by the degree of binding of porphyrin in a complex with metal, by the structure of the mixed coordination sphere, and by the mutual influence of the ligands; the compounds under study behave as catalases in living systems. Investigation of the interactions between porphyrin molecules and protein surroundings in biological systems is an important research area in modern biochemistry. These interactions may proceed in various ways, e.g., through the formation of covalent bonds, by ionic association, as donor–acceptor interactions, or through hydrogen bonds between separate covalent-bound fragments. Chapter 4 gives a systemic description of complexation properties of porphyrins and their ability to discern charged particles and small organic molecules. Depending on the molecular structure and medium, NH bonds of the coordination nucleus of porphyrins are considered as localized or partially localized ones. Their delocalization may be caused by internal molecular effects, such as polarization of the molecule by substituents or by specifically nonplanar conformations of the macrocycle during solvation in solvents containing electron–donor components, by polymerization, sorption or transition to the solid state. The activation of NH bonds in porphyrin molecules is associated with drastic changes in the coordination nucleus reactivity. Chapter 5 makes an analysis of the causes and pathways of activation of NH bonds with regard to the classification of porphyrins by their capability for NH activating. Some reliable quantitative criteria for determining the degree of NH activity were suggested for porphyrins and their analogues. Activation of NH bonds may manifest itself not only in the reactivity of porphyrin coordination nuclei but may lead to the reorganization of the π-chromophore and reacting sites and to tautomeric processes. Products of this reorganization are good models of supramolecular biological systems. Chemical activation of porphyrins is considered separately as a potent tool of controlling the activity of molecules along with biological aspects. Chapter 6 reviews the latest achievements in the chemistry of macroheterocyclic compounds – structural analogs of porphyrin and hexapyrins. The aromas of various macrocyclic molecules and their fragments were investigated in accordance with the geometric (EN, GEO, and HOMA) and magnetic (NICS) criteria based on experimental data and results of DFT quantum-chemical calculations. The coordination and biological properties of macroheterocyclic compounds and their complexes with metal cations are considered. Chapter 7 presents the results of studies on the kinetics of photochemical reactions of biologically active pigments and related compounds with molecular oxygen. Chapter 8 is devoted to the search of correlations of thermodynamic characteristics (the Gibbs energy and enthalpy and entropy constituents of the Gibbs energy) of solvation of molecules of drugs prepared by traditional experimental methods and their diffusive properties and biopharmaceutically important properties. Chapters 9–11 deal with biodamage of materials (adhesion of microorganisms on the surface of materials), aspects of controlled release from polymer films and transport of water as a structurally sensitive process that characterizes the morphology of biodegradable polymer systems. Chapters 12–14 discuss the development of an image analysis based method (direct tracking) for measuring the diameter of electrospun fiber, consider new trends and developments in image analysis of pore size distribution and present a detailed review of different

Foreword

3

aspects of electrospinning biodegradable and biocompatible natural nanofibers. This collection of papers dealing with the results of studies on various classes of biologically active compounds will add to understanding the problems of their reactivity and the nature of processes occurring in living objects with their participation. Editors: Prof. Gennady E. Zaikov Institute of Biochemical Physics, Russian Academy of Sciences Moscow, Russia; [email protected] Prof. Tatyana N. Lomova Institute of Solution Chemistry, Russian Academy of Sciences Ivanovo, Russia; [email protected]

4

Foreword

- нет ссылки 57 в списке лит. -

1

Synthetic and Natural Bacteriochlorins: Synthesis, Properties and Applications M.A. Grin and A.F. Mironov Lomonosov Moscow State Academy of Fine Chemical Technology, 86 Vernadsky Prospekt, Moscow, 119571, Russia; email: [email protected]

This chapter considers natural bacteriochlorins, their properties, isolation and chemical modifications. Part of the discussion is given to synthetic bacteriochlorins, with consideration of common and specific methods of producing these substances, their spectral and other physicochemical characteristics. Special attention is paid to the use of natural and synthetic bacteriochlorins in developing newgeneration photosensitizers for photomedicine.

Introduction In the recent decade, attention of scientists, working in the field of developing new photosensitizers (PS) for photodynamic therapy (PDT) of cancer, has been focused on compounds with intensive absorption in the range of 770 up to 850 nm. The use of PS with this therapeutic window of absorption opens new possibilities for the diagnostics and treatment of malignant neoplasms. The light with these wavelengths scatters weakly and, therefore, can penetrate deeper into the tissue (Scheme 1) [1]. This is of special significance in pigmented tumors, for instance, melanoma. Besides, it is important that accessible and cheap semiconductor lasers can be used for this range. Such compounds include derivatives of synthetic and natural bacteriochlorins. It is known that in porphyrin systems two peripheral double bonds in opposite pyrrole rings (B and D) are cross-conjugated, and their presence is not required for aromaticity to be preserved.

6

Penetration into tissues, mm

M.A. Grin and A.F. Mironov

nm Porphyrins

Chlorins

Bacteriochlorins

Scheme 1 Dependence of the penetration of light into tissues on the wavelength.

In the reduction of one bond (dihydroporphyrins–chlorins) or both bonds (tetrahydroporphyrins–bacteriochlorins), aromaticity is preserved, and the change of symmetry leads to a batochromic shift of the Q band. Figure 1 shows the real spectra of three compounds: porphyrin, chlorin and bacteriochlorin, which have the same m-hydroxyphenyl substituents in meso-positions of the macrocycle. Bacteriochlorins intensively absorb in the near IR region of the spectrum (λmax = 760–780 nm, ε = 4×104 –1×105 M –1 cm –1) [2] and, therefore, possess optimal properties for their use as photodynamic agents. Besides, they generate active oxygen species (AOS) with a high quantum yield, which depends on the nature of the central metal and peripheral substituents [3]. However, the macrocycle with two reduced double bonds is chemically unstable, therefore, derivatives of bacteriochlorophyll a are apt to be oxidized to respective chlorins and porphyrins. This fact largely restricts introduction of PS of bacteriochlorin series into clinical practice. A high hydrophobicity and related low solubility in polar solvents also complicate their use in medicine. The main works on the chemical conversions of bacteriochlorophyll a (Bchl a) aim to increase the chemical stability and to develop water-soluble forms of bacteriochlorin photosensitizers. Works on the chemical modification of bacteriochlorophyll a are comparatively a few. Major research in this field has been done at the Weizmann Institute of Science, Rehovot, Israel and Photodynamic Therapy Center, Roswell Park Cancer Institute, Buffalo, USA. Our group also does intensive research in this field, and major results will be presented in this review. There are two approaches to production of bacteriochlorins. The synthetic way includes the reduction of double bonds in pyrrole rings B and D in porphyrins, and the semisynthetic route is when bacteriochlorophylls isolated from natural sources are modified to increase their stability, improve the spectral characteristics and solubility in polar solvents.

1

Synthetic Bacteriochlorins

There are two approaches to the reduction of porphyrins to chlorins and bacteriochlorins. The first includes catalytic hydrogenation or treatment with metals in an alcohol medium

Synthetic and Natural Bacteriochlorins: Synthesis, Properties and Applications

1

nm

2

nm

3

nm

Figure 1 Electronic spectra of porphyrin (1), chlorin (2) and bacteriochlorin (3).

7

8

M.A. Grin and A.F. Mironov R

R

N

NH

R

R

R

A

HN

N

N

NH

HN

N

R

R

4

B

R

R

R

N

N

R

A

R

N

N

Zn N

R

R

Zn N

N

N

R

R

5

A: p-MeC6H4SO2NHNH2, K2CO3/Py, t [N2H2] B: Zn(OAc)2

Scheme 2 Reduction of porphyrins and their metal complexes with diimide.

(boiling with sodium in amyl alcohol). The other approach is related to the use of diimide NH=NH, which is formed from p-tosylhydrazine, as a reducing agent. When porphyrins are reduced as free bases, a mixture of respective chlorin and bacteriochlorin 4 is formed, whereas the reduction products of the zinc complex of porphyrin are chlorin and isobacteriochlorin 5 (Scheme 2) [4]. This approach was realized by R. Bonnett et al. in the reduction of 5,10,15,20tetrakis(m-hydroxyphenyl)porphyrin (m-THPP) 1 to form respective chlorin (m-THPC) 2 and bacteriochlorin (m-THPBC) 3 [5]. Though the molar extinction coefficients and λmax of the long-wavelength absorption band increase in the sequence porphyrin > chlorin > bacteriochlorin, the photophysical properties of the reduced species, including the quantum yields of the triplet and singlet oxygen, differ insignificantly [6]. The photodynamic efficiency in the given sequence increases at each stage of reduction,. Thus, at a tumour photonecrosis depth of 5 mm the doses of PS introduced decrease from 6.25 mmol/kg for porphyrin to 0.75 mmol/kg for chlorin and 0.39 mmol/kg for bacteriochlorin [5]. However, in in vivo experiments bacteriochlorin, possessing a high photodynamic activity, proved to be much less stable as compared with the chlorin analogue (Foscan® ); its considerable amount was oxidized in cells within 24 h. An interesting approach to the synthesis of di- and tetrahydroporphyrin derivatives was proposed by Callot et al. [7], who showed that, during the action of diazoacetic-acid methyl ester on tetraphenylporphyrin (TPP) 6, carbene formed in the reaction attacks the double bonds in the pyrrole rings B and D to form cycloaddition products chlorin 7 (λmax = 650 nm) and bacteriochlorin 8 (λmax = 720 nm) (Scheme 4).

9

Synthetic and Natural Bacteriochlorins: Synthesis, Properties and Applications OH

HO

N

NH

HN

N

OH

OH

OH HO

2

N

NH

+ HN

N

B

C

OH OH

A OH

1

HO

N

NH

HN

N

OH

OH

3

A: p-MeC6H4SO2NHNH2, K2CO3/Py, t [N2H2] B: [N2H2] C: ɨ-chloranil

Scheme 3 Reduction of 5,10,15,20-tetrakis(m-hydroxyphenyl)porphyrin with diimide. Ph

Ph

H

CO2Me

Ph

H

N

NH

A Ph

Ph HN

N

N

NH

H Ph

Ph

N

NH

+

H H Ph

Ph

HN

N

CO 2Me H

HN

N

H MeO2C Ph

Ph

H

H

7

6

Ph

8

A: N2CHCO2Me, CuI

Scheme 4 Interaction of TPP with diazoacetic-acid methyl ester.

Another method of bacteriochlorin synthesis is based on the treatment of porphyrins with osmium tetroxide. The reaction proceeds similarly to the above-described reduction of porphyrins with diimide: interaction of a free base with OsO4 yields tetrahydroxybacteriochlorin 9, whereas osmylation of the Zn complex produces mainly isobacteriochlorin 11.

10


NH

N

N

N

NH

A

HN

N

N

N

B

Zn N

N

HN

HO HO

10 C

A

A HO HO

OH OH NH

N

N

N

N

NH

Zn N

N N

HN

HN

N

HO HO HO

O

HO

9

11

C

A O

OH OH

O NH

N

N

C

N

NH

N

NH

+

HN

N

HN

N

HN

Et

O

O

12

O

13

Scheme 5 Interaction of octaethylporphyrin and its Zn complex with OsO4.

Pinacoline regrouping of vicinal tetrahydroxybacteriochlorins formed leads to an isomeric mixture of ketobacteriochlorins 12 and 13 (Scheme 5) [8]. Later on, R. Pandey et al. used this approach to produce vicinal dihydroxy- and ketobacteriochlorins from natural chlorins: methyl ester of mesopyropheophorbide a 14 and trimethyl ester of mesochlorin e6 17 (Scheme 6) [9]. The first to be formed in the course of the reaction are osmate complexes, which include the pyridine molecule; their reductive splitting by hydrogen sulfide leads to respective vicinal diols 15 and 18. Despite the improved spectral characteristics of the latter due to the batochromic shift of the Q band to the red region, their photodynamic activity in vivo is lower than that of respective keto derivatives 16 and 19. This is, apparently, due to a more hydrophobic character of the keto group, which enables ketobacteriochlorins to stay longer in tumor cells. Another example of producing bacteriochlorins from chlorophyll a derivatives is osmylation of 3-formyl-3-devinylpurpurin 18 methyl ester 20 to form dihydroxybacteriochlorin 21, that as the result of the pinacoline regrouping yields a mixture of 7-oxo- and 8-oxobacteriochlorins 22 and 23, which is the result of two variants of migration of alkyl groups in pyrrole B (Scheme 7) [10].

11

Synthetic and Natural Bacteriochlorins: Synthesis, Properties and Applications O OH OH N

NH

H

N

HN

N

H

HN

N

H

HN

H O

O

O

14

N

NH

B

H

H MeO2C

N

NH

A

16

MeO2C

15

MeO2C

O

OH OH NH

H

N

N

NH

A

HN

H

N

H

MeO2C

N

N

H

HN

H

CO2Me

CO 2Me CO2Me

CO2Me

17

NH

B

HN

H

CO2Me CO2Me

MeO2C

N

MeO2C

18

19

A: OsO4/Py, H2S; B: H2SO4

Scheme 6 Oxidation of methyl ester of mesopyropheophorbide a 14 and trimethyl ester of mesochlorin e6 and regrouping of vicinal diols. CHO

CHO

NH

PMe

N

N

HN

O

O

OH OH

NH

A

N

HN

N

PMe O

O

20

21

B O

CHO

O

O

CHO O

NH

PMe

N

N

HN

O

O

786 nm

22

NH

O

PMe

N

N

HN

O

O

777 nm

23


Scheme 7 Oxidation of 3-formyl-3-devinylpurpurin 18.

O

12

M.A. Grin and A.F. Mironov H

H B

B

O

N

NH

N

N

O

N

NH

A

D

B N

HN

D

Me

P

HN

O

D HO HO

PMe

PMe

PMe

26

PMe

27 O

O

A

O

B

B NH

N D

HN

PMe

O

N

A

N C P

O

N

NH

HN

N

D HO HO

Me

28

O

N

NH

HN C OH

P

Me

P

29

Me

PMe

P

Me

OH

30

PMe = CH2CH2CO2CH3


Scheme 8 Effect of electron-acceptor substituents in porphyrins on the regioselectivity of OsO4 oxidation.

R. Pandey et al. studied the effect of electron-acceptor substituents in porphyrins and chlorins on the regiospecificity of OsO4 oxidation [11]. The presence of an electronacceptor substituent in one of the pyrrole rings of the porphyrin macrocycle was shown to direct hydroxylation to the opposite pyrrole. Thus, the reaction with 3-acetyldeuteroporphyrin IX 24 leads to the hydroxylation of ring C 25, whereas a similar reaction with 8-acetyldeuteroporphyrin IX 26 yields a vicinal diol in ring D 27. The reaction of 3,8-diacetylporphyrin 28 with OsO4 showed no stereoselectivity, and a mixture of diols 29 and 30 was formed (Scheme 8). A similar study was carried out on chlorins, the treatment of which by OsO4 leads exceptionally to the hydroxylation of the pyrrole ring B. The character of the substituents in the macrocycle also affects the progress of the pinacoline regrouping [12]. An interesting regularity was found to exist between the total number of electron-acceptor groups in the molecule of vicinal diol and the selectivity of formation of 7- or 8-ketobacteriochlorins. Thus, irrespective of the arrangement of one carbonyl group (the pentanone exocycle or the formyl group), in the macrocycle 31 under acidic conditions 8-ketobacteriochlorin 32 is always formed, whereas the presence of two electron-acceptor groups (131-keto and 3-formyl) in chlorin 33 leads to a mixture of 7- and 8ketobacteriochlorins 34 and 35. The presence of the anhydride or imide exocycle conjugated with the macrocycle, as well as of three carbomethoxy groups in the lower part of the chlorin macrocycle 36 and 37 also leads to a mixture of 7- and 8-ketobacteriochlorins with various ratios of the isomers (Scheme 9). Bacteriochlorins can also be obtained from divinylporphyrins by the Diels–Alder reaction involving two vinyl groups (Scheme 10) [13, 14]. In turn, the latter are obtained from vicinal dihydroxy derivatives by boiling in benzene in the presence of p-toluenesulfo acid. A “double” Diels–Alder reaction by both vinyl groups in 8,18- 38 and 3,13- 40 in

13

Synthetic and Natural Bacteriochlorins: Synthesis, Properties and Applications OH OH N

NH

N

H

A

HN

N

H

O

B

HN

N

H

O

O

O

31

CO2Me

HN

H

H

H

N

NH

N

NH

CHO

CHO

32

CO2Me

CO2Me

OH OH

Me

N

H

O

N

N

NH

N

NH

A

B

HN

N

H

35

HN

+ O

H

H

O

O

N

CO2Me

33

CO2Me

34

OH OH NH

N

NH

A

N

O

N

B

+ H

N

HN

H

H

N

H

CO2Me

CO2Me MeO2C

36

CHO

CHO

NH

N

NH

OH OH N

N

HN

H

H

37

+

HN

H

CO2Me CO2Me

MeO2C

N

O N

B

A H

N

CO2Me

CO2Me MeO2C

O

HN

O

CO2Me

N

CO2Me MeO2C


Scheme 9 Effect of electron-acceptor substituents in chlorins on the progress of the pinacoline regrouping.

14

M.A. Grin and A.F. Mironov CO2Me

CO2Me

MeO 2C

NH

N

NH

A

N

HN

N

N

HN

CO2Me

MeO2C

38

CO2Me

39

CO2Me

CO2Me

MeO2C CO2Me NH

N

N

A

Me

HN

NH

N

N

HN

CO2Me CO2Me

CO2Me

40

CO2Me

41

A: dimethyl ester of acetylenedicarboxylic acid (DMAD), diazabicycloundecene (DBU)

Scheme 10 A Diels–Alder reaction with divinylporphyrins.

divinylporphyrins with various dienophiles, e.g., with dimethyl ester of acetylenedicarboxylic acid (DMAD) made it possible to produce bacteriochlorins 39 and 41, in whose spectra the Q band is shifted to the region of 800 nm. A similar approach was used by Pandey et al. to produce bacteriochlorin from methyl ester of 3-ethyl-7,8-dihydroxypurpurine 18 42 [15]. Boiling of the latter in o-dichlorobenzene led to methyl ester of 8-vinyl-3-ethylpurpurine 18 43 with a 60% yield. The above reaction with various dienophiles (TCE and DMAD) yielded adducts 44 and 45 (Scheme 11). Similar adducts were obtained for chlorin 48; the DMAD adduct was present as cis50 and trans- 51 isomers (Scheme 12). A close reaction is intramolecular cyclization of Ni complexes of 5,10- or 5,15-bis(vinylformyl)porphyrins 52 to form bacteriochlorins with two six-membered exocycles conjugated with the main macrocycle 53 (Scheme 13) [16]. The latter proved unstable and were oxidized in air.

2

Natural Bacteriochlorins and Their Chemical Modifications

Bacteriochlorophylls represent an independent group of natural chlorophylls and are widespread in nature, mainly in numerous photosynthesizing bacteria [17–19]. The pigments differ by the extent of hydrogenation of the macrocycle and by the character of the substit-

15


CN NC CN CN OH OH N

NH

H

N

HN

H O

O

MeO2C

A

NH

H

O MeO2C

42

N

N

HN

H O

O

NH

B H

N

HN

H O

O

O

MeO2C

43

N

O

44

C H+

MeO2C

CO2Me

H

NH

MeO2C

N

NH

N

HN

H O

O

B:

N

O

N

HN

H O

O

MeO2C

47

NH

N

D H

MeO2C

CO2Me

MeO2C

H H

E H

H

CO2Me

45

H

O

N

HN

H O

O

O

MeO2C

46

A: ɨ-dichlorobenzene, boiling B: tetracyanoethylene (TCE) ɋ: DMAD D: Et3N E: DBU

Scheme 11 Interaction of 8-vinylpurpurine 18 with dienophiles.

uents. Several modifications of bacteriochlorophylls are known. Thus, bacteriochlorophylls a and b have been isolated from purple bacteria; green bacteria were the source of bacteriochlorophylls a, c, d and e; sulfur bacteria, of bacteriochlorophylls c, d and e; bacteriochlorophylls g were isolated from some types of photosynthesizing bacteria. The mentioned bacteriochlorophylls are usually divided into two sufficiently large groups (Scheme 14). The first group, which includes bacteriochlorophylls a, b and g, is characterized by the presence of the tetrahydroporphyrin macrocycle 54 and, as the alkoxy radical R4 for the first two, the residues of phytol (a), geraniol (b) and 2,10-phytadienol; and for the third, farnesol (d) and geranyl geraniol. The second group with the dihydroporphyrin macrocycle 55, for which the name of chlorobium chlorophylls is also used, includes bacteriochlorophylls c, d and e. These bacteriochlorophylls are characterized by the presence of the pentanone ring, α-hydroxyethyl group in position 3, methyl substituent at the δ-meso-carbon atom and etherifying alcohol R4 – 2,6-phytadienol (CH3)2CH(CH2)3CH(CH3)(CH2)3C(CH3)=CH (CH2)2C(CH3)=CHCH2OH and 2,16,20-phytatrienol (CH3)2C=CH(CH2)2C(CH3)=CH (CH2)2CH(CH3)(CH2)3-C(CH3)=CHCH2OH.

16

M.A. Grin and A.F. Mironov CN NC OH OH N

NH

N

NH

N

H

H

HN

H

N

NH

B

A

49

48 C

CO2Me CO2Me

MeO2C H

CO2Me

CO2Me

N

NH

E

H

H

MeO2C

MeO2C

NH

N

CO2Me

MeO2C

H

D

N

NH

50

51

A: ɨ-dichlorobenzene, boiling B: TCE ɋ: DMAD D: Et3N E: DBU

Scheme 12 Interaction of 8-vinyl-8-deethylmesochlorin p6 with dienophiles.

N

N

CHO

A

N

Ni N

CHO

N Ni

N

N

N

53

52 A: HCl

Scheme 13 Intramolecular cyclization of bis(vinylformyl)porphyrins.

Pigments of the first group have intensive absorption bands in the near IR region. Of special interest are bacteriochlorophylls a and b, whose maxima are not only shifted to the red region, but also have the highest extinction coefficients. Of the two, Bchl a is usually taken as the initial material to develop novel photosensitizers for photodynamic therapy (PDT) of cancer and other possible photomedical applications. Bacteriochlorophyll a is a porphin derivative, which contains the pentanone ring (exocycle) condensed with the tetrahydroporphyrin macrocycle, various peripheral substituents


17

(CH3)2C=CH(CH2)2C(CH3)=CH(CH2)2C(CH3)=CH(CH2)2C(CH3)=CHCH2OH (a); (ɋH3)2ɋH(ɋH2)3ɋ(ɋH3)=CH(CH2)2CH(CH3)(CH2)3-C(CH3)=CHCH2OH (b); (CH3)2C=CH(CH2)2-C(CH3)=CH(CH2)2C(CH3)=CHCH2OH (c) R1

HO

H

R1

H

R2 N

R2

R3

N

N

Mg

Mg

R5 N

N

N

N

N R3

H

H H

H H CO2Me

R4O2C

O

O R4O2C

Bacteriochlorophyll ɚ: R1=COCH3 R2=H, R3=C2H5; Bacteriochlorophyll b: R1=COCH3 R2 + R3= (=CHCH3); Bacteriochlorophyll g: R1= - CH=CH2 R2 + R3= (=CHCH3);

Bacteriochlorophyll c: R1=R3=R5=CH3 R2=C2H5; Bacteriochlorophyll d: R1=CH3, R2=C2H5 - C5H11, R3= C2H5;R5=H Bacteriochlorophyll e: R1=CHO;R2=C2H5 - C5H11 R3= C2H5;R5=CH3

Scheme 14 Major types of natural bacteriochlorophylls.

and the central atom of magnesium. To enumerate hydrogen and nitrogen atoms forming the molecule of Bchl a and its derivatives, in this review we use the IUPAC nomenclature. Derivatives of natural Bchl a can be divided into two groups [20]. The former includes compounds, which contain the exocyclic fragment: O

R1

H H

N

N Mg

H

N

N

H

bacteriochlorophyll ɚ: Ɇ=Mg, R1=Me, R2=COOMe, R3= phytyl; bacteriopheophetin a: Ɇ=2H, R1=Me, R2=COOMe, R3= phytyl; bacteriopheophorbide a: Ɇ=2H, R1=Me, R2=COOMe, R3=H; bacteriopyropheophorbide a: Ɇ=2H, R1=Me, R2=R3=H; phytyl – a residue of the alcohol phytol

H R2

O

HO

R3O2C

The second group includes derivatives containing no exocyclic fragment:

18

M.A. Grin and A.F. Mironov O

R1

H

bacteriochlorin ɟ6: R1=Me, R2=R4=COOH, R3= CH2COOH

H NH

N

bacteriochlorin p: R1=Me, R2=R3=R4=COOH H

HN

trimethyl ester of bacteriochlorin p: R1=Me, R2=R3=R4=COOMe

H R4

R3

R2

Sources for production of Bchl a are biomass of the purple bacteria Rh. sphaeroides, Rh. roseapersiana and Rh. capsulata. At our laboratory, Bchl a is isolated from the biomass of Rh. capsulata [21], which contains no other bacteriochlorophylls; this greatly facilitates the isolation and purification of the main pigment [22, 23]. Based on Bchl a, various groups made research, the aim of which was to obtain stable individual bacteriochlorins with spectral characteristics not inferior to initial bacteriochlorophyll and in some cases even exceeding them. New photosensitizers should be less hydrophobic as compared with initial bacteriochlorophyll and be sufficiently well soluble in polar solvents. It is also desirable for these compounds to have functional groups, which enable adding other bioactive molecules to them to increase the tropicity to cancer cells and targeted intracellular transport. The first step on the way to solving these problems was to include an additional anhydride exocycle into the main macrocycle, which led to the increase of stability of the pigment [24]. For this, bacteriochlorophyll a 56 was extracted from biomass of the purple bacteria Rh. capsulata and then, without isolation and additional purification, was oxidized in an alkaline medium by oxygen of the air. The subsequent treatment with hydrochloric acid led to the formation of the anhydride ring and production of bacteriopurpurine 57 (Scheme 15).

C

O

C

H

O H

H

Biomass of Rhodobacter capsulatus

N

A

N

N

H

N

N

HN

H

H H39C20O2C

N

B

Mg H

H NH

H

O

HO2C

O

O

O

H3CO2C 56

57

Ⱥ: isopropanol; B: 1. O2/KOH, 2. HCl

Scheme 15

Production of bacteriopurpurine.

Transformation of the pentanone ring into the anhydride cycle, which includes oxidation of carbon atom C-132 in the exocycle by oxygen of the air in alcoholic solutions of chlorophyll a, was observed well back at the beginning of the 20th century by the founder

19

Synthetic and Natural Bacteriochlorins: Synthesis, Properties and Applications O

O

O B

A N

NH

A

N

NH

N

N

Mg H

N

N

D

C

N

H

HN

H

H

CO2Me O

CO2Me O PhytylO2C

PhytylO2C

56

HR

B

CO2Me O MeO2C

58 O

O

NH

N HN

N

H

61.R= 62.R=

OH OH

O

NH

D

HN

N

H

N

NH

C

HN

N

H

N

N HN

Me H HO HOOC

O

MeO2C

HOOC

63

N

O ROOC

57a. R=H 57b. R=Me 57c. R=Pr

CO2Me O MeO2C

59

E

O

NH

H CO2Me O

O

O

O

N

NH

HN

O

O MeO2C

60

NH

N

N

N

HN

O

CO2Me CO2Me

64 F

N HN

CO2Me CO2Me ROOC

65

A: 0.1% HCl; B: 5% H2SO4/MeOH; C: O2 of the air; D: KOH/CH3OH; E: CH2N2; F: KOH/CH3OH, CH2N2

Scheme 16 Chemical transformations of Bchl a in acidic and alkaline media.

of chlorophyll chemistry, Nobelist R. Willstätter [25]. He introduced the term “allomerization” for the autooxidation process, which is catalyzed by bases [26]. It is established at present that the mechanism of allomerization includes the formation of enol in the pentanone ring, which is then oxidized to form lactone (unstable chlorin), which, in turn, is transformed into the anhydride cycle conjugated with the chlorin macrocycle (purpurine 18). The occurrence of the exocycle, leading to an increase of the conjugation chain in the molecule, causes a batochromic shift of the absorption bands Qx (545 nm) and Qy (818 nm) and the emergence of a purple-red stain, with which the name of the compound (purpurine) is associated [26]. The first to observe allomerization of bacteriochlorophyll a was H. Fischer in 1938 [27]. Recently, A. Scherz et al. reported the formation of 132-hydroxy allomers in a metanol solution of Bchl a [28]. The oxidation process of Bchl a in BP 57 was studied in detail at the laboratory of R. Pandey (Scheme 16) [29]. The treatment of bacteriopheophetine a 58 with a solution of 5% H2SO4 in methanol yielded, along with the required bacteriopheophorbide methyl ester

20

M.A. Grin and A.F. Mironov C A

modification of the acetyl group

O

NH

Bacteriochlorophyll ɚ

D H

H

B

N

N

H

C HN E

H etherification

HO2C

O

O

O modification of the anhydride cycle

Scheme 17 Possible chemical modifications of bacteriopurpurine.

59, a mixture of oxidation products. Analysis of the latter showed it to consist of chlorin 60 and a mixture of diastereomers of 132-(R/S)-hydroxyderivatives 61 and 62 (epimers). The presence of the hydroxyl group in the pentanone ring was reliably proven by the spectra of 1 H NMR; in this case, the ratio between R and S isomers was 1:4. These detailed studies have shown that the S epimer, in which the carbomethoxy group in position 132 and the propionic acid residue in position 17 of the macrocycle are directed to different sides relative to the plane of the macrocycle, is thermodynamically more stable under acidic conditions [30, 31]. However, work with derivatives of Bchl a in acidic media is strongly complicated due to the rapid oxidation of the pigment. The authors also observed allomerization in the presence of alkali. Thus, bacteriopheophorbide 59 in an air-bubbled solution of KOH–propanol is allomerized to form “unstable bacteriochlorin” 63. The mechanism of allomerization under alkaline conditions is unknown, but an indirect proof of the formation of intermediate 63 can be production of bacteriochlorin with the glyoxalic acid residue in position 15 of the macrocycle 64 at the action of diazomethane on bacteriopheophorbide 59. “Unstable bacteriochlorin” 63 is converted into BP in the evaporation of the solvent; both free acid 57a and ester 57c can be formed. As BP 57a possesses a low solubility in organic solvents, it is expedient to convert it into methyl ester 57b, which facilitates chromatographic purification. To increase the yield of BP, our laboratory developed a method of its production without isolating bacteriopheophorbide as an intermediate. In this case, allomerization of Bchl a in an air-bubbled KOH–isopropanol solution takes longer (1.5–2 h), apparently, due to the presence of carotenoids, which are radical traps [32]. For bacteriopurpurine, a number of chemical transformations are possible (Scheme 17). They include the modification of the acetyl group with its reduction to an α-hydroxyethyl group and its conversion into a vinyl group; conversion of the anhydride cycle into the imide cycle; etherification of the propionic-acid residue by alcohols and serine. Bacteriopurpurine is stable only in neutral and acidic media; in the presence of bases, there occurs a rapid opening of the anhydride cycle, which leads to the recovery of the main spectral band in the electronic spectrum. Stability was increased and the spectral characteristics were improved by producing cyclic imides of chlorins and bacteriochlorins by substitution of oxygen in the exocycle for the nitrogen atom. The reaction was adjusted on methyl ester of purpurine 18 66 (Scheme 18) [33, 34]. The latter reacted to hexylamine at room temperature, giving a mixture of


NH

H

N

HN

N

NH

A H

H O

N

HN

N

H O

R1

O

CO2Me

CO2Me

21

R2

67

a. R1 = CO2H, R2 = CONH(CH2)5CH3 b. R1 = CONH(CH2)5CH3, R2 = CO2H

66

B

NH

N

NH

N

C H

HN

N

H

N

HN

R H

H O CO2Me

N C6H13

69

O

X CO2Me

O

Y

68

a. X = O, Y = N(CH2)5CH3 b. X = N(CH2)5CH3, Y = O

A: C6H13NH2; method 1 – B: DCA; ɋ: DBU; method 2 – from 67 to 69: 1. CH2N2, 2. KOH/CH3OH

Scheme 18 Synthesis of cyclic imides in the chlorophyll a series.

isomeric amides in positions 131 67a and 151 67b at a ratio of 6:1 with a yield of 95%. In order to obtain cyclic isoimides, the free carboxyl group of amides was activated using two techniques. In the first technique, a mixture of amides was treated with DCC to form two isoimides (λmax = 690 and 696 nm) at a ratio of 6:1 with the total yield of 96%. In one of them, the nitrogen of hexylamine is in position 131 68a; in the other, in position 151 68b. The mixture of isoimides at the action of diazabicycloundecene (DBU) in toluene in an alkaline medium at 60° is converted into the required cycloimide 69 with a low yield. The second approach includes etherification of intermediate amides by diazomethane, followed by the treatment of formed esters by a KOH methanol solution. In this case, cycloimide 69 is formed with the yield of more than 80%. The authors note that the substitution of DBU by a stronger KOH or NaOH base leads to a further increase of the yield. Optimized conditions were used for the synthesis of bacteriochlorin derivatives (Scheme 19) [34]. The mixture of amides 70a and 70b, obtained as the result of opening the anhydride cycle of bacteriopurpurine by hexylamine, yields unstable carbodiimide

22


O

H

H H

H NH

H

NH

N

H

HN

N

R1

O

O

HN

N

H

H O

N

A

CO2C3H7

CO2C3H7 1

R2

70 2

a. R = CO2CH3, R = CONH(CH2)5CH3 b. R2 = CONH(CH2)5CH3, R2 = CO2CH3

B O

O

H

H

H NH

H

N

NH

N

C H

HN

N

H

H

N

HN

H O

N R

CO2C3H7

72. R = C6H13

O

X

O

Y

CO2C3H7

71 a. X = O, Y = N(CH2)5CH3 b. X = N(CH2)5CH3, Y = O

A: C6H13NH2; method 1 – B: DCA; ɋ: DBU; method 2 – from 70 to 72: 1. CH2N2, 2. KOH/CH3OH

Scheme 19

Synthesis of cyclic imides in the bacteriochlorophyll a series.

derivatives, which are rapidly converted to the more stable cyclic isoimides 71a and 71b. The latter were separated chromatographically into individual isomers (λmax = 804 and 796 nm) at a ratio of 6:1. The base-catalyzed intramolecular cyclization of isoimides led, with a yield of 45%, to cycloimide 72 (λmax = 822 nm). Similar results were obtained with amides, whose carboxyl groups were methylated prior to the treatment by the base. In an alkaline medium, along with the synthesis of the required cycloimide 72, there occurred the oxidation of 12-CH3 group to form minor products: 12-formyl (4–6%) 75 and 12-hydroxymethyl (2–4%) 76 derivatives (Scheme 20). The authors believe that this unusual oxidation is the consequence of enolization leading to tautomers 73 and 74, which are apt to oxidation in air. In turn, keto-enolic tautomerism in cycloimide is possible owing to the strong electron-acceptor effect of the imide exocycle conjugated with the main tetrapyrrole system [35]. An unusual bacteriochlorin 78, having a Qy band at 849 nm, was obtained from 12hydroxymethylcycloimide 76 under acidic conditions (Scheme 21). The 2D ROESY spectra showed the interaction of 10-H meso-proton with the adjacent 12-CH2 group and the absence of 8-H proton. The authors assume cycloimide 78 to have the structure depleted of 8-H proton to shift the double bonds inside the macrocycle. As the mechanism, they propose

23

Synthetic and Natural Bacteriochlorins: Synthesis, Properties and Applications H

HN

N

HN

N

A B N

O

H N

O

O

Hexyl

PrO2C

H

HN

N

H

O

74

HN

N

HN

N

O

Hexyl

PrO2C

73

72

N

O

Hexyl

PrO2 C

CH2OH

CHO and N

O

N

O

O

Hexyl

PrO2C

O

Hexyl

PrO2C

76

75 A: KOH; B: HCl

Scheme 20 A possible mechanism of formation of 12-formyl- and 12-hydroxymethyl cycloimide of bacteriochlorin p in an alkaline medium. O

O

O H

NH

N

N

NH

A

HN

N

N

NH

HN

N

CH2OH

O

N Hexyl

PrO2C

76

N

H

HN

CH2

O

N

O

Hexyl

PrO 2C

77

H

O

N

O

O

Hexyl

PrO2C

78

+

A: H

Scheme 21 Transformations of 12-hydroxymethylcycloimide of bacteriochlorin 76 in an acidic medium.

elimination of the hydroxyl group under acidic conditions to form a carbocation of the benzyl type 77, which is stabilized due to the abstraction of the proton at C-8 [29]. In contrast with natural bacteriochlorophyll a, the isoimide and imide analogues are more stable, and their spectral characteristics compare favourably to respective precursors. The photodynamic activity of the derivatives of chlorin p6 and bacteriochlorin p can be further increased by introducing trifluoromethyl groups into cycloimide molecules (Scheme 22) [36]. Fluorine-containing substituents enhance the solubility of compounds in lipids, which increases the transport rate of such molecules through lipid membranes. The authors note that the activity of PS in vivo is affected not only by the nature of the substituents but also by how they are arranged in the macrocycle. Thus, cycloimide of chlorin p6 79b, containing a bis-(trifluoromethyl)benzyl grouping at the nitrogen atom of the macrocycle, evokes a more efficient inhibition of tumour growth as compared with the isomer containing this

24


NH

NH

A H

HN

N

H

N

N

N

HN

O

N

B H

H

H O

O

O

R

CO2Me

N

HN

NH

O

N

H

HN

O

N

O

R

R

N

a. R = H

N

a. R=CH3 b. R=CF3

D H

N

79

R

OR

N

NH

CO2Me

R=CH3 R=CF3

C

NH

Butyl

H

O

CO2Me

O

N

HN

O

N

H2C

H

CO2Me

O

Butyl

CO2Me

O

CF3

Butyl

b. R =

H2 C

80 CF3

A: 3,5-dimethylbenzylamine or 3,5-bis(trifluoromethyl)benzylamine; B: HBr/CH3COOH, C4H7OH; C: C4H9NH2; D: HBr/CH3COOH, 3,5-dimethylbenzyl alcohol or 3,5-bis(trifluoromethyl)benzyl alcohol

Scheme 22 Synthesis of fluorine-containing cyclic imides of chlorin p6.

substituent in the upper part of the chlorin macrocycle 80b (respectively, 100 and 66% of tumour regression in 90 days). Studies of the relation between the structure and activity were continued on cycloimides of bacteriochlorin p. For this, bacteriopurpurine methyl ester was treated with 3,5-bis(trifluoromethyl)benzylamine. Along with required cycloimide, a Schiff base was formed; the base proved to be labile in in vivo experiments. However, during the reduction of the C=N bond, stable cycloimide 81 with the Qy band in the region of 796 nm was formed [37]. CF3

NH

CF3

H NH H

N

N

HN

O

N

H O

CF3

CO2Me

81 81

CF3

25

Synthetic and Natural Bacteriochlorins: Synthesis, Properties and Applications NOH

O

NOH H

H NH H

N

HO2C

H O

N

H H

H NH

A H

HN

N

N

H NH

A H

HN

N

H O

HO2C

O

O

O

57

HN

H O

HO2C

O

C or D NOH

O

NOR

N HN

NH

A H

H NaO2C

H H

H NH

N

H NH

N H

N

R1O 2C

H O

HN

H COONa COONa

84

2

H

H

N

O N OH 83

82 B

H

N

NaO2C

COONa COONa

85

N HN

O N OR3

86 1

a: R = H; R2 = R3 = COCH3 b: R1 = R3 = CH3; R2 = H

A: NH2OH.HCl, Py; B: NaOH/CH3OH; C: Ac2O; D: CH2N2, (C2H5)2O

Scheme 23 Interaction of bacteriopurpurine with hydroxylamine.

Our laboratory was the first to perform the synthesis of cycloimides of chlorins and bacteriochlorins by acting with highly nucleophilic agents hydroxylamine and hydrazine hydrate on bacteriopurpurine. Initially, for modification of bacteriopurpurine we used our earlier proposed method of converting purpurine 18 to N-hydroxycycloimide [38, 39]. Herewith, the presence of the acetyl group in bacteriopurpurine instead of the vinyl substituent in purpurine 18 leads to the emergence of an additional reaction centre. Conducting the reaction of bacteriopurpurine with hydroxylamine, it was shown (Scheme 23) that oxime 82 was initially formed [40]. To prove this, we performed chemical conversions, which included the opening of the anhydride cycle in oxime 82 and the treatment of bacteriochlorin p 84 by hydroxylamine in pyridine. The identity of the products obtained indicates that in the treatment of BP by hydroxylamine the first to enter into the reaction is the acetyl group. Subsequent studies of the interrelation between bacteriopurpurine and hydroxylamine have shown that, when excess reagent is used and time is increased up to 10 h, the second molecule of hydroxylamine enters into the reaction with the anhydride cycle to yield oxime of N-hydroxycycloimide of bacteriochlorin p 83. The latter compound can be obtained both from bacteriopurpurine 57 and from intermediate oxime 82. The course of the reaction was followed chromatographically and spectrally by the change of position of the long-wave Qy band. In the first three hours, the maximum at 818 nm shifted to 792 nm, which corresponds to the formation of oxime 82, after which it gradually returned to the long-wave region (812 nm). We believe that at this stage oxime interacts with the second molecule of hydroxylamine to form compound 83. Thus, a derivative of hydroxamic acid was first obtained in

26

M.A. Grin and A.F. Mironov OH N

N

H

H

HO

NH

N

N

NH

H

N

HN

N

H

HN

H

H

H

H O

O

N

CO2Me

O

N

CO2Me

OMe

O

OMe

Scheme 24 Syn- and anti-isomers of oxime 86b.

O

O

H

NH

H

NH

A H

HN

H

57

N

CO2H CO2H CO2H

E H

C or D

OH

H

H

H NH

F

N NH

H

HN

HN

88

87

CO2H

N

N

H

H NH

N

CO2H CO2H

H

N

HN

N

H

N

HN

H

H O CO2R

H

NH

B

HN

N

O

O

CO2H

H

H

N

H

H O

H

H

N

N

OH

H

H

N

O

O

OR1

O

O

H O

CO2R

91

90

a. R=H, R1=H b. R=CH3, R1=H G c. R1=R2=CH3 d. R=CH3, R=Ts

a. R=H b. R=CH3 c. R=C2H5

CO2H

O

O

89

A: NaOH, CH3OH; B: NaBH4; C: HCl, dioxan; D: TsCl, C5H5N; E: TsOH, CHCl3; F: NH2OH HCl, Py; G: CH2N2, Et2O; H: TsCl, Py

Scheme 25 Synthesis and modifications of 3-vinyl-3-deacetylbacteriopurpurine 90.

the bacteriochlorophyll a series. The derivative, having a mobile hydrogen atom in its composition, readily enters into the acylation and alkylation reactions [41]. Thus, the treatment with acetic anhydride leads to the formation of diacetate 86a, which proved rather labile and in storage was decomposed into monoacetate and the initial compound 83. The treatment of oxime of hydroxamic acid 83 by diazomethane yields N-methoxycycloimide 86b with a good yield. This compound proved much more stable than acetate and showed a good photodynamic activity in in vitro and in vivo experiments. Two interesting facts were discovered in the course of the studies. The acidity of the hydrogen atom


27

in the oxime function is decreased so much that the function is not involved in the reaction with diazomethane. Besides, comprehensive analysis (TLC, mass spectra) of N-methoxy derivative 86b showed the substance to be a mixture of two isomers with very close values of Rf. The isolated isomers were characterized by the 1H NMR spectra, including 1D NOE spectroscopy; based on the data obtained, it was concluded that oxime 86b existed in the form of two stereoisomers (sin- and anti-) (Scheme 24). Such isomers are formed in the interaction of ketones with hydroxylamine [42]. As the acetyl group in BP strongly complicates the progress of the reaction with hydroxylamine, it was decided to convert it into the vinyl group. The vinyl group in compound 90 was obtained from the α-hydroxyethyl group in alcohol 88 under the action of p-toluenesulfo acid (Scheme 25) [43]. The acetyl group was reduced not on bacteriopurpurine 57, but on bacteriochlorin 87, as the treatment of bacteriopurpurine with sodium boron hydride had been earlier shown to contribute to the conversion of the anhydride exocycle into the δ-lactonic one [44]. Triacid 87 obtained from BP was reduced by sodium boron hydride to respective alcohol 88, which has an absorption maximum at 740 nm. The treatment of the latter by hydrochloric acid in dioxane led only to the closure of the anhydride cycle, but not to the dehydration of the α-hydroxyethyl group. Compound 89 was also observed to be formed at the action of tosylchloride on alcohol 88; if the reaction was continued for more than 2 h, the predominant product formed was O-tosylate. A more efficient method of producing the vinyl group was developed based on the treatment of alcohol 88 by p-toluenesulfo acid. Herewith, it was found that, along with the formation of the vinyl group, there occurs the closing of the anhydride exocycle in compound 90, which leads to the recovery of the main spectral band Qy to the region of 783 nm. Interestingly, if the reaction was performed in chloroform, the product was obtained as free acid 90a, whereas addition of methyl or ethyl alcohols directly to the reaction medium led to the rapid etherification of propionic acid residue to form respective methyl 90b or ethyl 90c esters. This activity of the carboxyl group is, evidently, due to the formation of the mixed anhydride of the residue of propionic acid and p-toluenesulfo acid. 3-Vinyl-3-deacetylbacteriopurpurine 90a is a structural analogue of purpurine 18 and, for this reason, it was of interest to perform the above-described reaction with hydroxylamine to assess the effect due to the conversion of the acetyl group into the vinyl group. The treatment of compounds 90a and 90b by hydroxylamine produced respective N-hydroxycycloimides 91a and 91b. Although the absolute values of the wavelengths of the Q band in the spectra of N-hydroxycycloimides 86b and 91c are very close to 812 and 809 nm, the relative change of the long-wave absorption in the reaction products strongly differs as compared with the initial substances. In the case of bacteriopurpurine, the formation of oxime by the acetyl group leads to a virtually complete levelling down of the effect from the introduction of the imide exocycle. As the result, the terminal cycloimide is only 6 nm worse in the long-wave absorption than the initial pigment. In the case of the vinyl derivative of bacteriopurpurine, transformation of the anhydride cycle to the imide cycle gives a significant increment of the wavelength of the Q band up to 27 nm. Recently, Sasaki and Tamiaki [45] synthesized a series of derivatives of bacteriopyropheophorbide 92, which contain various substituents in pyrrole A. The vinyl group was obtained by a different method as compared with that described above. The high yield of compound 94 was achieved at the expense of the outgoing mesyl group. Further on, the vinyl group was oxidized to the formyl group 95 using OsO4 and the subsequent splitting of diol NaIO4 (Scheme 26). As the reactivity of the formyl group in

28

M.A. Grin and A.F. Mironov 0

NH

Rhodobacter sphaeroides (purple bacterium)

A

N

C

B

BChl - a

HN

N

O CO2Me 92 OH

NH N

H

Ph

D

N

N

HN

O CO2Me 93

NH

N

E

HN

0

NH N

N HN

O CO2Me

O CO2Me

94

95

Scheme 26 Synthesis of bacteriopyropheophorbide derivatives with various substituents in pyrrole A.

the chlorin series has been well studied, the authors carried out a number of similar transformations in the bacteriochlorin series, which ncluded the oxidation and reduction of the formyl group to the carboxyl and hydroxymethyl groups to form compounds 96 and 97, as well as the Knoevenagel reaction with dicyanomalonic ester, which led to compound 98 (Scheme 27). Studies of N-hydroxycycloimides of bacteriochlorin have shown that the hydroxyl group can be successfully used to produce alkyl-substituted derivatives. At the same time, introduction of acyl radicals by this way is less promising due to the lability of such derivatives. In this connection, we developed a method of producing N-aminocycloimides of bacteriochlorin p [46]. As is known, hydrazine and its derivatives possess a high nucleophilicity with respect to sp2-carbon atoms [47]. For this reason, anhydrides of acids are convenient reagents for acylation of hydrazine. In the case of BP 57, we have shown (Scheme 28) that its treatment by hydrazine hydrate in pyridine initially leads to monohydrazide, evidently, in the form of two isomers 99a and 99b. The acetyl group in this case is converted to hydrazone [45]. In the subsequent treatment of the reaction mass by HCl, there occurs the intramolecular cyclization to form an additional six- or seven-membered cycle 100a or 100b. Hydrazone at this stage was converted to initial ketone. It is known that during the acylation of hydrazine after the addition of the first acyl group the subsequent acylation, due to a partial deactivation of the first nitrogen atom, usually occurs by the second amino group [48]. Thus, in the interaction of hydrazine with anhydrides of aromatic dicarboxylic acids symmetric hydrazides are formed with the increase of the initial five-membered cycle to the six-membered cycle (Scheme 29) [49]. In the case of bacteriopurpurine, this could lead to the formation of hydrazide with six100a or seven-membered 100b cycles. The use of 2D heteronuclear resonance and the study


29

O 0

0

HO

NH

NH

N

O

O CO2Me

HN

N

HN

N

N

CO2Me

96 C

A OH

0

H

NH

N

B

N HN

N

HN

N

NH

O CO2Me

O CO2Me 97

95 D

E

CN OH NC

NH

N

N

N Zn

N

HN

N

O CO2Me

N

O CO2Me

98

A: NH2SO3H/NaClO2, THP, 2-methyl-2-butene; B: t-BuNH2*BH3, CH2Cl2; C: AcOH/EDC*HCl/DMAP, CH2Cl2; D: malonic acid dinitrile, Et3N, THP; E: Zn(OAc)2, CH2Cl2/CH3OH

Scheme 27

Chemical transformations of 3-formyl-3-deacetylbacteriopyropheophorbide.

of 1H NMR spectra of the obtained compound at various temperatures, as well as chemical modifications including the formation of Schiff bases, alkylation and acylation of the amino group confirmed the six-membered structure of the exocycle 100a. The developed method of producing N-aminocycloimide derivatives of bacteriochlorophyll a is distinguished by simplicity and high yields [50].

30

M.A. Grin and A.F. Mironov N NH2

O H

H H

H

N

NH

NH

A H

N

HN

H

H

2 COR1 COR

O

O

HN

N

H O

N

CO2H

CO2H

57

99 a: R1=NHNH2; R2=OH b: R1=OH; R2=NHNH2

B,ɋ O

O

H

H

H

H NH

H

NH

N

H

HN

N

N

HN

N

H

H O CO2Me

N

O

NH2

O

N N

CO2Me

H H

O

100b

100a .

A: N2H4 H2O, pyridine; B: 1N HCl; C: CH2N2, Et2O

Scheme 28 Reaction of bacteriopurpurine with hydrazine hydrate. O

O O R

O

+

NH

NH2 NH2 R

NH O

Scheme 29 Formation of cyclic hydrazides in the interaction of hydrazine with anhydrides of aromatic dicarboxylic acids.

3

Amphiphilic and Water-soluble Derivatives of Bacteriochlorins

Natural bacteriochlorins are known to be distinguished with an increased hydrophobicity [2]. To be successfully used in PDT, they should have a more balanced ratio of hydrophobic and hydrophilic substituents in the macrocycle. Usually, this is achieved by introduction of one, two or three carboxyl groups, amino acid residues or hydroxyl-containing functions into the molecule [51–54]. Thus, introduction of serine by the residue of propionic acid by enzyme re-etherification of Bchl a makes it possible to significantly increase its solubility in water. Such conjugates of bacteriochlorophyllide a (Bchlfd a) with serine preserve the photophysical properties of Bchl a, generate AOS with a high yield, but, unfortunately, are subject to photooxidation, demetallation in a weakly acidic medium and biodegradation, which restricts their use in clinic. Substitution of the central atom of Mg by Pd, etherification of the propionic acid

31


O H

H H

N

N

A

Mg H

H N Mg

N

H

H

N

O

N

H

HO2C

HO2C MeO2C

N

O

101

O

H2N

O

MeO2C

102

B

O

O

H

H

H NH

N

H N

C

N Pd

H

N

HN

H

HO2C

O O

MeO2C

N

103

N

H

HO2C

O H2N

H

O H2N

O O

MeO2C

104

A: serine; B: HCl; C: Pd(CH3COO)2

Scheme 30

Production of complexes of bacteriochlorophyll a derivatives with serine.

residue and re-etherification of the carbomethoxy group in the pentanone ring lead to stable derivatives of [Pd]-Bchlfd a with a high photodynamic activity [55]. The quantum yield of AOS for these compounds is sufficiently high – from 1 in nonpolar solvents down to 0.5 in aqueous solvents. Bacteriochlorophyllide a 101 and its complex with serine 102 showed a high anti-tumour activity on the M2R mouse melanoma cell line (LD50 0.2–0.5 µM) (Scheme 30) [56]. Substitution of Mg by Pd 104 increased the photodynamic activity, which led to a decrease of the value of LD50 to 0.01–0.03 µM. Besides, owing to the high fluorescence of 102 in cancer tissue as compared with healthy tissue (the selectivity reaches 8–10), the latter, along with PDT, can be also used for fluorescent diagnostics (FD). A series of new negatively charged water-soluble derivatives of Bchl a was obtained at the Weizmann Institute of Science, Israel. These compounds proved promising for vascular-targeted photodynamic therapy (VPT). The damage of vessels providing for the blood supply of the tumour was found to be the essential factor of its necrosis. VPT is efficient in the treatment of hard tumours, as well as non-tumour processes associated with increased vascularization, e.g., age-related macular degeneration. Regression and necrosis of the tumour occur as the result of both the death of cancer cells and due to the occlusion or perforation of tumour vessels. However, hemorrhagic necrosis of the tumour, caused by the above-named water-soluble derivatives of bacteriochlorophyll a, can have undesirable consequences, especially when tumours are localized in such vital organs as bronchi and lungs. The method of VPT leads to necrosis of the central part of the tumour, which can spread to 95% of tumour’s volume. However, cancer cells on the periphery of the tumour can fail to

32


O

N

N

N

M N

N

CO2CH3 R

N Pd

N

O CO2-K+

O

N

O CO2CH3

NH

SO3- K+

109

56:M=Mg, R=phytyl 105:M=2H, R=OH 106:M=Pd, R=OH (WST09, Tookad) 107:M=Pd, R= O-succinimide-SO3-Na+a+ 108:M=Pd , R=NH-(CH2)3-SO3-Na+ SO3- K+ N

O

N

N

N

N

O

NH CO2CH3 SO3- K+

110

N

O

NH SO3- K+

O

111

N

NH

CO2CH3 SO3- K+

N Pd

N

N

N

N

O

N Pd

Pd

NH

O

SO3- K+

CO2-K+

O NH CO2CH3

SO3- K+

112: M = Pd 113: M = 2H 114: M = CuII 115: M = Zn 116: M = MnIII

Scheme 31. Negatively charged PS based on bacteriopheophorbide and its metal complexes.

die and grow again. In this connection, the VPT method is not universal and can be considered only in combination with chemotherapy, radiotherapy and photodynamic therapy. Water-soluble negatively charged PS were obtained by aminolysis of the pentanone exocycle in bacteriopheophorbide and its metal complexes (Scheme 31). Their hydrophilic properties and capability of aggregation in aqueous solutions was studied [57]. Amphiphilicity was assessed by the octanol/water distribution coefficient (P). The effect of peripheral substituents on the ability of obtained compounds to be solved in water and polar organic solvents was shown. It was found that the opening of the exocycle in bacteriopheophorbide significantly increased the hydrophilicity of bacteriochlorins and, as a consequence, enhanced their ability to be solved in polar organic solvents, including methanol, ethanol, DMFA and DMSO. Besides, compounds with the open cycle 109–116 are well soluble in aqueous solutions (PBS) – up to 40 mg/ml, which is much greater than the solubility of compounds with the pentanone ring, but having polar substituents on the periphery of the macrocycle (compound 108, 4 mg/ml). The coefficient P strongly increases (1:19) at the substitution of the central metal ion Pd(II) by Mn(III) 116. At the same time,


33

a decrement of the hydrogen chain in the alkyl substituent 109 as compared with 112 does not in practice affect the distribution coefficient. All compounds with the open exocycle in aqueous solutions form aggregates (2–8 molecules), which dissociate upon dilution, inclusion into micelles, as well in solutions with physiological concentrations of serum albumin. Among new PS of the bacteriochlorin series developed at the Weizmann Institute of Science in cooperation with Steba-Biotech (France) and Negma-Lagards (France), we should note the second-generation Tookad, which is a Pd complex of bacteriopheophorbide 106. Tookad exhibits a high activity to such tumours as rat glioma [58], prostate cancer [59], HT29 human colon carcinoma in experiments on laboratory animals. The main mechanism of action includes tumour vessel damage, which leads to hypoxy and necrosis of the tumour. At present, the preparation is at the second stage of clinical tests, which are carried out at medical centres of Canada, Europe and Israel. As Tookad is poorly soluble in aqueous solutions (octanol:water = 24:1), it is introduced as a suspension with Cremophor. At the action on the Pd complex of bacteriopheophorbide by taurine (2-sulfoethylamine) there occurs the opening of the pentanone cycle and the formation of dianion (code name WST11) 112 [60]. The preparation possesses a good solubility in phosphate buffer (up to 50 mg/ml), where it is present in the shape of small aggregates. In blood serumcontaining solutions, it is subjected to deaggregation and occurs as monomers in a complex with serum albumin (BSA) and high-density lipids.

Absorption

O

N

N Pd

N

nm

N

O N CO2CH3 H CO2H

SO3Na

112

To increase the hydrophilicity of cycloimides of bacteriochlorin p, we synthesized derivatives, which in pyrrole A contain hydroxyl-containing substituents attached to the macrocycle by an ether bond (Scheme 32). For this, initial bacteriopurpurine 57 was transformed into bacteriochlorin p 87, the acetyl group in which was reduced by NaBH4 88. The subsequent closing of the anhydride cycle using TFA 89 and the treatment by hydroxylamine and diazomethane yielded 3-(α-hydroxymethyl)-3-deacetyl-N-hydroxymethylcycloimide of bacteriochlorin p 117. The hydroxyl group in pyrrole A was activated by means of trifluoroacetic acid anhydride; the obtained trifluoroacetate 118 was condensed with the methyl ester of ethylene glycol, di- and triethylene glycol and glycerol to yield compounds 119–122. Experiments in vitro on HeLa and A549 cell lines and in vivo on lymphoma mice have shown that the phototoxicity of the above-named compounds is 20 times as high as that of the initial bacteriopurpurine [61]. The cause of the so high activity of preparations obtained is due to the increased selectivity to cancer cells (tropicity 8–13) and the significant quantum yield of the generation of singlet oxygen (0.54–0.57). Analysis of the distribution of PS in tissues showed the largest accumulation in tumour vessels, leading to hypoxy and necrosis of the tumour, which

34

M.A. Grin and A.F. Mironov O OH

C CF3

O

NH

N

H NH

N

H

HN

N

H

HN

N

O

O

N

O

O

N

CO2Me

OMe

CO2Me

OMe

CO2Me 117

118

119

HN

N

H

H

H O

N

B

A H

H

H

H NH

R1

H

H

119 - 122

121

O

O

N OMe

O

O

O

O

OH

HO 120

O

122

OH

O

OH

O

A: (CF3CO)2O; B: 1. CH2(OH)CH2OMe, 2. HO(CH2CH2O)2H, 3. HO(CH2CH2O)3H, 4. HOCH2CH(OH)CH2OH

Scheme 32 Synthesis of cycloimides of bacteriochlorin p with polar substituents in pyrrole A. O

O

O H

H H

NH

H

N

NH

N

H

HN

O

N

H

N

NH

HN

H

H

H MeO2C

H H

N NH2

O

MeO2 C

N

N

HN

H O

N

O

HN

O

MeO2C

O

O

N

O

HN

C

C

N

N

I

A: isonicotinic acid chloroanhydride, Py; B: CH3I, boiling

Scheme 33 Synthesis of cationic cycloimide of bacteriochlorin p.

is absolutely consistent with the data by Israeli investigators for the action mechanism of bacteriochlorophyll a derivatives. Much less is known of the introduction of positively charged substituents into bacteriochlorins. However, just these photosensitizers may prove to be the most efficient in photodynamic antimicrobial therapy. The method is based on the inactivation of viruses, bacteria, yeasts and protozoa by active oxygen species, which are generated by photosensitizers in illumination [62]. Available literature data show that this method is considerably behind photodynamic

35


O

H

NH

O

H

H

H

H

N

NH

H

N

NH

N

C HN

N

H

H

H

N

HN

H

H O CO2Me

O

O

N

O

HN

O

CO2Me

N

O

O

N

CO2Me

H 3C

N

A

B O

O

NH

N

HN

O

N

H

O

NH

N

HN

O

N

O

N

O

H

OH

N

N

HN

O

N

H CO2Me

HN

H

N

H

H

H

H

H

N

O

H

H

NH

O

NHMe

127

125 N

CO2Me

HN

H

N

123

H

N

CH3

CO2Me

O

NH2

100 N

N

123 a

126

Scheme 34 Chemical transformations of cycloimide 123.

therapy of cancer by the level of fundamental elaboration and practical application. There are only separate data on the photosensitization of nonpathogenic yeasts in the presence of porphyrins and phthalocyanines [63–65]. As is known, the outer surface of bacteria carries a negative charge, in connection with which the most efficient coupling to bacterial cells and the photodynamic action is expected from cationic photosensitizers. Such photosensitizers based on N-cycloimides of bacteriochlorin have been synthesized at our laboratory. Initially, it was planned to use for these purposes the N,N-dimethylamino derivative. However, the treatment of it by excess methyl iodide or dimethyl sulfate failed to lead to the quaternization of the nitrogen atom [46]. A more successful effort was introduction into the N-amino derivative 100 of the isonicotinic acid residue 123 followed by the quaternization of the nitrogen atom in the pyridine ring 124 (Scheme 33) [66]. For this, cycloimide 100 was treated with isonicotinic acid chloroanhydride in pyridine. We showed hydrazide obtained to exist in two isomeric forms 123 and 123a, which are formed owing to keto-enolic tautomerism or due to the absence of free rotation around the bond C(O)–N.

36


O

H

H

H NH

H

H

N

A

N

HN

O

N

O

HN

O

NH

B

100

N

H

H

N

HN

H

CO2Me

O CO2Me

128

N

O

HN

O

129 N N

C

C

O

O

H

H

H NH

H

H

N

NH

N

HN

O

N

O

H2N

O

N

H

H

N

HN

H

CO2Me

130

O CO2Me

I

131

N

N

O

H2N

O

N

I

A: 6-methylnicotinic acid chloroanhydride, pyridine; B: 6-methylpicolinic acid chloroanhydride, pyridine; C: CH3I, boiling

Scheme 35 Synthesis and quaternization of cycloimides with residues of pyridinecarboxylic acids.

O

N

O

N

O

H N Me

Figure 2 Formation of a five-membered ring in hydrazide 129.

To establish the structure of the isomers, chemical modification of cycloimide 123 was carried out; it included the treatment of the latter by diazomethane (Scheme 34). As the result, two substances with the same molecular mass but different chromatographic mobility were obtained, which enabled their isolation in the individual form. A more mobile isomer was assigned the structure of N-methylhydrazide 125; and the other isomer, of O-methylimidate 126.


37

The structure of N-methylhydrazide 125 was confirmed by countersynthesis. For this purpose, N-aminocycloimide 100 was treated with a stoichiometric amount of methyl iodide, and N-monomethylaminocycloimide 127 was condensed with isonicotinic acid chloroanhydride. The obtained compound by its molecular mass and chromatographic mobility proved to be identical to the fast-moving isomer 125. The other isomer 126 proved rather labile and rapidly degraded to form N-aminocycloimide 100. The 1H NMR spectra of the fast-moving isomer 125 preserved the signal doubling similar to that observed for initial hydrazide 123. This phenomenon is, apparently, due to the existence in amides of one more type of isomerism, which emerges owing to the absence of free rotation around the bond C(O)–N. As the rotation barrier for such isomers is comparatively low, it can be overcome by increasing the temperature. Indeed, during the recording of the 1H NMR spectra for compounds 123 and 125 at 50° the signals merged. At the same time, in isomer 126, whose 1H NMR spectrum was recorded in the first hour after isolating the substance, there was no such signal doubling, which corresponds to the presence of only one Z or E isomer. The cationic photosensitizer 124 is a much more hydrophilic compound as compared with initial bacteriochlorins 100 and 123. In this connection, it is characterized by a better solubility in water–alcoholic solutions, which makes it more promising for medical applications. Along with isonicotinic acid, other pyridinecarboxylic acids, as well as more complex compounds of the quinoline series were used for the synthesis of cationic photosensitizers [66]. Interaction of N-aminocycloimide 100 with chloroanhydrides of 6-methylnicotinic and 6-methylpicolinic acids occurred similar to the reaction with isonicotinic acid, and compounds 128 and 129 were obtained with high yields (Scheme 35). However, an attempt of the quaternization of the obtained hydrazides showed significant differences in their reactivities. In the case of the hydrazide of 6-methylnicotinic acid 128, quaternization was successful, and cationic cycloimide 130 was obtained with a high yield. In contrast, compound 131 was not formed even at multi-hour boiling of 6-methylpicolinic acid hydrazide 129. Apparently, this is due to the formation of the five-membered cycle (Fig. 2) at the expense of the intramolecular hydrogen bond between nitrogen atoms of the heterocycle and amide hydrogen. The data of the 1H NMR spectra, which lack the proton signal doubling, are also in favour of the rigid fixation of the pyridine ring in the plane of the amide bond. We observed similar differences in chemical activity also for hydrazides with residues of quinolinecarboxylic acids (Scheme 36). Condensation of N-aminocycloimide 100 with chloroanhydrides of 2- and 6-quinolinecarboxylic acids yielded hydrazides 132 and 133, of which only compound 132 was converted into quaternary base 134. A logical follow-up of the above-described research into the synthesis of hydrophilic photosensitizers of the bacteriochlorin series was the development of the routes of synthesizing zwitterionic cycloimides, which have the N-methylpyridine and carboxyl groups in the lower part of the macrocycle 137. The methyl ester of propionic acid residue in cycloimide 124 was hydrolyzed to introduce a negative charge spatially close to the cationic grouping (Scheme 37). However, the traditional method of hydrolyzing esters in an alkaline medium was unacceptable due to the lability of the imide exocycle under such conditions. Therefore, hydrolysis of cycloimide methyl ester was performed in an HCl/dioxane medium in an argon atmosphere. Another route of producing zwitterionic cycloimides assumed the use of

38


O

H

H

H NH

H

N

H

A

N

HN

O

N

O

HN

O

100

NH

B

HN

N

H

H

N

H

CO2Me

O CO2Me

132

N

O

HN

O

133 N

N

C

C

O

O

H

H

H NH

H

H

N

NH

N

HN

O

N

O

HN

O

H

H

N

N

HN

O

N

O

HN

O

H

CO2Me

CO2Me

N

135

134

I

I N

A: 6-quinolinecarboxylic acid chloroanhydride, pyridine; B: 2-quinolinecarboxylic acid chloroanhydride, pyridine; C: CH3I, boiling

Scheme 36 Synthesis and quaternization of cycloimides with residues of quinolinecarboxylic acids.

free acid in the synthesis (Scheme 38). For this purpose, N-aminocycloimide of bacteriochlorin with the residue of propionic acid 138 was taken as the initial compound. All stages of the synthesis, including acylation by the chloroanhydride of isonicotinic acid 139 and the subsequent quaternization of the nitrogen atom 136, were performed at the presence of an unprotected carboxyl group in the molecule. This route proved to be less successful, and the total yield did not exceed 18%. The obtained betaine derivative of cycloimide of bacteriochlorin 137 possessed a sufficiently high solubility in water, which made it possible to abandon Cremophor usually

39


O

O

H

H H

NH

H H

N

NH

H

N

NH

A H

N

HN

O

N

N

HN

O

N

CO 2H

H 2N

H

H

ɋ

124

H

O

H 2N

O

N

HN

N

H

H

CO 2Me

N

B

136

O O

N

I

O

N

CO2

O

H 2N

137

I

O

N

A: (CH3)2CO/HCl, Ar, t; B: KOH/H2O; C: HCl

Scheme 37

Synthesis of zwitterionic cycloimide of the bacteriochlorin series.

O

O

H

H H

NH

H NH

N

N

A H

N

B H

HN

N

H

H HO2C

HN

O

N

138

HO2C

O

O

NH2

N

O

HN

O

139

O

O

N

H

H H

NH

N

H NH

C

N

D H

N

H

HN

N

H HO2C

HN

H O

O

O

N

O

O

HN

O

N

O

HN O

136

137 N

x

N

Scheme 38 Synthesis of zwitterionic PS.

used for dissolution of hydrophobic compounds and possessing an intrinsic toxicity [67, 68].

40


Introduction of monosaccharides into bacteriochlorin derivatives also makes it possible to regulate the amphiphilicity of photosensitizers. Besides, carbohydrate-containing porphyrins are capable of selective accumulation in neoplastic tissues, by specifically interacting with receptors on the surface of tumour cells [69–71]. We have performed the synthesis of a new carbohydrate-containing photosensitizer based on cycloimide of bacteriochlorin p. As the initial compound, we took the above-described cycloimide containing the residue of isonicotinic acid 123. The carbohydrate component – 6-dehydroxy-6iodo-D-galactopyranoside – was obtained by the scheme including the reaction of D-galactose peracetate with bromoethanol under conditions of acidic catalysis, substitution of the bromine atom by iodine by boiling with sodium iodide in acetone and the subsequent elimination of protective groupings under the action of 0.1 M solution of sodium methylate. Introduction of the carbon fragment into the pigment molecule was performed by way of quaternization of the nitrogen atom of the pyridine ring by the above-described derivative of galactose. Glycoconjugate 140, as well as its acetylated analogue were obtained with a high yield (70%) using 1-O-(2-iodoethyl)2,3,4,6-tetra-O-acetyl-β-D-galactopyranoside. Due to the positive charge and the hydrophilic carbohydrate residue, the compound obtained is well soluble in water and, as shown by preliminary biological tests, can be considered as a promising photosensitizer for photodynamic therapy of cancer. O H H NH

HN

N

H

N

H O

N

CO 2Me

OH

O

N O

HO

O

HN

O

X

OH OH

140

Besides the development of PS themselves for PDT of cancer, of great interest in the recent years is the development of means of delivering PS inside cancer cells to damagesensitive targets. Targeted intracellular transport can help to achieve a significant enhancement of photodynamic action, which makes it possible to reduce the concentration of the drug. Some of such endocyted conjugates are capable of accumulating in lysosomes or penetrating into the nuclei. The group of Prof. A.S. Sobolev (Moscow) develops targeted delivery systems, which include the model ligand (insulin, alpha-melanocyte-stimulating hormone), photosensitizer (chlorin e6, bacteriochlorin p), nuclear localization signal of SV-40 large T-antigen, and, as a carrier, E. coli hemoglobin-like protein (Fig. 3) [72–75]. Earlier studies by this laboratory have shown that the conjugate of bacteriochlorin p


His - tag

D -Tox

HMP

NLS

41

D - MSH

His-tag – (His)6; D-Tox – translocation domain of diphtheria toxin; HMP – E.coli hemoglobin-like protein; NLS – nuclear localization signal of SV-40 large T-antigen; α-MSH – alpha-melanocyte-stimulating hormone. The pocket for porphyrins is in HMP. Figure 3

Design of a targeted PS delivery system.

with the modular recombinant transporter (MRT) carrying the alpha-melanocyte-stimulating hormone for its delivery to melanoma cell nuclei has a phytocytotoxicity exceeding this parameter in free photosensitizers 230 times [76]. Recent works by this group showed [77] that the binding of bacteriochlorin p with MRT carrying the epidermal growth factor (EGF) increased the activity of PS of the order of 1000 times (IC50 = 3000 nM for bacteriochlorin p and IC50 = 4.2 nM for the conjugate) for human carcinoma A431 cells. The authors explain this increased activity by an enhanced expression of EGF receptors on the surface of cancer cells.

4

Conclusion

This review has a chemical trend, as it was written by organic chemists. It does not include numerous works on the biological tests of bacteriochlorophyll a derivatives. It is beyond doubt that these compounds are in extreme demand by oncologists and specialists of other medical fields (dermatology, ophthalmology) in view of their prospective use as photosensitizers for photodynamic therapy. The first works on the synthesis of bacterio-derivatives, in which our group was also involved, appeared 10–12 years ago; now some of these compounds are at various stages of clinical tests. All this is indicative of rapid developments in this field of the chemistry of tetrapyrrole compounds. It is perfectly evident that for the successful use of PDT in clinical practice health processionals should have a series of PS with various therapeutic absorption windows. Derivatives of bacteriochlorophyll a, which absorb in the near IR region of the spectrum, have their own field of use, where other PS prove little efficient.

References 1. D. Dolphin, Can. J. Chem., 72, 1005 (1994). 2. R. Bonnett, Chemical Aspects of Photodynamic Therapy, Gordon and Breach Science Publishers, UK. 3. C. Musewald, G. Hartwich, F. Pollinger-Dammer, H. Lossau, H. Scheer and M.E. Michel-Beyerle, J. Phys. Chem. B, 102, 8336–8342 (1998). 4. H.W. Whitlock (Jr.), R. Hanauer, M.Y. Oester and B.K. Bower, J. Am. Chem. Soc., 91, 7485 (1969). 5. R. Bonnett, R.D. White, U.-J. Winfield and M.C. Berenbaum, Biochem. J., 261, 277–280 (1989). 6. R. Bonnett, P. Charlesworth, B.D. Djelal, S. Foley, D.J. McGarvey and T.G. Truscott, J. Chem.

42


Soc. Perkin Trans., 2, 325–328 (1999). 7. H.L. Callot, A.W. Johnson and A. Sweeney, Chem. Soc. Perkin Trans., 1, 1424 (1973). 8. P.K. Chang, S. Cotiriou and W. Wu, J. Chem. Soc., Chem. Commun., 1213 (1986). 9. R.K. Pandey, F.-Y. Shiau, A.V. Sumlin, T.J. Dougherty and K.M. Smith, Bioorg. Med. Chem. Lett., 4, 1263 (1994). 10. A.N. Kozyrev, R.K. Pandey, C.J. Medforth, G. Zheng, T. J. Dougherty and K.M. Smith, Tetrahedron Lett., 37, 747–751 (1996). 11. R.K. Pandey, M. Isaac, I. MacDonald, C.J. Medforth, M.O. Senge, T.J. Dougherty and K.M. Smith, J. Org. Chem., 62, 1463 (1997). 12. R.K. Pandey, F.-Y. Shiau, M. Isaac, S. Ramaprasad, T.J. Dougherty and K.M. Smith, Tetrahedron Lett., 33, 7815 (1992). 13. R K. Pandey, F.-Y. Shiau, K. Ramachandran, T. J. Dougherty and K. M. Smith. J. Chem. Soc. Perkin Trans., 1, 1377 (1992). 14. P. Yon-Hin, T. P. Wijesekera and D. Dolphin, Tetrahedron Lett., 32, 2875 (1991). 15. G. Zheng, A.N. Kozyrev, T.J. Dougherty, K.M. Smith and R.K. Pandey, Chem. Lett., 1119, (1996). 16. A.R. Morgan, D. Skalkos, G.M. Grabo, R.W. Keck and S.H. Selman, J. Med. Chem., 34, 2126 (1991). 17. U. Eisner, J. Chem. Soc., 3461 (1957). 18. Chlorophyll, ed. by H. Scheer, CRC Press Jnc. (1991). 19. J. Deisenhofer and H. Michel, The Photosynthetic Reaction Centre of Purple Bacteria, Moscow (1990) (translated from German). 20. Chemical Encyclopedia, Bolshaya Rossiyskaya Entsyklopediya Publishers: Moscow, vol. 5, pp. 572–579 (1998) (in Russian). 21. A.F. Mironov and A.V. Efremov, Russian Federation Patent No 2,144.085, 12 July (1996) (in Russian). 22. A.A. Tsygankov and T.V. Laurinavichene and I.N. Gogotov, Biotechnol. Tech., 8, 575–578, (1994). 23. A.A. Tsygankov, T.V. Laurinavichene, V.E. Bukatin, I.N. Gogotov and D.O. Hall, Biochem. Microbiol., 33, 485–490 (1997). 24. A.F. Mironov, A.N. Kozyrev and A.S. Brandis, Proc. SPIE, 1922, 204 (1992). 25. R. Willastatter and A. Stoll, Untersuchungen über Chlorophyll, Springer: Berlin (1913). 26. G.R. Seely, in: Chlorophylls, ed. by L.P. Vernon and G.R. Seely, Academic Press: New York, London, pp. 67–119 (1966). 27. H. Fischer, R. Lambrecht and H.Z. Mittenzwei, Physiol. Chem., 1, 253 (1939). 28. G. Hartwich, L. Fiedor, I. Simonin, E. Cmiel, W. Schafer, D. Noy, A. Scherz and H. Scheer, J. Am. Chem. Soc., 120, 3675–3683 (1998). 29. A.N. Kozyrev, Y. Chen, L.N. Goswami, W.A. Tabaczynski and R.K. Pandey, J. Org. Chem., 71, 1949–1960 (2006). 30. M.R. Waielewski and W.A. Svec, J. Org. Chem., 45, 1969–1974 (1980). 31. A. Osuka, S. Marumo, Y. Wada, I. Yamazaki, T. Yamazaki, Y. Shirakawa and Y. Nishimura, Bull. Chem. Soc. Jpn., 68, 2909–2915 (1995). 32. P. Hynninen, in: Chlorophylls, ed. by H. Scheer, CRC Press: Boca Raton, pp. 145–209 (1991). 33. R.K. Pandey, F.-Y. Shiau and A.B. Sumlin, Bioorg. Med. Chem. Lett., 4, 1263–1267 (1994). 34. A.N. Kozyrev, G. Zheng and C.F. Zhu, Tetrahedron Lett., 37, 6431–6434 (1996). 35. A.N. Kozyrev, T. J. Dougherty and R. K. Pandey, Chem. Commun., 481–482 (1998). 36. A.L. Gryshuk, A. Graham, S.K. Pandey, W.R. Potter, J.R. Missert, A. Oseroff, T.J. Dougherty and R.K. Pandey, Photochem. Photobiol., 76, 555–559 (2002). 37. A.L. Gryshuk, Y. Chen, W.R. Potter, A. Oseroff and R.K. Pandey. J. Porph. Phthalocyan., 8, 671 (2004). 38. A.F. Mironov and V.S. Lebedeva, Tetrahedron Lett., 39, 905 (1998). 39. A.F. Mironov, V.S. Lebedeva, R.I. Yakubovskaya et al., Proc. SPIE, 3563, 59–67 (1999). 40. A.F. Mironov, M.A. Grin, A.G. Tsiprovskiy, J. Porphyrins Phthalocyanines, 6, 358–361 (2002). 41. A.F. Mironov, M.A. Grin, A.G. Tsiprovskiy et al., Bioorg. Khimiya, 29, 214–221 (2003) (in Russian). 42. D. Barton and W.D. Wallis, General Organic Chemistry, (Russian translation, ed. by N.K. Kochetkov), Moscow: Khimiya Publishers, vol. 2, pp. 521–523 (a); vol. 3, pp. 61–77 (b) (1982) (in Russian).


43

43.A.F. Mironov, M.A. Grin, D.V. Dzardanov, K.V. Golovin and Y.K. Shim, Mendeleev Commun., 205–206 (2001). 44. A.F. Mironov, A.V. Efremov, O.A. Efremova, R. Bonnett and G. Martinez, J. Chem. Soc. Perkin Trans., 1, 3601 (1998). 45. S. Sasaki and H. Tamiaki. J. Org. Chem., 71 (7), 2648–2654 (2006). 46. A.F. Mironov, M.A. Grin, A.G. Tsiprovskiy et al., J. Porphyrins Phthalocyanines, 7, 707–712 (2003). 47. D.E. Remy, R.E. Whitfield and N.L. Needles, J. Chem. Soc. Chem. Commun., 681–695 (1967). 48. A.R. McCarthy, W.D. Ollis, A.N. Barnes et al., J. Chem. Soc. (B), 1185–1193 (1969). 49. K. Belnaik, E. Domagaliva and H. Hopkala, Roczniki Chem., 41, 831–843 (1967). 50. A.F. Mironov, M.A. Grin, A.G. Tsiprovskiy et al., Russian Federation Patent 2,223.274, 10 February (2004) (in Russian). 51. M. Hoebeke, H.J. Schuitmaker, L.E. Jannink et al., Photochem. Photobiol., 66, 502–508 (1997). 52. A. Scherz, J. Salomon and L. Fiedor, EP Appl., 584552 (1994). 53. L. Fiedor, V. Rosenbach-Belkin, M. Sai and A. Scherz, Plan. Physiol. Biochem., 34, 393–398 (1996). 54. H. Stiel, K. Teuchner, D. Leupold, H. Sheer, Y. Salomon and A. Scherz, Photochem. Photobiol., 72, 204–209 (2000). 55. A. Scherz, Y. Salomon, A. Brandis and H. Scheer, PCT Patent WO00/33833 (2000). 56. I.G. Meerovich, I.Yu. Kubasova, N.A. Oborotova, G.A. Meerovich, S.A. Demura, A. Brandis, V. Rosenbach-Belkin, A.Yu. Baryshnikov and A. Scherz, Proc. SPIE, 5973, 121–131 (2005). 57. 58. S. Schreiber, S. Gross, A. Brandis, A. Harmelin, V. Rosenbach-Belkin, A. Scherz and Y. Salomon. Int. J. Cancer, 99, 279–285, (2002. 59. N.V. Koudinova, J.H. Pinthus, A. Brandis, O. Brenner, P. Bendel, J. Ramon, Z. Eshhar, A. Scherz and Y. Salomon, Int. J. Cancer, 104, 782–789 (2003). 60. O. Mazor, A. Brandis, V. Plaks, E. Neumark, V. Rosenbach-Belkin, Y. Salomon and A. Scherz, Photochem. Photobiol., 81, 983–993 (2005). 61. G.V. Sharonov, T.A. Karmakova, R. Kassies, A.D. Pljutinskaya, M. Refregiers, R.I. Yakubovskaya, M.A. Grin, A.F. Mironov, J.-C. Maurizot, P. Vigny, C. Otto and A.V. Feofanov, Free Radicals in Biology and Medicine, 40, 407–419 (2006). 62. M. Wainwright, J. Antimicrob. Chemother., 42, 13–28 (1998). 63. Z. Malik, I. Hanania and J. Nitzan, Photochem. Photobiol., 5, 281–293 (1990). 64. T. Ito, Photochem. Photobiol., 34, 521–524 (1991). 65. M.G. Strakhovskaya, A.F. Mironov and A.M. Seregin, Dokl. Akad. Nauk, 384, 263–266 (2002) (in Russian). 66. A.F. Mironov, M.A. Grin, A.G. Tsiprovskiy, R. Titeev, E. Nizhnik and I. Lonin, Mendeleev Commun., 5, 204–207 (2004). 67. A. Grichin, A. Feofanov, T. Karmakova, N. Kazachkina, E. Pecherskih, R. Yakubovskaya, A. Mironov, M. Egret-Charlier ana P. Vigny, Photochem. Photobiol., 73, 267 (2001). 68. A. Feofanov, A. Grichin, T. Karmakova, V. Lebedeva, A. Filyasova, R. Yakubovskaya, A. Mironov, M. Egret-Charlier ana P. Vigny, Photochem. Photobiol., 75, 633 (2002). 69. K.R. Adams, M.C. Berenbaum, R. Bonnett et al., J. Chem. Soc. Perkin Trans., 1, 1465–1470 (1992). 70. A.A. Aksenova, Yu.L. Sebyakin and A.F. Mironov, Bioorg. Khimiya, 26 (2), 126–129 (2000) (in Russian). 71. G. Zheng, A. Graham, M. Shibata, M.R.J. Missert, A.R. Oseroff, T.J. Dougherty and R.K. Pandey, Org. Chem., 66, 8709–8716 (2001). 72. T.V. Akhlynina, D.A. Jans, A.A. Rozenkranz et. al., J. Biol. Chem., 272, 20328–20331 (1997). 73. T.V. Akhlynina, D.A. Jans, N.V. Statsyuk et al., Int. J. Cancer, 81, 734–740 (1999). 74. A.A. Rozenkranz, D.A. Jans and A.S. Sobolev, Immunol. Cell. Biol., 78, 452–464 (2000). 75. A.S. Sobolev, D.A. Jans and A.A. Rozenkranz, Prog. Bioph. Mol. Biol., 73, 51–90 (2000). 76. A.A. Rozenkranz, V.G. Lunin, P.V. Gulak, O.V. Sergienko, M.A. Shumiantseva, O.L. Voronina, D.G. Gilyazova, A.P. John, A.A. Kofner, A.F. Mironov, D.A. Jans and A.S. Sobolev, FASEB J., 17, 1121–1123 (2003). 77. E.O. Artemenko, D.G. Gilyazova, A.A. Rozenkranz, V.G. Lunin, O.V. Sergienko, K.N. Timofeyev, M.A. Grin, A.F. Mironov, A.B. Rubin and A.S. Sobolev, Mol. Meditsina, 4, 43–47 (2005) (in Russian).

44


- после 25-ой схемы идет 27; две схемы 33

2

meso-Phenylporphyrins as Synthetic Models of Natural Porphyrins: Synthesis and Modification A.S. Semeykin, S.A. Syrbu and O.I. Koifman Ivanovo State University of Chemistry and Technology, 1 F. Engels Prospect, Ivanovo, 153000, Russia email: [email protected]

Methods of synthesis of meso-mono-, di-, three- and tetraphenylporphyrins as well as methods of introducing the supplement substituents and their further modification are considered. Possible ways for fine regulation of the physicochemical properties of porphyrins are shown.

Introduction Porphyrins and metalloporphyrins are widespread in nature and are of great biological significance. Their most important representatives are chlorophyll, which, as part of the protein– lipid complex, performs the initial stage of photosynthesis in green plants, and blood heme, which, in combination with the protein globin, performs the reverse binding and transportation of oxygen to living cells. Metalloporphyrins are also present in some enzymes – catalases, peroxidases, etc. At present, interest in porphyrins and their metal complexes is extremely great. Results obtained in studies of the chemistry and properties of these compounds are discussed in detail in a number of monographs [1 – 10]. Real ways have been outlined for practical applications of porphyrins and their metal complexes as efficient catalysts of oxygen electroreduction, oxidation of sulfur dioxide in electrochemical synthesis of sulfuric acid and fuel hydrogen, polymerization regulators of acrylates, as photooxidation sensitizers, medical preparations and analytical reagents, semiconductors and model compounds in studies of such biological processes as photosynthesis,

46

A.S. Semeykin, S.A. Syrbu and O.I. Koifman

reversible binding of oxygen, enzyme catalysis [6]. The successful development of these trends depends on reliable methods of synthesis and modification of porphyrins possessing the required physicochemical properties and resistant to aggressive media and reagents. The broad use of porphyrins in engineering, technology and medicine is restrained by the low accessibility of most porphyrins, many of which are obtained with very low yields. In this connection, of special interest and topicality are issues of the chemistry of synthetic porphyrins. In this context, one’s attention is attracted by porphyrins containing in mesopositions aryl substituents, which can be subjected to various chemical transformations. At present, the properties of natural porphyrins are modelled mainly using their synthetic analogues meso-tetraphenylporphyrins 1, readily obtained by condensation of commercially available pyrrole with benzaldehydes. However, in some cases they are not very convenient, as, unlike natural porphyrins, they have no alkyl or pseudoalkyl substituents in β-positions of the porphyrin cycle, whereas meso-positions are substituted. On the other hand, sufficiently accessible octaalkylporphyrins 2 are not always convenient, either, as they have no active groups, which can be changed to impart them with various physicochemical properties. Therefore, of great interest are porphyrins, which combine the advantages of these two classes, such, for instance, as 5,15-diaryloctaalkylporphyrins 3 and 5-aryloctaalkylporphyrins 4. These porphyrins are produced mainly by condensation of α-unsubstituted linear derivatives of pyrroles with benzaldehydes. R

Ar

R

R

NH

R

N

Ar

NH

R

R

Ar

R

R

N

NH

R

Ar

R

R NH

N

N

Ar N

HN

N

N

HN

R

Ar

1

1

R

R R

R

2

HN

N

R

R R

HN

R

R

3

R R

Ar

R

4

Synthesis of meso(5,10,15,20)-tetrasubstituted porphyrins

meso-Tetrasubstituted porphyrins were first obtained by Rotmund, who found [11] that meso-substituted porphyrins 5 were formed in the interaction of pyrrole with aldehydes (Scheme 1). Condensation of acetaldehyde with pyrrole leads, with a low yield, to mesotetramethylporphyrin (5, R=Me), and of pyrrole with formaldehyde to porphin (V, R=H) with a negligible yield of 0.03% [12]. In the subsequent works by Rotmund [12, 13], the yield of porphyrins was significantly improved due to the performance of the reaction in pyridine in a sealed ampoule at a high temperature (140–240°C). Condensation of pyrrole with benzaldehyde yielded meso-tetraphenylporphin (5, R=Ph, H2TPP) [13], which is formed with a sufficiently high yield and is readily isolated from the reaction mixture [13]. Later, the works [14, 15] showed the yield of H2TPP to increase upon addition of zinc acetate to the reaction mixture. The obtained zinc complex of tetraphenylporphin (ZnTPP) is then transferred by the action of a mineral acid to free porphyrin. However, the yield of H2TPP even under these conditions, optimal for the condensation reaction in pyridine, does not exceed 18%. At present, 2,4,6-trimethylpyridine (collidine) [16, 17] and quinoline are also used as a reaction medium besides pyridine; they boil at a higher temperature than pyridine (171

meso-Phenylporphyrins as Synthetic Models of Natural Porphyrins

47

R NH

o

+

RCHO

140-240 C

N H

N

R

R N

HN R 5

Scheme 1

and 237 against 115°C, respectively), which makes it possible to perform the condensation reaction at an atmospheric pressure and in the presence of oxygen of the air as an oxidant. By the Rotmund’s method, the reaction is performed at high concentrations of the reacting substances (more than 1 M), but a comparatively low yield of meso-substituted porphyrins restricts its application. Methods of high-temperature synthesis have been modified to date, which sometimes makes it possible to completely avoid the use of the solvent. Thus, the reaction of a mixture of pyrrole and benzaldehyde in the presence of a metal salt (in the absence of a solvent) in a sealed ampoule at 150– 250°C produces H2TFP with a yield exceeding 50% [18]. Treatment of a mixture of pyrrole, benzaldehyde and silica gel or ceolite in a microwave oven leads after a chromatographic purification to H2TFP with a yield reaching 9% [19–21]. Introduction of pyrrole into the ampoule containing aldehyde in the gas phase (200–250°C) in a air atmosphere leads to the production of some meso-substituted porphyrins with yields of up to 23% [22]. Recently [23] tetraphenylporphins were reported to be obtained with yields of more than 20% in a condensation reaction in 2,4,6-trichlorophenol during the boiling in the presence of air at reagent concentrations of about 0.6 mol/kg. It is difficult to assign this reaction to the Adler’s method (see further) due to the small acidity of this phenol (pKa = 6.1). Besides, it is of interest that manganese complexes of many even di-ortho-substituted tetraphenylporphyrins are obtained with yields exceeding 50%, which implies the presence of a template effect on the Mg(II) ion. It should be noted that synthesis of meso-substituted porphyrins in basic media, undeservedly forgotten in the recent years due to the discovery of acid-catalysis methods, has its specific areas of application, such, for instance, as synthesis of meso-phenylporphyrins having in phenyl rings groups labile in an acidic medium or containing in meso-positions some heterocyclic residues (e.g., furanic or pyrrolic). Based on the works [13, 24] (their authors obtained with a small yield H2TPP at a prolonged boiling of a mixture of pyrrole and benzaldehyde in a methanol– pyridine medium), Adler [25], when studying the effect of various solvents, found that the condensation reaction is catalyzed by acids more efficiently than by bases. This conclusion was confirmed in the work by Treibs [26]. Adler [25] determined that when performing the reaction in aerated, boiling, acid-containing organic solvents the yield of H2TPP by the spectrophotometric estimate reached 40%. At present, the condensation reaction of pyrrole with aldehydes (Scheme 2) in acid-containing medium in the presence of the air is one of the major methods of producing meso-substituted porphyrins (the Adler’s method). The reagents mainly used as acidic solvents are acetic acid [25, 27], propionic acid [28, 29]; mixed solvents: pyridine-acetic acid [26], benzene-chloroacetic acid [25], toluene-p-toluene sulfo

48


R

+

NH

air H+

RCHO

N H

N

R

R N

HN R 5

Scheme 2

acid [30], xylene-chloroacetic acid [31– 33] and some others. The reaction is usually performed at the boiling temperature of the solvent, sometimes at the passage of air through the reaction mixture [26]. It has been found [34] that the yield of porphyrins depends on the temperature of the reaction medium, and the largest yield is obtained at temperatures close to 140°C. At lower temperatures, the porphyrin formation rate is small, and at higher temperatures the rate of oxidation of porphyrins produced is high. An optimal concentration of the reagents in the reaction medium is of the order of 0.2– 0.4 mol/l. Finally, there are new modifications of the Adler’s method for the synthesis of mesosubstituted porphyrins: condensation of pyrrole with aldehydes in a medium of dimethylformamide (dimethylsulfoxide)-aluminium trichloride (yield of H2TFP, 30%) [35], or in propionic acid with microwave irradiation (yield of meso-tetraarylporphyrins, 20–43%) [36]. The reaction of anisic aldehyde with pyrrole at 120°C in propionic acid containing 30% nitrobenzene, which in this case is an oxidant, made it possible to obtain a respective porphyrin with the yield of 45% [37]. The yields of 5–20% were observed for a number of other arylaldehydes [37, 38]. HH

R NH

H H

N

R

R N

HN R 6

It should be noted that under conditions of the reaction of acidic condensation, respective chlorins 6 are also formed besides porphyrins, and in some cases chlorins become the main reaction products [39]. However, they can easily be transformed into respective porphyrins by the treatment with benzoquinone derivatives: para-chloranil (p-CA) or 2,3dichloro-5,6-dicyanobenzoquinone-1,4 (DDQ) [40–43]. The use of a mixture of valeric acid and nitrobenzene at 160°C for the condensation of pyrrole with benzaldehydes makes it possible to significantly increase the yield of porphyrins, which contain no chlorins [44]. Isolation of porphyrins causes no problems, when they are crystallized during the

49


R

H R

+ R-CHO N H

CF3COOH or BF3etherate R CH Cl , 25oC H 2

2

NH HN NH HN R H 7

H DDQ R R

N

NH

R N

HN R 5

Scheme 3

cooling of the reaction mass in a sufficiently pure form [26, 29]; in other cases, the solvent is driven off under vacuum [45] or with steam [31] and the residue is chromatographed. In the case of using acids as solvents the reaction mixture can be neutralized by a solution of ammonia, the residue be filtered and porphyrin also be isolated by chromatography on a suitable sorbent [46]. The Adler’s method proved itself to be excellent for preparative-scale synthesis of porphyrins from aldehydes, which are relatively stable. The possibility to obtain quickly and easily an approximately 20% yield of such porphyrins at reagent concentrations of up to 0.4 M makes the Adler’s method sufficiently convenient. The use of the Adler’s method is limited in syntheses of meso-substituted porphyrins from aldehydes having substituents, which do not withstand the action of acids at high temperatures, many 2,6-disubstituted benzaldehydes and many aliphatic aldehydes. A comparatively recently developed new method of synthesizing meso-substituted porphyrins under mild conditions [47, 48] consists in the condensation of pyrrole with aldehydes in chloroform or methylene chloride in the presence of trifluoroacetic acid or boron trifluoride etherate in an inert atmosphere at room temperature to respective porphyrinogen 7 with subsequent oxidation of the reaction mixture by a stoichiometric amount of DDQ or p-CA (the Lindsey method) (Scheme 3). There is an example, where oxidation of porphyrinogen 7 is performed by hydrogen peroxide in acetic acid [49]. Besides the chemical methods, meso-alkylsubstituted porphyrinogens, which are more resistant to oxidation, can be converted to porphyrins photochemically [37]. The reaction, as found in [47], is sensitive to reagent concentrations. The highest yields of H2TPP (35–40%) are obtained in the reaction of 10 mM benzaldehyde and pyrrole each, and the yield decreases approximately two times at reagent concentrations of 100 mM each and 1 mM each. The decrease of the yield at a higher concentration of the reagents can be partially reduced by increasing the amounts of acid catalyst. For instance, at 100 mM benzaldehyde and pyrrole each, the yield of H2TPP was 23% with 1 mM BF3 etherate; 30%, with 3.2 mM and 29% with 10 mM BF3 etherate. These magnitudes are still lower than the 35– 40% yield obtained at reagent concentrations of 10 mM each in CH2Cl2 with BF3 etherate or trifluoroacetic acid as a catalyst [50]. The reaction is also sensitive to the concentration of acid catalyst: for the reagent concentrations of 10 mM each, BF3 etherate is efficient at a concentration of 1 mM, whereas trifluoroacetic acid is required at a much higher concentration of 20– 50 mM [47]. Addition of sodium or ammonium chlorides was found to significantly increase the yield of porphyrins [51, 52]. By the example of tetra(mesityl)porphin (5 R = 2,4,6-trimethylphenyl; H2TMP), it has been shown [53– 55] that, for synthesis of tetraphenylporphyrins containing electron-donor

50


substituents in both ortho-positions of the phenyl rings, at the stage of condensation it is required to use BF3 etherate as a catalyst in the presence of a cocatalyst, 0.75% ethanol. This method was improved to obtain gram amounts of H2TMP [56, 57]. Cocatalysts in the synthesis of H2TMP and similar porphyrins can be, besides ethanol, also ethylene glycol, 2-methoxyethanol [55] or methanol [58]. Syntheses of tetraphenylporphyrins containing electron-acceptor substituents in both ortho-positions of the phenyl rings do not require the use of cocatalyst. Application of 2,2-dimethoxypropane as a dehydrating agent also leads to an increase of the yield of H2TMP [58]. A further improvement of the method was to increase the reagent concentrations and to use oxygen of the air as an oxidant of porphyrinogen in the catalysis by iron phthalocyanine and p-CA [51]. This slightly decreases the yield of the end product, but makes it possible to use the method in large-scale syntheses [59]. Mild conditions of the reaction at the stages of condensation and oxidation enable the use of the Lindsey method for a broad range of aldehydes. This method is of special significance for synthesis of spatially hindered porphyrins [54, 57, 60] and meso-alkylporphyrins [61, 62], which are obtained by other methods with low yields. The yield of porphyrins can reach 50% depending on the aldehyde. At present, this is the most widely used method of synthesizing meso-substituted porphyrins. There are several modifications of this method: use of pyrrole with aldehydes of clays, such as monomorillonite and the like as condensation catalysts [20, 21, 63, 64]. The activity of clays is ascribed to their porosity but not acidity, and their efficiency as condensation catalysts is associated with stabilization of porphyrinogens in clay nanotubes [21, 65]. For synthesis of porphyrins with polar groups, which are obtained by the Lindsey method with low yield [47, 66], the method of micellar synthesis was proposed [67], which consists in conducting the condensation reaction in hydrochloric-acid solution of dodecylsulfate used as a surfactant, followed by oxidation of the reaction mixture by a solution of p-CA in tetrahydrofurane. This method makes it possible to obtain tetraphenylporphins having carboxy, hydroxy, acetamido, ether and ester groups in the phenyl rings. R

R R NH N R

N

NH

HN

N R

R

D,E,E,E- 8

N HN R

R

D,E,D,E- 8 R

R NH N R

R

R

R

N

NH

HN

R

D,D,E,E- 8

N

N HN

R

D,D,D,D- 8

51


The Lindsey method has the mildest reaction conditions. The yields are in most cases higher than those for the Adler’s method, though the reaction at reagent concentrations from 0.01 M up to 0.1 M requires large amounts of solvent to be removed. The method is applicable in the case of aldehydes having substituents unstable to the action of acids, 2,6-disubstituted benzaldehydes, aliphatic aldehydes and little-accessible expensive aldehydes (for production of porphyrins from which high yields are of great importance). meso-Tetraphenylporphyrins, having in one of ortho-positions of the phenyl rings bulky substituents, are formed as a statistical mixture of atropisomers 8 owing to the impossibility of the turn of teh phenyl rings; in some cases, when substituents are polar, the mixture can be separated chromatographically [68, 69]. The synthesis mechanism of meso-substituted porphyrins in an acidic medium (Scheme 4) includes the interaction of pyrrole by the carbonyl group of aldehyde to form pyrrylcarbinol 9. Further, it is protonated, yielding stabilized carbocation 10, which then

+

H+

H+

RCHO

N H

N H 9

pyrrole

CH R

N H

+

CH R

N H 10

RCHO H+

N H R N H H 11

NH

HN

H RCHOR H R

NH HN

CHOH R

N H R N H H

H

CH OH R n

N H

12 H R

H R

R H

+

CHOH R

NH

H R

HN

H R

R H

NH

R

HN

NH HN

NH

H R OH

H R OH

H

H

R

N H H pyrrole R H R

NH

HN

NH

H R

HN

NH

1. cyclization 2. oxidation

NH

NH

HN

R

R N

R

R

R

HN

NH

N

R

R N

R N

R

NH

13

N

R N

HN R

R

5

14

R N H

R

N NH

HN

R

15

Scheme 4

reacts to the additional molecule of pyrrole to form meso-substituted dipyrrolylmethane 11. Further the reaction is repeated with the buildup of the polypyrrole up to 12, which is cyclized at n = 4 to porphyrinogen 7, oxidized further by oxygen of the air (Adler’s method) or by benzoquinone derivatives (the Lindsey method) to meso-tetrasubstituted porphyrin 5.

52


Using the Alder’s method, the linear polypyrrole chain can be also partially oxidized to the closure into the cycle. Depending on the reaction conditions, relative concentrations of the reagents and catalyst used the end result of the cyclization and subsequent oxidation can differ. Thus, conditions were determined, at which, besides porphyrins 5, also corroles 13 can be formed (excess of pyrrole, yield up to 10%) [70], N-confused porphyrins 14 (yield up to 10%) [71, 72] and tetraphenylsapphirins 15 (yield up to 1%) [73, 74] (Scheme 4). The condensation reaction of pyrrole with aldehydes is also used for the synthesis of compound structures containing porphyrin fragments. Thus, in condensation of dialdehydes, in which phenyl rings via ortho-positions are connected by a sufficiently long dioxymethylene chain [R = –O(CH2)nO–, n ≥ 4], with pyrrole, a mixture of three isomeric belted porphyrins 16–18 is formed; herewith, the most strained porphyrin 16 is not formed if the chain is too small (n = 4) [75–77]. At an even smaller chain length (n = 2, 3 and 4 for meta-positions), under the condensation conditions, dimeric sandwiched porphyrins 19 are formed [78, 79]. R

R

R

O

R

O

O O

O NH N

O

O

N

NH

HN

O

N

O

N

NH

HN

N

N HN O

O

R

O

16

17 R

O

R

18 R

O

O O N NH

N

HN

NH

N

HN

N

O O R

O

R

O

19

Condensation of various-structure tetraaldehydes with pyrrole yielded the so called capped porphyrins 20– 23 [33, 80– 85]. Often, syntheses of this kind are made using high-dilution methods [75, 76]. All listed porphyrins at small lengths of connecting bridges have a distortion of the plane structure of the porphyrin macrocycle, caused by the tension of the bonds between phenyl rings in meso-positions and meso-substituted macrocycle. Moreover, sandwiched dimers 19 having an even number of methylene links (n = 2, 4) are distorted more than their analogues with an odd number (n = 3) [78]. It should be noted that such porphyrins also have hindrances in conducting the reactions by the central atoms of nitrogen on one or both sides of the macrocycle, i.e., they possess the properties of both distorted and hindered porphyrins. An increase of the length of connecting bridges to greater than an extreme value leads to the disappearance of the distortion of the porphyrin macrocycle due to the elimination of

53

meso-Phenylporphyrins as Synthetic Models of Natural Porphyrins R

R

R

R

R R

R R NH N

R

R

N

NH

HN

N

R

R

N

NH

HN

N

20

21 NH

N HN

22 NH

O

O HO O

O

O NH N

N HN

O

23

the strain from the phenyl–macrocycle link [86]. By condensation of mono-4-formylphenyltriphenylporphyrins with pyrrole, pentameric porphyrins 24 were synthesized [87, 88]. R R X X

NH

R N

N

HN

X

NH

X=

N

N HN R

X R R 24

The adduced exotic structures are usually obtained with yields not exceeding 10%; however, synthesis of respective polyaldehydes presents no large problems in most cases, which makes such porphyrins interesting in many biological and catalytic studies of porphyrin systems. The first intermediate compounds in the condensation of pyrrole with aldehydes are pyrrylcarbinols 9. Therefore, of great interest are studies of the possibility of using these compounds for synthesis of meso-substituted porphyrins. Three methods of synthesis of pyrrylcarbinols have been developed to date (Scheme

54


RCOCl

I N MgX

N H 25

RCON(CH3)2

II N H

POCl3

R LiAlH4 (NaBH4) O

RLi CHR OH

N H 9

H N H 26

III

O

Scheme 5

5). The first two methods are based on the initial production of 2-ketopyrroles: the reaction of the pyrrole Grignard reagent with chloroanhydride of the acid yields 2-ketopyrrole 25 and 3-ketopyrrole with yields of ~60 and 15%, respectively (I) [89]; an alternative method of producing 2-ketopyrroles consists in the interaction of N,N-dimethylamides with pyrrole according to Wilsmeyer in the presence of phosphorus oxychrolide (II) [90– 92]. 2-Ketopyrroles 25 are then reduced by LiAlH4 or NaBH4 to pyrrylcarbinols 9. The direct one-stage method of producing pyrrylcarbinols is performed by the reaction of pyrrole-2-carboxyaldehydes 26 with lithium or magnesium organic compounds (III) [90– 92]. The last two methods have no drawback of the first method – nonselectivity of substitution in acylation of pyrrole. Besides, it has been found [93, 94] that pyrrylcarbinols with perfluoroalkyl residues are easily formed in condensation of hydrated perfluoroaldehydes with pyrrole in the presence of alkali. Pyrrylcarbinols 9 are self-condensed, yielding meso-substituted porphyrins under conditions of Rotmund’s reaction (Scheme 6) [95]. The yields of porphyrins in this reaction are comparable with those for condensation of pyrrole with aldehydes. Besides, coloured intermediate compounds observed in condensation of pyrrole with aldehyde, are the same as in self-condensation of respective pyrrylcarbinol. These results prove that pyrrylcarbinols 9 are indeed intermediate products in the condensation reaction in an acidic medium [95]. R

N H

CHR OH

H+ [O]

N

NH R

R N

HN R 5

9

Scheme 6

meso-Substituted porphyrins are usually synthesized under conditions similar to those of the Adler’s reaction [90-92, 97, 96]; in some cases, the yield of porphyrins is observed to be significantly increased upon addition of zinc acetate to the reaction mixture [98]. Sometimes, the two-stage Lindsey method is also employed [93, 94, 99]. The method using pyrrylcarbinols is mainly applied to produce meso-alkylporphyrins and hindered porphyrins. Of interest is the fact that bulky pyrrylcarbinols obtained from 2-isobutyryl or 2-pyvaloylpyrrole can not be converted to respective porphyrins [98].


1.1

55

Synthesis of meso-tetra-β-octasubstituted porphyrins

Besides pyrrole, its 3,4-disubstituted derivatives can enter into the condensation reaction with aldehydes to form, after oxidation, meso-tetra-β-octasubstituted porphyrins possessing a distorted porphyrin cycle owing to the steric interaction of substituents in adjacent mesoand β-positions. The properties of porphyrins are mainly determined by two factors. One of them, electronic, is due to the occurrence of a broad aromatic π-conjugation contour embracing the macrocycle. The other, geometric, is represented by a plane tetrapyrrole system possessing a rigidity and relative stability to strain. At present, however, there are numerous data, which show that the porphyrin macrocycle is sufficiently flexible and is capable of the existence in non-plane conformations [100– 102]. These properties of porphyrin molecules are extensively studied, as the distortion of the porphyrin cycle can play a significant role in biological photosynthetic and redox systems [103–105]; besides, the conformation distortions of the porphyrin chromophore can serve as a tool of fine tuning of its physicochemical properties. We should note the distinction of hindered porphyrins from distorted ones. The peculiarity of the former is in shielding the central cavity of the porphyrin molecule by bulky substituents from attacks of various reagents. This spatial hindrance strongly affects the reactivity of porphyrins, but has a minor effect on their physical properties. In contrast to these compounds, distorted porphyrins have a non-plane deformed macroring, owing to which fact both the chemical and physical characteristics of these porphyrins significantly change as compared with analogues possessing the plane structure of the porphyrin core. Distortion of the porphyrin cycle can in an ideal form produce two extreme – saddled and ruffled – conformations [102]. The data of X-ray diffraction studies show that most symmetric distorted porphyrins and their complexes have a saddled (or close to it) conformation, and only in sufficiently rare cases, for some metal complexes, the pure ruffled conformation is realized [102, 106]. _

0

+

+

_ 0 + +

_ 0 saddled

+ _

+ NH N _ + _ + N HN

+

_ NH N 0 0 0 0 N HN

_ 0 _

_ _

_ +

+

+

ruffled

The most known and basic method of distortion of the porphyrin cycle is introduction of substituents to its periphery to adjacent meso- and β-positions; the distortion rises with their number and size increasing. Thus, in condensation of benzaldehyde with pyrrole and its 3,4-disubstituted analogue under the action of boron trifluoride etherate, a mixture of porphyrinogens was obtained; condensation of the mixture by DDQ led to six possible porphyrins 27–32 (Scheme 7) [107– 109]. The X-ray diffraction analysis of specimens of these porphyrins showed [107] the deformation of the macrocycle to grow from almost plane tetraphenylporphin 27 (A = Ph) to strongly distorted dodeca-substituted porphyrin 32 (A = Ph, B = Et, Ph). Condensation of 3,4-diphenylpyrrole with formaldehyde and benzaldehyde under the

56


32 (A = Ph, B = Et, Ph). B ACHO +

B 1. BF3 etherate

+ N H

2. DDQ

N H B

A

B

A

B NH

NH A

N

A

B

N

A

27-32

N

NH

A

A

HN

N

N

A

A

HN

N

HN B

A

27 A

B

B

B

B

A 30

N

A

A

HN

B

B B

NH A

N

HN

B

B

N

A

A N

A 29 A

B B

NH

N

A

B

B

B NH

A 28 A

N

HN

B A 31

B

B B

A 32

B

Scheme 7

action of HBr in ethanol, followed by oxidation of the mixture of porphyrinogens by DDQ in boiling toluene, produced six possible porphyrins 32– 37 (A = B = Ph) (Scheme 8) [110] with the deformation increasing from plane β-octaphenylporphin 33 to dodecaphenylporphyrin 32. Similar results were obtained in condensation of 3,4-disubstituted pyrroles and their 2-hydroxymethyl derivatives 38 with benzaldehydes in methanol containing HBr followed by the oxidation of the mixture of porphyrinogens by p-CA in THF (Scheme 8) [111]. Of interest is the fact that, proportionally to the growth of distortion of the porphyrin cycle, the batochromic shift of all bands in absorption spectra of these porphyrins also rises [107, 109– 111]. Scheme 8

B

B

ACHO + CH2O + N H

or B

B

B

B

+

ACHO + N H

1. HBr - EtOH 32–37 2. DDQ

N H 38

CH2OH

1. HBr - EtOH 32–37 2. DDQ

57


B

B

B

B

B

B NH N

NH

HN

N B

B B

B B

B NH

B

B

B

B

B

B

B B B

NH A

N

A

A

HN

N B

B

A 35 A

B

N

B

36

HN B

B

A N

N

B

B NH

HN

B

34 A

B

N A

N

N B

B

B NH

HN

B

B

A

B

N

B

B 33 A

B B

N

B

B

A

HN B

B

B

A 32

B

37

B

The most strongly distorted dodeca-substituted porphyrins 32 are obtained by condensation of aldehydes with 3,4-disubstituted pyrroles; methods of their synthesis differ little from those for tetraphenylporphins (Scheme 9). In the main, these are two known modifications of the method: condensation of pyrrole with aldehyde in organic acid-containing boiling solvent in the presence of oxygen of the air (Adler’s method) [27, 112, 113], or condensation under the action of a catalyst (acid, boron trifluoride etherate) under mild conditions to porphyrinogen followed by its oxidation without isolation (or with isolation) to porphyrin by benzoquinone derivatives (the Lindsey method) [114– 120]. The high resistance of porphyrinogens to oxidation makes it possible to isolate these compounds in pure form and then oxidize them to porphyrins. Scheme 9 B B

B

B

B NH

[O] + N H

B

A

ACHO

H+

N

A

A N

HN

B

B B

A

B

32

It has been found [121–123] that the strong distortion of the porphyrin cycle also results from the occurrence of bulky substituents, such, e.g., as tert-butyl or adamantyl in meso-positions in the absence of substituents in β-positions of the cycle. In production of distorted β-alkylsubstituted porphyrins 32 (B = Alk), it is better to use the condensation in acid-containing alcoholic solvents followed by the isolation of porphyrinogen and its subsequent oxidation by DDQ in THF. For synthesis of β-phenylporphyrins 32 (B = Ar), the condensation is better to be conducted in boiling acetic acid followed by oxidation of porphyrinogen or without its oxidation in boiling with DDQ, or else with

58


isolation and subsequent oxidation in boiling toluene also using DDQ. The use of the Lindsey method [120, 124] is justified in the synthesis of porphyrins 32 having o-disubstituted phenyl rings in meso-positions. It should be noted that condensation between benzaldehyde and pyrrole nonsymmetrically substituted in β-positions produces a statistical mixture of four randomeric porphyrins (39– 42). B

A

C

A

B B C

N

A C

B

B

C

B

C

C N

HN A C

B

B N

NH

A

40 (type II)

B

C N

A

C

N

C

A

39 (type I)

B

HN

A

A

A

C N

A

C NH

C

HN

B A

A N

NH

C

B

A

N

NH

B

C

HN

A

B

A C

B

41 (type III)

B

42 (type IV)

Interesting is the fact that condensation of 3,4-dialkylpyrrole with alkylaldehydes followed by oxidation of the reaction mixture by benzoquinone does not lead to distorted dodeca-alkylporphyrins, but is stopped at the stage of respective porphodimethenes 43 [119]. In this case, distortion of the macrocycle skeleton is so great that the oxidation potential of oxidants used is not sufficient for its aromatization. On the contrary, practically nondistorted porphyrins 44, 45 are obtained with high yields [119, 125].

, B H

A

B

B

A

R

B NH

N

A

NH A

N

HN

B

R

N

A

NH

A N

HN

N

B

N HN

R B

A H B A,B = Alk 43

1.2

A A = Alk 44

R 45

Synthesis of nonsymmetric meso-tetrasubstituted porphyrins

Condensation of aldehyde with pyrrole makes it possible to obtain porphyrin having four identical meso-substituents. However, there are many applications, which require various substituents in meso-positions of the porphyrin cycle. One of the simplest approaches to this goal is the so called mixed-aldehyde condensation (Scheme 10). The reaction of pyrrole with a mixture of aldehydes makes it possible to obtain a mixture of six porphyrins 46– 51, which can be separated by means of thin-layer chromatography [126], and in the case of a polar group in one of the aldehydes, using column chromatography [127]. The expected ratio of porphyrins in mixed-aldehyde condensation is set by the binominal distribution. At an aldehyde ratio of 1:1, the distributions are as follows: A4, 6.25%; A3B, 25%; cis-A2B2, 25%; trans-A2B2, 12.5%; AB3, 25%; B4, 6.25%. This result assumes the equal reactivity of both aldehydes at all stages of the porphyrin formation reaction. A3B

59


porphyrins are obtained the most often in mixed-aldehyde condensation. The yield of A3B porphyrin is maximal at a ratio of the reacting aldehydes 3:1: A4, 31.64%; A3B, 42.19%; cis-A2B2, 14.06%; trans-A2B2, 7.03%; AB3, 4.69%; B4, 0.39%. A larger content of aldehyde A increases the yield of A4 porphyrin at a decrease of the absolute amount of A3B porScheme 10 + ACHO + BCHO

46–51

N H A

B N

NH A N

A

A A4 46 B N HN

B trans-Ⱥ2ȼ2 (ȺȼȺȼ) 49

N HN

N

NH B

B Ⱥȼ3 50

HN

A cis-Ⱥ2ȼ2 (ȺȺȼȼ) 48 B

A N

B N

HN

NH A

N

A

A A3B 47 B

A

N

NH A

N

HN

NH

N

NH A

B

B

B N

HN B ȼ4 51

phyrin. However, the ratio of the reacting aldehydes, which gives the highest isolated yield of A3B porphyrin, depends on the real reactivities of these aldehydes, as well as on the ease of separation of the porphyrin mixture [128]. When a mixture of strongly-polar hard-ofaccess aldehyde and low-polar aldehyde is used to produce AB3 porphyrins, it is sometimes more advantageous to use a large excess of low-polar aldehyde, so that the strongly-polar one reacts more completely. Herewith, symmetrically substituted B4 porphyrin is mainly produced, which is, however, easily separable chromatographically from the required AB3 porphyrin [129–131]. Using this method, a number of dimeric porphyrins linked via mesopositions by various spacers was synthesized [131–136]. In practice, isolated yields of AB3 porphyrins in Adler’s method are mainly at the level of 5%, with an approximate total yield of the porphyrin mixture 20% and a statistical distribution of the mixture. Reactivities of aldehydes in mixed-aldehyde condensation sharply differ in most cases; for this reason, namely the Lindsey method is the most suitable for such syntheses, or a modified Adler method, in which the first stage of the reaction is performed in the absence of oxygen of the air to avoid oxidation of porphyrinogen formed by more reactive aldehyde and to achieve an equilibrium between porphyrinogens, which is set by the ratio of the initial aldehydes. Sometimes, the reaction is conducted with two aldehydes of close reactivities or with

60


aldehydes stable under the synthesis conditions. Further, by modification of substituents, they are imparted with required polarity or structure [20, 97, 137]. Of interest is the fact that aldehydes, which do not yield meso-tetrasubstituted porphyrins, often enter into mixed-aldehyde condensation as one of the components [95, 138, 139]. Mixed-aldehyde condensations are usually restricted by the use of two reagents. Three-component condensations are very rare. Condensation with three different aldehydes and pyrrole made it possible to obtained a porphyrin mixture, from which 5-(2-hydroxyphenyl)-15-(2-nitrophenyl)-10,20-di-p-tolylporphin was isolated with the yield of less than 0.1% [95]. Joint condensation of benzaldehyde, terephthalaldehyde and 4-hydroxybenzaldehyde with pyrrole made accessible monohydroxyporphyrin dimer isolated as a mixture of isomers with the 0.7% yield [140]. By condensation of a mixture of trialdehyde 52 and aldehydes 53 with pyrrole by the Adler’s method, capped porphyrins with additional covalently bound extra ligand 54 was synthesized, though with low yields, but in a minimal number of stages (Scheme 11) [141]. Scheme 11 O

O

O

O

O

O

O O O

N H

+

O

+

CHO

O O

R

OHC

O

O air

O

R NH

EtCOOH

OHC

O

N

HN

O

CHO N

O

N

O

52

53

O

N

54

Mixed-pyrrole condensations can also be performed. Thus, at a ratio of 3:1, a mixture of pyrrole and pyrrole containing a carboxyl group, which is attached to it via polymethylene bridge 55, reacts to benzaldehyde (or p-toluyl aldehyde) in boiling propionic acid. The further treatment of the reaction mixture makes it possible to obtain porphyrin having one β-substituted pyrrole residue, which can provide for the attachment of this porphyrin to polymer 56 (Scheme 12) [142]. Mixed-pyrrole condensations are used much more rarely Scheme 12 H3C

(CH2)nCO 2H +

3 N H

+ N H

CH3

CHO

N NH

4

(CH2)nCO 2H HN

N

CH3

H3C

55

CH3

56


61

than mixed-aldehyde ones, undoubtedly, due to the lower accessibility of substituted pyrroles. A more promising synthesis of nonsymmetric meso-tetrasubstituted porphyrins is the method, which makes use of preliminarily obtained meso-substituted dipyrrolylmethanes 11. Condensation of unsubstituted pyrrole with aldehydes leads to a mixture of oligomeric products: dipyrrolylmethanes 11, tripyrranes 57, bilanes 12 (n = 4), cyclic porphyrinogens 7 and higher linear and cyclic oligomers. However, conducting this reaction at a significant excess of pyrrole makes it possible to stop it at the stage of meso-substituted dipyrrolylmethanes 11 (Scheme 13). A large number of methods was reported for direct condensation leading to dipyrrolylmethanes [93, 143– 154]. As condensation catalysts, use is usually Scheme 13 H R

RCHO

+

NH excess

R

NH

BF3 etherate H

NH

TPAA or

+

R

NH HN

H

NH

11

Ⱥ

ɛ

57

ɛ

made of trifluoroacetic acid or boron trifluoride etherate; the reaction is run at room temperature in a solvent or without it. Removal of excess pyrrole under vacuum yields raw dipyrrolylmethane, which is recrystallized [147], chromatographed [147] or distilled [146, 155] to obtain a pure product. The isolation of respective tripyrrane 57 from nonpurified dipyrrolylmethane has been also reported [155]. In a similar way, using excess pyrrole, meso-substituted dipyrrolylmethane 11 was synthesized from pyrrylcarbinol 9 with the yield of 95% (Scheme 14) at the catalysis by trifluoroacetic acid or BF3 etherate [147]. Scheme 14

N H

9

CHOH R

+

NH excess

NH

R PhCHO TPAA or H BF3 etherate

NH 11

Condensation of meso-substituted dipyrrolylmethane 11 with aldehyde leads to ABAB(trans)-type meso-tetrasubstituted porphyrins 58 [156–158]; the use of two aldehydes in condensation makes it possible to obtain ABAC-type porphyrin 59; however, in a mixture with two porphyrins ABAB and ACAC [149, 159– 161], if one of aldehydes has a substituent similar to dipyrrolylmethane one, the formed product is mainly BA3-type porphyrin 60 [162– 164] (Scheme 15). In condensation of a mixture of two meso-substituted dipyrrolylmethanes with aldehyde similar to one of the meso-substituents of dipyrrolylmethane, the product obtained is mainly BA3-type porphyrin 60, and using aldehyde with a substituent not similar to meso-

62


Scheme 15 B NH

B N

A N

NH

A

BCHO

A

CCHO H

HN

BCHO

NH

A

NH

N

C ABAC + ABAB+ACAC 59

N

A HN

B ABAB 58

11 ACHO BCHO B NH

N A

A N

HN A BA3 + ABAB+A4 60

substituents of dipyrrolylmethanes, ACBC porphyrin 61. However, in both cases other porphyrins are also formed (Scheme 16); their chromatographic separation is possible only in the occurrence of polar groups in substituents. Scheme 16 A

C NH

N B

A N

HN

CCHO A

NH

H

NH

C ACBC + ACAC+BCBC 61

HN + HN

B H

ACHO

NH

N

A

B N

HN A BA3 + ABAB + A4 60

It should be noted that for such kind of syntheses, the suitable method is only the low-temperature Lindsey method, as acidolysis of meso-substituted dipyrrolylmethanes in acid media at high temperatures and regroupings leads in the subsequent oxidation to the formation of a compound mixture of porphyrins (Scheme 17). However, in [165] it was shown that during the condensation in propionic acid in the presence of zinc acetate only trans-porphyrin 58 without regrouping was formed, which is, possibly, due to oxidation processes prior to the closure of the cycle. This reaction conducted by the two-stage Lindsey method forms a mixture of isomeric porphyrins, though with a large yield [162, 165]

63


Scheme 17

A

NH

H

NH

H+

BCHO

+

CH + A

N H

H+

N H

pyrrole B

NH

H

NH

+

CH B

N H

+

H

A number of studies [157, 158] have shown dipyrrolylmethanes having spatially hindered or perfluoroalkyl substituents in meso-positions to be the most resistant to acidolysis. . The condensation reactions using meso-substituted dipyrrolylmethylcarbinols 62 instead of a mixture of aldehyde and dipyrrolylmethane make it possible to avoid acidolysis due to the impossibility of forming nonstabilized dicarbocation (Scheme 18). Scheme 18 B CHOH NH

A H

H+

B + CH NH

A H

NH

H+

+

+

HC A

NH

CH + B

N H

N H

62

Scheme 19 presents some possible routes of using meso-substituted dipyrrolylcarbinols 62 and dipyrrolyldimethylcarbinols 63 for synthesis of nonsymmetrically meso-tetrasubstituted porphyrins of various types. Of special interest is the latter route, which makes it possible to obtain totally nonsymmetric meso-tetrasubstituted porphyrins 64 [166– 168]. Scheme 19 B CHOH

B

NH

A

NH

H

A

NH

N

H

NH

HN +

NH

HN

HOHC D

HN

B ABAB 58

62 B CHOH

A

N

A

B

C H

NH

N

A

C N

HN

D ABCD + ABAB+CDCD 64

64


B CHOH NH

A

B HN

NH

C

H

NH

C

H

HN

N

CHOH B 63

11

NH

B HN

NH

NH

C

+

H

HN

B ABCB 65

B CHOH A

N

A

+

C

H

HN

N

CHOH D 66

N

A HN

D ABCD 64

11

ɛ To obtain carbinols 62 and 63, use can be made of the methods applied for synthesis of pyrrylcarbinols (Scheme 5): acylation of the magnesium derivative of dipyrrolylmethane by acylchlorides or according to Wilsmeyer to acyldipyrrolylmethane followed by its reduction to carbinol or interaction of respective carbaldehyde with aryllithium derivatives. A new convenient method of synthesis of carbinols has been developed [169, 170]. The reaction of o-mercaptophenol with carboxylic acid or its derivatives in the presence of BF3 etherate yields benzoxathiol tetrafluorate 67, a masked equivalent of aroyl. Benzoxathiol

B +

O

Scheme 20 BO

B S -

BF4 S

NH

+ A NH

67

pyridine CH3CN

NHO A NH

11

NH

HBF4 OEt2 HgO

A NH

HBF4 OEt2 HgO

B OH

NH

LiAlH4

S

B O 69

B

BCOOH + OH

S

A

68

BF3 Et2O

SH

NH

66 ɩpyridine CH3CN

B O

NH A

NH A

NH

B OH

LiAlH4

NH

O NH A NH

OH B

62

63

O B

65


tetrafluorate readily reacts in acetonitrile containing one equiv. pyridine at room temperature with dipyrrolylmethane 11 to yield 5- 68 or 5,5′-disubstituted dipyrrolylmethane 69 (the use of another benzoxathiazol tetrafluorate at the 2nd stage ultimately leads to nonsymmetrically substituted dimethylcarbinol 66), whose oxidative hydrolysis to acyldipyrrolylmethanes and subsequent reduction yields required carbinols 62 or 63 [166] (Scheme 20). There are methods of synthesis of nonsymmetric meso-tetrasubstituted porphyrins using tripyrranes ([3+1] condensation) [171] (Scheme 21) and bilanes [166] (Scheme 22). The reaction of pyrrole with benzoxathiazole tetrafluoroborate 67 successively yields 2-substiScheme 21 O A +

+

O

BF4 S

N H

A

A

S

-

pyridine CH3CN

HgO NH BF3 OEt2

S

NH

A

A

O

A

O LiAlH4

NH

O

A 71

67 70 B HO HN

B

A N NH

TFA, 25oC

HO HN

excess pyridine

N H NH H N

B

A

B

A

OH

A

1. TFA 2. DDQ

N

OH

73

72

Scheme 22 OH

O NH A NH

1. EtMgBr A 2. BCOCl

NH

B

B

B

LiAlH4

NH

BF3-Et2O A ɩɢɪɪɨɥ excess ɢɡɛɵɬɨɤ pyrrole

A NH

NH

HN

NH

HN

OH

O B

B 11

B 74

63 B NH

N

N

HN

A

1. CCHO, TFA 2. DDQ C

B 65

NH

66


tuted pyrrole and then 2,5-disubstituted pyrrole 70 unlike the repeated substitution in position 4 in the acylation reaction. The use of dicarbinol 71, obtained after the reduction, for synthesis of tripyrrane 72 in excess pyrrole makes it possible to perform further the 3+1 condensation to obtain cis-A2B2 porphyrin 73. The use of dipyrrolylmethane dicarbinol 63 makes it possible to obtain bilane 74 for synthesis of ABCB-type meso-tetrasubstituted porphyrin 65 [166]. Thus, the use of linear meso-substituted oligopyrroles makes it possible to obtain individual nonsymmetric meso-tetrasubstituted porphyrins without impurities of isomeric products. However, these methods require sufficiently complex syntheses of initial oligopyrroles.

2

meso(5,15)-Disubstituted Porphyrins

5,15-Diphenylporphyrin and its β-alkylsubstituted analogues are of interest because these compounds have some common features peculiar of 5,10,15,20-tetraphenylporphyrin, namely the presence of two meso-phenyl groups, with some features of β-alkylporphyrins owing to the occurrence of unsubstituted meso-positions, and porphin, in the absence of β-substituents; herewith, such a porphyrin is not so inaccessible as the totally unsubstituted initial porphin. 5,15-Diphenylporphyrins 4 of symmetric structure (trans-) have been known since rather long in the form of octaalkylsubstituted derivatives, but in the form of β-unsubstituted analogues they appeared only recently, when unsubstituted dipyrrolylmethane 75 [26, 93, 172, 173] and meso-phenyldipyrrolylmethanes 11 [147, 172, 174] became relatively accessible. Dipyrrolylmethane 75, which has neither meso- nor β-substituents, is obtained in three stages from pyrrole and thiophosgene via thioketone 76 [175], which by the action of hydrogen peroxide in an alkaline medium is converted to ketone 77 [175], reduced to 75 under the action of diborane with the total yield of about 40% [173, 176] (Scheme 23). Recently, it was found that 75 could be obtained by direct desulfonation of thioketone 76 by Reney nickel or sodium boron hydride [172]. Besides, it was possible to perform a small-scale condensation of polyformaldehyde (paraform) with excess pyrrole in the presence of boron trifluoride etherate or trifluoroacetic acid, which yields dipyrrolylmethane 75 with the yield of 40% in one stage [177, 178]. Proceeding from the features of structure of meso(5,15)-disubstituted porphyrins, several possible routes for their synthesis can be presented (Scheme 24). All these methods Scheme 23 O

S CSCl2 NH

N H

HN

H2O2 OH-

NH

76 (CH2O)n BF3

Ni or NaBH4

NH HN 75

77 B2H6

HN

67


were tested for synthesis of trans-substituted porphyrins, but the most applicable are routes A and B, as routes C and D require an additional stage of producing formyldipyrrolylmethane. However, the use of route D makes it possible to obtain porphyrins nonsymmetrically substituted via phenyl rings without impurities of other porphyrins. Scheme 24 R

R

CHO

R

R

H

NH

Ar

NH

NH +

ArCHO

A

NH

R

R

R

R

N

H

B +

Ar

N

R 80 R

HN

R

NH

R

R NH

R 78 R X

C

R

Ar

R R

HC(OMe)3

Ar 4

R D

R

NH

H

NH

Ar

NH

X = H, CO2H

R

R CHO

H

HN

Ar

CHO

R

X

R 81

R 79

R

HN +

R

X

X

R 79

It should be noted that β-substituted dipyrrolylmethanes used for synthesis of 5,15-substituted porphyrins via routes A, B and D should have a symmetric system of substitution relative to the meso-carbon atom, otherwise two isomeric porphyrins are formed. Synthesis of symmetric (relative to the meso-carbon atom) meso-substituted β-substituted dipyrrolylmethanes 86 used to produce meso-substituted porphyrins consists in acidcatalyzed condensation of aldehydes with α-unsubstituted pyrroles 82, in which the second α-position is substituted by a readily removed group (usually, carbethoxyl or carboxybenzyl) [165, 179, 180] (Scheme 25). Then follows the removal of protective groups Scheme 25 B

B B

NH A

COOR

A

COOR H+

H

NH

NH

H2(Pd/C) R=Bz H

NH

OH- R=Et

A t

A

A

83

COOH

A C

NH

CCHO C

B

COOH

COOR

C

NH

H

NH

A

B

B

B

84

85

86

by, respectively, alkaline hydrolysis 84 (R = Et) or hydrogenolysis 84 (R = Bz) [173] with subsequent decarboxylation 85 by heating in high-boiling solvents, such as ethylene glycol [165], ethanolamine, DMFA [173], diethylformamide [181, 182]. For carbethoxy derivatives 84 (R = Et), the protective grouping could be removed in one stage by heating in an aqueous alkaline solution in a sealed ampoule at 180°C [80] or by boiling in alkaline ethylene glycol in an inert atmosphere [183].

68


Symmetric meso-unsubstituted dipyrrolylmethanes 88 are obtained by self-condensation of pyrrylcarbinyl cations 87 in binary solvents containing water, with removal of formaldehyde, often with an almost qualitative yield [126, 184, 185] (Scheme 26). This method is the most convenient for synthesis of such dipyrrolylmethanes. Scheme 27 A

+

RO2C

N H 87 H2O B

A RO2C

B

N H

CH2

OH B H2C

B A

B A

+

NH HN

A

A

-ɋH2O CO2R RO2C

RO2C

B

NH HN CO2R 88

CH2OH

Pyrrylcarbinyl cations 87 are sufficiently stable due to the delocalization of positive charge by the electron-rich pyrrole nucleus (Scheme 27); as the result, they are readily obtained from various precursors, such as 89, by the SN1 mechanism. The broad diversity of precursors and ease of formation of pyrrylcarbinyl cations makes it possible to obtain and use them in various nucleophilic solvents, such as alcohols, carboxylic acid and even water. The most stable α-acetoxymethylpyrroles give high yields of dipyrrolylmethanes in boiling in ethanol or methanol containing 1% hydrochloric acid [185– 187] or in boiling in aqueous acetic acid [188].

A R

A

B

N H

CH2X

-

-X

R

B

..

N H

A +

CH2

R

87

B +

N H

A CH2

R

+

..

N H

Scheme 28 B CH2

89 +

+

X = -Br, -Cl, -OCH3, -OH, -OCOCH3, -NR3, -NHR2 etc.

R = CO2Et(Bz)

The most widespread synthesis of 5,15-diphenylporphyrins 4 consists in condensation of α,α′-unsubstituted dipyrrolylmethanes 78 with aldehydes in the presence of acid (route A) and is similar to that of meso-tetraphenylporphins. The condensation reaction of dipyrrolylmethane 78 (R = Et) with arylaldehydes in benzene containing a catalytic amount of trifluoroacetic acid and oxidation by oxygen of the air yielded trans-substituted porphyrins 4 with 30–40% yields and with only traces of monoarylporphyrins 3; however, in boiling in propionic acid containing zinc acetate, only monoarylporphyrins 3 with the yield of 15–25% were formed [189]. It follows from these data that oligomeric polymethylenepyrroles and cyclic porphyrinogens initially formed in condensation via route A exist in acidified solutions as an equilibrium mixture with respective pyrrylcarbinyl cations, which acquire stability owing to the

69


delocalization of charge on the pyrrole ring. Pyrrylcarbinyl cations, e.g., 90, can be inculcated via the pyrrole– methylene carbon link, followed by the release of the alternative pyrrylcarbinyl cation 91, which can lead to the removal of the aryl substituent from mesoposition and isomerization of β-alkyl substituents in porphyrins obtained in the subsequent oxidation of porphyrinogens (Scheme 29). A H

Scheme 29 Ar

B NH

H H

A

+

H

+

Ar

NH

CH2

N H 90

B A H

B

B

H

Ar

A H

+

Ar

N H

B H H A H

NH

Ar

NH

A

+

H

+

H

B H H

H Ar

NH

H H

N H

B H H

B H H NH

A

+

CH Ar

N H 91

A

A

B

B

H

B A

A

+90

B H H

+

H Ar

N H

A

Ar

A

-91

NH NH

A

H

NH

H

NH

B B H H

B H H

A H

Ar

Initially, the synthesis was conducted under conditions similar to the production of tetraphenylporphins by a one-stage method in solvents containing acid [26]; however, further studies [190] showed the best results to be provided for by the use of a two-stage method with intermediate formation of porphyrinogen 93 from 92 under the action of acid, and the subsequent oxidation of 93 to porphyrin 94 by benzoquinone derivatives (Scheme 30). Condensation of 5,5′-unsubstituted tetraalkyldipyrrolylmethanes 92 (A, B = Me, Et) with benzaldehydes in methanol in the presence of strong inorganic or organic acid in an Scheme 30 A

A

H R

A

B

B NH +

B

+

H

NH

HN

NH

HN

B

B

[O]

B A 92

A

R NH

N

N

HN

RCHO

NH B

A

B A

H R 93

A

B

B A

R 94

A

70


inert atmosphere followed by isolation and oxidation of the formed porphyrinogen 93 by benzoquinone derivatives yields 5,15-diphenylporphyrins 94 with a high yield (up to 60%) [190– 194]. Studies of the conditions for carrying out this reaction in methanol with organic acid additions [193] showed the yield of diphenylporphyrins not to depend in practice on acid used (CF3COOH or CCl3COOH). However, the formation rate of porphyrinogen decreases significantly in passing from trifluoroacetic acid to benzoic acid. The reaction of condensation and oxidation can be conducted similar to the Lindsey method for meso-tetraphenylporphins without isolating intermediate porphyrinogen in methylene chloride or chloroform. In this case, we can use sufficiently high concentrations of the reacting components (in contrast to the Lindsey method). Studies have shown that the reaction is best run in chloroform with an addition of chloro- or trichloroacetic acid followed by the oxidation of the reaction mass by o-chloranil, p-CA or DDQ [195]. The effect of the length of alkyl substituents in initial dipyrrolylmethanes 92 on the yield of 5,15-diphenyloctaalkylporphyrins 94 was studied [183]. The occurrence of small-size substituents in 3,3′-positions of dipyrrolylmethane (B = H, Me, Et; A = H, Me) has an insignificant effect on the yield of porphyrins; however, the presence of more bulky groups (B = propyl, butyl, amyl, hexyl and especially strongly benzyl; A = Me) decreases it considerably. These facts indicate that the occurrence of bulky substituents in 3,3′-positions of dipyrrolylmethane prevents the formation of conformation 92a, which is required for the condensation reaction to be performed. In this case, they are mainly in the energetically more advantageous transoid form 92b. B H H

B H H

B

NH

A A

A NH HN 92a

NH

B

A

92b

Introduction of the methyl group in 4,4′-positions of dipyrrolylmethane 92 (A = Me) little affects the yield of porphyrin 94 (it even slightly increases); however, its substitution by the ethyl group (A = Et) leads to a decrease of the yield due to steric hindrances of the condensation reaction. In passing from formaldehyde to acetic aldehyde, the yield of porphyrins 94 increases significantly. Transition to benzaldehyde little affects the yield. The electronic nature of substituents and their position in initial benzaldehydes have a small effect on the yield of porphyrins 94, but the occurrence of two substituents in ortho-positions of the phenyl ring leads to a strong decrease of the yield [183]. This two-stage method is at present used sufficiently broadly also for synthesis of strapped 95 [196-199] and cyclophane dimeric porphyrins 96 [198, 200, 201], as well as compound systems containing porphyrin fragments [202–207]. Interestingly, strapped porphyrins 95 are formed in condensation of dimeric benzaldehydes with a certain, larger-than-minimal, length of the linking group [ortho-; R = –(CH2)5-]; yields are sufficiently high. At the length of the substituting group close to a minimum, strapped porphyrins possess a strained distorted dome-like structure, which determines their main physicochemical properties. At the length of the substituting group less than a minimum [ortho-, meta-; n = 3, 4], cyclophane dimeric porphyrins 96 with a low yield are formed. These porphyrins possess an unstrained structure, unlike cyclophane

71


B A

O

A

O

B

N B

O

A

A

O

(CH2)n A

N

N

95

A B B

NH

B

O

HN

B B

(CH2)n

HN

A

N A

A

N

NH

A

N

NH

R B

B

HN

A B

O

A B

96

dimeric porphyrins with four bridges 19; however, they exist in the form of three fixed atropisomers with different distances between the reaction centres of the porphyrin nuclei α,α,α,α-, α,α,α,β- and α,β,α,β- 97. Other possible isomers are not formed due to too large strains in bridge groups [200, 201]. NH N

NH N

N

NH

HN

N

NH

N

N

HN

N

N

HN

NH

N

N

HN

NH

Į,Į,Į,ȕ-97

Į,Į,Į,Į-97

HN N

HN N

Į,ȕ,ȕ,Į-97

Condensation of bis(3-formylbenzyl)sulfide with 4,4′-dimethyl-3,3′-diethyldipyrrolylmethane in acetonitrile in the presence of trichloroacetic acid, with subsequent oxidation by p-CA [208] yields mainly 5,15-porphodimethene 98 (yield, 65%) and only a small amount of dimeric porphyrin 99 (yield, 3%). In the process of condensation, use of a mixture of two aldehydes leads to three porphyrins (Scheme 31); herewith, the required porphyrin 100 by statistical considerations is obtained in the amount of 50% of the porphyrin mixture, which is easier separated than the

NH S

N

CH2

CH2

N HN

H2C

CH2

S NH H

N

N HN

H

S

H2C

CH2 NH N

98

99

N HN

72


mixture of six porphyrins formed in condensation of two aldehydes and pyrrole (Scheme 10), so this method is also used sufficiently widely [207, 209–214]. It should be noted that this method sometimes makes use of a large excess of one, more accessible aldehyde, so that the other aldehyde is used more completely [215, 216]. Other examples include the synthesis of rotoxanes based on porphyrins [217–219] and oligoporphyrins [207, 220]. Scheme 31 C

C

D

H A

D NH NH

H

B-CHO

D NH

[O]

N

D C 92

H B

HN

D

D C

N + A2 + B2

NH HN

D

C

A

D

D NH HN

+

A-CHO +

C

C

D B C 100 (AB)

C

C

As it was pointed out earlier, the reaction (route A) is complicated by that it forms, to a greater or smaller extent, impurities of monophenylporphyrins due to partial regrouping in an acidic medium of dipyrrolylmethanes or intermediate tetrapyrroles [190] (Scheme 29). In some cases, monophenylporphyrins become the main or even sole reaction products, especially at elevated temperature [183, 189, 190], however, their yield is lower than that of diphenylporphyrins. To rule out regrouping processes, of great interest is the use of a method similar to Rotmund’s synthesis of diphenylporphyrins 94. The method consists in the condensation of dipyrrolylmethanes 101 with aldehydes in a medium of high-boiling heterocyclic solvents (pyridine, quinoline, collidine) in the presence of coordination agents (zinc acetate) (Scheme 32) [221]. In the case of using pyridine, the process is run at 180°C under pressure using nitrobenzene as an oxidant, and the application of higher-boiling quinoline makes it possible to conduct the reaction at boiling in the presence of air. The yields of phenyl-substituted porphyrins in this reaction are 10–20%; however, on the one hand, the method rules out regroupings proceeding in an acidic medium, and on the other it enables using more accessible α,α′-dicarboxydipyrrolylmethanes 101 (X = COOH) stable to oxidation. Besides, the method is of interest in condensation with aldehydes unstable in an acidic medium. Another, the most known, method of synthesis of 5,15-diphenylporphyrins 94, first used by Baldwin [222], consists in condensation of α,α′-unsubstituted or α,α′-dicarboxymeso-substituted dipyrrolylmethanes 102 with trimethoxy(ethoxy)methane in the presence of trichloroacetic acid or trifluoroacetic acid in chloroform or methylene chloride [152, 172, Scheme 32 A

A X

B

NH + RCHO NH B

X A 101

A

R

B

B N

Zn(OAc)2 pyridine, nitrobenzene B X=H, COOH

N Zn

N

N B

A

R

A

73


Scheme 33 B

B X

A H

NH

R

NH

B

A H+

A NH

H

+ HC(OMe)3

A

B

R

N

N

R

H

HN

B

A N

NH

R N

A

X=H, COOH

B 102

A

R [O]

A

X

B

HN

A

A

B

B

B

103

ɛ

/

94

ɮ

/

221, 223– 226] (Scheme 33) (route B). An oxidant of intermediate porphodimethene 103 can be oxygen of the air or 1,4-benzoquinone added at the end of the synthesis [223, 225, 226] or DDQ [152]. The reaction proceeds via formylation of 102, condensation of formyldipyrrolylmethenes to porphodimethene 103 and its subsequent oxidation to porphyrin 94, being, in this way, a simplified variety of routes C or D. In [225], the yield of porphyrin 94 was shown to strongly depend on the extent of drying of initial α,α′-dicarboxylic acid of dipyrrolylmethane, whereas the use of α,α′-unsubstituted dipyrrolylmethane makes this reaction less dependent on its conditions. Application of the benzoquinone oxidant at the last stage increases the yield of porphyrins as compared with the use of air oxidation. Method B is widely used for synthesis of belted porphyrins 95 [179, 222, 227] (Scheme 33) and cyclophane dimeric porphyrins 88 [200] (Scheme 34). In this case, selfcondensation of bis-dipyrrolylmethane 104 or 105 is performed. Scheme 33 R

R

B O

NH

HOOC

+ HC(OMe)3

NH

A

B

B COOH

A

A

HN

H+

A

B NH N

A

A

HOOC

B

O

[O]

HN

COOH

O

HN

B

B

A

N

O

A B

104

95

Scheme 34 B

B COOH

A O

O

COOH + HC(OMe)3 COOH

B B A

A

NH

NH

O

NH

A (CH2)n

A

NH

N

H+ [O]

A

B 105

A B B

N A

O

HN

O

HN

B

A B

96

(CH2)n A

N

NH

O

A

N

B B

(CH2)n

NH COOH

B

74


Route D makes it possible to obtain trans-disubstituted porphyrins 108 having various meso-substituents; however, this route requires two different dipyrrolylmethanes 106 and 107, each of which can not self-condense [228] (Scheme 35). Scheme 35 B A NH

R

H

NH

HN

R1

A B

D

H [O]

R1

R N

HN D

A B

C 107

106

N

NH

+

D

OHC

C

A

D

HN +

H

B

C OHC

C 108

Asymmetrically substituted diphenylporphyrins 108, besides, can be obtained by condensation of 1,19-unsubstituted 10-substituted biladienes-a,c 109 with aldehydes. An example of this kind of synthesis is presented in Scheme 36 [229]. Initial biladienes 109 can be obtained by condensation of meso-substituted dipyrrolylmethanes 106 with 2-formylpyrroles 110. Scheme 36 B

B

A

A

R

NH

H

+ NH OHC

A

C

D

106

+

H N H

B

3

C NH HN

R

D

+

D

+

D

A C 110

N

NH

R1

R

H

NH HN

C

A

R 1-CHO

+

H

B 109

B

N

HN D

A B

C 108

Synthesis of meso(5)-Phenyl-β-octaalkylporphyrins

There are several methods of synthesis of 5-phenyloctaalkylporphyrins 3. Earlier, it has been shown [189, 190] that monophenylporphyrins are formed, owing to regroupings in an acidic medium (Scheme 29), as an impurity in condensation of α-unsubstituted dipyrrolylmethanes with aldehydes; herewith, in some cases they are the main products of this reaction, though their yield is lower than that of respective diphenylporphyrins. The most currently widespread method of synthesis of 5-substituted porphyrins 113 consists in condensation of meso-substituted α-unsubstituted dipyrrolylmethanes 106 with 5,5′-diformyldipyrrolylmethanes 111 in alcohols or methylene chloride under the action of strong acids (hydriodic, perchloric or p-toluenesulfo); the oxidant of intermediate porphodimethene 112 is oxygen of the air or benzoquinone derivatives (Scheme 37) [207, 225, 230–240]. 5,5′-Diformyldipyrrolylmethanes 111 are produced by formylation of α-unsubstituted dipyrrolylmethanes 92 by a mixture of POCl3 –DMFA according to Wilsmeyer [173] or a mixture of ortho-formic ester and trifluoroacetic acid [241].

75


Scheme 37 C

A

A

OHC

B H

D

HN

NH

B H

+

A

C B

D

H

NH

R

N

N

HN

NH

B

D

OHC A

C

106

111

D NH

[O]

N

HN

B

HN

B

D A

N

R

+

R

C

D A

C

C

112

113

Another, similar, method is condensation of α-unsubstituted meso-phenyldipyrrolylmethanes 106 with 5,5′-dimethoxymethyldipyrrolylmethenes 114 in boiling benzene followed by oxidation by benzoquinone derivatives [180] (Scheme 38). A lower yield of porphyrins is compensated for by the accessibility of 5,5′-dimethoxymethyldipyrrolylmethenes 114, which are obtained by solvolysis of pyrroles 115 in a hot solution of hydrobromic acid (48%) and formic acid (88-90%), which leads with high yields (80– 90%) to dipyrrolylmethenes 116 symmetric relative to the meso-bridge carbon. Dipyrrolylmethenes 116 are further brominated in a methanol solution to 114 [116, 242, 243] (Scheme 39). The method is complicated by that, along with 5-phenylporphyrins 113, it forms a sufficiently large amount of octaalkylporphyrins unsubstituted in meso-positions. However, the performance of this synthesis in the absence of acid catalysts excludes the possibility of regroupings of β- and meso-substituents [180]. Scheme 38 C

A MeOH2C

B H

NH

R

NH

+

B

A D

B

HN Br+ HN

N

R N

HN

B

C 114

A 106

D NH

D

MeOH2C

C

D A

C 113

Scheme 39 C

C C H3C

D

D N H

HBr CO2R HCOOH R = H, Et, tBu

115

CH3

D

NH 1. Br2 Br+ 2. MeOH NH D CH3 C 116

CH2OMe

D

NH Br+ NH CH2OMe C 114

76


Scheme 40 B

B

A NH HN 2Br+ NH HN

B A

+

A +

+

RCHO

H or OH

C D

A NH

-

N

C

B

R N HN

C

D

C D

D

117

113

Scheme 41 B

C X

D

A B

NH

A

+ N H

NH D

B NH HN 2Br+ NH HN

+

H CHO

C

X

X= H, COOH

C 101

A

+

C D

109

D 117

Scheme 42 B

B

A NH HN 2Br+ NH HN

B A

+

A NH

N

NH HN C D

117

A OH-

C D

B

C

C D

D 118

Yet another widespread method of synthesis of 5-phenylporphyrins 113 is condensation of benzaldehydes with 1,19-diunsubstituted biladienes-a,c 117 in alcohols at acid catalysis [234, 238, 244] or base catalysis (Scheme 40). In this case, there are no limitations on the symmetry of β-substituents in the porphyrin cycle, as their set is given by the initial biladiene-a,c. Initial biladienes are sufficiently accessible and are obtained in acidcatalyzed interaction of dipyrrolylmethanes 101 with formylpyrroles 109 (Scheme 41). An interesting feature of this synthesis of 5-phenylporphyrins is that respective corrole 118 can be formed besides the required 5-phenylporphyrin 113; herewith, the amount of the corrole is determined by the porphyrine formation rate. Corroles 118 are the only products in self-condensation of 1,19-diunsubstituted biladienes 117 in the presence of bases during the action of weak oxidants on them [246] or in illumination [247] (Scheme 42).

77


4

4.1

Reactions of Introduction and Modification of Substituents in Phenyl Rings of meso-Phenyl-substituted Porphyrins Introduction of substituents into phenyl rings of meso-phenylporphyrins

As β-positions of the porphyrin macroring in reactions of electrophilic substitution are more active than phenyl rings, only some examples of direct electrophilic substitution in mesoaryl groups are known. In all such cases, the porphyrin cycle is deactivated to the electrophilic attack by protonation via intracyclic atoms of nitrogen. In 1962, Winkelman, in studies of the ability of porphyrins to be selectively accumulated in tumour cells, first performed the sulfonation of tetraphenylporphin (H2TPP) and obtained water-soluble meso-tetra(4-sulfophenyl)porphin 119 (yield, ~70%) (Scheme 43) [248]. The reaction was performed at 100°C; concentrated sulfuric acid was used as a sulfonating agent. Subsequently, it was found that, along with 119, porphyrin with sulfo groups was also formed in only three phenyl rings [249]. Scheme 43 HO3S

SO3H

N NH

N HN

H2SO4 o

NH

HN

100 C N

N

SO3H

HO3S 119

Isolation of compound 119 from the reaction mixture is associated with a number of problems, however, is possible via its sodium or ammonium salts by way of multiple resedimentation from a methanol– acetone solution [250], via the insoluble potassium salt [251], as well as by dialysis [252]. To investigate the liquid-crystalline properties of H2TPP derivatives, the authors of [253] conducted the sulfonation of covalently bound linear dimer of H2TPP to hexasulfophenyl derivative 120 with the yield of 90%. The reaction was carried out by heating porphyrin in concentrated sulfuric acid in a sealed ampoule on a boiling water bath for 4 h. Similarly, sulfonation of mono-, di- and tetraphenyl-β-octamethylporphins was carried out [254]. The results obtained show that the reaction is general irrespective of the number of phenyl residues and the presence or absence of β-alkyl groups. Electron-donor groups in phenyl rings facilitate sulfonation. Thus, in sulfonation of β-octabromo-meso-tetramesitylporphin by oleum at 120°C all eight free positions of mesomesityl groups are sulfonated. This porphyrin was assumed to be used as a water-soluble catalyst in oxygen transfer reactions [255, 256]. Production of sulfonated derivatives with hydroxy- and methoxy- groups in ortho- and para-positions of phenyl rings [257– 259] and meso-tetrakis(2-thienyl)porphin 121 was described [260]. These porphyrins were assumed to be used in analytical purposes. It is interesting to note that sulfation of meso-tetrakis-

78


(2,6-dimethylphenyl)porphin proceeds in meta-positions of phenyl rings, forming 122 [43], but not in para-positions as it was assumed earlier [261]. Sulfonation of mono-p-nitrophenyltriphenylporphin by oleum leads to 5,10,15-tri(4-sulfonylphenyl)-20-(4-nitrophenyl)porphin, which was purified by chromatography on Sephadex G-10 [262]. HO3S

SO3H

OCH2CH2CH2CH3O

N

N NH

HN

NH

HN N

N

HO3S

SO3H

HO3S

SO3H

120 HO3S

SO3H

SO3H

S

Me

S

Me SO3H

N HN

NH

N

Me NH

HN

N

Me

S S

Me

N

HO3S

SO3H

HO3S

Me

Me

Me

121

HO3S

122

The authors of [263], using the reaction of H2TPP to chlorosulfonic acid at room temperature, obtained respective tetrachlorosulfonic derivatives. Chlorosulfonation of mesotetrakis(2,6-dichlorophenyl)porphin under the same conditions yields a mixture of mono-, di- and tri-para-sulfonyl derivatives. Chlorosulfonation can be carried out with a high yield at 100°C [263]; besides, tetrachlorosulfonic derivatives of H2TPP can be obtained by treating salts 119 with excess thionyl chloride in DMFA at 50°C for 3 h [264]. Chlorosulfonyl groups are easily converted to free sulfo acid, sulfamides or sulfo esters at the nucleophilic attack with water, ammonia (amines) or alcohols, respectively [263, 264] (Scheme 44). Scheme 44 ClO2S

SO2Cl H2O N NH

HN

SO2OH

HNR2

SO2NR2

HOR

SO2OR

N

ClO2S

SO2Cl

79


To obtain models of cytochrome P-450, porphyrin 123 was obtained by the treatment of the sulfochloride derivative with respective amine [265, 266]. RO2S Cl

Cl

RO2S N

Cl NH Cl

Cl HN Cl

N

Me R = NH(CH2)3 Si OEt OEt SO2R

Cl

Cl

SO2R 123

Of the other reactions of electrophilic substitution in phenyl rings of phenylporphyrins, only the nitration of H2TPP by the action on it of excess fuming nitric acid in chloroform is described [262]; herewith, the formed product is 5-(p-nitrophenyl)-10,15,20-triphenylporphin 124 with the yield of 55% (Scheme 45). An increase of the nitration time and concentration of nitrating reagent leads to a mixture of mono- and di(nitrophenyl)porphyrins. Porphyrin 124 can also be obtained with the yield of 80% by the treatment with a solution of H2TPP in methylene chloride by acetic anhydride, trifluoroacetic acid and HNO3 in the presence of montmorillonite K10 [267]. The work [268] studied the effect of the nature of substituent on the nitration reaction. The most readily nitrated were found to be H2TPP derivatives having electron-donor substituents in meta-positions of phenyl rings (nitration proceeds in para-position). Occurrence of an electron-donor substituent in para-positions leads, under conditions of the reaction, mainly to oxidative destruction of the porphyrin cycle. Electron-acceptor substituents in meta-positions hinder the nitration reaction, and in their occurrence in para-positions the reaction does not proceed. Scheme 45 NO2

N

N NH

HN

HNO3 0

CHCl 3, 0-5 C

NH

HN N

N

9

124

4.2

Modification of functional groups in phenyl rings of meso-phenylporphyrins

Many derivatives of phenyl-substituted porphyrins are difficult to synthesize by condensation of pyrrole or its linear derivatives with respective benzaldehydes. In this connection, of interest is to consider methods, which make use of a modification of substituents in

80


benzene rings of porphyrins obtained with high yields. Of great interest are porphyrins containing in their phenyl rings active substituents, such as oxy or amino groups, as they are capable of various chemical conversions. By direct synthesis, these porphyrins are obtained with a low yield or are not formed at all. 4.2.1

meso-Oxyphenylporphyrins and their modification

meso-Oxyphenylporphyrins are formed in condensation reactions with low yields and large amounts of hard-to-separate impurities. Protection of the oxy group in respective oxybenzaldehydes by acylation [127] or sulfonylation [269] leads to a significant increase of the yields of respective porphyrins, which, by way of hydrolysis in an alkaline medium, can sufficiently easily yield required meso-oxyphenylporphyrins. However, a more promising method is hydrolysis of readily available meso-methoxyphenylporphyrins (Scheme 46). Scheme 46 OCOR OH OH OSO2R BBr3

-

OH

RHal K2CO 3

OR

OCH3

Hydrolysis of methoxyphenylporphyrins by the general method (by 48% hydrobromic acid) showed this method to be little efficient, in view of its long duration and incomplete conversion. The use of anhydrous aluminium chloride in boiling chlorobenzene as demethylating agents gives no satisfactory results, either, as under the stringent conditions characteristic of this method, porphyrins are subjected to a significant breakdown. Good results are obtained using pyridine or aniline hydrochlorides in boiling as demethylating agents [270]. This method, however, is only applicable in the synthesis of the most stable meta- and para-oxysubstituted tetraphenylporphyrins, whereas ortho-oxysubstituted tetraphenylporphyrins, as well as monooxyphenyl- and dioxyphenylporphyrins are subjected to a significant destruction under conditions of this reaction. A more suitable demethylating agent for synthesis of oxyphenylporphyrins is 60% hydrobromic acid (boiling in inert atmosphere), the use of which makes it possible to significantly increase their yield [270]. At present, a mild demethylating agent – boron tribromide in methylene chloride at –20°C in inert atmosphere – is used for hydrolysis of methoxyphenylporphyrins [271– 276]. However, it has been shown [277, 278] that the demethylation reaction can, without decreasing the yield, be conducted at room temperature in the air in chloroform. The reaction of alkylation of oxyphenylporphyrins by haloalkanes is of certain interest, as it enables one-stage one-precursor production of porphyrins possessing diverse physicochemical properties or having active groups at the periphery of the molecule. These active groups can interact with the active central part of the porphyrin macrocycle or serve for attaching the molecule to various substrates.

81


Alkylation of oxyphenylporphins is performed at present mainly in dimethylformamide in the presence of the basic agent potassium carbonate [279– 281]. The reaction proceeds at room temperature for labile oxyphenylporphyrins [279, 282] or during the boiling for stable derivatives of tetraphenylporphin, which significantly decreases the interaction time [281]. The yield of alkoxy derivatives of porphyrins is on average 80–95%. Alcohols are not in practice used for this reaction, as both initial porphyrins and interaction products solve in them poorly. The discussed method was used to synthesize various alkoxy-substituted porphyrins, well soluble in nonpolar organic solvents [281, 282], as well as tetra(carboethoxymethyleneoxyphenyl)porphyrins 125, whose hydrolysis forms tetra(carboxymethyleneoxyphenyl)porphyrins 126 soluble in alkaline solutions [283] (Scheme 47). Scheme 47 OH ClCH2CO2Et

OCH2CO2H

OCH2CO2Et 1. OH2. H+

K2CO3 125

126

Metalloporphyrins in nature function in chromoproteins, whose protein groups strongly affect the properties of metal complexes [284]. Therefore, of great importance is the synthesis and study of porphyrins, which at the periphery of the molecule have active functional groups capable of interacting with the central reaction centre of the porphyrin macrocycle. Synthesis of such compounds based on modifications of natural porphyrins is rather complicated and includes a large number of stages [223]. Bonds formed after attaching residues with active groups are not too strong (they are mainly amide ones or ester ones) [223, 285–287]. For synthesis of such compounds, it is of interest to use synthetic porphyrins with oxyphenyl groups capable of forming stable ether bonds. The amount of active groups per one porphyrin molecule rarely exceeds one, so the most appropriate oxyphenylporphyrins for these purposes are monooxyphenylporphyrins: oxyphenyltriarylporphyrins 127a, which are rather easily obtained by condensation of pyrrole with a mixture of oxybenzaldehyde and benzaldehyde [47, 126, 127], or meso-oxyphenyl-β-alkylporphyrins 128a closer to natural phorphrins. R

A B

NH

B

N

Ar

A

A

R NH

A

R

B

OH

B

N

NH

aR=

N NO2

Ar N

HN

N

HN

N

C

Ar

127

C D

D

128

bR =

HN

B

B A

R

A

NH2 ɫR=

129

A generalized scheme of the synthesis of porphyrins with active groups by the alkylation reaction is presented in Scheme 48. As is seen in the scheme, the attachment of residues with active groups to porphyrin

82


Scheme 48 O(CH2)nX Br(CH 2)nX OH

A HX B

O(CH2)nBr

Br(CH 2)nBr

ɏ - residue with an active group can proceed via two routes: alkylation of oxyphenylporphyrin by haloalkane with an active group (route A), or preliminary alkylation of oxyporphyrin by excess α,ω-dibromoalkane to form ω-bromoalkoxyphenylporphyrin with its subsequent interaction under similar conditions with a compound having an active group (route B). Route A is more preferable, as it is shorter and provides a higher yield of the end product (with respect to initial oxyphenylporphyrin). However, it is not always possible to obtain and purify haloalkane having an active functional group. In this connection, route B is mainly used [279, 288]. 4.2.2

meso-Aminoporphyrins and their modification

High yields of aminophenylporphyrins in condensation reactions of aminobenzaldehydes with pyrrole and its derivatives are hard to expect. Aminobenzaldehydes are known to be very unstable; e.g., p-aminobenzaldehyde polymerizes in an acidic medium and m-aminobenzaldehyde exists only in diluted solutions or as a complex with tin dichloride dehydrate. The literature gives a mention of only the direct synthesis of tetrakis(4-aminophenyl)porphin [26] with the nonreproducible yield of about 1%. Protection of the amino group by acylation leads to a significant (up to 10%) increase of the yield of porphyrin; however, other isomeric aminophenylporphyrins could not be obtained in this way. Proceeding from the above-said, a more promising approach is production of aminophenylporphyrins by reducing respective nitrophenylporphyrins, which are obtained with sufficiently high yields by the condensation of nitrobenzaldehydes with pyrrole and its derivatives. As the sole reducer, this reaction makes use at present of tin dichloride dihydrate in hydrochloric acid or in polar solvents with its addition. This method was proposed by Collman [289] and found wide use. For tetra(nitrophenyl)porphyrins, reduction is done at a 1.5-fold excess of tin dichloride dihydrate in concentrated hydrochloric acid at a temperature of 70–80°C, which enables an almost qualitative yield of tetra(aminophenyl)porphyrins [290– 293]. Mono- 127b, 128b and disubstituted 129b nitrophenylporphyrins require more mild conditions of reduction, so in this case the reduction reaction is conducted at room temperature [294, 295] and in a methanol medium, which transfers the obtained aminophenylporphyrins into a solution and contributes to the complete reduction of initial nitrophenylporphyrins. The amino group of aminophenylporphyrins is very active and reacts to many reagents. Of special interest is the diazotization reaction, which is widely used for synthesis of aromatic compounds with various functional groups. This reaction has advantages in that

83


only a small amount of groups prevent it, in contrast with reactions using Grignard reagents. Application of the diazotization reaction enables production of an interesting and important class of organic compounds – azo dyes. Tetra(aminophenyl)porphins are readily diazotized by sodium nitrite in aqueous solutions of mineral acids [296–298]. The obtained diazonium salts are rather stable. Their noticeable decomposition with evolution of nitrogen is observed only at a temperature higher than room temperature. The porphyrin cycle under conditions of the diazotization reaction is stable; however, diazotized tetra(2-aminophenyl)porphin during the heating yields compounds of nonporphyrin character. Owing to this, ortho-substituted tetraphenylporphins are formed with low yields and a large amount of hard-to-separate impurities. However, the azocoupling reaction of diazotized tetra(2-aminophenyl)porphin to phenol successfully proceeds on the cold to form tetra(2-oxyphenylazophenyl)porphin. Diazotized tetra(3- and 4-aminophenyl)porphins during the heating yield tetra(oxyphenyl)porphins as well as enter into the Sandmeyer reaction and the azocoupling reaction, which makes it possible to obtain tetra(halogenophenyl)porphins and azo dyes based on porphyrins, which are impossible to obtain in any other way (Scheme 49). Scheme 49 F I HBF4 KI Cu NO2

SnCl 2 HCl

NH2 HNO

+

N2

2

H+ HR

Cu 2X2 X=Cl,Br,CN t

X

OH

N=NR

Diazotization of mono- 127c, 128c and disubstituted 129c aminophenylporphyrins followed by conversion of the obtained diazonium salts makes it possible to synthesize substituted porphyrins, which are inaccessible using other routes of synthesis. The use of sodium nitrite as a diazotizing agent in aqueous solutions of the acids in this case does not lead to positive results, as such aminophenylporphyrins are not soluble under these conditions. However, aminophenylporphyrins are diazotized by amylnitrite in a mixture of chloroform and acetic acid [298]. The obtained solutions of diazonium salts enter into the Sandmeyer reaction and the azocoupling reaction; the required basicity of the medium (pH = 8–9) was created in this case by triethylamine. Methylation of aminoporphyrins by iodomethane in DMFA in the presence of 2,6lutidine makes it possible to obtain cationic trimethylaminophenylporphyrins with the yield of about 70% [299]. Tetrakis(para- and meta-trimethylaminophenyl)porphins are soluble in water within a wide range of pH. Using potassium carbonate as bases, the reaction is stopped at the stage of formation of dimethylaminophenylporphyrins [299] (yield, about 80%) (Scheme 50). Thus, combining the methods of synthesis of meso-phenylporphyrins and modification of substituents in their phenyl rings makes it possible to obtain compounds with practically any predetermined physicochemical properties and a required set of substituents.

84


Scheme 50 N(CH3)2 CH3I NH2

K2CO 3 CH3I

2,6-lutidine

+

N(CH3)3

The chemistry of porphyrins and their analogues is rapidly developing, so this review does not pretend to be a complete coverage of the literature on the subject. It can be recommended as a strategy for the synthesis and modification of required meso-substituted porphyrins based on the methods developed at present the most.

References 1. B.D. Berezin, Coordination Compounds of Porphyrins and Phthalocyanine, Moscow: Nauka, 280 pp. (1978) (in Russian). 2. G.P. Gurinovich, A.N. Sevchenko and K.N. Solovyev, The Spectroscopy of Chlorophyll and Related Compounds, Nauka i Technika: Minsk, 517 pp. (1968) (in Russian). 3. The Porphyrins and Metalloporphyrins, ed. by K.M. Smith, Elsevier: New York, etc., 590 pp. (1975). 4. The Porphyrins, ed by D. Dolphyn, New York: Acad. Press, 7 v. (1978–1979). 5. Porphyrins: Structure, Properties, Synthesis, ed. by N.S. Enikolopyan, Nauka: Moscow, 333 pp. (1985) (in Russian). 6. Porphyrins: Spectroscopy, Electrochemistry and Applications, ed. by N.S. Enikolopyan, Nauka: Moscow, 384 pp. (1987) (in Russian). 7. M.R. Tarasevich and K.N. Radyushkina, Catalysis and Electrocatalysis by Porphyrins, Nauka: Moscow, 168 pp. (1982) (in Russian). 8. B.D. Berezin and N.S. Enikolopyan, Metalloporphyrins, Nauka: Moscow, 160 pp. (1988) (in Russian). 9. Advances of the Chemistry of Porphyrins, in four volumes, ed. by O.A. Golubchikov, Chemistry Research Institute, St.-Petersburg University: St.-Petersburg (1997, 1999, 2001, 2004) (in Russian). 10. The Porphyrins Handbook, in 20 volumes, ed. by K.M. Kadish, K.M. Smith and R. Guilard, Acad. Press: New York (2000–2005). 11. P. Rothemund, J. Amer. Chem. Soc., 57 (7), 2010 (1935). 12. P. Rothemund, J. Amer. Chem. Soc., 61 (9), 2912 (1939). 13. P. Rothemund, J. Amer. Chem. Soc., 63 (1), 267 (1941). 14. R.H. Ball, G.D. Dorough and M.A. Calvin, J. Amer. Chem. Soc., 68 (11), 2278 (1946). 15. J.H. Priesthoff and C.V. Banks, J. Amer. Chem. Soc., 76 (3), 937 (1954). 16. M.M. Williamson, C.M. Prosser-McCartha, S. Mukundan (Jr.) and C.L. Hill, Inorg. Chem., 27, 1061 (1988). 17. O. Bortolini, M. Ricci, B. Mennier, P. Frant, I. Ascone and J. Goulon, Nouv. J. Chem., 10 (1), 39 (1986). 18. D.B. Sharp, U.S. Patent 3.076.813 (1963). 19. A. Petit, A. Loupy, P. Mallard and M. Momenteau, Synth. Commun., 22 (8), 1137 (1992). 20. P. Laszlo and J. Luchetti, Chem. Lett., 3, 449 (1993). 21. M. Onaka, T. Shinoda, Y. Izimi and E. Nolon, Chem. Lett., 1, 117 (1993). 22. C.M. Drain and X. Gong, Chem. Commun., 2117 (1997). 23. G.A. Mirafzal, H.M. Bosse and J.M. Summer, Tetrahedron Lett., 40, 623 (1999). 24. D.V. Thomas and A.E. Martell, J. Amer. Chem. Soc., 78 (7), 1335 (1956). 25. A.D. Adler, E.R. Longo and W. Shergalis, J. Amer. Chem. Soc., 86 (15), 3145 (1964). 26. A. Treibs and H. Haberle, J. Liebigs Ann. Chem., 718, 183 (1968).


85

27. D. Dolphin, J. Heterocycl. Chem., 2, 275 (1970). 28. A.D. Adler, F.R. Longo, J.D. Finarelli, J. Goldmacher, J. Assour and L. Korsakoff, J. Org. Chem., 32, 476 (1967). 29. J.B. Kim, J.J. Leonard and F.R. Longo, J. Amer. Chem. Soc., 94 (11), 3986 (1972). 30. M.J. Crossley, P. Thordarson, J.P. Bannerman and P.J. Maynard, J. Porphyrins Phthalocyanines, 2 (6), 511 (1998). 31. A.S. Semeykin, O.I. Koifman and B.D. Berezin, Khim. Geterotsykl. Soed., 6, 798 (1986) (in Russian). 32. N.E. Kagan, D. Mauzerall and R.B. Merrifield, J. Am. Chem. Soc., 99 (16), 5484 (1977). 33. S. Banfi, F. Montanari, M. Penso, V. Sosnovskikh and P. Vigano, Gazz. Chim. Ital., 117 (11), 689 (1987). 34. A.S. Semeykin, N.G. Kuzmin and O.I. Koifman, Zhurn. Prikl. Khim., 6, 1426 (1988) (in Russian). 35. C.-C. Guo, X.-T. He and G.-Y. Zhon, Chin. Org. Chem., 11 (4), 416 (1991). 36. S.M.S. Chauhan, B.B. Sahoo and K.A. Srinivas, Synth. Commun., 31 (1), 33 (2001). 37. A.M. Rocha Gonsalves, J.M.T.B. Varejao and M.M. Pereira, J. Heterocyclic Chem., 28 (3), 635 (1991). 38. V. Sol, J.C. Blais, G. Bolbach, V. Carre, R. Granet, M. Guilloton, M. Spinro and P. Krausz, Tetrahedron Lett., 38, 6391 (1997). 39. K. Ohta, M. Ando and I. Yamamoto, J. Porphyrins Phthalocyanines, 3 (4), 249 (1999). 40. G.H. Barnett, M.F. Hudson and K.M. Smith, J. Chem. Soc. Perkin Trans., 1 (14), 1401 (1975). 41. G. H. Barnett, M.F. Hudson and K.M. Smith, Tetrahedron Lett., 2887 (1973). 42. K. Rousseau and D. Dolphin, Tetrahedron Lett., 48, 4251 (1974). 43. M.F. Zipplies, W.A. Lee and T.C. Bruice, J. Am. Chem. Soc., 108 (15), 4433 (1986). 44. R.A.W. Johnstone, M.L.P.G. Nunes, M.M. Pereira, A.M. Rocha Gonsalves and A.C. Serra, Heterocycles, 43, 1423 (1996). 45. L. Czuchajowski and M. Lozynski, J. Heterocycl. Chem., 25 (1), 349 (1988). 46. M.A. Torrens, D.K. Straub and L.M. Epstein, J. Am. Chem. Soc., 94 (12), 4160 (1972). 47. J.S. Lindsey, I.C. Scheriman, H.C. Hsu, P.C. Kearney and A.M. Marguerettaz, J. Org. Chem., 52 (5), 827 (1987). 48. J.S. Lindsey, H.C. Hsu and I.C. Schreiman, Tetrahedron Lett., 27, 4969 (1986). 49. A.M.d’A. Rocha Gonsalves, M.M. Pereira, A.C. Serra, R.A.W. Johnstone and M.L.P. Nunes, J. Chem. Soc. Perkin Trans., 1, 2053 (1994). 50. J.S. Lindsey, K.A. MacCrum, J.S. Tyhonas and Y.-Y. Chuang, J. Org. Chem., 59 (3), 579 (1994). 51. F. Li, K. Yang, J.S. Tyhonas, K.A. MacCrum and J.S. Lindsey, Tetrahedron, 53 (37), 12339 (1997). 52. B.J. Littler, Y. Ciringh and J.S. Lindsey, J. Org. Chem., 64 (8), 2864 (1999). 53. J.S. Lindsey and R.W. Wagner, J. Org. Chem., 54 (4), 828 (1989). 54. R.W. Wagner, D.C. Lawrence and J.S. Lindsey, Tetrahedron Lett., 28 (27), 3069 (1987). 55. R.W. Wagner, F. Li, H. Du and J.S. Lindsey, Org. Proc. Res. Dev., 3, 28 (1999). 56. M.K. Safo, G.P. Gupta, F.A. Walker and W.R. Scheidt, J. Am. Chem. Soc., 113 (15), 5497 (1991). 57. A.W. Van der Made, E.J.H. Hoppenbrauwer, R.J.M. Nolte and W. Drenth, Rec. Trav. Chim. Pays-Bas., 107 (1), 15 (1988). 58. M. Kihn-Botulinski and B. Meuier, Inorg. Chem., 27 (1), 209 (1988). 59. M.S. Chorghade, D. Dolphin, D. Dupre, D.R. Hill, E.G. Lee and T.P. Wijesekera, Synthesis, 1320 (1996). 60. S. Banfi, F. Montanari and S. Quici, J. Org. Chem., 53 (12), 2863 (1987). 61. A.M. Rocha Gonsalves and M.M. Pereira, J. Heterocycl. Chem., 22 (3), 931 (1985). 62. S.V. Vodzinsky, PhD (Chemistry) Thesis, Odessa State University: Odessa, 21 pp. (1990) (in Russian). 63. M. Onaka, T. Shinoda, Y. Izumi and E. Nolon, Tetrahedron Lett., 34, 2625 (1993). 64. T. Shinoda, Y. Izumi and M. Onaka, J. Chem. Soc. Chem. Commun., 1801 (1995). 65. T. Shinoda, M. Onaka and Y. Izumi, Chem. Lett., 493 (1995). 66. J.S. Lindsey, in: Metalloporphyrin-Catalayzed Oxidations, ed. by F. Montanari and L. Casella, Kluwer Academic Publishers: The Netherlands, p. 49 (1994).

86


67. R.P. Bonar-Law, J. Org. Chem., 61 (11), 3623 (1996). 68. J.A.S. Cavaleiro, G.W. Kenner and K.M. Smith, J. Chem. Soc. Perkin Trans. 1, 2478 (1973). 69. E.K. Gottwald and E.F. Ullman, Tetrahedron Lett., 3071 (1969). 70. R. Paolesse, L. Jaquinod, D.J. Nurco, S. Mini, F. Sagone, T. Boschi and K.M. Smith, Chem. Commun., 1307 (1999). 71. P.J. Chmielewski, L. Laios-Grazynski, K. Rachlewicz and T. Glow, Angew. Chem., 106, 805 (1994); Angew. Chem. Int. Ed. Engl., 33, 779 (1994). 72. H. Furuta, T. Asano and T. Ogawa, J. Am. Chem. Soc., 116, 767 (1994). 73. G.R. Geier (III) and J.S. Lindsey, J. Porphyrins Phthalocyanines, 6 (3), 159 (2002). 74. P.J. Chmielewski, L. Latos-Grazynski and K. Rachlewicz, Chem. Eur. J., 1, 68 (1995). 75. U. Simonis, F.A. Walker, P.L. Lee, B.J. Hanquet, D.J. Meyerhoff and W.R. Scheidt, J. Amer. Chem. Soc., 109 (9), 2659–2668 (1987). 76. D. Reddy and T.K. Chandrashekar, J. Chem. Soc. Dalton Trans., 619–625 (1992). 77. M. Ravikanth, D. Reddy, A. Misra et al., J. Chem. Soc. Dalton Trans., 1137 (1993). 78. O.A. Golubchikov, S.G. Korovina, E.M. Kuvshinova et al., Zhurn. Org. Khim., 24 (11), 2378 (in Russian) (1988). 79. E.M. Kuvshinova, O.A. Golubchikov, and B.D. Berezin, Zhurn. Obshch. Khim., 61 (8), 1799 (1991) (in Russian). 80. J. Almog, J.E. Baldwin, R.L. Dyer and M. Peters, J. Amer. Chem. Soc., 97 (1), 226 (1975). 81. J.E. Baldwin, J.H. Cameron, M.J. Crossley, I.J. Dadley, S.R. Hall and T. Klose, J. Chem. Soc. Dalton. Trans., 8, 1739 (1984). 82. K. Ma, C. Slebobodnick and J.A. Ibers, J. Org. Chem., 58 (23), 6349 (1993). 83. W.F.K. Schnatter, O. Almarsson and T.C. Bruice, Tetrahedron, 47 (41), 8687 (1991). 84. H.-Y. Zhang, A. Blasko, J.-Q. Yu and T.S. Brince, J. Amer. Chem. Soc., 114 (17), 6621 (1992). 85. B. Garcia, C.-H. Lee, A. Blasko and T.C. Bruice, J. Am. Chem. Soc., 113 (21), 8118 (1991). 86. M. Momenteau, B. Loock, J. Muspelter and E. Bisagni, Nouv. J. Chim., 3 (2), 77 (1979). 87. O. Wennerstrom, H. Ericsson, I. Raston et al., Tetrahedron Lett., 30 (9), 1129 (1989). 88. Yu.V. Ishkov and Zh.V. Grushevskaya, Sci. Conf. of Young Scientists, Odessa State University: Odessa, p. 90 (1989) (in Russian). 89. H. Volz and M. Hassler, Liebigs Ann. Chem. 2, 171 (1989). 90. H. Volz and H. Schaffer, Chem.-Ztg., 109 (9), 308 (1985). 91. H. Volz, M. Hassler and H. Schaffer, Z. Naturforsch., 41b (10), 1265 (1986). 92. H. Volz and M. Hassler, Z. Naturforsch., 43b (8), 1043 (1988). 93. S.G. DiMagno, R.A. Williams and M.J. Therien, J. Org. Chem., 59 (23), 6943 (1994). 94. J.G. Goll, K.T. Moore, A. Ghosh and M.J. Therien, J. Am. Chem. Soc., 118 (35), 8344 (1996). 95. R.G. Little, J. Heterocycl. Chem., 18 (4), 129 (1981). 96. H. Volz and G. Herb, Z. Naturforsch., 39b (10), 1394 (1984). 97. R.G. Little, J. Heterocycl. Chem., 18 (4), 833 (1981). 98. Y. Kuroda, H. Murase, Y. Suzuki and H. Ogoshi, Tetrahedron Lett., 30 (18), 2411 (1989). 99. G. Pozzi, F. Montanari and S. Quici, Chem. Commun., 69 (1997). 100. J.L. Hoard, in: Porphyrins and Metalloporphyrins, ed. by K.M. Smith, Elsevier: Amsterdam, Chapter 8, p. 317 (1975). 101. W.R. Scheidt, in: The Porphyrins, ed. by D. Dolphin, Academic Press: New York, 3, Chapter 10, p. 463 (1979). 102. W.R. Scheidt and Y. Lee, J. Struct. Bonding (Berlin), 64, 1 (1987). 103. J.A. Shelnutt, X.-Z. Song, J.-G. Ma et al., Chem. Soc. Reviews, 27, 31 (1998). 104. W. Jentzen, J.-G. Ma and J.A. Shelnutt, Biophys J., 74, 753 (1998). 105. J.-G. Ma, M. Laberge, X.-Z. Song et al., Biochem., 37, 5118 (1998). 106. D.J. Nurco, C.J. Medforth, T.P. Forsyth et al., J. Amer. Chem. Soc., 118, 10918 (1996). 107. M.O. Senge and W.W. Kalisch, Inorg. Chem., 36 (26), 6103 (1997). 108. W.W. Kalisch and M.O. Senge, Tetrahedron Lett., 37 (8), 1183 (1996). 109. J. Takeda and M. Sato, Tetrahedron Lett., 35 (21), 3565–3568 (1994). 110. J. Takeda and M. Sato, Chem. Lett., 2233 (1994). 111. S.A. Syrbu, T.V. Lyubimova and A.S. Semeykin, Khim. Geterotsykl. Soed., 10, 1464 (2004) (in Russian). 112. N.S. Dudkina, P.A. Shatunov, E.M. Kuvshinova, S.G. Pukhovskaya, A.S. Semeykin and O.A. Golubchikov, Zhurn. Obshch. Khim., 68 (12), 2042 (1998) (in Russian).


87

113. B. Evans, K.M. Smith and J.-H. Fuhrhop, Tetrahedron Lett., 5, 443 (1977). 114. K.M. Barkigia, M.D. Berber., J. Fajer, C.J. Medforth, M.W. Renner and K.M. Smith, J. Amer. Chem. Soc., 112 (24), 8851 (1990). 115. O. Finikova, A. Cheprakov, I. Beletskaya and S. Vinogradov, Chem. Commun., 3, 261 (2001). 116. S. Ito, T. Murashima, H. Uno and N. Ono, Chem. Commun., 1661 (1998). 117. N.G. Kuzmin, A.S. Semeykin and O.I. Koifman, USSR Author’s Certificate 1574603, Byul. Izobr., 24 (1990) (in Russian). 118. T.D. Lash and P. Chandrasekar, J. Am. Chem. Soc., 118, 8767 (1996). 119. C.J. Medforth, M.O. Senge, K.M. Smith, L.D. Sparks and J.A. Shelnutt, J. Amer. Chem. Soc., 114 (25), 9859 (1992). 120. J. Takeda and M. Sato, Chem. Pharm. Bull., 42, 1005 (1994). 121. T. Ema, M.O. Senge, N.Y. Nelson, H. Ogoshi and K.M. Smith, Angew. Chem. Int. Ed. Engl., 33 (18), 1879 (1994). 122. W. Jentzen, M.C. Simpson, J.D. Hobbs, X. Song, T. Ema, N.Y. Nelson, C.J. Medforth, K.M. Smith, M. Veyrat, M. Mazzanti, R. Ramasseul, J.-C. Marchon, T. Takeuchi, W.A. Goddard III and J.A. Shelnutt, J. Am. Chem. Soc., 117 (45), 11085 (1995). 123. M.O. Senge, L. Bischoff, N.Y. Nelson and K.M. Smith, J. Porphyrins Phthalocyanines, 3, 99 (1999). 124. J.A. Hodge, M.G. Hill and H.B. Gray, Inorg. Chem., 34 (4), 809–812 (1995). 125. T.D. Lash, K.A. Bladel, C.M. Shiner, D.L. Zajeski and R.P. Balasubramaniam, J. Org. Chem., 57 (18), 4809 (1992). 126. J.A. Anton and P.A. Loach, J. Heterocycl. Chem., 12 (3), 573 (1975). 127. R.G. Little, J.A. Anton, P.A. Loach and J.A. Ibers, J. Heterocycl. Chem., 12 (2), 343 (1975). 128. J.S. Lindsey, S. Prathapan, T.E. Johnson and R.W. Wagner, Tetrahedron, 50, 8941 (1994). 129. S. Noblat, O. Dietrich-Buchecker and J.-P. Sauvage, Tetrahedron Lett., 28 (47), 5829 (1987). 130. I. Tabushi, K.-I. Sakai and K. Yamamura, Tetrahedron Lett., 21, 1821 (1978). 131. I. Tabushi, S.-I. Kugimiya, M.G. Kinnard and T. Sasaki, J. Amer. Chem. Soc., 107 (14), 4192 (1985). 132. J.P. Collman, D.A. Tyvoll, L.L. Ching and H.T. Fish, J. Org. Chem., 60 (7), 1926 (1995). 133. J.P. Collman, H.T. Fish, P.S. Wagenknecht, D.A. Tyvoll, L.-L. Ching, T.A. Eberspacher, J.I. Brauman, J.W. Bacon and L.H. Pignolet, Inorg. Chem., 35, 6746 (1996). 134. K. Kohata, H. Higashio, Y. Yamaguchi, M. Koketsu and T. Odashima, Bull. Chem. Soc. Jpn., 67, 668 (1994). 135. H. Meier, Y. Kebuks and S.-I. Kugimiga, J. Chem. Soc. Chem. Commun., 14, 923 (1989). 136. I. Tabushi and T. Sasaki, Tetrahedron Lett., 23 (18), 1913 (1982). 137. D. Hammel, C. Kautz and K. Müllen, Chem. Ber., 123 (6), 1353 (1990). 138. Z. Gross and I. Toledano, J. Org. Chem., 59, 8312 (1994). 139. E. Rose, M. Soleithavoup, L. Christ-Tommasino and G. Moreau, J. Org. Chem., 63, 2042 (1998). 140. U. Rempel, B. von Maltzan and C. von Borczyskowski, Chem. Phys. Lett., 245, 253 (1995). 141. U. Orth, H.-P. Pfeiffer and E. Bretimaier, Chem. Ber., 119 (11), 3507–3514 (1986). 142. H. Kamogawa, T. Nakata and S. Komatsu, Bull. Chem. Soc. Jpn., 64, 2300 (1991). 143. G. Casiraghi, M. Cornia, F. Zanardi, G. Rassu, E. Ragg and R. Bortolini, J. Org. Chem., 59, 1801 (1994). 144. M. Cornia, S. Binacchi, T. Del Soldato, F. Zanardi and G. Casiraghi, J. Org. Chem., 60 (16), 4964–4965. 145. G. Casiraghi, M. Cornia, G. Rassu, C. Del Sante and P. Spanu, Tetrahedron, 48, 5619 (1992). 146. D. Hammel, P. Erk, B. Schuler, J. Heinze and K. Müllen, Adv. Mater., 4, 737 (1992). 147. C.-H. Lee and J.S. Lindsey, Tetrahedron, 50 (39), 11427 (1994). 148. T. Mizutani, T. Ema, T. Tomita, Y. Kuroda and H. Ogoshi, J. Am. Chem. Soc., 116 (10), 4240 (1994). 149. N. Nishino, R.W. Wagner and J.S. Lindsey, J. Org. Chem., 61 (21), 7534 (1996). 150. J.P. Nagarkatti and K.R. Ashley, Synthesis, 186 (1974). 151. G. Shipps (Jr.) and J. Rebek (Jr.), Tetrahedron Lett., 35, 6823 (1994). 152. B. Vaz, R. Alvarez, M. Nieto, A.I. Paniello and A.R. DeLera, Tetrahedron Lett., 42 (42), 7409 (2001). 153. S.J. Vigmond, K.M.R. Kallury and M. Thompson, Anal. Chem., 64, 2763 (1992).

88


154. T.P. Wijesekera, Can. J. Chem., 74, 1868 (1996). 155. C. Bruckner, E.D. Sternberg, R.W. Boyle and D. Dolphin, Chem. Commun., 1689 (1997). 156. I.M. Dixon and J.-P. Collin, J. Porphyrins Phthalocyanines, 5 (7), 600 (2001). 157. Y. Suga, T. Arimura, S. Ide, T. Nishioka, H. Sugihara, S. Murata and H. Tsuzuki, J. Chem. Res. Synop., 11, 512 (2000). 158. G.R. Geier (III), B.J. Littler and J.S. Lindsey, J. Chem. Soc. Perkin Trans., 2 (5), 701 (2001). 159. T. Akiyama, H. Imahori, A. Ajawakon and Y. Sakata, Chem. Lett., 907 (1996). 160. M. Ravikanth, J.-P. Strachan, F. Li and J.S. Lindsey, Tetrahedron, 54, 7721 (1998). 161. R.W. Wagner, T.E. Johnson and J.S. Lindsey, J. Amer. Chem. Soc., 118 (45), 11166 (1996). 162. G.A. Baker, F.V. Bright, M.R. Detty, S. Pandey, C.E. Stilts and H. Yao, J. Porphyrins Phthalocyanines, 4 (7), 669 (2000). 163. E.N. Durantini, J. Porphyrins Phthalocyanines, 4 (3), 233 (2000). 164. A.R. Genardy and D. Gabel, J. Porphyrins Phthalocyanines, 6 (6), 382 (2002). 165. J.-I. Setsune, M. Hashimoto, K. Shiozawa, J. Hayakawa, T. Ochi and R. Masuda, Tetrahedron, 54, 1407 (1998). 166. C.-H. Lee, F. Li, K. Iwamoto, J. Dadok, A.A. Bothner-By and J.S. Lindsey, Tetrahedron, 57, 11645 (1995). 167. D.M. Wallace, S.H. Leung, M.O. Senge and K.M. Smith, J. Org. Chem., 58, 7245 (1993). 168. D.M. Wallace and K.M. Smith, Tetrahedron Lett., 31, 7265 (1990). 169. M. Barbero, S. Cadamuro, L. Degani, R. Fochi, A. Gatti and V. Regondi, J. Org. Chem., 53, 2245 (1988). 170. M. Barbero, S. Cadamuro, I. Degani, R. Fochi, A. Gatti and V. Regondi, Synthesis, 1074 (1986). 171. P.-Y. Heo and C.-H. Lee, Bull. Kor. Chem. Soc., 17, 515 (1996). 172. C. Bruckner, J.J. Posakony, C.K. Johnson, R.W. Boyle, B.R. James and D. Dolphin, J. Porphyrins Phthalocyanines, 2 (6), 455 (1998). 173. R. Chong, P.S. Clezy, A.J. Liepa and A.W. Nichol, Austral. J. Chem., 22, 229 (1969). 174. S.J. Vigmond, M.C. Chang, K.M.R. Kallury and M. Thompson, Tetrahedron Lett., 35, 2455 (1994). 175. P.S. Clezy and G.A. Smithe, Austral. J. Chem., 22, 239 (1969). 176. J.A. Ballantine, A.H. Jackson, G.W. Kenner and G. McGillivray, Tetrahedron, Suppl. 7, 22, 241 (1966). 177. K.-T. Oh, J-W. Ka, J.-Y. Park and C.-H Lee, Bull. Kor. Chem. Soc., 18, 222 (1997). 178. Q.M. Wang and D.W. Bruce, Synlett., 1267 (1995). 179. J.E. Baldwin, T. Klose and M.K. Peters, J. Chem. Soc. Chem. Communs., 21, 881 (1976). 180. C.K. Chang and I. Abdalmuhdi, J. Org. Chem., 48 (26), 5388 (1983). 181. A.R. Battersby, E. Hunt, E. McDonald, J.B. Paine (III) and J. Saunders, J. Chem. Soc. Perkin Trans., 1, 1008 (1976). 182. A.R. Battersby, E. Hunt, M. Ihara, E. McDonald, J.B. Paine (III), F. Satoh and J. Saunders, J. Chem. Soc. Chem. Commun., 994 (1974). 183. A.S. Semeykin, S.A. Syrbu and T.V. Lyubimova, Zhurn. Obshch. Khim., 71 (10), 1747 (2001) (in Russian). 184. H. Fischer and H. Orth, Justus Liebigs Ann. Chem., 489, 62 (1931). 185. A.F. Mironov, T.R. Ovsepyan, R.P. Evstigneeva and N.A. Preobrazhenkii, Zh. Obshch. Khim., 35, 324 (1975) (in Russian). 186. A.W. Johnson, I.T. Kay, E. Markham, R. Price and K.B. Shaw, J. Chem. Soc., 3416 (1959). 187. M.B. Berezin, A.S. Semeykin and A.I. Vyugin, Zhurn. Fiz. Khim., 70 (8), 1364 (1996) (in Russian). 188. P.S. Clezy and A.J. Liepa, Austral. J. Chem., 23, 2443 (1970). 189. H. Ogoshi, H. Sugimoto, T. Nishiguchi, T. Watanabe, Y. Matsuda and Z.-I. Yoshida, Chem. Lett., 1, 29 (1978). 190. M.J. Gunter and L.N. Mander, J. Org. Chem., 46 (23), 4792 (1981). 191. Y. Aoyama, T. Kamohara, A. Yamagishi et al., Tetrahedron Lett., 28 (26), 2143 (1987). 192. M.J. Gunter and B.C. Robinson, Austral. J. Chem., 43 (11), 1839–1860 (1990). 193. G. Li, S. Wu and Y. Te, Youji Huaxue, 4, 300 (1985). 194. R. Young and C.K. Chang, J. Amer. Chem. Soc., 107 (4), 898 (1985). 195. A.S. Semeykin, T.V. Lyubimova and O.A. Golubchikov, Zhurn. Prikl. Khim., 66 (3), 710


89

(1993) (in Russian). 196. E.M. Kuvshinova, S.G. Pukhovskaya, A.S. Semeykin and O.A. Golubchikov, Zhurn. Obshch. Khim., 74 (10), 1733 (2004) (in Russian). 197. S.V. Zaitseva, S.A. Zdanovich, A.S. Semeykin and O.I. Koifman, Zhurn. Neorg. Khim., 50 (11), 1919 (2005) (in Russian). 198. A. Osuka, F. Kobayashi, T. Nagata and K. Maruyama, Chem. Lett., 2, 287 (1990). 199. S.L. Springs, A. Andrievsky, V. Kral and J.L. Sessler, J. Porphyrins Phthalocyanines, 2 (4–5), 315 (1998). 200. N.Zh. Mamardashvili, A.S. Semeykin, and O.A. Golubchikov, Zhurn. Org. Khim., 29 (12), 2445 (1993) (in Russian). 201. N.Zh. Mamardashvili, S.A. Zdanovich and O.A. Golubchikov, Zhurn. Org. Khim., 32 (6), 934 (1996) (in Russian). 202. A.K. Burrell, D.L. Officer and D.C.W. Reid, Angew. Chem. Int. Ed. Engl., 34, 900 (1995). 203. T. Nagata, A. Osuka and K. Maruyama, J. Am. Chem. Soc., 112 (8), 3054 (1990). 204. A. Osuka, N. Tanabe, R.-P. Zhang and K. Maruyama, Chem. Lett., 1505 (1993). 205. A. Osuka, S. Nakajima, T. Nagata et al., Angew. Chem., 103 (5), 579 (1991). 206. A. Osuka, S. Nakajima and K. Maruyama, J. Org. Chem., 57 (26), 7355 (1992). 207. J.L. Sessler, V.L. Capuano and A. Harriman, J. Am. Chem. Soc., 115 (11), 4618 (1993). 208. N. Ono, H. Kaziro and K. Maruyama, Bull. Chem. Soc. Jpn., 64 (11), 3471 (1991). 209. J.I. Bruce, J.-C. Chambron, P. Kolle and J.-P. Sauvage, J. Chem. Soc. Perkin Trans., 1 (10), 1226 (2002). 210. A. Osuka, B.-L. Liu and K. Maruyama, J. Org. Chem., 58, 3582 (1993). 211. J. L. Sessler, B. Wang and A. Harriman, J. Am. Chem. Soc., 117 (2), 704 (1995). 212. M.R. Wasielewski, M.P. Niemezyk, W.A. Svec and E.B. Pewitt, J. Am. Chem. Soc., 107 (19), 5562 (1985). 213. M.R. Wasielewski, G.L. Gaines (III), M.P. O’Neil, W.A. Svec and M.P. Niemezyk, J. Am. Chem. Soc., 112 (11), 4559 (1990). 214. M.R. Wasielewski, D.G. Johnson, M.P. Niemezyk, G.L. Gaines III, M.P. O’Neil and W.A. Svec, J. Am. Chem. Soc., 112, 6482 (1990). 215. A. Osuka, H. Yamada, K. Maruyama, N. Mataga, T. Asahi, M. Ohkouchi, T. Okada, I. Yamazaki and Y. Nishimura, J. Am. Chem. Soc., 115 (21), 9439 (1993). 216. M. Ohkohchi, A. Takahashi, N. Mataga, T. Okada, A. Osuka, H. Yamada and K. Maruyama, J. Am. Chem. Soc., 115 (25), 12137 (1993). 217. J.-C. Chambron, V. Heitz and J.-P. Sauvage, J. Chem. Soc. Chem. Commun., 1131 (1992). 218. J.-C. Chambron, V. Heitz, J.-P. Sauvage, J.L. Pierre and D. Zurita, Tetrahedron Lett., 36, 9321 (1995). 219. J.-P. Collin, A. Harriman, V. Heitz, F. Odobel and J.-P. Sauvage, Coord. Chem. Rev., 148, 63 (1996). 220. A. Osuka, S. Marumo, N. Mataga, S. Taniguchi, T. Okada, I. Yamazaki, Y. Nishimura, T. Ohno and K. Nozaki, J. Am. Chem. Soc., 118 (1), 155 (1996). 221. N.Zh. Mamardashvili, A.S. Semeykin, and O.A. Golubchikov, Zhurn. Org. Khim., 29 (6), 1213 (1993) (in Russian). 222. J.E. Baldwin, M.J. Crossley, T. Klose, E.A. O’Rear (III) and M.K. Peters, Tetrahedron, 38 (1), 27 (1982). 223. J.P. Collman, A.O. Chong, G.B. Jameson, R.T. Oakley, E. Rose, E.R. Schmittou and J.A. Ibers, J. Am. Chem. Soc., 103, 516 (1981). 224. A. Lecas, J. Levisalles, Z. Renko and E. Rose, Tetrahedron Lett., 25 (15), 1563 (1984). 225. K. Maruyama, T. Nagata and T. Osuka, J. Phys. Org. Chem., 1, 63 (1988). 226. M.O. Senge, C.J. Medforth, T.P. Forsyth, D.A. Lee, M.M. Olmstead, W. Jehtzen, R.K. Pandey, J.A. Shelnutt and K.M. Smith, Inorg. Chem., 36 (6), 1149 (1997). 227. J. Weiser and H.A. Staab, Tetrahedron Lett., 26 (49), 6059 (1985). 228. H. Tamiaki, A. Kiyomori and K. Maruyama, Bull. Chem. Soc. Jap., 67, 2478 (1994). 229. N.Zh. Mamardashvili, O.A. Golubchikov, G.M. Mamardashvili and W. Dehaen, J. Porph. Phthalocyan., 6 (7–8), 476 (2002). 230. I. Abdalmuhdi and C.K. Chahg, J. Org. Chem., 50 (3), 411 (1985). 231. J.P. Collman, J.E. Hutchison, M.A. Lopez et al., J. Amer. Chem. Soc., 114 (25), 9869 (1992). 232. R. Duibard, M.A. Lopes, A. Tabard et al., J. Amer. Chem. Soc., 114 (25), 9877 (1992).

90


233. D. Heiler, G. Mc Lendon and P. Rogalskyj, J. Amer. Chem. Soc., 109 (2), 604 (1987). 234. A. Osuka and K. Maruyama, Chem. Lett., 825 (1987). 235. A. Osuka and K. Maruyama, J. Amer. Chem. Soc., 110 (13), 4454 (1988). 236. A. Osuka, K. Maruyama, I. Yamazaki and N. Tamai, J. Chem. Soc. Chem. Commun., 18, 1243 (1988). 237. A. Osuka, K. Ida and K. Maruyama, Chem. Lett., 5, 741 (1989). 238. A. Osuka, H. Tomita and K. Maruyama, Chem. Lett., 7, 1205 (1988). 239. J.L. Sessler and S. Pierind, Tetrahedron Lett., 28 (52), 6569 (1987). 240. J.L. Sessler and M.R. Johnson, Angew. Chem., 99 (7), 679 (1987). 241. P.S. Clezy, C.J.R. Fookes and A.J. Liepa, Austral. J. Chem., 25, 1979 (1972). 242. H. Fischer, P. Halbig and B. Walach, Justus Liebigs Ann. Chem., 452, 268 (1927). 243. G.M. Trofimenko, A.S. Semeykin, M.B. Berezin and B.D. Berezin, Zhurn. Koord. Khim., 22 (6), 505 (1996) (in Russian). 244. D. Karris, A.W. Johnson and R. Caete-Holmes, Bioorg. Chem., 9, 63 (1980). 245. A.M. Shulga and G.P. Gurinovich, Dokl. Akad. Nauk BSSR, 25, 55 (1981) (in Russian). 246. D. Dolphin, A.W. Johnson, J. Zeng and P. van der Brock, J. Chem. Soc., 9, 880 (1966). 247. A.W. Johnson and I.T. Kay, J. Chem. Soc., 3, 1620 (1965). 248. J. Winkelman, Cancer Res., 22 (5), 589 (1962). 249. R.F. Pasternak, P.R. Huber, P. Boyd et al., J. Amer. Chem. Soc., 94 (13), 4511 (1972). 250. E.B. Fleisher, J.M. Palmer and A. Srivastava, J. Amer. Chem. Soc., 93 (13), 3162 (1971). 251. A. Srivastava and M. Tsutsui, J. Org. Chem., 38 (11), 2103 (1973). 252. C.A. Busby, R.K. Dinello and D. Dolphin, Can. J. Chem., 53 (11), 1554 (1975). 253. V.V. Bykova, N.V. Usoltseva, A.S. Semeykin et al., Zhurn. Org. Khim., 35 (4), 632 (1999) (in Russian). 254. V.S. Radnyuk, T.V. Lyubimova and A.S. Semeykin, Abstracts of the XIX All-Russian Chugaev Conf. on the Chemistry of Complex Compounds, Ivanovo, p. 171 (1999) (in Russian). 255. P. Hoffman, G. Labat, A. Robert and B. Meunier, Tetrahedron Lett., 31, 1991 (1990). 256. D. Mandon, P. Ochsenbein, J. Fisher et al., Inorg. Chem., 31, 2044 (1992). 257. A.A. Fernandes, C.M. Stinson and A. Shamim, Pakistan J. Sci. Ind. Res., 30 (9), 643 (1987). 258. X. Chen, F. Tang and C. Wang, Phys. Test. Chem. Anal., 26 (2), 70 (1990). 259. F. Tang, X. Chen and C. Wang, Chem. Reagents, 9 (1), 29 (1987). 260. S. Tong and G. Sun, Chem. Reagents, 14 (1), 7 (1992). 261. W.A. Lee, M. Gratzel and K. Kalyanasundaram, Chem. Phys. Lett., 107 (3), 308 (1984). 262. W.J. Kruper, T.A. Chamberlin (Jr.) and M. Kochanny, J. Org. Chem., 54 (11), 2753 (1989). 263. A.M. d’A. Rocha Gonsalves, R.A.W. Johnstone, M.M. Pereira et al., Heterocycles, 43, 829 (1996). 264. V.V. Morozov, A.S. Semeykin, B.G. Gnedin and B.D. Berezin, Khim. Geterotsykl. Soed., 6, 770 (1988) (in Russian). 265. K.J. Ciuffi, H.C. Sacco, J.B. Valim et al., Non-Cryst. Solids, 247, 146 (1999). 266. M.S.M. Moreira, E.A. Vidato, O.R. Nascimento and Y. Iamamoto, ICPP-2, Kyoto, Japan, O-7, p. 223 (2002). 267. L. Jaquinod, in: The Porphyrin Handbook, Acad. Press: New York, 1, Pt. 5, p. 202 (2000). 268. S.A. Syrbu and A.S. Semeykin, Abstracts of the I-st Int. Conf. on the Topical Problems of Chemistry and Chemical Engineering, Ivanovo, p. 81 (1997). 269. V.I. Melnik, PhD (Chemistry) Thesis, Odessa State University: Odessa, 21 pp. (1979) (in Russian). 270. A.S. Semeykin, O.I. Koifman, B.D. Berezin and S.A. Syrbu, Khim. Geterotsykl. Soed., 10, 1359 (1983) (in Russian). 271. L.R. Mildrom, J. Chem. Soc. Perkin Trans., 1 (10), 2535 (1983). 272. A.C. Chan, J. Dalton and L.R. Mildrom, J. Chem. Soc. Perkin Trans., 2 (6), 707 (1982). 273. J. Dalton and L.R. Mildrom, J. Chem. Soc. Chem. Commun., 14, 609 (1979). 274. E. Tsuchida, E. Hasegava, T. Komatsu et al., Chem. Lett., 3, 389 (1990). 275. S. Matile, T. Hansen, A. Storster and W.D. Wogjon, Helv. Chim Acta, 77 (4), 1087 (1994). 276. M. Momenteau, F. Le Bras and B. Looch, Tetrahedron Lett., 35 (20), 3289 (1994). 277. S.A. Syrbu, A.S. Semeykin and B.D. Berezin, USSR Author’s Certificate 1684284, Byul. Izobr., 38 (1991) (in Russian).


91

278. S.A. Syrbu and A.S. Semeykin, Zhurn. Org. Khim., 35 (8), 1262 (1999) (in Russian). 279. R.G. Little, J. Heterocycl. Chem., 15 (2), 203 (1978). 280. M. Momenteau and B. Loock, J. Mol. Catal., 7, 315 (1980). 281. A.S. Semeykin, O.I. Koifman, G.E. Nikitina and B.D. Berezin, Zhurn. Org. Khim., 54 (7), 1599 (1984) (in Russian). 282. B.D. Berezin, A.S. Semeykin, G.E. Nikitina et al., Zhurn. Fiz. Khim., 59 (9), 2226 (1985) (in Russian). 283. S.A. Syrbu, A.S. Semeykin, B.D. Berezin and O.I. Koifman, Khim. Geterotsykl. Soed., 10, 1373 (1989) (in Russian). 284. Methods and Advances of Bioinorganic Chemistry, ed. by K. MacOliff, Mir: Moscow, 416 pp. 1978 (Russian translation). 285. L. Ding, G. Etemad-Moghadam, S. Cros et al., J. Med. Chem., 34 (3), 900 (1991). 286. P. Kus, G. Knerr and L. Gzuchajowski, Tetrahedron Lett., 31, 5133 (1990). 287. X. Jiang, P.K. Pandey and K.M. Smith, J. Chem. Soc. Perkin. Trans., 1, 1607 (1996). 288. S.A. Syrbu, A.S. Semeykin, B.D. Berezin and O.I. Koifman, Khim. Geterotsykl. Soed., 8, 781 (1987) (in Russian). 289. J.P. Collman, R.R.Gagne, T.R. Halbert et al., J. Amer. Chem. Soc., 95 (23), 7868 (1973). 290. A.S. Semeykin, O.I. Koifman and B.D. Berezin, Khim. Geterotsykl. Soed., 10, 1354 (1982) (in Russian). 291. A.S. Semeykin, O.I. Koifman and B.D. Berezin, Izv. Vuz. Khim. Khim. Tekhnol., 28 (11), 47 (1985) (in Russian). 292. F. Tang, L. Wang and Z. Chai, Chem. Reagents, 15 (6), 324 (1993). 293. X. Wu, Z. Chen and Z. Ziang, J. Wunan Univ. Natur. Sci. Ed., 4, 30 (1993). 294. S.E. Gribkova, V.N. Luzgina and R.P. Evstigneeva, Zhurn. Org. Khim., 29 (4), 758 (1993) (in Russian). 295. A. Palka and L. Czuchajowski, Chem. Lett., 3, 547 (1994). 296. A.S. Semeykin, O.I. Koifman and B.D. Berezin, Khim. Geterotsykl. Soed., 4, 486 (1986) (in Russian). 297. A.S. Semeykin, O.I. Koifman and B.D. Berezin, Izv. Vuz. Khim. Khim. Tekhnol., 24 (5), 566 (1981) (in Russian). 298. S.A. Syrbu, A.S. Semeykin and B.D. Berezin, Khim. Geterotsykl. Soed., 11, 1507 (1990) (in Russian). 299. S.A. Syrbu, A.S. Semeykin and T.V. Syrbu, Khim. Geterotsykl. Soed., 5, 668 (1996) (in Russian).

92


- ссылки 7 и 14 одинаковые - упоминается табл. 6 (a таблиц всего 5)

3

The Mechanism of Catalytic Action of the Coordination Centres of Catalase Synthetic Models T.N. Lomova1, M.E. Klyueva2 and M.V. Klyuev3 1Institute

of Solution Chemistry, Russian Academy of Sciences, 1 Akademicheskaya Street, Ivanovo, 153045, Russia; email: [email protected] 2Ivanovo State University of Chemistry and Technology, 7 F. Engels Prospect, Ivanovo, 153000, Russia; email: [email protected] 3Ivanovo State University, 39 Ermak Street, Ivanovo, 153025, Russia; email: [email protected]

This chapter presents the results of studying the catalytic activity of manganese(III) and copper(II) porphyrins alkyl- and phenyl-substituted at β- and meso-macrocycle positions in hydrogen peroxide decomposition in DMFA–KOH–H2O. The ion-molecular mechanism of the process with kinetically significant stages of twoelectron metalloporphyrin oxidation and further partial reduction as well as acidbase peroxide equilibrium reactions is stated. Effective catalysis of hydrogen peroxide decomposition is determined by the degree of porphyrin binding, mixed coordination sphere structure and mutual influence of the ligands. The compounds studied are shown to act similar to naturally occurring catalases.

Keywords: porphyrins, complexes, hydrogen peroxide, disproportionation reaction, catalysis, mechanism, catalase models

Introduction Studies of chelate complexes of copper as catalysts of the hydrogen peroxide decomposition reaction [1] showed the common factors for individual reaction mechanisms to be

94

T.N. Lomova, M.E. Klyueva and M.V. Klyuev

coordination of peroxide particles on the metal atom and availability of two free coordination sites for the catalytic action of copper(II) compound to be realized. Copper(II) porphyrins (CuP), as polychelate macrocyclic complexes, are of interest in both aspects. First, they contain a coordinatively unsaturated copper atom of the 3d 9 electron configuration in the environment of the cyclic aromatic ligand; this copper atom forms direct N→Cu dative π-bonds along with coordination σ-interaction. A similar π-interaction is also possible in the case of addition of axial ligands containing unshared electron pairs. Second, CuP satisfy the requirement of the two-side access of reagents to the coordination core. Prospectivity of using metalloporphyrins with a two-charge metal cation (MP) for catalyzing the disproportionation reaction of hydrogen peroxide is evident at present [2, 3]. However, the required coordinative unsaturation of metal in metalloporphyrins is the greater, the higher the formal charge of the metal atom is [4, 5]. For this reason, studies of acidoporphyrin metal complexes of the oxidation degree greater than two (X)n–2MP is also rather promising. A suitable model in the given case are manganese(III) porphyrins. Thus, in [3], monomeric and covalently bound dimeric manganese porphyrinates were found to exhibit a catalase activity in the presence of the basic nitrogen of imidazole under conditions of phase-transfer catalysis. However, no rigorous theory has been developed to predict the reactivity of new coordination compounds, porphyrin complexes of copper and manganese including, and to explain the differences in their catalytic activity with respect to hydrogen peroxide decomposition. Several attempts have only been made to study the effect of the ligand environment of the central ion on its catalytic activity and the H2O2 degradation mechanism [6, 7]. In this situation, of special significance is to obtain new information on the mechanism of the catalytic action of complex compounds with regulated coordination saturation of the central atom. The latter is realized in the case of metalloporphyrins by targeted functional substitution of porphin in the complexes and rational choice of acidoligands X – in mixed (X)n–2MP complexes. The results obtained by the authors of this chapter in studies of copper(II) porphyrin complexes with various degrees of substitution by β- and meso-positions of the macrocycle (formulas 1–4) as catalysts of the hydrogen peroxide decomposition reaction in the DMFA–KOH–H2O system (reaction 1) are considered in [8, 9]. This chapter presents the major regularities and conclusions pertaining to the catalytic properties of copper porphyrins with respect to reaction (1). For manganese complexes, new data are presented on the catalysis of the disproportionation reaction (1) of manganese(III) porphyrins with a regularly changing structure (formulas 5–9) [10]. The catalytic reactions involving manganese(III) tetraphenylporphyrins 10 and 11 were studied earlier [8, 9]. 2H2O2 = 2H2O + O2↑ .

(1)

Octaethylporphin H2OEP and meso-tetraphenylporphin H2TPP were synthesized by the known methods [11, 12]. meso-Phenyl-substituted octaethylporphyrins monophenyloctaethylporphin H2MPOEP, 5,10-diphenyloctaethylporphin H25,10DPOEP, 5,15-diphenyloctaethylporphin H25,15DPOEP and 5,10,15,20-tetraphenyloctaethylporphin H2TPOEP were synthesized by A.S. Semeykin (Ivanovo State University of Chemistry and Technology) [12]. Complexes of porphyrins with manganese(III) were obtained by the reaction of respective porphyrin with metal salt according to Adler’s method [13]. As complex formers, we used MnCl2 ·4H2O (AR grade), Mn(AcO)2 ·4H2O (AR grade). (SCN)MnOEP 7 was

The Mechanism of Catalytic Action of the Coordination Centres of Catalase Synthetic Models R1

R

R

R

R N

N Cu

R4 N

R2

95

CuOEP, R1 = R2 = R3 = R4 = H, R = C2H5 Cu5,10DPOEP, R1 = R2 = C6H5, R3 = R4 = H, R = C2H5 CuTPOEP, R1 - R4 = C6H5, R = C2H5 CuTPP, R1 = R2 = R3 = R4 = C6H5, R = H

(1) (2) (3) (4)

(Cl)MnOEP, R1 = C2H5, R2 = R3 = R4 = R5 = H, X = Cl (AcO)MnOEP, R1 = C2H5, R2 = R3 = R4 = R5 = H, X = AcO (SCN)MnOEP, R1 = C2H5, R2 = R3 = R4 = R5 = H, X = SCN (Cl)MnMPOEP, R1 = C2H5, R2 = C6H5, R3 = R4 = R5 = H, X = Cl (Cl)Mn5,15DPOEP, R1 = C2H5, R2 = R4 = C6H5, R3 = R5 = H, X = Cl (Cl)MnTPP, R1 = H, R2 = R3 = R4 = R5 = C6H5 (AcO)MnTPP, R1 = H, R2 = R3 = R4 = R5 = C6H5

(5) (6) (7) (8) (9) (10) (11)

N

R

R R

R

R3

X

R1

R1

R5 R1 N

R1 R1

R1

N

N

R4

R2

Mn N

R1

R3

R1

obtained by treating a solution of (AcO)MnOEP 6 in DMFA with an excess aqueous solution of NaSCN.

1

Substituted Copper(II) Porphyrins as Catalysts of the Hydrogen Peroxide Disproportionation Reaction

To carry out reaction (1), an aqueous solution of hydrogen peroxide was added to a prepared solution of the copper complex and KOH in DMFA. The concentrations of the complexes, KOH and H2O2 were varied within the limits of 10 –6 –10 –4, (0.18–3.54)·10 –2, 2.98–5.97 mol/l, respectively. The rate of reaction (1) was determined volumetrically by measuring the volume of evolving oxygen. For copper porphyrin complexes studied, the catalysis of H2O2 decomposition is observed not within all catalyst concentration ranges used. Thus, for CuDPOEP, as for CuCl2 and Cu(AcO)2, the catalytic effect of the rate increase (W) as compared with the noncatalyzed reaction is observed only at concentrations greater than 1·10 –5 mol/l and increases at an increase of CCuP. For reactions catalyzed by CuOEP and CuTPOEP the observed order of the catalyst concentration is close to zero: W is practically independent of CCuP. Among metalloporphyrins studied, significant catalytic activities in the H2O2 decomposition reactions are manifested by CuTPP and CuDPOEP. Their presence in the reaction mixture increases the reaction rate and decreases the activation energy three- to fourfold as compared with the noncatalytic process. The activity of CuOEP and CuTPOEP complexes is comparable with copper salts. The complete kinetic equations (2) and (3) for complexes 2 and 4, CuCl2, Cu(AcO)2 and for complexes 1 and 3, respectively, were determined:

dCO2 / dτ = k ⋅ CH2O2 ⋅ CKOH ⋅ CCuP ⋅ (CO2 )0 = k ⋅ CH2O2 ⋅ CKOH ⋅ CCuP ,

(2)

dCO2 / dτ = k ′ ⋅ CH2O2 ⋅ CKOH ⋅ (CCuP )0 ⋅ (CO2 )0 = k ′ ⋅ CH2O2 ⋅ CKOH .

(3)

Using the UV-visible spectroscopy method, π-cation-radical forms of metalloporphyrins 1 and 3 are identified as end products of the reaction with H2O2 in the presence of KOH.

96


fast

fast slow

H2O + O2 Scheme 1 A scheme of elementary reactions in the copper(II) porphyrin–hydrogen peroxide–KOH system.

For the reaction catalyzed by Cu5,10DPOEP (compound 2), the spectral characteristics in the course of conversion are similar to complexes 1 and 3. In reaction mixtures, where CuTPP is present as a catalyst, attempts to record the electronic absorption spectra were unsuccessful due to the vigorous evolution of oxygen. However, after CuTPP was isolated from the reaction mixture to CHCl3 the electronic absorption spectra reflected the absorpI I tion of initial CuTPP (λmax = 538.5 nm) and its π-cation-radical form (λmax = 700 nm and II λmax = 664 nm). According to the spectral data, CuTPP and Cu5,10DPOEP (complexes 2 and 4) in the absence of alkali do not enter into a chemical interaction with H2O2, unlike 1 and 3, as well as in contrast with similar manganese(III) complexes considered below. If the reaction with hydrogen peroxide is assumed to run to form axial complexes with the catalyst (Scheme 1), as is the case in the functioning of natural catalase [1, p. 466], which is in complete agreement with the experimental kinetic equations (2, 3), the above difference in reactivity can be explained by different stabilities of the above axial complexes. The catalytic activity of metalloporphyrins in the H2O2 decomposition reaction, the rate of which, according to Scheme 1, is W = k4 ·K, is explained by the ability of copper cation in the complex to coordinate ligands additionally to the fifth coordination site and by low ionization potentials of the coordinated porphyrin macrocycle. Inertness of complexes 2 and 4 in the reaction with H2O2 in the absence of KOH is due to their poor capability of axial coordination: a weak holding of OH· ligands in (·OH)CuP+· eliminated in the course of reduction at the slow stage (Scheme 1) leads to a sharp decrease of the activation energy in the case of CuTPP (27±3 kJ/mol) and CuDPOEP (17±2 kJ/mol) as compared with other CuP, for which E changes within the limits of (39±3)–(73±7) kJ/mol [9].

2

Kinetic Regularities and Mechanisms of Peroxide Decomposition Reactions in the Presence of Acido Complexes of Highly Substituted Manganese Porphyrins

As it was already noted, considering the role of axial coordination processes, one expects an increased catalytic activity in passing to manganese complexes with porphyrins, where

The Mechanism of Catalytic Action of the Coordination Centres of Catalase Synthetic Models

97

the efficient charge of the central atom is not completely compensated for in coordination and strong binding of the aromatic macrocyclic ligand. After elucidating the potential of manganese complex compounds as catalysts of redox processes and understanding the role of these compounds in medicine, agriculture and other fields, interest of researchers in them permanently rises. Consideration of the role of manganese in enzyme as well as model systems simulating the functional properties of various enzymes [14] showed the significance of compounds of this element in vital processes. Of especial importance are complexes of manganese with porphyrins, as in their presence many reactions proceed under milder conditions and in media, where simpler derivatives of manganese are inert or nonspecific [14]. Using complexes of manganese with various derivatives of porphin and various coordination centres, varying solvents and temperature, it is possible to achieve selective processes of hydrogen peroxide decomposition, thus, modelling the action of catalases. The most significant results in this direction were obtained in [15–20]. In the course of hydrogen peroxide decomposition catalyzed by metalloporphyrins and metallophthalocyanines, the catalyst is observed to be partially (up to 30–50%) degraded or its activity to be inhibited by an excess of bases added into the system, such as imidazole [3, 21]. The data of the previous section suggest redox processes of the catalysis in the course of the H2O2 decomposition reaction. For this reason, studies of the reactivity of porphyrin catalysts at the action of hydrogen peroxide on them are topical. In [22, 23] published well back in 1960s, metallophthalocyanines have been shown to be subjected to oxidative destruction under the action of H2O2 in an acidic medium. Pheophytin and its complexes with metals are also oxidized but much slower. Since that time, conversion processes of metalloporphyrins directly at the action of peroxide have not been studied systematically. In studies of the kinetics of H2O2 decomposition in the presence of manganese(III) tetra(2,6-dimethyl-3-sulfonatophenyl)porphin at pH 7.6–12.1, a conclusion was made of the equilibrium coordination of the hydrogen peroxide molecule by the catalyst and the subsequent slow formation of the π-cation-radical form of oxomanganese(IV) porphyrin or oxomanganese(V) porphyrin [15]. Formation of oxo complexes (O)MnIVP, [(O)MnIVP]+·, (O)MnVP, readily passing into one another, is also noted when using manganese porphyrin–peroxide systems for oxidation and epoxidation of aliphatic and aromatic compounds [16–20]. Thus, interaction of peroxyacetic acid with tetra(2,6-dichloro-4-R-phenyl)porphyrin (R = CH3O, H, Br, Cl, NO2) complexes of manganese(III) in an acetonitrile–acetic acid mixture leads to the formation of an intermediate “catalyst–oxidant” complex [18]. In catalytic reactions of oxidation of cis-stilbene and naphthalene, one observes an irreversible conversion of an intermediate complex [(HOAc)MnRTDCPP](X) formed in the substitution of anion X by a molecule of HOAc in the inner coordination sphere, into a mixture of two high-valent oxomanganese complexes. The complexes are assumed to have the composition (X)(O)MnVRTDCPP and [(O)MnIVRTDCPP]+· (X) [19]. The first complex is responsible for epoxidation of cis-stilbene; the second, for hydroxylation of naphthalene. In the medium of CH3CN, formation of a stable complex is registered by the appearance of an absorption at 415 nm in the reaction of (Cl)MnTDCPP with peroxyacetic acid, whereas in a medium of CH2Cl2 such a complex proves to be short-lived and rapidly breaks down to destroy the macrocycle [20]. Conclusions on the conversions of catalysts in the cited works were mainly made based on the analysis of the conversions of reagents or special compounds – traps – as in [15], while the state of the catalyst itself – metalloporphyrin – was not studied. In this review, we analyze the results of studies of the direct reaction of manganese porphyrins with the oxidant H2O2. Using the spectrophotometric and kinetic methods, we studied the reac-

98


tions of (chloro)manganese(III) octaethylporphin (Cl)MnOEP (compound 5) with hydrogen peroxide in water–DMFA media at a temperature of 288–308 K. 2.1

Kinetics and Reaction Mechanism of Oxidation of Manganese(III) Porphyrins by Hydrogen Peroxide

The kinetics of the reactions of mixed acidotetraphenylporphin complexes of manganese(III) 10 and 11 with hydrogen peroxide in DMFA was studied earlier [8, 9, 24]. The crucial role in the coordination of H2O2 by the porphyrin complex (equation 4), from which the redox process begins, was shown to belong to the coordination unsaturation of the manganese atom, which is mainly regulated by the state of the links with the coordinated macrocycle. The structure of the macrocycle and extent of its aromaticity can be considered as a means of controlling the oxidation process. Data accumulation on the porphyrin structure–reactivity relationship with respect to the oxidation of hydrogen peroxide is at the very starting point.

(Х)(H2O2)MnTPP .

(Х)MnTPP + H2O2

(4)

According to the spectrophotometry data [8, 9], the end products of the reaction of complexes 10 and 11 with hydrogen peroxide are not the same depending on the peroxide concentration ranges, which are 0.017–0.5, 0.67–1.0 and 1.26–3.98 mol/l. The products are identified as an oxidized form with the localization of electron deficit at the aromatic macrocycle – the π-cation-radical of manganese(III) tetraphenylporphin in the first interval of H2O2 concentrations and the form oxidized at Mn atom – manganese(IV) tetraphenylporphin – at higher concentrations of peroxide. In all cases, kinetic measurements revealed the first order by initial metalloporphyrin. Experimentally found complete kinetic equations for the H2O2 concentration range of 0.017–0.1 and 1.26–3.32 mol/l are written down as equations (5) and (6): − dC(Cl)Mn III TPP / dτ = k v1 ⋅ C(Cl)Mn III TPP ⋅ CH 2O2 ,

(5)

−1/ 2 . − dC(Cl)Mn III TPP / dτ = k v3 ⋅ C(Cl)Mn IIITPP ⋅ CH O

(6)

2

2

Here kv1 and kv3 are the true rate constants of the reaction with hydrogen peroxide respectively for the first and third concentration ranges of peroxide. For the overall reaction described by kinetic equation (5), the following sequence of elementary stages is proposed, with account for which the rate constant in equation (5) is equal to kv1 = k2 ·K1: (Cl)MnIIITPP + H2O2

k

K1

(Cl)(H2O2)MnIIITPP ,

.

2 (Cl)(H2O2)MnIIITPP ⎯⎯→ О=MnIII(TPP+ ?) + Н2О + Cl- .

(7) (8)

The concentration constant of equilibrium (7) at 298 K, found by the method of

99


log I b

a log CH2O2

CH2O2, mol/l Figure 1 A titration curve (a) and a dependence of log I on log CH2O2 (b) (R2 = 0.98) at the titration of the complex of (Cl)MnIIITPP in DMFA by hydrogen peroxide.

spectrophotometric titration in this work (Fig. 1) and calculated by formula (9), is K1 = (20±3) l·mol –1. Hence, the rate constant k2 of an elementary reaction (8) at 298 K is equal to 1.95·10 –3 s –1. K =

(Ap − A0 )/(A∞ − A0 ) 1 , ⋅ 0 1 − (Ap − A0 )/( A∞ − A0 ) ( С H O − C (Cl)M nTPP ⋅ (Ap − A0 )/( A∞ − A0 )) 2 2

(9)

where A0, A∞, Ap are the optical densities of the solutions of initial (Cl)MnTPP, of the complex (Cl)(H2O2)MnIIITPP and of an equilibrium mixture at the working wavelength of 468 nm. The scheme of the reactions conforming to kinetic equation (6) is more complex due to the appearance of perhydroxyl HO2– in kinetically significant amounts at the increase of the concentration of hydrogen peroxide: K1

H2O2

−

H+ + HO 2 ,

k

−

(10)

−

2 (Cl)MnIIITPP + HO 2 ⎯⎯ → (Cl)(HO 2 )MnIIITPP fast ,

K3

−

(Cl)(HO 2 )MnIIITPP

+•

−

+•

(TPP+·) ) + ОН- , (Сl)O=MnIV(TPP

k

4 (TPP+·) ) + HO 2 ⎯⎯→ (Сl)O=MnIV(TPP O=MnIVTPP + HCl + O2 slow ,

ОН- + H2O2

K5

−

HO 2 + H2O .

(11)

(12)

(13)

(14)

It has been experimentally found that, due to the low concentration of manganese porphyrin in the form of (Cl)O=MnIV(TPP+·), there is no gas evolution in the experiment, and the O2 formed is, evidently, in solution. The rate constant in equation (6) includes the equilibrium constants (10), (12) and (14):

100


k3 = k4 ⋅ K3 ⋅ K5 ⋅ K1−1/ 2 .

(15)

The numerical value of K1 is known: 2.4·10 –2 at 298 K [25]. The values of kv1 and kv3, found by optimizing the dependences of the efficient rate constant of the reaction of (Cl)MnIIITPP with H2O2 on the concentration of the latter, are given in Table 1 [8]. Table 1 True rate constants, E and ∆S≠, of the reaction of (Cl)MnTPP with H2O2. CH2O2 range, mol/l

T, K

kv ·102 (a), s–1 ·mol –1 ·l

E, kJ/mol

∆S≠, J/(mol·K)

(Cl)MnTPP 0.017–0.10

1.26–3.32

288 298 308 288 298 308

1.41 3.90 9.45 0.72 1.34 2.51

70±1

–44±3

46±1

–133

54±1

–87±3

(AcO)MnTPP 0.013–0.052

(a)k

v1

288 298 308

7 15 31

and kv 3 for the CH2O2 ranges of 0.017–0.10 and 1.26–3.32, respectively.

The overall reaction of (Cl)MnTPP with H2O2 at all concentrations of the latter can be presented as Scheme 2. At low concentrations of H2O2 the reaction is limited by the stage of two-electron oxidation of manganese porphyrin and H2O2 reduction on metal in (Cl)(H2O2)MnTPP (k2′); at high concentrations, by the stage of two-electron reduction of π-cation radical (Cl)O=MnIV(TPP+·) up to O=MnIVTPP and Cl – and oxidation of HO–2 (k3′) with evolution of O2 and H+.

HO

Scheme 2 An overall scheme of the reaction of (Cl)MnTPP with H2O2.


101

Under conditions of H2O2 excess, the complex O=MnIVTPP oxidized by the central atom of manganese is stable in solution in time. It proved that the acetate complex (AcO)MnIIITPP (compound 11) reacts with H2O2 much faster than the chloride complex (Cl)MnIIITPP (Table 3). Already at CH2O2 equal to 1.0 mol/l the reaction is instantaneous, the reaction product in this case being O=MnIVTPP. The order for [H2O2] for this reaction at low CH2O2 (0.013–0.052 mol/l) is close to unity (0.72), which indicates the common character of its mechanism with that for (Cl)MnIIITPP (Scheme 2, route K1′ → k2′). The rise of reactivity in the case of (AcO)MnIIITPP is explainable by a stronger binding of acidoligand AcO – as compared with Cl – . In the medium of DMFA–H2O, where the reaction is run, in the initial complex at the coordination of H2O2 there occurs the substitution of the DMFA molecule coordinated to the sixth coordination site. Owing to the effect of trans-influence of AcO – , substitution of DMFA is easier in the (AcO)MnIIITPP complex than in (Cl)MnIIITPP. Besides, owing to the possibility of transferring the electronic effect of acidoligand via the transition metal with a partially filled d-shell to the macrocycle, the macrocycle donates the electron easier to form π-cation radical in (AcO)MnIIITPP than in (Cl)MnIIITPP. Manganese(III) porphyrins studied possess a catalase activity in an alkaline medium. The results presented above show the ease of forming oxidized forms of complexes and the absence of destruction of the macrocycle in the interaction with H2O2. This makes manganese(III) porphyrins rather promising for their use as models of natural catalases. New data on the spectral and kinetic studies of the reaction of (chloro)manganese(III) octaethylporphin (Cl)MnOEP with hydrogen peroxide in water–DMFA media show that what was said above regarding catalase models also pertains to the complex of manganese(III) with another porphin derivative octaethylporphin. The kinetics of the reaction of (Cl)MnOEP with H2O2 in an H2O–DMFA medium was studied using a spectrophotometric method within the temperature range of 288–308 K. The optical density of the solutions in the course of the reaction was measured at a wavelength of 458 nm. The initial concentration of H2O2 in water was 17.4 ± 0.1 mol/l. Interaction of (Cl)MnOEP with H2O2 at temperatures close to room temperature is accompanied with the change of its electronic absorption spectra. Herewith, the isobestic points at 394, 446, 478, 523 and 582 nm are preserved (Fig. 2), which indicates the mutual conversion of two stable stained compounds, one of which is the initial (Cl)MnOEP, the form of existence of which differs at different concentrations of peroxide, as is seen from the further discussion. The intensity of the absorption bands of initial (Cl)MnOEP at 367, 458 and 541 nm decreases, and at 434, 517 and 726 nm goes up. The end product was identified using literature data [26–28]. The spectrum of the end product corresponds to π-cation radical of manganese(III) octaethylporphin (Fig. 2). The pattern of spectral changes presented in Fig. 2 is characteristic of the entire range of H2O2 concentrations studied, which distinguishes the octaethylporphin complex from above-considered (Cl)MnTPP, for which the similar reaction was observed to yield various end products at different CH2O2. The first order of the reaction of (Cl)MnOEP with H2O2 with respect to metalloporphyrin is demonstrated by the data of Fig. 3. The efficient rate constants of the reaction are given in Table 2. The order of the reaction with respect to [H2O2] is variable. Experimentally, we found three ranges of peroxide concentrations: 0.01–0.07, 0.2–0.52 and 1.00–2.22 mol/l, each of which was observed to have its own order for hydrogen peroxide (n) (Figs. 4 and 5): by the T – log CH2O2, n is, respectively, 1/2, 0 and –1/2. slope of the lines in the coordinates log kef The zero order is evident due to the absence of the dependence of kef on CH2O2 (Table 2).

102


λ, nm

Figure 2 Electronic absorption spectra of (Cl)MnOEP in the course of the reaction with H2O2. 298 K, CH2O2 = 1.0 mol/l. ln

τ, s

Figure 3 A dependence of ln(C0 /Cτ) – τ for the reaction of (Cl)MnOEP with H2O2. C(Cl)MnIIIOEP = 2·10 –5 mol/l. CH2O2 = 0.33 (1), 0.49 (2), 0.05 (3), 1.35 (4), 0.03 (5), 0.01 (6) mol/l. T, K: 308 (1), 303 (2), 298 (3), 293 (4, 5, 6). R2 = 0.99.

The considered results make it possible to write down the experimentally found complete kinetic equations for the first, second and third reacting sequences, respectively: , −dC(Cl)Mn III OEP / dτ = kv1 ⋅ C(Cl)Mn IIIOEP ⋅ C1/2 Н О

(16)

−dC(Cl)Mn III OEP / dτ = k v2 ⋅ C(Cl)Mn III OEP ,

(17)

2

2

−1/2 . − dC(Cl)Mn IIIOEP / dτ = k v3 ⋅ C(Cl)Mn IIIOEP ⋅ CН О 2

2

(18)


103

Table 2 Efficient kinetic parameters of the reaction of (Cl)MnOEP with H2O2 in DMFA. kef ·104, s –1

CH2O2, mol/l 0.01 0.03 0.05 0.07 0.20 0.33 0.49 0.52 1.00 1.35 1.72 2.22

293 K

298 K

303 K

308 K

2.9±0.2 5.0±0.1 6.1±0.2 8.6±0.5 10.2±0.4 10.3±0.6 9.8±0.5 10.0±0.7 9.5±0.2 8.5±0.3 7.3±0.2 6.1±0.3

4.6±0.1 8.1±0.3 10.0± 0.4 12.2±0.4 16.4±0.7 16.4±0.6 15.8±0.4 15.9±0.8 13.1±0.5 10.9±0.3 9.9±0.2 8.1±0.2

7.2±0.3 12.3±0.2 14.6±0.5 18.8±0.8 25.0±1.0 24.0±1.3 20.7±0.6 22.5±1.1 17.8±0.5 16.3±0.6 14.9±0.5 11.4±0.5

11.0±0.4 17.8±0.3 22.9±0.4 25.7±1.0 32.6±0.9 31.3±0.8 31.7±0.6 30.3±1.0 25.8±0.5 22.3±0.5 19.7±0.6 16.6±0.5

E′, kJ/mol

∆S≠ ′, J/(mol·K)

67±1 63±2 65±2 56±2 59±4 56±4 57±4 55±3 50±2 49±3 51±2 50±2

–91±3 –100±7 –91±7 –120±7 –107±13 –118±13 –115±13 –121±10 –140±7 –144±10 –138±7 –143±7

log kTe

-log CH2O2 T on –log C Figure 4 A dependence of log kef H2O2 for the reaction of (Cl)MnOEP with H2O2 within the H2O2 concentration range of 0.01–0.07 mol/l. T, K: 308 (1), 303 (2), 298 (3), 293 (4). R2 = 0.98.

log kTe

log CH2O2 Figure 5 A dependence of log kefT on log CH2O2 for the reaction of (Cl)MnOEP with H2O2 within the H2O2 concentration range of 1.00–2.22 mol/l. T, K: 308 (1), 303 (2), 298 (3), 293 (4). R2 = 0.98 (1, 3, 4), 0.92 (2).

104


The numerical values of the true rate constants kv1 and kv3, found by optimizing the dependences presented in Figs. 4 and 5, and the reaction activation parameters are given in Table 3. Table 3 True rate constants, activation energy and entropy of the reaction of (Cl)MnOEP with H2O2 in DMFA. CH2O2 range, mol/l

T, K

kv ·103, s–1 ·mol –1 ·l

0.01–0.07

293 298 303 308 293 298 303 308

3.0±0.1 4.6± 0.1 7.2±0.1 10.4±0.4 0.95±0.02 1.27±0.02 1.8±0.1 2.55±0.02

1.00–2.22

E, kJ/mol

∆S≠, J/(mol·K)

63±1

–85±3

50±2

–140±6

Considering the spectral data (Fig. 2) and taking into account equation (16), the following scheme of elementary stages can be proposed for the first range of H2O2 concentrations. The scheme is supported by comparing experimental equation (16) with the theoretical equation deduced for the system of equations (19–21). K1

H2O2 −

(Cl)MnIIIOEP + НО 2 −

(Cl)(НО 2 )MnIIIOEP

−

Н+ + НО 2 , K2

(19) −

(Cl)(НО 2 )MnIIIOEP ,

k3 . ⎯⎯→ O=MnIII(OEP + • ) + ОН- + Cl- slow .

(20)

(21)

Perhydroxyl anion present in the system at equilibrium concentrations in accordance with the equilibrium (19) is added reversibly (reaction (20)) to the coordinatively unsaturated Mn atom, thus forming the donor–acceptor bonds –O – → Mn. Hydrogen peroxide is coordinated by manganese(III) octaethylporphin not in the form of the H2O2 molecule as in the case of (X)MnTPP complexes, but as a stronger nucleophile HO–2 . The cause is, apparently, a decrease of the positive charge δ+ at the Mn atom in transition from the tetraphenylporphin complex to the octaethylporphin complex due to the electron–donor action of eight β-alkyl substituents and the elimination of the electron–acceptor effect of the mesophenyl groups. The perhydroxyl ion activated owing to the coordination at Mn donates the oxygen atom to metalloporphyrin with abstraction of OH – in the slow reaction (21) of two-electron oxidoreduction. (Cl)MnOEP passes into the oxidized form O=MnIII(OEP+·), which is well identifiable by the electronic absorption spectra (Fig. 2). Due to the low concentration of HO–2 , no interaction of the oxidized form of the complex with the second molecule of peroxide (as is the case in all catalytic processes of H2O2 disproportionation) is observed. Coordination of HO–2 (reaction (20)) evidently occurs via the sixth coordination site;


105

CH2O2, mol/l Figure 6

A curve of the titration of the complex (Cl)MnIIIOEP by hydrogen peroxide in DMFA.

acidoligand Cl remains to stay in the coordination sphere. This is confirmed by the example of (X)MnTPP, where the rate of the process was found to strongly depend on the nature of X (see above). The rate equation for the limiting stage has the form:

− dC(Cl)(HO − )Mn IIIОЕP / dτ = k3 ⋅ C(Cl)(HO− )Mn IIIОЕP . 2

2

(22)

After expressing C(Cl)HO–2)MnIIIOEP via the equilibrium constant K2 and then CHO–2 via the equilibrium constant K1, we obtain equation (23):

− dC(Cl)(HO − )Mn IIIОЕP / dτ = − dC(Cl)Mn IIIOEP / dτ = 2

k3 ⋅ K 2 ⋅ C(Cl)Mn IIIОЕP ⋅ C1/2 ⋅ K11/ 2 . H O 2

2

(23)

The first equality in equation (23) is written with account for equilibrium (20). Thus, from the comparison of equations (20) and (23), we have

k ν1 = k3 ⋅ K11/2 ⋅ K 2 ,

(24)

where K1 = 2.4·10 –12 at 298 K [25]. To determine the numerical value of K2, we carried out the spectrophotometric titration of the solution of (Cl)MnIIIOEP in DMFA by hydrogen peroxide (Fig. 6). The titration curve has one stage. From the slope of the curve in the coordinates “the logarithm of the ratio of the equilibrium concentration of (Cl)(HO–2)MnIIIOEP to the equilibrium concentration of (Cl)MnIIIOEP (indicator ratio) vs the logarithm of CH2O2 to the abscissa axis”, the number of added H2O2 is 1/2. The concentration constant of the equilibrium of coordination of molecular peroxide, which can be equated to the thermodynamic constant with account for large dilutions in the reaction system, is, at 298 K, (2.2±0.3)·10 –5 mol –1/2 ·l1/2. Hence, the equilibrium constant (equation 20) K2 = (9±1)·106 mol –3/2 ·l3/2 provided that for the above constant K1 the value is given in molar scale, which does not explicitly follow from the results of [25]. The rate constant of the el–4 –1 ementary reaction (21) at 298 K is k3 = kv1/(K1/2 1 · K2) = (3.3±0.3)·10 s . Note that the dimensionality of the constant k3 corresponds to equations (21) and (22), which confirms the condition put forward above.

106


The third range of H2O2 concentrations with experimental kinetic equation (18) needs the account for the following five reactions: K1

H2O2 −

(Cl)MnIIIOEP + НО 2 −

(Cl)(НО 2 )MnIIIOEP

−

Н+ + НО 2 ,

(25)

k2 − ⎯⎯→ (Cl)(НО 2 )MnIIIOEP fast , K3

ОН- + Н2О2

.

(Cl)O=MnIV(OEP + • ) + ОН- , K4

−

НО 2 + Н2О ,

(26)

(27)

(28)

.

k5 2(Cl)O=MnIV(OEP+ .) + HO2-⎯⎯→ 2O=MnIII(OEP + • ) + HCl + O2 + Cl- slow . (29) Due to the low concentration of HO–2 , no gas evolution is observed in the experiment, and O2 formed is, evidently, in solution. The first stage involving manganese porphyrin (26) – coordination of hydrogen peroxide in the form of HO–2 to the sixth coordination position of the manganese atom – occurs at higher concentrations of H2O2 rapidly and irreversibly. The transfer of two electrons from the readily polarized porphyrin macrocycle and metal cation to the coordinated molecule of peroxide proceeds in the equilibrium stage (27). Probably, this reaction is made equilibrium by the stability of π-cation radical of (Cl)O=MnIV(OEP+·) decreased as compared with O=MnIII(OEP+·) (equation (21)), and the presence of higher concentrations of negative ions HO–2 and OH –, which prevent the ionization of acidoligand Cl – . The latter remains in the coordination sphere of π-cation radical (Scheme 3). Finally, the reaction with the second H2O2 (in the form of HO–2) becomes possible; the reaction proceeds slower than reaction (26) and limits the process. The reduction of π-cation radical of manganese(IV) porphyrin does not proceed to the end, that is to the relatively stable O=MnIVOEP, but stops at the one-electron reduced form O=MnIII(OEP+·), which is registered in the experiment as a relatively stable end product. This is one more distinction from the (Cl)MnIIITPP–H2O2 system, where the end product is a two-electron reduced form. We used the word “relatively” when assessing the stability. The cation-radical form O=MnIII(OEP+·) upon isolation into chloroform after the reaction of (Cl)MnOEP with H2O2 and washing off excess H2O2 still passes into the MnIV complex: bands at 394, 450 (arm), 514, 546 and 667 nm appeared in the spectrum, which is indicative of the formation of O=MnIVOEP. In turn, this compound, if kept still in solution for 24 h, slowly passes into manganese(III) porphyrin with a characteristic band in the region of 473 nm (the fact known in the chemistry of manganese porphyrins [26]). The rate equation for the limiting stage has the form: − dC(Cl)O = Mn IV (OEP+ .• ) / dτ = k5 ⋅ C(Cl)O = Mn IV (OEP+.• ) ⋅ CHO − . 2

(30)

After successively expressing C(Cl)=MnIV(OEP+·) via the equilibrium constants K3 (equation 31) and K4 (equation 32), we obtain equation (33):

107


Scheme 3 Assumed structure of π-cation radical of (Cl)O=MnIV(OEP+·). −1 C(Cl)O= Mn IV (OEP +•. ) = K3 ⋅ C(Сl)(HO− )Mn IIIOEP ⋅ CОН −,

(31)

2 −1 CОН − = СHO − ⋅ K4 ,

(32)

2

2

−2 C(Cl)O= Mn IV (OEP +•.) = K3 ⋅ C(Сl)(HO− )Mn IIIOEP ⋅ K4 ⋅ CHO − . 2

(33)

2

With account for the equilibrium (25), we express CHO2–: . CHO− = K11/ 2 ⋅ C1/2 H O 2

2

(34)

2

Upon substitution of the expressions for concentrations, (33) and (34), into equation (30), we obtain equation (35), which is consistent with the found experimental equation (18): − dC(Cl)O = Mn IV (ОЕP +•. ) / d τ = − dC(Cl)Mn III OEP / d τ = k5 ⋅ K 3 ⋅ K 4 ⋅ C (Cl)(HO − )Mn III OEP ⋅ C −1 − = k5 ⋅ K 3 ⋅ K 4 ⋅ K11/ 2 ⋅ C (Cl)Mn III OEP ⋅ C H−1/2 O . (35) 2

HO 2

2

2

Thus,

k ν3 = k5 ⋅ K3 ⋅ K 4 ⋅ K11/ 2 ,

(36)

where K1 = 2.4·10 –12 at 298 K. Change of details of the mechanism of the reaction of the octaethylporphin complex with H2O2 as compared with complexes with H2TPP is due to the difference in the electronic state of the aromatic ligand. The proof is the result of comparing the quantitative characteristics of the reactions. The (Cl)MnTPP reacts with H2O2 much faster than (Cl)MnOEP. The true rate constant is by an order higher in the case of (Cl)MnTPP both at low and high concentrations of H2O2, and the reaction of (Cl)MnTPP with H2O2 has more negative values of ∆S≠ (Tables 3 and 5). In the case of (Cl)MnTPP, the π-cation radical form is formed only at low concentrations of H2O2, and at CH2O2 > 0.1 mol/l the end product of the reaction is manganese(IV)

108


tetraphenylporphin – the form oxidized at Mn atom. In the case of (Cl)MnOEP, the end product of the reaction is π-cation radical of manganese(III) octaethylporphin within the entire range of H2O2 concentrations. Hence, it follows that, by changing the electronic state of the macrocycle and of Mn–N bonds, we can achieve optimal parameters for the coordination reaction of H2O2 molecules and HO2– ions at various stages of the reaction with peroxide and then use them in the development of synthetic catalases. 2.2

Kinetics of peroxide disproportionation in the presence of manganese(III) porphyrins

This section presents the results of studies of the kinetics of H2O2 homogeneous decomposition reaction by the volumetric and spectrophotometric methods in the presence of manganese(III) complexes with octaethylporphin unsubstituted in meso-positions and mesophenyloctaethylporphins and acidoligands Cl – , AcO – , SCN – (compounds 5 – 9). The reaction was carried out in a DMFA–KOH–H2O medium at temperatures of 343–363 K in a reactor equipped with a jacket for thermostatting and a magnetic stirrer, under conditions of intensive stirring. An aqueous solution of hydrogen peroxide was added to a prepared solution of the manganese complex and KOH in DMFA. The concentrations of the complex, KOH and H2O2 were varied within the range of 10 –6 –10 –5, 0.73–3.65 and 2.61–4.79 mol/l, respectively. The initial concentration of H2O2 in water (17.43±0.06 mol/l) was determined by iodometric titration. Table 4 presents the experimental values of the rate constants for the hydrogen peroxide decomposition reaction of formal zero order, found as a slope of the linear portion of the dependence of VO2 – τ to the positive direction of the abscissa (Fig. 7). The figure also presents the catalytic activity (A) obtained by dividing the value of W on the known concentration of the catalyst – the manganese complex. Table 5 presents activation energies determined by the slope of the curves in the coordinates log W – 1/T, and activation entropies calculated by formula (37):

ml

τ, min Figure 7 A dependence of the volume of evolving O2 on time in the course of the H2O2 decomposition reaction. C(Cl)MnMPOEP = 1.08·10 –5 (1), C(Cl)MnOEP = 2.77·10 –5 (2), C(AcO)MnOEP = 2.25·10 –5 (3), C(Cl)Mn5,15DPOEP = 3.87·10 –5 mol/l (4), CH2O2 = 3.49 mol/l, CKOH = 1.5·10 –2 mol/l. T = 343 K.


109

Table 4 Decomposition rates of H2O2 (W) and catalytic activity (A) of metalloporphyrins at 343 K. Catalyst

CH2O2, mol/l(a)

W, ml O2/min

A, s –1

(Cl)MnOEP

2.61 3.49 4.36 4.79

0.45±0.02 0.83±0.02 1.26±0.03 1.54±0.08

0.28±0.01 0.52±0.01 0.78±0.02 0.96±0.05

(AcO)MnOEP

2.61 3.49 3.87 4.36 4.79 2.69 3.14 3.58 4.03 4.48

0.40±0.01 0.74±0.01 1.08±0.05 1.35±0.08 1.45±0.07 0.44±0.01 0.55±0.02 0.74±0.02 1.07±0.03 1.31±0.07

0.67±0.02 1.25±0.02 1.82±0.08 2.3±0.1 2.45±0.11 0.57±0.01 0.72±0.02 0.97±0.02 1.40±0.04 1.72±0.09

(Cl)MnMPOEP

0.45 0.72 1.08

0.97±0.03 1.45±0.04 1.86±0.06

12.7±0.4 12.0±0.3 10.2±0.3

(Cl)Mn5,15DPOEP

2.58 3.44 3.87 4.30 4.73

0.33±0.01 0.73±0.02 0.83±0.02 1.12±0.04 1.31±0.04

0.20±0.01 0.45±0.01 0.51±0.01 0.69±0.02 0.80±0.04

Catalyst

CKOH ·10 –2, mol/l(b)

W, ml O2/min

A, s –1

(Cl)MnOEP

0.73 1.46 2.19 2.56 2.92 3.65

0.45±0.02 0.83±0.02 1.28±0.03 1.42±0.05 1.68±0.09 2.0±0.1

0.28±0.01 0.52±0.01 0.79±0.02 0.88±0.02 1.04±0.05 1.26±0.06

(AcO)MnOEP

0.73 1.46 2.19 2.56 2.92 3.65

0.47±0.01 0.74±0.02 1.09±0.06 1.51±0.08 1.62±0.08 2.14±0.09

0.79±0.02 1.25±0.02 1.8±0.1 2.6±0.1 2.7±0.1 3.61±0.15

(Cl)Mn5,15DPOEP

0.78 1.55 2.33 2.71 3.10 3.88

0.40±0.01 0.73±0.02 1.21±0.06 1.4±0.1 1.52±0.08 2.0±0.1

0.24±0.01 0.45±0.01 0.74±0.03 0.86±0.06 0.93±0.05 1.25±0.06

(SCN)MnOEP

110


Table 4 Decomposition rates of H2O2 (W) and catalytic activity (A) of metalloporphyrins at 343 K. (continued) Catalyst

Ccatalyst ·105, mol/l(c)

W, ml O2/min

A, s –1

– (Cl)MnOEP

– 0.095 0.19 0.38 2.77 5.11

0.62±0.02 0.76±0.01 0.83±0.02 0.84±0.06 1.58±0.06 2.10±0.06

– 47±1 26±1 13±0.6 3.3±0.1 2.4±0.1

(AcO)MnOEP

0.07 0.14 0.35 2.25

0.74±0.01 0.74±0.04 0.73±0.03 1.29±0.04

63±1 31±2 12±0.5 3.4±0.1

(SCN)MnOEP

0.09 0.18 0.32 0.45 2.71

0.73±0.03 0.73±0.02 0.75±0.03 0.78±0.05 1.48±0.05

48±2 24±0.6 14±0.5 10.3±0.6 3.2±0.1

(Cl)MnMPOEP

0.45 0.72 1.08

0.97±0.03 1.45±0.04 1.86±0.06

12.7±0.4 12.0±0.3 10.2±0.3

(Cl)Mn 5,15DPOEP

0.097 0.48 0.97 3.87 8.53

0.76±0.02 0.73±0.03 0.76±0.02 1.05±0.02 1.52±0.03

46±1 9.0±0.4 4.6±0.1 1.61±0.03 1.06±0.02

MnCl2

0.60 1.43 5.95 11.90

1.26±0.04 1.3±0.1 1.4±0.3 1.7±0.2

12.4±0.4 5.5±0.5 1.5±0.3 0.8±0.1

(a)C –2 –5 KOH = 1.5·10 mol/l, C(X)MnP = 0.2·10 mol/l, (b)C –5 H2O2 = 3.49 mol/l, C(X)MnP = 0.2·10 mol/l, (c)CH2O2 = 3.49 mol/l, CKOH = 1.5·10 –2 mol/l.

Table 5 Catalytic H2O2 decomposition reaction rates at various temperatures and activation parameters. C(X)MnP = 0.2·10 –5 mol/l, CH2O2 = 3.49 mol/l, CKOH = 1.5·10 –2 mol/l. Catalyst

T, K

W, ml O2/min

A, s –1

E, kJ/mol

∆S≠, J/(mol·K)

(Cl)MnOEP

343 348 353 358 363

0.83±0.02 1.16±0.08 1.74±0.12 2.27±0.15 2.75±0.20

0.52±0.01 0.71±0.05 1.05±0.07 1.35±0.08 1.61±0.11

64±4

–67±13

111


Table 5 Catalytic H2O2 decomposition reaction rates at various temperatures and activation Table 5 (continued) parameters. C(X)MnP = 0.2·10 –5 mol/l, CH2O2 = 3.49 mol/l, CKOH = 1.5·10 –2 mol/l. (AcO)MnOEP

343 348 353 358 363

0.74±0.01 0.95±0.03 1.38±0.08 1.86±0.10 2.44±0.09

1.25±0.02 1.58±0.05 2.26±0.13 3.01±0.16 3.90±0.15

63±2

–71±7

(SCN)MnOEP

343 348 353 358 363

0.74±0.02 0.91±0.06 1.27±0.04 1.80±0.15 2.46±0.18

0.97±0.02 1.17±0.07 1.62±0.05 2.26±0.18 3.05±0.22

65±3

–65±10

(Cl)Mn5,15DPOEP

343 348 353 358 363

0.73±0.01 1.00±0.05 1.61±0.10 2.23±0.15 3.53±0.20

0.45±0.01 0.60±0.03 0.96±0.06 1.31±0.09 2.04±0.11

82±3

–16±10

∆S ≠ = 19.1 ⋅ log W +

E ± ∆E − 19.1 ⋅ log T − 205 . T

(37)

It is seen from Table 4 that the decomposition rate of hydrogen peroxide increases with the rise of the concentrations of the manganese(III) complex, KOH and H2O2 in the system on condition of a significant excess of the other reagents as compared with the complex. There is a satisfactory correlation between W and the catalyst concentration, which points to the first order of the reaction with respect to this component (Fig. 8). A similar dependence for CH2O2 is not linear. Figure 9 demonstrates a correlation in the logarithmic coordinates log W – log CH2O2 with the slopes of the curves close to 2 (n = 2.01 for (Cl)MnOEP, n = 2.26 for (Cl)Mn5,15DPOEP). A similar dependence for COH – is also linear, with the slope α close to unity (Fig. 10). Thus, according to the data of the experiment, we can write down the equation for the reaction rate: W, ml O2 / min

Ccatalyst, 10-5, mol/l Figure 8 A dependence of the H2O2 decomposition rate on the concentration of (Cl)MnMPOEP (1), (Cl)MnOEP (2), (AcO)MnOEP (3), (Cl)Mn5,15DPOEP (4). CH2O2 = 3.49 mol/l, CKOH = 1.5·10 –2 mol/l, T = 343 K. R2 = 0.97–0.99.

112


logW

logW logCKOH logCH2O2

Fig. 9

Fig. 10

Figure 9 A dependence of log W on log CH2O2 for the reaction of H2O2 decomposition in the presence of (Cl)MnOEP (1) and (Cl)Mn5,15DPOEP (2). C(X)MnP = 0.2·10 –5 mol/l, CH2O2 = 3.49 mol/l, CKOH = 1.5·10 –2 mol/l, T = 343 K. R2 = 0.99. Figure 10 A dependence of the logarithms of the rates of H2O2 decomposition in the presence of (Cl)MnOEP (1) and (Cl)Mn5,15DPOEP (2) on the logarithms of KOH concentration. C(X)MnP = 0.2·10 –5 mol/l, CH2O2 = 3.49 mol/l, CKOH = 1.5·10 –2 mol/l, T = 343 K. R2 = 0.99.

2 dCO2 / dτ = k ⋅ Ccatalyst ⋅ СН ⋅ COH - ⋅ CO0 . 2О 2

(38)

2

The numerical values of the true rate constants k are calculated as arithmetic means of three true constants obtained in optimizing the dependences in coordinates W–Ccatalyst, log W–log CKOH and log W–log CH2O2 at integer values n, respectively, 1, 1 and 2. The constants k for (Cl)MnOEP, (AcO)MnOEP and (Cl)Mn5,15DPOEP are, respectively, (1.73±0.06), (1.64±0.17) and (1.49±0.07) l4 ·mol –3 ·s –1. A special spectrophotometric study of the interaction of (Cl)MnOEP with H2O2 in a DMFA–H2O medium in the absence of alkali, the results of which are considered in the previous section, made it possible to reveal in the entire range of H2O2 studied (0.01–3.74 mol/l) a manganese(III) porphyrin → π-cation radical of manganese(III) porphyrin transition (Fig. 2). With account for the data considered, we can write down the assumed scheme of conversions in the course of the catalytic reaction of H2O2 decomposition (equations (39)–(43)). (Х)MnIIIOEP + H2O2 (Х)(H2O2)MnIIIOEP

(Х)(H2O2)MnIIIOEP ,

(39)

. k2 ⎯⎯→ O=MnIII(OEP + • ) + H2O + Х- fast ,

H2O2 + OH-

.

O=MnIII(OEP + • ) + HO 2−

.

(HO 2 )O=MnIII(OEP + • ) −

K1

K3

K4

−

HO 2 + H2O,

(41)

.

(HO 2 )O=MnIII(OEP + • ), −

(40)

k5 ⎯⎯→ [MnIIIOEP]+ + H2O + O2 slow.

(42) (43)


113

For the limiting stage (43), we can write down the kinetic equation: −dC(HO- )O=Mn III (OEP+•.) / dτ = dCO2 / dτ = k5 ⋅ C(HO- )O=Mn III (OEP+•.) . 2

2

(44)

With account for the equilibrium (42), we express C(HO2-)O=MnIII(OEP+·) : C(HO- )O=Mn III (OEP+•.) = K4 ⋅ CO=Mn III (OEP+•.) ⋅ CHO- . 2 2

(45)

We write down equalities (46) –(48) for reactions (41), (40) and (39), respectively: СHO- = K3 ⋅ CH 2O2 ⋅ COH - . 2

(46)

СO=Mn III (OEP+•.) = C(Х)(H O )Mn IIIОЕP . 2 2

(47)

C(Х)(H O )Mn IIIОЕP = K1 ⋅ C(Х)Mn IIIOEP ⋅ CH 2O2 . 2 2

(48)

Let us perform the substitution of the expressions for CHO-2 and CO=MnIII(OEP+·) also into equation (45): 2 C(HO- )O=Mn III (OEP+•.) = K 4 ⋅ K1 ⋅ K3 ⋅ C(Х)Mn III OEP ⋅ CH ⋅ COH 2 O2 2

(49)

After substituting the expression for C(HO-2 )O =MnIII(OEP+·) (equation (49)) into (44), we obtain equation (50), which is consistent with experimentally found equation (38): 2 dCO2 / dτ = k5 ⋅ K 4 ⋅ K1 ⋅ K3 ⋅ C(Х)Mn IIIOEP ⋅ CH ⋅ COH - . 2 O2

(50)

From the comparison of equations (38) and (50), it follows that k = K1 · K3· K4· k5. In accordance with equations (39)–(43), the catalytic reaction of H2O2 decomposition proceeds with the participation of the catalyst, which changes its redox states, and two peroxide units (in the form of the H2O2 molecule and perhydroxyl anion HO2– ). At the first stage, H2O2 is coordinated to the sixth coordination site (X)MnIIIOEP; as the result, the O–O bond, which is not too strong as it is (200 kJ/mol), is activated [25]). The coordinated molecule of H2O2 breaks down to evolve H2O and to eliminate two electrons from metalloporphyrin. The latter passes into a π-cation radical form, being oxidized via the macrocycle and metal, which remains triple-charged only formally (O=MnIIIOEP+·) R [O=MnIVOEP]+), as, for instance, in the known stable oxidized forms of diphthalocyanines of lanthanides LnIIIPc2· [29]. The second molecule of peroxide is added as a more nucleophilic particle HO–2. The formed π-cation radical of the complex, being in a highly coordinated state, slowly reduces with abstraction of molecular O2. It should be emphasized that no manifestations of the radical mechanism of H2O2 decomposition have been revealed in support of the considered mechanism, in particular, no slow stages of reaction initiation. Variation of the ligand, which is in the fifth coordination position, with the view to confirm the provisions of the mechanism of the catalytic action of manganese(III) porphyrins, gave the following results. The catalytic activity of manganese(III) octaethylporphins (Table 6) is essentially not dependent on the nature of acidoligand (Cl, AcO, SCN), which

114


experimentally confirms the proposed stage-wise ion-molecular mechanism (equations (39)–(43)). The rate constants of the limiting stage, k5, do not depend on the nature of X (equation (43)), and the constant K1 is present in the expression for the rate (50) as a constant value, apparently differing for different X insignificantly. In earlier works [9], we proposed another form of recording the intermediate oxidized state of the catalyst formed from (X)(H2O2)MnIIIP (P, dianion of H2TPP). The one-electron oxidized form of (Cl)(OH·)MnIII(TPP+·) formed in abstraction of OH – from coordinated H2O2 was assumed. Evidently, this particle is unstable and should pass into a form with bidentate coordinated oxygen O=MnIII(TPP+·). Among the manganese(III) porphyrins studied, the highest catalytic activity in the H2O2 decomposition reaction is manifested by (Cl)MnMPOEP; the lowest, by the diphenyl-substituted complex. The activity of (Cl)Mn5,15DPOEP is comparable with that of MnCl2 (Table 6). Thus, introduction of one meso-phenyl group into (Cl)MnOEP leads to a significant activation of the catalase ability of the complex. This result is expectable within the framework of the expounded mechanism: a decrease of electronic density in the macrocycle at the substitution by phenyl group contributes to an increase of the positive charge at manganese, which leads to the reduction of (HO–2)O=MnIII(OEP+·) in the limiting stage (43). However, introduction of the second phenyl group (transition (Cl)MnMPOEP → (Cl)Mn5,15DPOEP) not only fails to increase the catalytic activity, but decreases it to the level of manganese salts (Table 6). This could be due only to the worsening of the coordination conditions of the second molecule of peroxide in trans-position with respect to O2– owing to the distortions, which, apparently, the (Cl)Mn5,15DPOEP molecule experiences like similar copper(II) complexes. According to [30, 31], the coordination centre of the copper complex of octaethylporphin remains planar at the introduction of one meso-phenyl substituent and is significantly distorted in Cu5,15DPOEP. Thus, the catalase activity of manganese(III) porphyrins can be both enhanced and reduced to zero depending on the structure of the macrocycle. Probably, understanding of the mechanism of the process would facilitate prediction of this property for other metalloporphyrins and catalytic systems.

3

Conclusion

The catalytic-action mechanism of manganese(III) porphyrins expounded above repeats in many ways the mechanism of action of natural catalases. The general regularities of conversions involving the enzymes hemoproteins have been discussed already in the monograph [25]. The enzyme catalase, which has a prosthetic group of iron protoporphyrin, is bound to the protein by means of one coordination link (one coordination site of Fe) and via propionic acid residues. H2O2 in the amount of several micromoles per litre is either decomposed or reacts with other particles. With iron protoporphyrin, hydrogen peroxide forms, apparently owing to coordination to the sixth coordination position, complexes of three types: Fe·H2O2, Fe·OOH and Fe·OOH –, the last of which is active as catalase. This complex further reacts with another molecule of H2O2 or other active particles. Thus, the expounded conclusions regarding the catalytic-action mechanism of manganese(III) porphyrins obtained in our work basically correspond to the data for the mechanism of natural catalases. Complexes of natural catalases and peroxidases have not been isolated individually due to their instability. The use of simpler models, as shown by our study, makes it possible to describe intermediate complexes spectrally. Besides, a possibility appears to


115

regulate and, which is more important, increase the catalytic activity of catalases by modifying the aromatic moiety of their molecules. The work was supported by grants from the Programme for Basic Research of the Russian Academy of Sciences “Elaboration of Methods for Producing Chemical Substances and Development of Novel Materials” (No 8, 2006); from the Analytical Departmental Target-oriented Programme “Development of Higher-School Scientific Potential” (2006–2008). Activity 2: Scientific Methodological Provision of the Development of Higher-school Science Infrastructure”; from RNP 2.2.1.1.7181 “Development of the Mechanisms of Integrating Ivanovo State University and Institute of Problems of Chemical Physics RAS”; and from the Russian Foundation for Basic Research 06-03-96343-r.

References 1. G. L. Eichorn (ed.), Inorganic Biochemistry, Elsevier: New York, vol. 2 (1973). 2. I. Naruta, M. Sasayama and K. Ishichara, Zhurn. Org. Khim., 32 (2), 233 (1996) (in Russian). 3. O.V. Cheremenskaya, A.B. Solovyeva, G.V. Ponomarev and S.F. Timashev, Zhurn. Fiz. Khim., 75 (10), 1787 (2001) (in Russian). 4. M.Yu. Tipugina, T.N. Lomova and T.A. Ageyeva, Zhurn. Obshch. Khim., 69 (3), 459 (1999) (in Russian). 5. M.Yu. Tipugina and T.N. Lomova, Zhurn. Neorg. Khim., 47 (7), 1085 (2002) (in Russian). 6. M.R. Tarasevich and K.A. Radyushkina, Catalysis and Electrocatalysis by Metalloporphyrins, Nauka: Moscow, p. 168 (1982) (in Russian). 7. A.Ya. Sychev and V.G. Isak, Coordination Compounds of Manganese in Catalysis, Shtiinca: Kishinev, p. 321 (1990) (in Russian). 8. M.E. Klyueva, T.N. Lomova and M.V. Klyuev, in: Peroxides at the Beginning of the Third Millennium. Synthesis, Properties and Application, ed. by V.L. Antonovsky, O.T. Kasaikina and G.E. Zaikov, Nova Science: New York, pp. 143–166 (2004). 9. T.N. Lomova, M.V. Klyuev, M.E. Klyueva, E.N. Kiseleva and O.V. Kosareva, Ross. Khim. Zhurn., XLVIII (4), 35–51 (2004) (in Russian). 10. E.N. Kiseleva, M.E. Klyueva and T.N. Lomova, Abstracts of the VIII-th Scientific SchoolConference on Organic Chemistry, p. 139, Kazan’, 22–26 June (2005) (in Russian). 11. A.D. Adler, F.R. Longo and J.D. Finarelli, J. Org. Chem., 32, 476 (1967). 12. N.S. Dudkina, P.A. Shatunov, E.M. Kuvshinova, S.G. Pukhovskaya, A.S. Semeykin and O.A. Golubchikov, Zhurn. Obshch. Khim., 68 (12), 2042 (1998) (in Russian). 13. A.D. Adler, F.R. Longo, F. Kampas and J. Kim, J. Inorg. Nucl. Chem., 32, 2443 (1970). 14. A.Ya. Sychev and V.G. Isak, Coordination Compounds of Manganese in Catalysis, Shtiinca: Kishinev, p. 321 (1990) (in Russian). 15. P.N. Balasubramanian, E.S. Schmidt and T.C. Bruce, J. Am. Chem. Soc., 109 (25), 7865 (1987). 16. R.D. Arasasingham, G.-X. He and T.C. Bruce, J. Am. Chem. Soc., 115, 7985 (1993). 17. J.T. Groves and M.K. Stern, J. Am. Chem. Soc., 109 (12), 3812 (1987). 18. S. Banfi, M. Cavazzini, G. Pozzi, S.V. Barkanova and O.L. Kaliya, J. Chem. Soc. Perkin Trans., II, 871 (2000). 19. S. Banfi, M. Cavazzini, G. Pozzi, S.V. Barkanova and O.L. Kaliya, J. Chem. Soc. Perkin Trans., II, 879 (2000). 20. S. Banfi, M. Cavazzini, G. Pozzi, S.V. Barkanova and O.L. Kaliya, J. Chem. Soc. Perkin Trans., II, 1577 (1997). 21. O.A. Golubchikov and B.D. Berezin, Usp. Khim., 55 (8), 1361 (1986) (in Russian). 22. B.D. Berezin and G.V. Sennikova, Kinet. Kataliz, 9 (3), 528 (1968) (in Russian). 23. B.D. Berezin and G.V. Sennikova, Zhurn. Fiz. Khim., 43 (10), 2499 (1969) (in Russian). 24. E.N. Kiseleva, M.E. Klyueva and T.N. Lomova, Abstracts of the IX-th Int. Conf. on the Chemistry of Porphyrins and their Analogues, Ivanovo, p. 106 (2003). 25. W.C. Schumb, C.N. Satterfield and R.L. Wentworth, Hydrogen Peroxide, ACS Monograph, Reinhold Publishing Corp.: New York, 757 pp. (1955). 26. H. Volz and W. Müller, Chem. Ber. Recueil, 130, 1099 (1997). 27. L. Kaustov, M.E. Tal, A.I. Shames and Z. Gross, Inorg. Chem., 36 (16), 3503 (1997).

116


28. N. Carnieri, A. Harriman, G. Porter and K. Kalyanasundaram, J. Chem. Soc. Dalton Trans., 7, 1231 (1982). 29. P.I. Moskalev, USSR Author’s Certificate 525318, Byul. Izobr. 4 (1978) (in Russian). 30. O.A. Golubchikov, S.G. Pukhovskaya and E.M. Kuvshinova, Advances of Porphyrin Chemistry, vol. 4, ed. by O.A. Golubchikov, Chemistry Research Institute, St.-Petersburg University: St.-Petersburg, p. 45 (in Russian). 31. M.E. Klyueva, T.N. Lomova, E.E. Suslova, and A.S. Semeykin, Teor. Eksp. Khimiya, 39 (5), 299 (2003) (in Russian).

4

Complexation of Porphyrins with Ions and Organic Molecules N.Zh. Mamardashvili1, V.V. Borovkov2, G.M. Mamardashvili1, Y. Inoue2 and O.I. Koifman1 1Institute

of Solution Chemistry, Russian Academy of Sciences, 1 Akademicheskaya Street, Ivanovo, 153045, Russia [email protected] 2Entropy Control Project, ICORP, Japan Science and Technology Agency, Kamishinden 4-6-3, Toyonaka-shi, Osaka 560-0085, Japan [email protected], [email protected]

Among the great diversity of host–guest interactions, a special place is occupied by the complexation processes of porphyrins and related chromophores with ions and organic molecules of various nature. This chapter considers the major principles, driving forces and factors, which enable an efficient control of supramolecular systems formed in this process.

Keywords: porphyrin, complexation, conformational correspondence, conformity, multipoint binding, chirality, circular dichroism

Introduction At present, the complex-forming properties of porphyrins are understood to be both the ability of porphyrins to enter into complexation reactions with metal cations to form metalloporphyrins and the ability of porphyrins to form supramolecular complexes. While metalloporphyrin formation processes have been fruitfully studied for many decades, works on supramolecular complexes of porphyrins having no less than two binding points, appeared in the literature only in the recent 10 to 15 years. Presented material generalizes and

118

N.Zh. Mamardashvili, V.V. Borovkov, G.M. Mamardashvili, Y. Inoue and O.I. Koifman

analyzes the literature data and own results by the authors on supramolecular complexes of porphyrins, showing the main trends of the development of the modern supramolecular chemistry of porphyrins.

1

Complexation of Porphyrins with Ions

One of the essential coordination properties of porphyrins and their analogues is their ability to coordinate virtually with all ions of metals in the periodic system of elements to form especially strong intracomplex salts. Complexation of porphyrins with metal cations, according to [1], occurs as the result of bimolecular collision of the ligand molecule and solvated salt: activation of solvation complexes of octahedral structure as the determining factor includes a partial decomposition of the first coordination sphere of metal, i.e., formation of two coordination vacancies in cis-position of the octahedral complex. The reaction, as a rule, is of the first kinetic order with respect to porphyrin and metal salt. Salts of two-charge ions of d-metals interact the most readily with porphyrins in a medium of weakly coordinating organic solvents – carboxylic acids and lower alcohols. The rate of the reaction very little depends on the anionic composition of a salt. Thus, very strong complexes, in which anion is bonded by an ionic-covalent bond (PdCl2, HgI2 etc) react with porphyrins at rates characteristic of almost all ionic salts ( (AgNO3, Cu(NO3)2, FeCl2 etc). This indicates that not anions but molecules of a solvating solvent go from the coordination sphere of a salt in transition state. Anion of the salt leaves the coordination sphere beyond the peak of the potential barrier. It is seen in Table 1 that small and acyclic molecules (CH3OH, C2H5OH, CH3COOH, acetone) are eliminated more readily and, as a rule, with lower energy and entropy of activation; large and cyclic molecules (pentanol, ethyl acetate, dioxane, pyridine) are hard to eliminate. In solvents of the first group, metalloporphyrins are formed much faster and easier than in those of the second group. Optimal solvents for complexation are monocarboxylic acids. At a high temperature, a good medium to form metalloporphyrins is pyridine. Structural peculiarities of porphyrin molecules also have a strong effect on the rate of the process. At the stage of activation of the reaction, the porphyrin structure undergoes very significant changes, in particular, loses two central hydrogen atoms as protons, which Table 1 Rate constants (k) and activation parameters (E and ∆S≠ ) of the complexation of copper(II) cation with chlorophyllic acid in organic solvents at 298 K [1]. Solvent Methanol Ethanol n-Pentanol Acetone Dioxane Acetic acid DMFA Pyridine Quinoline Ethyl acetate

k, l/(mol·s)

E, kJ/mol

∆S ≠, J/(mol·K)

4.45 1.65 0.55 0.90 0.05 5.30 0.04 0.01 0.03 0.51

58 44 113 46 72 44 79 96 100 61

–17 –96 126 –67 –25 –79 –8 37 63 –63

Complexation of Porphyrins with Ions and Organic Molecules

119

under the action of an electric field of the cation pass into solution and bind with solvent molecules. Therefore, both the rate of formation of metalloporphyrin and activation parameters of the process depend on the extent of change of the N-H bond covalence and electron density at atoms of nitrogen of the porphyrin coordination core under the influence of substituents in the molecule. Table 2 shows how great that influence is. The formation rates of metal complexes, depending on the structure of porphyrin, change hundreds of times; and activation energies, from 41.2 up to 115 kJ/mol. Table 2 Rate constants and activation parameters of the complexation reaction of porphyrins with copper(II) cation in ethanol at 298 K [1]. Porphyrin Chlorophyll Porphin Ethioporphyrin I Ethiochlorin II Tetraphenylporphin Protoporphyrin Mesoporphyrin Deuteroporphyrin Hematoporphyrin Phylloporphyrin Rhodoporphyrin Tetrabenzoporphin (a) Tetrazaporphin (a) (a)Solvent:

k, l/(mol·s)

E, kJ/mol


0.35 0.55 1.94 2.08 2.05 2.43 5.44 3.88 2.26 13.5 0.25 3.92 1000

60.7 55.3 56.1 47.3 58.6 78.7 69.5 60.7 75.7 41.2 76.0 58.7 115

–50 –67 –50 –75 –42 –63 –4 –37 –8 –75 –11 –84 118

pyridine

It should also be noted that, as compared with complex-formation processes with open (not macrocyclic) ligands, the formation rates of metalloporphyrins are extremely low. This is due to the rigid planar structure of the ligand, to the shielding of the coordination core by the adjacent atoms and π-electronic cloud. This spatial shielding of the reaction site is typical of the formation processes of rigid aromatic macrocycles and forms the basis of the macrocyclic effect. It is the rigidity of the ligand, determined by the aromaticity of porphyrin, that does not enable it to take on the conformation suitable for rapid coordination. In the course of interaction, it is not the ligand that has to adjust to the coordination sphere of the salt, but, the other way round, the salt to the ligand, and the reagent is primarily the salt. Upon introduction of an alkyl group, the porphyrin molecule is strongly polarized, which contributes to the solvation of the molecule as a whole. As the result, the transition state of meso-alkyl-substituted porphyrin becomes more probable [1]. The coordination rate of phylloporphyrin 1 by Cu(II) cations in ethanol is three times as large as that of pyrroporphyrin 2. meso-Ethyl substitution of octaethylporphyrin (compound 3) increases the rate of the reaction with zinc cation in acetonitrile 6.6-fold (Table 3). Introduction of a hexyl group into hexamethyldiethylporphyrin 4 increases the rate of coordination of the molecule by copper(II) and zinc(II) cations in propanol by approximately an order [2]. That is, with the length of alkyl in meso-position (Alk = CH 3, C2H5, C6H13) increasing, the rate of the process rises 3- to 10-fold.

120


Table 3 Kinetic parameters of the coordination reaction of octalkylporphyrins (1–6) with various substituents in meso-position, copper and zinc cations in alcohols at 298 K [3]. Porphyrin

mesoSubstituent

c°salt·103, mol·l–1

k, l·mol–1 ·s–1

E, kJ/mol


CH3 H C6H13 C6H5 C2H5 C6H13 C6H13 C6H5

0.92(Cu(OAc)2) 2.61(Cu(OAc)2) 0.35(Cu(OAc)2) 0.37(Cu(OAc)2) 1.15(Zn(OAc)2) 0.33(Zn(OAc)2) 0.24(Zn(OAc)2) 0.20(Zn(OAc)2)

48.5 15.5 26.1 30.2 25.0 23.4 29.0 55.0

10.0 7.0 44.2 48.7 35.0 42.1 34.4 38.2

–18.7 –32.3 –81.9 –64.9 –10.4 94.6 –111.1 –94.7

1 2 5 6 3 (a) 4 5 6 (a) Data

are given for acetonitrile.

A single meso-phenyl substitution (compound 6) with fixed position of the phenyl group leads to an almost the same effect as hexyl substitution (compound 5). What is more, in the case of Zn(II) the reactivity of the porphyrin molecule strongly changes (~ 2-fold) as compared with the similar process involving Cu(II). The works [4–7] studied the change of the complex-forming ability of 3,7,13,17-tetramethyl-2,8,12,18-tetrabutylporphyrin (7) during its substitution at two meso-positions: 1) by phenyl fragments (compound 8); 2) by phenyl fragments containing electron-donor (CH3O) and electron-acceptor (NO2) substituents (compounds 9–12); by alkyl groups of various lengths (compounds 13, 14). The complexation kinetics of disubstituted octalkylporphyrins with copper(II) and zinc(II) cations was studied in ethanol, acetic acid, acetonitrile and pyridine. The complexation reaction of porphyrins (7–14) with copper and zinc cations in these solvents is described by the first-order kinetic equation with respect to porphyrin, usual for monomeric porphyrins. As the order of the reaction with respect to copper acetate in acetic acid [8], ethanol and dimethylformamide [9] is equal to 0.5, the total order of the reaction with respect to the reagents is 1.5. Tables 4–9 present the 1.5-order rate constants calculated using equation (1): (1) k

1.5

/

cCu (OAc)

(1)

2

R5

R4 R3 NH

N

N

HN

R6

R7 R9 1-6

1 : R1 =CH2CH2COOH, R2= R3= R5 = R7 = R9 = Me, R4= R6=Et, R8=H, 2 : R1 =CH2CH2COOH, R2= R3= R5 = R7 = Me, R4= R6=Et, R8=R9 =H 3 : R1 = R2= R3= R4 = R5= R6= R7 = R8= R9=Et 4 : R1 = R2= R3= R6 = R 7 =R 8 =Me, R4 = R5=Et, R9 =C 6H13

R2 R1

= kef

R8

5 : R1 = R4 = R5 =R 8 = Me, R2= R3 =R 6=R 7 =Et, R9 =C6H13 6 : R1 = R4 = R5 = R8 = Me, R2= R3 =R6 =R7=Et, R9 =Ph

121

Complexation of Porphyrins with Ions and Organic Molecules 7 : R = Bu, R' = H

R

R H3C NH

CH3

N

R'

R' N

HN

9 : R = Bu, R' = PhOMe-p 10 : R = Bu, R' = PhOMe-m 11 : R = Bu, R' = PhOMe-o

H3C

CH3 R

8 : R = Bu, R' = Ph

R

12 : R = Bu, R' = PhNO2-p 13 : R = Me, R' = Et 14 : R = Me, R' = C3H7

7-14

In the complexation reaction with copper(II) acetate the substituents of benzene fragments little affect the reactivity of porphyrins 9–12. The presence of methyl groups in β-positions of the macrocycle excludes the possibility of rotation of benzene fragments, which, being located in the plane perpendicular to the plane of porphyrin, do not enter into conjugation with the macrocyclic π-electronic system. Nevertheless, there is a tendency to an increase of the complexation rate under the influence of electron-donor substituents of benzene residues. Apparently, the substituents, by changing the electon density in a benzene fragment, affect the electronegativity of carbon atom of porphyrin macrocycle’s methine bridge. Further on, the effect of the substituent is passed by means of σ-bonds C–C and C–N to the reaction centre. Introduction of electron-donor substituents (para-CH3O 9, meta-CH3O 10) into benzene fragments of 5,15-diphenylporphyrin 8 insignificantly accelerated the reaction (Tables 4, 5). Table 4 Kinetic parameters of the complexation reaction of 5,15-diarylporphyrins (8–12) with copper(II) cation in acetic acid at 298 K: cCu(OAc)2 = 3.047·10–3 mol/l; cºH2P = 4.021·10–5 mol/l. No

meso-Substituent

k1.5 ·102, l0.5 ·mol–0.5 ·s–1

8 9 10 11 12

C6H5 C6H5(n-OCH3) C6H5(m-OCH3) C6H5(o-OCH3) C6H5(n-NO2)

5.1±0.1 5.2 ±0.2 5.3±0.15 8.4±0.1 11.0±0.2

E, kJ/mol 97±2 89±1 87±2 89±1 77±0.5

∆S ≠, J/(mol·K) 70±6 43±4 37±7 43.5±4 3.5±3

Table 5 Kinetic parameters of the complexation reaction of 5,15-diarylporphyrins (7–12) with copper(II) cation in acetonitrile at 298 K: cCu(OAc)2 = 3.760·10–3 mol/l; cºH2P = 3.521·10–5 mol/l. No

meso-Substituent

7 8 9 10 11 12

H C6H5 C6H5(n-OCH3) C6H5(m-OCH3) C6H5(o-OCH3) C6H5(n-NO2)

l0.5 ·mol–0.5 ·s–1

k1.5 ·102,

E, kJ/mol


17.1±0.05 9.2 ±0.05 10.3±0.05 10.1±0.05 5.1±0.05 5.8±0.05

46±1 77±2 78±1 71±1.5 85±0.5 76±2

–113±3 87±6 –10±3 –34±5 7±3 –21±7

122


Table 6 Kinetic parameters of the complexation reaction of 5,15-diarylporphyrins (7–12) with copper(II) cation in pyridine at 298 K: cCu(OAc)2 = 4.956·10–3 mol/l; cºH2P = 2.04·10–5 mol/l. No

meso-Substituent

7 8 9 10 11 12

H C6H5 C6H5(n-OCH3) C6H5(m-OCH3) C6H5(o-OCH3) C6H5(n-NO2)

l0.5

k1.5 ·102, ·mol–0.5 ·s–1 0.16 0.08 0.09 0.08 0.09 0.11

E, kJ/mol 78±1.0 98±4 98±1.5 94.5±2 94±2 86.5±1.0

∆S ≠, J/(mol·K) 3±3 70±10 70±5 58±8 57±6 32±4

Table 7 Kinetic parameters of the complexation reaction of 5,15-diarylporphyrins (7, 8, 10, 11) with copper(II) cation in ethanol at 298 K: cCu(OAc)2 = 1.211·10–3 mol/l; cºH2P = 4.520·10–5 mol/l. No

meso-Substituent

7 8 10 11

H C6H5 C6H5(m-OCH3) C6H5(o-OCH3)

l0.5

k1.5 ·102, ·mol–0.5 ·s–1

E, kJ/mol


219 ± 2 117±1.2 130±1.4 128±2

37±3 44±3 45±1 43±4

–59±8 –103±8 –101±3 –108±4

The electron-acceptor substituent (para-NO2, 12) slightly slows down the reaction. Against the background of pronounced effects of benzene fragments’ substituents in porphyrins 7--12, the role of solvation effects increases. Thus, in pyridine (Table 6), diphenylporphyrin 8 and its ortho-methoxy derivative 11 have almost the same reactivity (α,α-atropisomer was studied, in which both methoxy groups are on one side of the porphyrine macrocycle), whereas in acetonitrile (Table 5) the derivative 11 reacts with copper(II) cation two times as slow as the initial compound 8. In passing from meso-unsubstituted porphyrin 7 to dialkylporphyrins 13, 14, the reactivity of the macrocycle increases (Tables 8, 9). The rate constant in ethanol changes from 219·10 –2 l/mol·s (compound 7) to 271·10 –2 l/mol·s (compound 13). Probably, alkyl substituents, possessing a positive induction effect, increase the electron density at tertiary atoms of nitrogen and facilitate the interaction of metal cation with porphyrin in transition state. A decrease of the activation energy (~ by 7.0 kJ/(mol·K)) in meso-alkyl substitution of porphyrin in ethanol is due to an increase of solvation of the salt--porphyrin system’s transition state during the emergence of alkyl substituents in the molecule. It is known that introduction of two ethyl groups in meso-position of octamethylporphyrin causes an increase of solubility of the molecule in benzene by two orders of magnitude. The effect of substituents is more pronounced in acetic acid (Tables 4, 9), which specifically solvates the reaction centre of porphyrin and decreases its reactivity in the interaction with copper(II) cation. The strength of the hydrogen bonds between nitrogen atoms of porphyrin and hydrogen atoms of acid is reduced at the introduction of acceptor substituents into 8. This causes an increase of the reaction rate (Table 4). At 298 K, the rate constant increases from 5.1·10 –2 (8) to 11·10 –2 l/(mol·s) (para-nitroderivative 12). As compared with diphenylporphyrin, the activation energy of the nitroderivative decreases by ~20 kJ/mol; and activation entropy, by 67 J/(mol·K). Introduction of electron-donor alkyl groups leads to a decrease of reactivity of the macrocycle (Table 9).

123


Table 8 Kinetic parameters of the complexation reaction of porphyrins (7,13, 14) with copper(II) and zinc(II) cations in ethanol at 298 K. No

mesoSubstituent

c°salt ·105, mol/l

k1.5 ·102, l0.5 ·mol–0.5 ·s–1

E, kJ/mol


7 13 14

H C2H5 C3H7

1.21 (Cu(OAc)2) 2.80 (Cu(OAc)2) 5.81 (Zn(OAc)2)

219±6 271±6 208±2

40±1.5 37±3 48±3

–59±9 –121±5 –89±12

The rate constant in the case of zinc complexes of porphyrins is of the second order (kv, l/mol · s). Table 9 Kinetic parameters of the complexation reaction of 5,15-diarylporphyrins (7, 13, 14) with copper(II) cation in acetic acid at 298 K: cCu(OAc)2 = 2.410·10–3 mol/l; cºH2P = 6.721·10–5 mol/l. No

meso-Substituent l

7 13 14

H C2H5 C6H13

k1.5 ·102, ·mol–0.5 ·s–1

E, kJ/mol


3.1±0.15 0.7±0.05 0.4±0.05

60±2 52±4 55±2

–54±6 –82±13 –73±7

0.5

The reactivity of porphyrins in complexation with metal cations strongly depends on the nature of a solvent. Low rates of formation of metal complexes in pyridine (Table 6) are determined by a high strength of the solvate coordination sphere of [Cu(OAc)2(Py)4]. The maximal values of k1.5 are observed in running the reaction in ethanol (Tables 7, 8), which forms a solvated complex [Cu(OAc)2(C2H5OH)4] unstable as compared with pyridine. Owing to the increase polarity of ethanol, the polar transition state of the salt–porphyrin system is subjected to efficient additional solvation in alcohol. For this reason, in ethyl alcohol as compared with pyridine and acetic acid, the values of energy and entropy of the coordination reaction have minimal values. A decrease of reactivity of porphyrins in acetic acid shows that the prevalent effect on the rate of the process is rended by blocking the porphyrin’s reaction centre due to the occurrence of hydrogen bonds. A significant increase of energy and entropy of activation in acetic acid as compared with ethanol has been noted (Tables 7–9). Despite the low donor strength of acetonitrile [10--12] forming labile solvated complexes of transition metals [13--16], the reaction rate of copper(II) cation with porphyrins is lower than in ethanol (Tables 7, 8). Probably, this is due to the decreased basicity of acetonitrile, incapable of efficient solvation of protons leaving the porphyrin’s reaction centre in transition state. Of special interest are porphyrins containing complex-forming cavities of various nature (e.g., calix[4]arenes). The presence of additional complex-forming fragments in such molecules enables their use in studies of complexation processes occurring without the direct participation of tetrapyrrole macrocycles [17, 18]. The works [19, 20] investigated porphyrin-calix[4]arenes 15, 16 with “activated” amide substituents, which are capable of interacting with anions by forming hydrogen bonds by NH groups of bridges. Analysis of the binding constants of porphyrin-calixarene conjugates 17--19 with ions I – , Br – , Cl – and NO3– (Table 10) shows that, irrespective of the conformation of the calix[4]arene moiety of the molecule, compounds 17--19 efficiently bind the spherical anions of small size. The bridge groups linking the calix[4]arene and

124


O

R

NH

NH

R

NH

N

N

NH

anion

R R

X

15, 16 O

P

NH

O

NH

NH OPrOPr

N

N

NH

O

O

NH

NH

OPr 17

NH

NH

O

O

NH

NH

OPr OPr

OPr OPr

18

P=

P

P

NH

NH

OPr

OPr

NH

P

P

NH

O

NH

P

16 : X=

NH

15 : X = H ,

OPr OPr

OPr

19

NH N

N HN

porphyrin fragments possess a sufficient flexibility for placing ions I – and Br – , irrespective of their location [cis- or trans-], on the upper rim of calix[4]arene. What is more, the propoxy groups located between porphyrin fragments of 1,3-alternate do not prevent the compolexation with anion. At the same time, significant differences in the binding constants for larger anions (I – and NO3– ) point to certain spatial restrictions for complexation. With the diameter of anion increasing, the binding constant of anions decreases [for porphyrin-calix[4]arene 19 KCl (6.9·105 l/mol) > KBr (6.9·104 l/mol) > KI (2.4·103 l/mol)]. Herewith, compound 19 exhibits a better complexing ability with respect to Cl – as compared


125

with the system having a similar structure of the calix[4]arene platform, but instead of tetrapyrrole fragments containing phenyl groups (KCl 4.6·103 l/mol) [20]. Table 10 Constants of binding Ka [l/mol] by porphyrin-calix[4]arenes (17–19) of various-nature anions in dichloromethane at 24ºC (creceptors ~ 1.5·10–6 mol/l).

2

Anion

17

18

19

Cl– Br– I– NO–3

6.3×10 3 1.2×10 3 240 820

5.8 × 10 5 7.8×10 4 8×10 3 3×10 4

6.9×10 5 6.9×10 4 2.4×10 3 1.3×10 4

Complexation of Porphyrins with Organic Molecules: The Thermodynamic Aspect

Axial coordination (extracoordination) of ligands is inherent in many complexes with a planar structure of the coordination node ML4, but on metalloporphyrins it is especially pronounced and peculiar [1, 21–23]. Formation of porphyrin--ligand extra complexes (their structure, composition, stability) depends on the nature of metal. In the case of Zn(II) prone to sp3 or d 2sp2 hybridization, but in a planar porphyrin molecule forced to take on a dsp2 configuration, electron-deficient 4pz and 4dz2 orbitals are formed, which enter into a donoracceptor interaction with electron-donor ligands. Zn-porphyrins are readily bind nitrogen-containing ligands and much weaker coordinate oxygen- and sulfur-containing ligands. But although oxygen coordination on zinc is less weak than nitrogen coordination, it also can be used as a nonselective anchor point in multipoint recognition [24, 25]. According to [26, 27], monomeric Zn-porphyrins are capable of binding only one axial ligand, thus forming a five-coordinated complex. At the same time, the authors of [25, 28, 29] note that in the case of weaker bases Zn-porphyrins can add two extra ligands, but with different strength. In coordination of oxygen-, sulfur- and nitrogen-containing ligands on monomeric Zn-porphyrin, the binding constant of Zn-porphyrin--ligand complexes (Ka) (association constant, extracoordination constant, stability constant -- there are different terms in the literature) very strongly depends on the basicity of this ligand. In a general case, the larger the basicity of a ligand, the stronger it is coordinated on metalloporphyrin. In the sequence ethanol < pyrrole < DMSO < pyridine < imidazole < piperidine the binding constant of Zn-tetraphenylporphin (Zn-TPP) in toluene increases by four orders [27]. This dependence is disturbed, e.g., in the case when the free electron pair of the ligand, taking part in coordination, is delocalized on adjacent substituents [30]. As seen from Table 11, the effect of the solvent and of the alkyl- and arylporphyrin structure on the process of axial coordination is much smaller than the effect of the nature of the axial ligand itself. The extracoordination process on metalloporphyrins is usually studied in a medium of “inert” solvents (benzene, toluene, dichloromethane). Works by Viugin et al. [31, 32] investigated extracoordination of molecules of various solvents for a large number of metalloporphyrins by the calorimetric method. Coordination enthalpies of the above “inert” solvents on Zn-porphyrins are close to zero. According to the data of Table 11, the binding constants of Zn-porphyrin--ligand complexes in passing from one “inert” solvent to another change

126


no more than fourfold. The largest values of Ka are observed in aromatic solvents of the type of benzene and toluene. The effect of alkyl and aryl substitution on the binding constants in porphyrins 20--31 is levelled out even more (no more than twofold), as the electronic influence of phenyl fragments and alkyl groups on the state of the central zinc ion is insignificant. At the same time, Table 11 Ka of some alkyl- and aryl-substituted monomeric Zn-porphyrins in monodentate ligands at 298 K. Zn-porphyrin 20

21 22 23 24 25 26

27 28 29

30

31 (a) Data

Monodentate ligand Piperidine Imidazole Quinoline Pyridine Pyridine Pyrrole Methanol Pyridine Pyridine Pyridine Pyridine 4-Me-Pyridine Pyrazole Methylimidazole Imidazole Pyridine DMFA Pyrazole Imidazole Pyridine Pyrazole Pyridine Pyridine Pyridine 4-Me-Pyridine 4-But-Pyridine Pyridine Pyridine Imidazole Pyridine Pyridine Imidazole Pyridine Pyridine Imidazole Pyridine

Solvent

Ka, mol–1 ·l

Reference

Toluene Toluene Toluene Toluene Toluene Toluene Toluene Benzene CHCl3 CHCl3 CHCl3 CHCl3 CHCl3 o-Xylene o-Xylene o-Xylene o-Xylene CH2Cl2 Benzene CH2Cl2 CH2Cl2 Benzene Toluene-methanol (2:1) Toluene Toluene Toluene CH2Cl2 CH2Cl2 Benzene Benzene CHCl3 Benzene Benzene CHCl3 Benzene Benzene

809000 24000 16900 5800 4790 200 4.6 3750 2525 617 900 1600 1500 110000 15000 1800 560 2600 51920 2890 800 10300 1380 5000 11000 14000 2800 5200 59740 4694 1136 33140 5586 2590 35450 5795

1 1 1 1 33 1 1 34 35 36 37 37 37 38 38 38 38 39(a) 1 40(b) 39 35 41 42 42 42 43(c) 43 44 45 45 44 45 45 1 45

are given at 295 K; (b) temperature not indicated; (c) data are given at 303 K.

127


in the sequence of metalloporphyrins 32--49 with different deformation degrees of the tetrapyrrole macrocycle the binding constant of monodentate basic nitrogens changes, depending on the nature of this macrocycle, by several orders (Table 12). The main factors determining the axial coordination process in these porphyrins are: a) schielding of the reaction centre of the tetrapyrrole macrocycle (on one or both sides); b) change of electron density on nitrogen atoms and of charge on Zn(II) cation as the result of Table 12 Ka of sterically hindered monomeric and dimeric Zn-porphyrins and monodentate ligands at 298 K. Zn-porphyrin 32 33 34 35 36

37 38 39 40 41

42 43 44 45 46 47 48 49

50 51 52 53 54 55 56 57

Monodentate ligand

Solvent

Pyridine Toluene Isoquinoline Toluene Pyridine Toluene Isoquinoline Toluene Pyrazole CH2Cl2 Pyridine Benzene o-Xylene 2-Methyl-imidazole Imidazole o-Xylene o-Xylene Pyridine o-Xylene DMFA Pyrazole CH2Cl2 Pyrazole CH2Cl2 Pyridine CH2Cl2 Pyridine CH2Cl2 Pyridine CH2Cl2 Imidazole CH2Cl2 2-Methyl-imidazole CH2Cl2 2-Phenyl-imidazole CH2Cl2 Pyridine CHCl3 Pyridine CHCl3 Pyrazole CH2Cl2 Pyridine CH2Cl2 Pyridine CH2Cl2 Pyridine CH2Cl2 Pyridine CH2Cl2 Pyridine Toluene 4-Methyl-pyridine Toluene Toluene 4But-Pyridine Pyridine CH2Cl2 Pyrazole CH2Cl2 Pyrazole CH2Cl2 Pyridine Toluene-methanol (2:1) Pyridine Toluene-methanol (2:1) Pyridine Benzene Pyridine Benzene Pyridine CH2Cl2

Ka, mol–1 ·l 23300 4.6×105 29700 9.6 ×105 1300 3100 460 930 240 42 180 590 1300 360 1990 1259000 195000000 2550000 14330 91855 2300 102 105 4100 960 16000 62000 730000 4500 660 690 2700 3200 1400 15000 1380

Reference 46 46 46 46 39 35 38 38 38 38 39 39 47 47 48 48 48 48 49 49 39 40 40 47 47 42 42 42 43 39 39 41 41 35 35 50

128

N.Zh. Mamardashvili, V.V. Borovkov, G.M. Mamardashvili, Y. Inoue and O.I. Koifman R8

R1

R R7

R2 N

N Zn

N

N N

Zn

N

R

R6

N N

R3 R5

R

R4

22, 23 22 : R1=R2=R3=R4=R5=R6=R7=R8=H R

23 : R1=R3=R5=R7=CH3, R2=R6=C2H5, R4=R8=(CH2)2COOCH3

20, 21 20 : R = H,

21 : R = CO2CH3-m

R1 R2

R

R1 R2

N N Zn N N

R1

N

CH3OOCCH2CH2

O

CH-m 27 : R = C2H5, R2= C 28 : R = (CH2)2COOCH3 , R2= C

CH2CH2COOCH3

29-31

24-28 24 : R1=C2H5, R2= COOCH3-m 25 : R1 = C4H9, R2 = OCH3-o O 26 : R1 = , R2= C

Zn N

N

R1

R

N

CH-m

29 : R= CH(OH)CH3 30 : R = CH2CH3 31 : R = CH=CH2

CH-m

deformation of the macrocycle; c) correspondence of the size of the cavity formed by sterically hindered porphyrin macrocycle to the size of extra ligand; d) emergence of additional interactions of the type of π−π, CH−π and others between metalloporphyrin (host molecule) and ligand (guest molecule). Coordination of axial ligands on picket-fence, strapped and dimeric porphyrins has its own features. The term “picket-fence” is used for porphyrins with peripheral bulky substituents. In most cases, these are derivatives of tetraphenylporphin containing an ortho-substituent in each benzene nucleus. Depending on the orientation of substituents relative to the plane of the tetrapyrrole macrocycle (α, on one side; β, on the other side), four atropisomers are distinguished: α,α,α,α- (α4), α,β,α,α-, α,α,β,β- (cis-α2) and α,β,α,β (trans-α2). Works by Imai and Kyuno investigated the axial coordination on picket-fence metalloporphyrins (32, 33) of pyridine, piperidine and quinoline in toluene and other noncoordi-

129

Complexation of Porphyrins with Ions and Organic Molecules t-Bu

O

t-Bu O t-Bu

O

NH

NH O

NH

t-Bu N N Zn H N

NH

NH

t-Bu O

t-Bu

O

N N N Zn N N

NH

NH

O

R

NH R

33

32

O

O Et

Et N

O

N

O OH Et OH O

O

R1

R1

HO HO

OH

N Zn N

OH Et

O

OH

O

O

O 34

R1

O

R1

O O OH

X O

Bu

Bu N N

Zn

O

N

O

N

O

Et

RO RO N N

Bu

Bu

Zn

O O

N N

Et

35, 36

Et

Et

37,38 35 : X = OCH3

36 : X =

37 : R = COCF3 38 : R = H

CH2

CH2 OCH3

nating solvents [46, 49]. The results obtained showed trans-α2 atropisomers to have a higher affinity to pyridine and quinoline than α2 atropisomers. As in the case of trans-α2 atropisomers both sides of the reaction centre of porphyrin are equally accessible for coordination, an increase of affinity to pyridine and quinoline is

130

N.Zh. Mamardashvili, V.V. Borovkov, G.M. Mamardashvili, Y. Inoue and O.I. Koifman O

O

N

N O

N

O

N

Zn

O

O O

N

O

Et

O

O

Et

O N N

N

N

Zn

N

O Et

Et

40

39 C5H11 C5H11

C5H11 C5H11

C5H11

C5H11

C5H11 O O

O

O

R

O

O

R

NR

O O

O

O

O

O

O

O

O O R

R

C5H11

R

N N Zn N N

N Zn N N

43, 44

41, 42

O

R=

NH

Et

41 : n = 1, 42 : n = 5 43 : n = 1, 44 : n = 5

(CH2) n

O

Et

45 : X=

O

O

O N

Zn

N

X Et Et

X

Et

N

Zn

O

N

Et

O

48 : X= O

O

O O O

Et

45-48

O

Et

47 : X=

N

N

O

46 : X=

N

N

N O

O N O

O O

determined by the emergence of weak CH-π interactions in the system considered, which is the cause of higher binding constants. In α4 atropisomers 32, both sides of the reaction centre of porphyrin are inadequate. Four ortho-substituents shield the reaction centre almost completely, and the ligands are coordinated only by the “open” side of the porphyrin macrocycle, without forming any additional bonds. .

Complexation of Porphyrins with Ions and Organic Molecules R

N

( )n

Zn

N

N

R R N N

R

N

N

R R

O

.

R N

(

N Zn N

)n

Zn

N N

RO

OR OR

RO N N

N Zn N

R

49 - 51

, n=2 O 50 : R = (CH2)2COOCH3 , n = 2 51 : R = (CH2)2COOCH3 , n = 4 49 : R=

131

52, 53 52 : R = H,

53 : R = COCF3

Strapped porphyrins include compounds, in which two diametrically opposite peripheral positions of the macrocycle are linked by bridges of various nature [51, 52]. The aim of the synthesis of belted porphyrins is to develop a diversity of structures modelling enzyme systems. The binding constant of strapped porphyrins is first of all determined by the shielding of the porphyrin site (on one or two sides) and by the correspondence of the size of the intramolecular cavity of metalloporphyrine to the size of coordinated ligand. If the bulky substituents do not shield the reaction centre (as in the case of 34), then Ka of such a porphyrin does not in practice differ from the respective value of a sterically not shielded porphyrin (Table 12) [53]. If the bulky substituent shields the reaction centre on one side, the binding constant depends primarily on the correspondence of the size of the intramolecular cavity of metalloporphyrin to the size of coordinated ligand. If there is no such a correspondence, the ligand is coordinated by only the “open” side of the strapped macrocycle. Herewitth, the binding constants of metalloporphyrins with various ligands can be lower than those of porphyrins without steric shielding, which is observed, e.g., in the case of porphyrins 34--40 (Table 12). The authors of [54--58] give no unambiguous explanation to these data. All investigated metalloporphyrins (35--40) with lower binding constants, occurring in the linking bridge, have aromatoc fragments, which, in our opinion, by entering into π – π-interaction with porphyrin macrocycles, lead to a decrease of charge at the Zn cation of the reaction centre. This is one of the main causes of lower Ka in porphyrins (35--40) as compared with respective metalloporphyrins without “shielding” substituents. If the size of the ligand corresponds to that of the intramolecular cavity of sterically shielded porphyrin, i.e., the ligand, by being coordinated at the Zn cation, can incorporate into the intramolecular cavity, in this case the binding constants increase significantly. Complexation studies of Zn-porphyrin-calix[4]arene conjugates 41--44 with organic molecules of various size and nature by the method of spectrophotometric titration showed complex 41 with the least distance between the porphyrin and calix[4]arene fragments to possess the greatest capability of axially binding N-methylimidazole and pyridine (Table 13) [34]. Complex 41 has four binding links, which, due to the strengthening of the molecule structure impart the hydrophobic cavity of the calix[4]arene fragment with a fixed shape, and organic molecules are bound depending on their size. Large-size molecules (nicotine, nicotinamides and 4-phenylpyridine) are bound on the ooo

132


Table 13 Ka (·103 l/mol) of nitrogen-containing ligands by porphyrin-calix[4]arenes (41–44) and Zn-tetraphenylporphyrin (Zn-TPP) in chloroform (cZn-porphyrins ~ 5·10–6 mol/l). Ligand

41

42

43

44

Zn-TPP

147

15

18

6.8

0.9

233

46

80

20

1.6

0.4

10

52

13

1.4

1077

140

290

79

1.5

0.7

0.3

62

19

0.3

0.4

13

15

1.6

0.3

0.2

16

20

41

1.4

N

N

N

N N O NH 2 N O N H

2

N

N N

outer side of the receptor with the binding constant equal to or even smaller than in Zn-TPP. Binding of smaller-size molecules (N-methylimidazole, pyridine, picoline) strongly depends on the shielding effect of the calix[4]arene cover and on the electron density at the donor atom of nitrogen. The presence of long and flexible binding links in receptor 42 leads


133

to the formation of a large-size cavity, and most molecules of the host are bound on the inner side of the receptor (excluding nicotine). Simultaneously, the shielding effect of the calix[4]arene “cover” is attenuated. The decrease of the number of binding links in complex 43 to two leads to an increase of the molecule’s flexibility and, as a consequence, to an increase of the distance between the porphyrin and calix[4]arene fragments. This enables the molecule of 4-phenylpyridine to penetrate into the inner cavity of the receptor. Porphyrin-calix[4]arene 43 binds nicotinamide 200 times stronger as compared with Zn-TPP and is the best receptor in the sequence of complexes 41--44 for binding nicotinamide, 4-phenylpyridine and isonicotinamide. An increase of the length of the binding links in compound 44 leads to a decrease of the shielding effect of the calixarene cover as compared with complex 43. In compound 44, the flexibility of the binding links enables the molecule to “fold” the binding links with a shift of the porphyrin and calix[4]arene fragments relative to one another. This decrease ofthe inner cavity leads to a decrease of the complexing ability of receptor 44 with small molecules as compared with complexes 41--43. An exception is nicotine, whose efficient binding can, probably, be explained by the additional formation of hydrogen bonds with amide groups of the binding links. In cyclophane dimers 45--53, each monomeric porphyrin fragment is also shielded on one side owing to the other porphyrin fragment, as the result of which Ka of the ligands by dimers 45, 47, 48 and 51 is lower than of their monomeric analogues. In coordination by 47, 49 and 50 of monodentate ligands to form complexes of composition 1:1, the binding constants are observed to increase as compared with monomeric analogues (in the case of the coordination of pyridine, Ka increases two- to threefold). Some authors associate this with electronic effects [34, 54]. In their opinion, pyridine, along with the formation of a σ-bond Zn--N, interacts with metal by the π-type with the transfer of electron density from dxz, dyz orbitals of Zn(II) to vacant π*-molecular orbitals of pyridine. In passing from monomeric porphyrin to dimeric one, the probability of this interaction increases due to the enhancement of the electron-donor properties of the deformed porphyrin macrocycle, which ultimately is accompanied with an increase of Ka of the pyridinate axial complex of cyclophane dimer. According to the opinion of other authors [59], an increase of the binding constant of one ligand by dimer as compared with monomer, as in the cases of spatially hindered porphyrins, is determined by the correspondence between the size of the inner cavity of dimeric porphyrin and that of the ligand. The better this correspondence, the greater the binding constant is. In the sequence of para-substituted pyridins, the best correspondence has para-tertbutylpyridine (Ka increases 50-fold). The second extra ligand is attached only from the outer side and its binding constant is much weaker. In coordination by cyclophane dimeric porphyrins of bidentate ligands (with two nitrogen atoms), if the geometric requirements of the interplane space correspond to the size of the ligand, a dimeric porphyrin--bidentate ligand complex of composition 1:1 is formed, in which the bidentate ligand is located in the inner cavity of dimeric porphyrin and had two binding points. Conditions of forming various-composition complexes, their stability constants and other regularities are described in works by Sanders by example of such ligands as pyrazine, 4,4′-dipyridyl and DABCO (1,4-diazo-bicyclo[2,2,2]octane) [53, 56, 60]. Zn-porphyrin with monodentate ligand forms a complex of composition 1:1 with the binding constant Ka (Scheme 1). In the case of bidentate ligand, a 1:1 complex is formed with the binding constant Ka,1 ≈ 2Ka (Scheme 2). Addition of the second ligand to form a 2:1 complex occurs

134


Scheme Scheme1 1 L N N

Zn

N N

Ka

+ L

N N

Zn

N N

Scheme 2

Scheme 2 N N

Zn

L

N

N

Ka,1

N

+

N

Zn

N N

L L L

Scheme Scheme 3 3 N N N N

Zn

N N

+

Zn

N

N

N

N

N N

L

Ka,2

L

Zn

L

L N N

Zn

N N

with Ka,2, which is related to Ka,1 as 4Ka,2 = aKa,1, where a is the “interaction” parameter, being a measure of interactin between two reaction centres (Scheme 3). In the case of bifunctional ligand and dimeric porphyrin, the situation becomes more complicated. Complexes can be formed both on the inner (Ka (in)) and outer ((Ka (out)) side of the dimer. The strongest complexes are formed in the case, when bifunctional ligand is coordinated by the inner cavity of dimer. Formation energy of this complex in the general Table 14 Ka of Zn-porphyrins with bidentate ligands. Zn-porphyrin

Ligand

Solvent

Ka, mol–1 ·l

Zn-monomeric porphyrins without steric shielding (binding constants of the complexes of composition 1:1) 20 23 24 25

Ethylene-diamine DABCO DABCO BiPy DABCO Pyrazine

26

Ethylene-diamine

CHCl3 CH2Cl2 CH2Cl2 CH2Cl2 Toluene-methanol (2:1) Toluene-methanol (2:1) CHCl3

18270 240000 16000 5400 5400 500 18970


135

Table 14 (continued) Ka of Zn-porphyrins with bidentate ligands. Zn-porphyrin

Ligand

27 28

BiPy BiPy

Solvent

Ka, mol–1 ·l

CH2Cl2 CH2Cl2

4700 8400

Sterically shielded Zn-monomeric porphyrins (binding constants of the complexes of composition 1:1) 37 38 39 40

BiPy BiPy DABCO DABCO

CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2

970 2.6×103 4.9×104 2.3×104

Zn-dimeric porphyrins (binding constants of the inner complexes of composition 1:1) 45 46 47

48

52 53

DABCO DABCO DABCO BiPy Pyrazine DABCO BiPy Pyrazine DABCO BiPy DABCO BiPy

CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2

7.4×107 1×107 1.3×106 3800 600 2.4×105 1800 690 2.0×105 1.7×104 8.0×107 2.5×106

case is not the doubled energy of forming a complex with monomer, owing to a whole range of factors: the chelate effect, interaction between reaction centres of bidentate ligand, conformation changes, as well as new interactions, which can emerge in the system. These components of the total binding energy of the dimeric porphyrin--bidentate ligand complex depend on the nature and conformations of linking bridges in dimeric porphyrin and can be rather significant. Depending on the nature of a linking bridge, the constant of binding of dimeric porphyrin with bidentate ligand by the two-site scheme can be higher, as compared with the monomeric analogues, from 2 up to 5,000 times (Table 14). Coordination of bidentate ligands by linear dimeric porphyrins depends on the flexibility of the linking bridge. In the case of dimer 54 with the rigid aromatic bridge, porphyrin fragments are fixed in space at a large distance one from another, and complexation with DABCO in CH2Cl2 proceeds by the type of 1:2 interaction with Ka ≈ 105 l/mol (each metalloporphyrin fragment of the dimer binds one molecule of bidentate ligand) [54, 61]. In dimer 55 with a more flexible linking bridge, complexation with DABCO proceeds to form an “internal” complex of composition 1:1 with Ka ≈ 107 l/mol (two metalloporphyrin fragments of the dimer bind one molecule of bidentate ligand). “Internal” complexes with DABCO are also formed by dimer 56 (Ka = 6·105 l/mol) [59, 62]. The authors of [55] note a high affinity to diamine in “flexible” dimer 57 with a crown-ether linking bridge; in the presence of barium perchlorate, the affinity increases. The recognizing ability of calixarene-porphyrin conjugates 58, 59 with respect to DABCO was studied in [19, 56]. It is noted there that compound 59 forms an internal 1:1

136

N.Zh. Mamardashvili, V.V. Borovkov, G.M. Mamardashvili, Y. Inoue and O.I. Koifman OH

N

N

N Zn N

N Zn N

N

N

54 O

O O

O

N

N

N Zn N

N Zn N

N

N

55 Ph N Zn N N

N N N

Zn

N N

O O O

N N

Zn

N

O

Ph

O

O

N

Ph N

56

Ph

N 57

N Zn N

Ph

Ph

complex, whereas 58 coordinates DABCO only on the outer side. A significant difference of the complexing ability of the calix[4]arene derivative 58 from thiacalix[4]arene derivative 59 is explained by the authors by differences in the size of respective calixarene cavities. A larger cavity in conjugate 59 corresponds better to the geometry of bidentate ligand containing two diametrically located nitrogen atoms, which leads to the formation

137


O

t

Bu

N N Zn N N

NH

O XX

t

Bu

t

Bu

O

R

o

R

X X t

Bu

O

N N Zn N N

NH

O

58-59 58 : X=CH2,

59 : X=S

R

R

N N Zn N N

.

R

R

R

60-62 60 : R = CH2CH3, n=3 61 : R = (CH2)2COOCH3, n=3 62 : R = (CH2)2COOCH3, n=4

R SiMe3

n

N N Zn N N Me3Si

R

R 63

n=4, R=CH2CH2COOCH3

of an intramolecular 1:1 complex with the binding constant Ka = (1.0±0.1) ·107 l/mol in chloroform at 294 K. Conjugate 58 containing the usual calix[4]arene fragment forms a complex by way of independent coordination of DABCO molecules by two tetrapyrrole macrocycles (each metalloporphyrin fragment coordinates a ligand molecule) to form a 2:1 complex. A detailed investigation of binding oligopyridyl ligands by di-, tri- and tetrameric porphyrins was carried out in [59--62]. A good geometrical correspondence of the size of porphyrin cavity and the size of ligand is observed in dimeric porphyrin 50 with BiPy (6 ·106 l/mol), trimeric porphyrins 60, 61 with Py3P (9 ·109 and 4·1010 l/mol, respectively), tetrameric porphyrin 62 with H2Py4P (2 ·1010 l/mol), which is slightly larger than in linear tetramer 63 with this ligand (7·109 l/mol) [43]. Dimeric, trimeric and tetrameric porphyrins synthesized according to an improved method have even more capacious intromolecular cavities than cyclophane dimers considered above. The distance between porphyrin fragments in dimer 51 is approximately 1.5 nm. This enables the zinc complex to be the host

138

N.Zh. Mamardashvili, V.V. Borovkov, G.M. Mamardashvili, Y. Inoue and O.I. Koifman R

R N N Zn N N R

R

n

64 n=3, R= CH2CH2COOCH3

N

N

N

N

Py2C2

BiPy

N

N

N

N N

N

N

N

Py3P

N

N

N

NH N N

HN

Py3C12H3

N

N

H2 Py4P

for a guest of respective geometry -- bis(4-pyridyl)ethine (Py2C2) (7·106 l/mol). Trimer 64 has a good geometric correspondence with Py3C12H3 (2·108 l/mol) [59]. Of other metalloporphyrins, Ni- and Ru-porphyrins are used the most widely as porphyrin receptors to nitrogen-containing ligands. The binding strength of basic nitrogens by the central metal atom decreases in the sequence Ru >> Zn >> Ni. The most strong complexes are formed by Ru-porphyrins [63--66]. In the case of Zn-porphyrins, molecular com-


H2P

139

H2P

N

H2P

Ru

N

H2P

H2P

H2P

65 +

N

R N

N +

N

N

N

+

N

M N

Zn N

N

R

N

67-69 67 : R=H, 68 : R=OH, 69 : R=OCH3

+

N

66 : M=H2, Zn, Co

plexes are formed based on two or three (maximum four) porphyrins; based on Ru-porphyrin complexes, compounds based on five, seven and nine porphyrins were produced. An example of such a compound is Ru-complex 65 [64]. In processes of molecular recognition of amino acids, sugars, quinones and purine bases by porphyrins, of great significance are hydrogen bonds, van-der-Waals and dispersion interactions. Recognition of amino acids in aqueous media was studied in [67, 68] by the example of tetrakis-(4-methylpyridyl)-porphyrin 66 and its zinc and cobalt complexes. In the case of metalloporphyrins, the main driving force of amino acid recognition is extracoordination of the amino group of amino acid to metalloporphyrin. Besides this, a weighty contribution is introduced by the interaction between the molecule of porphyrin and amino acid (ligand--ligand interaction) and dispersion interactions between carbonyl anion of the guest and pyridinium cation of the host. The value of dispersion interactions depends on pH [68]. Ligand--ligand interaction takes place in the case of amino acids with aromatic substituents phenylalanine and triptophane. This π--π interaction between the porphyrin molecule and aromatic moiety of amino acid stabilizes the complex [69], thus significantly increasing the binding constant (from 60 l/mol in glycine up to 1200 l/mol in triptophane).

140


According to [70--74], in organic solvents recognition of amino acids is performed owing to the coordination interaction via metal cation of the porphyrin complex with the simultaneous formation of H bonds at the periphery of the macrocycle. Porphyrinates, in which the hydro group participating in the formation of H bonds is in naphthalene fragments possess a high recognizing ability with respect to methyl esters of α-amino acids [72]. Based on studies of the complexing ability of porphyrins 67--69 forming and not forming H bonds, the authors assessed the contribution of amino acid derivatives into the total energy (∆G°total) of forming ZnP--amino acid methyl ester complexes: 1) contribution of axial coordination energy (∆G°ac); 2) contribution of H bonds (∆G°hb). While ∆G°ac in CH2Cl2 for 68 is ~15 kJ/mol, ∆G°hb is, depending on the nature of amino acid, within the limits of 1--6 kJ/mol. Of esters of aliphatic amino acids, porphyrin has a preference to amino acids with bulky groups. The largest value of ∆G°hb is in leucine (6.2 kJ/mol); the lowest, in glycine (0.8 kJ/mol). Though Ka in aliphatic and aromatic amino acids are commensurable, the high binding constants of aromatic amino acids are determined not by the formation of hydrogen bonds, but by additional π--π interaction (Table 15). Table 15 Association constants observed in binding of esters of amino acids and porphyrinates (68, 69) in chloroform at 288 K. ME of amino acid

Ka (68), l/mol

Ka (69), l/mol

Ka (68)/Ka (69)

Gly-OMe Ala-OMe Val-OMe Leu-OMe

3.46×103 2.23×103 8.07×103 1.09×104

9.15×102 3.29×102 3.51×102 2.72×102

3.8 6.8 23.0 40.0

Complexation of porphyrinates 70–76 with amino acid esters was studied in [74]. Formation of an additional hydrogen bond between the hydroxy group of porphyrinate and oxygen atom of amino acid (complex 77) can be judged by the ratio of the association constants (Ka) of porphyrins with hydroxy and methoxy groups of analogous structure. Thus, for three diphenylporphyrins 71, 73 and 75, the hydrogen bonds with all amino acids studied are formed only when the hydroxy group is in para-position (Table 16). The association constants (Ka′) of complexes with two binding points (donor-acceptor and hydrogen bonds) differ from the association constants (Ka′′) of complexes with one binding point (donoracceptor bond) two- to fourfold. As seen from the data of Table 17, the energy of axial coordination for all systems studied is within the limits of 22–24 kJ/mol, whereas the largest value of hydrogen bonds, which are formed in Zn-porphyrin–amino acid methyl ester X H3C H9C4

H

CH3 N

C4H9

N Zn

N

H9C4

N

C4H9 CH3

H3C

70 : X = H 71 : X = OH-p 72 : X = OMe-p 73 : X = OH-m 74 : X = OMe-m 75 : X = OH-o 76 : X = OMe-o

R

NH2 HO

Zn

77 X

70-76

CO2CH3 OH

141


complexes, are ~4 kJ/mol. The number of phenyl fragments with hydroxy groups also affects the value of Ka, but to a lower degree. The largest differences are observed in cases with the symmetric and nonsymmetric arrangement of phenyl rings. Table 16 Association constants observed in binding of esters of amino acids and porphyrinates (72--77) in toluene at 293 K. Zn-porph. (1)

Ka, l/mol

Zn-porph. (2)

Ka' /K''a

Ka, l/mol Gly-OMe

70 71 73 75

1750 6270 2730 1680

72 74 76

1420 1430 1400

Ka(71)/ Ka(72) Ka(73)/ Ka(74) Ka(75)/ Ka(76)

4.4 1.9 1.2

Ka(71)/ Ka(72) Ka(73)/ Ka(74) Ka(75)/ Ka(76)

2.9 1.8 1.2

Ka(71)/ Ka(72) Ka(73)/ Ka(74) Ka(75)/ Ka(76)

2.5 1.2 1.0

L-Ala-OMe 70 71 73 75

1020 2660 1680 1080

72 74 76

915 920 910 L-Leu-OMe

70 71 73 75

1660 2850 2480 1390

72 74 76

1120 2040 1380

Table 17 Energies of hydrogen bonds (∆Ghb), axial coordination (∆Gac) and the total free association energy (∆Gtotal) of porphyrin–amino acid ME in toluene. ∆Gtotal, kJ/mol

∆Gac, kJ/mol

∆Ghb, kJ/mol

∆Gtotal, kJ/mol

Gly-OMe 72 74 76

–27.2 –24.7 –23.6

–23.2 –23.2 –23.1

∆Gac, kJ/mol

∆Ghb, kJ/mol

∆Gtotal, kJ/mol

L-Ala-OMe –4.0 –1.6 –0.4

–25.1 –23.6 –22.5

–22.1 –22.2 –22.1

∆Gac, kJ/mol

∆Ghb, kJ/mol

L-Leu-OMe –3.0 –1.4 –0.4

–24.9 –24.5 –23.0

–22.6 –24.0 –23.0

–2.3 –0.5 0

∆Gtotal = −RT ln Ka (ppm−1 ) ,

(2)

∆Ghb = − RT ln( Ka′ / Ka′′ ) ,

(3)

∆Gac = ∆Gtotal − ∆Ghb .

(4)

The effect of hydroxyl groups, which are at both sides of porphyrin receptor 78, on its recognition properties to amino acides is considered in [75[. The largest binding constants (6.9·104 l/mol) are observed in the case of aspartic acid dimethyl ester with two –COOCH3 groups, whereas in leucine methyl ester the binding constant is four times as small (1.8·104 l/mol).

142


N

N

N NH2

N OH

OH

OH N N

Zn

OH

N N

N

N

OH OH

N Zn N N

78

N N R

NH2 N

79

The recognizing ability of porphyrin receptors with respect to purine bases was studied by the example of Rh(III) and Zn(II) complexes. Rh(III) complexes exhibit a very strong affinity to nucleobases, especially to adenine. For instance, meso-tetraphenylporphyrin Rh(III)·Cl has Ka in CH2Cl2 with 9-ethyladenine of the order of 107 l/mol, which in practice indicates the irreversibility of this process [76]. 9-Ethyladenine is coordinated on Rh(III) of the metalloporphyrin fragment by N-1 atom, which is indicated by X-ray data [71]. Unlike Rh(III)-porphyrins, the coordination process of nucleobases on Zn(II) porphyrins is a reversible process. Ka of Zn(II)porphyrin--purine base complexes of composition 1:1 are within the limits of 102 -104 l/mol [77]. Table 18

Constants of binding (Ka, l/mol) of sugars by porphyrinates.

Porphyrin

Sugars

Ka, l/mol

Solvent

T, K

Reference

37 37 37 37 76 76 76 76

Octyl-β-D-mannoside Octyl-α-D-mannoside Octyl-β-D-glucoside Octyl-α-D-glucoside Octyl-β-D-mannoside Octyl-α-D-mannoside Octyl-β-D-glucoside Octyl-α-D-glucoside

4.8 ×103 6.4 ×103 1.4 ×103 8.5 ×102 2.2 ×103 3.0 ×103 1.6 ×103 2.9 ×103

CH2Cl2 CH2Cl2 CH2Cl2 CH2Cl2 CHCl3 CHCl3 CHCl3 CHCl3

297 297 297 297 288 288 288 288

78 78 78 78 80 80 80 80

Porphyrins with two hydroxyl groups on one side of the macrocycle form complexes 79 with adenine; in these complexes, one of the adenine molecules is coordinated by cation of the tetrapyrrole fragment, and the other by the hydroxyl groups of the naphthalene fragments. Such porphyrins are called “bidentate”, and the binding constants of such complexes are several orders higher [77]. Complexes of sugars with Zn-porphyrins containing two nitrogen atoms at the periphery of macrocycle 80 were studied in [78--83]. The hydroxyl groups of sugars were found to be involved in the formation of bonds with: 1) Zn cation of the porphyrin’s reaction centre; 2) nitrogen atoms of the aryl substituent. The binding constants of these complexes are presented in Table 18. Besides Zn-porphyrins, some porphyrin ligands also possess a recognizing ability with

143

Complexation of Porphyrins with Ions and Organic Molecules OR

δ− N

+ δH

δ−

− δO

O

O

δ+H

O

O

H

H

δ− N

N N δ+Zn N

N

80

N NH HN

OR OR

N

OR OR

Br N X X Y X N N X N X

OR OR OR OR

Br

NH N N HN

N

Br

X X Y X N N X N X NH N N HN

Br

X = (CH2)6 , Y = CO(CH2) 4CO

82

81

N

N N

M

O

O

O

S

O

O S

O

O

N O

O N N

M

N N

O

83 : M= H2, Zn

respect to sugars [84-86]. Thus, Kral et al. made use for this purpose of such porphyrin receptors as macrocycles with 1,1′-binaphthosubstitution 81 or cyclic cryptand-porphyrin conjugates 82 [84--86]. The latter are more efficient in aqueous media. Increased interest in triple acceptor--donor--acceptor systems for studies of particular stages of photosynthesis led to a large number of works on quinone-substituted porphyrins. Of special interest are model compounds with well-defined geometry. Thus, in [87] it was found that the conformational arrangement of the quinone fragment in porphyrin 83 depended on the presence or absence of Zn cation in the tetrapyrrole macrocycle. In the presence of Zn, the quinone fragment is immediately over the reaction centre of porphyrin and enters into a donor-acceptor interaction with it. Porphyrins with four hydroxyl groups possess a recognizing ability with respect to various quinone substituents,, irrespective of whether it is metalloporphyrin or a free base.

144


R1

R1

R1

OH OH OH

NH N

N

-

OH

SO3

N -

SO N3

-

SO3

N 2+ Zn N N

N

R

+

-

SO3

HN

+

N+ R

+ N

84

R

85 : R1= O

O

,

R=

Especially high values of binding constants are observed in the case of porphyrin 84 and methoxy-substituted quinones. Ka for 2,3,5,6-tetramethoxy-4-benzoquinone is 3.5·104 l/mol [88]. As an example of supramolecular porphyrin-based complexes formed only at the expense of dispersion interactions, we can name complexes formed by tetracations of Zn-porphyrins and tetraanions of calix[4]arenes [89--90]. The high strength of complexes formed is indicated by high binding constants -- Ka of complex 85 in CHCl3 is 1.4·107 l/mol [89].

3

Complexation of Porphyrins with Organic Molecules: The Chirality Aspect

In the recent years, among the great diversity of host--guest interactions, special attention is attracted to complexation processes of porphyrins and related chromophores with organic molecules possessing a chirality [22, 91--116]. This section considers the main principles, mechanisms, driving forces and factors, which enable efficient control over the supramolecular systems formed. 3.1

Host–guest systems based on monomeric porphyrins

Two major types can be singled out among host--guest systems based on monomeric porphyrins: 1) those including achiral porphyrins or racemates (optically inactive compounds of equimolar amounts of antipodes); 2) those including porphyrins possessing a chirality. In the former case, there occurs an interaction between the host molecule of optically inactive porphyrin and the optically active guest molecule, which can be registered by various spectral methods of analysis. Despite the apparent simplicity of systems based on monomeric porphyrins, they can possess some very interesting properties. For instance, a number of highly substituted porphyrins (86 – 88) having a saddled shape owing to steric repulsion between adjacent substituents, exist as two enantiomers, which, being mutually converted, form a racemate mixture [117, 118]. It has been found that this equilibrium can be shifted towards one certain conformation (more than 98%) at the interaction with various enantiopure acids via the formation of hydrogen bonds between carboxy groups and nitrogen atoms of pyrrole fragments of the porphyrin cycle to form a 1:2 complex (one porphyrin and two


R3

Me

R1

Me

Me NH

Me

N

R2 N

NH

N

R

R

R N

HN

Me

145

HN

Me Me

R1

Me

R1

86-88 86 : R = Ph(OMe)2 -o,o; R1=R2=R3=Ph 87 : R = R2= Rh(OMe)2 -o,o; R1=R3=Ph 88 : R = R1=R2=Ph(OMe)2-o,o; R3=Ph

89 R=PhB(OH)2-o, R1 =Ph

Et

Et

Et

Et N

N

N

Zn N

N

N

Et

Et Et

90

Et

acids). These supramolecular complexes have a rather large signal of induced circular dichroism (CD) in the region of the Soret band, whose sign correlates with the relative value of substituents at the asymmetric centre and makes it possible to determine the absolute configuration. Although the authors do not discuss the electronic nature of observed optical activity, induced chirality exhibits the properties of memorizing the optical activity in the replacement of chiral acid by achiral acid. Another interesting example is the porphyrin sensor for D-lactulose, whose selectivity is determined by the correspondence in the arrangement of the reaction centres of sugar considered and two fragments of boric acid in cis-porphyrin 89, which leads to the formation of a 1:1 complex with a CD signal of one sign in the Soret band region [119]. The works [79, 120, 121] developed and investigated the complexation properties of porphyrin-containing sensors for saccharides 68, 78, 90, 91 and alkaloids 92. We would especially like to mention host--guest systems, which contain amino acids being natural building blocks. Thus, using trans-isomers of porphyrins 68 and 90, the authors of [75] developed a receptor for amino acids with a high selectivity with respect to aspartic acid methyl ester (the binding constant corresponds to 6.98·104 M –1). The conformational correspondence in this case creates conditions for three-point binding (one coordinational interaciton and two interactions to form hydrogen bonds). In all systems studied, a negative-positive CD signal of small intensity was found in the region of the Soret band, induced by electronic interaction through the space between porphyrin and carbonyl group of amino acid, though the influence of other factors is not to be ruled out, either. To expand

146

N.Zh. Mamardashvili, V.V. Borovkov, G.M. Mamardashvili, Y. Inoue and O.I. Koifman Ph

O(CH2)3OH Me

O(CH2)3OH N

HO(H2C)3

N

N

Ph

N

Et

PhOCH2COR-o

Zn

M

N

N

N

N

N

Me Ph

(CH2)3OH

Et

92 : R = OH, NH2

91 : M = Lu(NO3), Gd(NO3)(Cl)

R1

XOOC(H2C)n

N R1

R

M

XOOC(H2C)n

N

N

COOX

93 : M = Zn, R = R1 =

N

X = Me or K, n=1,4,10

R1 93-95 O

X Y

94 : M = Gd(acac), R = R1 =

O

O O

O

X

O

X = t-Bu, Me, OMe, F, Cl, Br; Y = H, OMe NHCO 95 : M = Er(acac), Gd(acac), Yb(acac); R =

,

R1 = Ph

the potentialities of using receptors for amino acids, the work [122] proposes a number of porphyrins (93), which in nonpolar organic solvents can be used as esters, and in aqueous media as products of their hydrolysis. It has been found that two factors contribute to the complexation: electrostatic interactions in coordination of the amino group of amino acid and Zn ion of the porphyrin cycle in organic solvents (enthalpic forces) and host--guest dispersion interactions (enthalpic force) in combination with the desolvation-controlled

147


Ph

Br

PhSO3H-p

PhOMe-o

Br NH

N

N

p-HO3SPh

PhSO3H-p N

HN

PhSO3H-p 96

Ph

Br

N

Ph

Zn N

N Br

Br Br

Ph

Br

97

process of complexation (entropic force). The latter predominate in water. In [123], recognition of the optical activity of zwitterionic amino acids was studied in a two-phase organic solvent--water system. For extraction from the aqueous phase, use was made of porphyrins 94, which formed 1:1 complexes with amino acids. Their CD spectra have two well-resolved opposite-sign signals, which correspond to a split Soret band and absorption bands in the visible range of the spectrum. The value of CD molar amplitude (A) strongly depends on the nature of solvent and the nature of substituents of phenyl rings. The largest sensitivity was found for aromatic solvents and substituent X=t-Bu, and the sign of induced CD is determined by chirality of amino acids. This method is proposed by the authors for determining the absolute configuration of α-amino acids. The sensitivity of this method was subsequently increased by including a crown ester fragment into porphyrin molecule 95 [123]. Simultaneous coordination of the CO2– 2 group of amino acid by the lanthanide reaction centre of porpyrin and the NH3+ group by the crown ester fragment in combination with the respective choice of metal significantly increases the extraction capacity and value of A. In [125], recognition of amino acids was done using the method of solid-phase optical detection. It is based on hypsochrome shifts in the visible spectrum of porphyrin 96 immobilized as a monolayer on a cellulose film, which occur owing to the interaction with amino acids. The extent of spectral changes depends on the structure of amino acids, which makes possible quantitative assays of amino acid in solutions. In the literature, there is an large body of information on supramolecular systems, in which both the guest molecule and the host molecule possess a chirality. On the whole, the effect of chiral recognition is based on the principle of three-dimensional correspondence according to the lock--key principle, when the complexing centres of the guest and host well correspond to each other, and usually requires the presence of three sites for respective interactions. Ideally, of two enantiomeric forms of the guest molecule, only one form would be efficiently bound by the respective host molecule to form a stable complex, whereas the other enantiomeric form woudl be bound weaker. This difference in binding can be established by various spectral methods. For instance, two diastereomeric complexes of porphyrin 97 and (S)-2-pyrrolidinemethanol were easily identified in [126] by the method of 1H NMR as a consequence of the stabilization of the optimal complex due to coordination and formation of hydrogen bonds. In another case, a direct correspondence was established between enantioselectivity in complexation and catalytic oxidation [127]. Thus, at the substitution of Fe by Al, 1,1′-binaphthyl derivative of porphyrin 98 was found to have clear distinctions in the electronic

148

N.Zh. Mamardashvili, V.V. Borovkov, G.M. Mamardashvili, Y. Inoue and O.I. Koifman O R

N H

O NH

N Cl Fe

N R= N

N

MeO MeO

HN O H N

R O

98

Ph

N Ph

N O

R

Zn

P

N

N

O

O 100 : R =

Ph 99-101

SiMe3 O O

99 : R =

P

O O O

SiMe3

P

O

101 : R =

absorption spectra in antipodal binding of epoxides and alcohols; these distinctions occur against the background of the linear dependence between respective differences in free energy for complexation and catalytic activity, which indicates the similar stereoselective mechanism for both processes. Binaphthyl fragments were also used to obtain optically active phosphite-containing porphyrins 99--101. They, in combination with various phosphorus-containing chiral and achiral ligands, were applied as supramolecular bidentate


149

102 : M=2H, R=R1=R2=R3= L-ValOMe or L-ThrOMe or L-TrpOMe

NHCOR3

103 : M=Ru(O2), R=R1=R2=R3=C(OMe)(CF3)Ph R2CONH N

105 : M =Zn, R=R1=Ph, + R2=R3=CH2N Me3

N M N

N

NHCOR 106 : M = Zn, R=R1=R2=R3=CH(NHBoc)Me

R1OCHN

107 : M = Zn, R=R1=R2=R3=CH(NHBoc)Pro 108 : M = Zn, R=R1=R2=R3=CH(NHBz)Gln

102, 103, 105-108

O

NO2

NH

N

HN

N

O

Zn N

NH

N

O O 2N

HN O

104

matrices in palladium-catalyzed asymmetric allyl alkylation of racemates of 1,3-diphenyl-2-propenyl acetate using dimethylmalonate as nucleophile [128]. The enantioselectivity was found to be determined not only by the absolute configuration of porphyrins and ligands, but also by such a structural feature of porphyrins as voluminosity (introduction of SiMe3 groups into porphyrin 101) and by the location of the substituent (ortho- and metain porphyrins 99 and 100). The work [129] investigated amino acid-modified derivatives of 5,10,15,20-tetraphenylporphyn 102, yielding with sugars complexes of composition 1:2. It is noted that these porphyrins exhibit no advantages in complexing ability as compared with the achiral analogue. At the same time, it is reported [130, 131] that the tetraphenylporphyrin derivative possessing a chirality (103) exhibits chiral recognition abilities with respect to amino acid esters. The largest value of enantiomeric excess (66%) was found for methyl ester of valine; what is more, predominantly the L-enantiomer is bound. The respec-

150


tive Zn-porphyrin complex (103) can also efficiently recognize the [S]-antipode of bulky 1-(1-naphthyl)ethylamine with a 2:4 enantiomeric binding ratio [132]. The further improvement of selectivity up to 7.5 (also for valine methyl ester) is achieved in the case of porphyrin 104 by way of respective coordination and hydrogen bonds between host and guest molecules [132]. The thermodynamic analysis revealed a significant change in the guest molecule owing to the coordination process, whereas the enthalpy factor determined mainly by the formation of hydrogen bonds indicates the stabilization of the complex as a whole. Two enantiomers of chiral water-soluble porphyrin 105 were also used for chiral recognition of a number of amino acids and dipeptides in an alkaline aqueous solution. The stereoselectivity changing from 1.2 up to 3.3 is based on a combined action of the effect of coordination, Coulomb and spatial interations [134]. The work [135] describes one of the most efficient ways of recognizing amino acid esters with an enantioselectivity of 21.54 (for methyl ester of phenylalanine) using amino acid-modified porphyrins 106--108 [135]. Enantioselectivity was shown to depend exponentially on temperature, and it was found that diastereoisomeric complexes exhibit various CD properties . In particular, the CD spectrum of porphyrin 106 in the presence of weakly bound methyl ester of L-analine is similar to the spectrum of individual 106 and represents a single negative Cotton effect in the Soret band region. On the other hand, strongly bound methyl ester of D-alanine gives a positive-negative signal determined by exciton interactions between porphyrin and the carbonyl group of amino acids. The importance of the model of three-site interactions for chiral recognition was additionally confirmed by low stereoselectivity (15%) of the Ru complex of porphyrin 109 in the binding of racemate isocyaanides (it is capable of forming respective coordination complexes, but has no additional interactions sites related to the formation of hydrogen bonds) [136]. In [137--141], by the example of chiral porphyrins 110, 111, it was clearly shown using the methods of X-ray diffraction analysis and 1H NMR that a combinaiton of one coordination interactin and two interactions by the type of forming hydrogen bonds provides for efficient chiral recognition of amino alcohols and amino acid esters as in the case of the Co complex of porphyrin 110. Moreover, this porphyrin is suitable for the determination of the enantiomeric excess in primary amines and aziridines by 1H NMR, yielding a linear dependence for a large number of enantiomeric compositions with a relative error of several percent. Porphyrin 110 was also found to feature the dependence of the conforMe R O

R

N

O

N

Me Me

N

M N

O R

O

COOMe

Me

O

O

N

COOMe

M N

O

O

109 : R =

MeOOC

N

N

R MeOOC Me

Me

O

O

Me Me 110 : M =Co, Fe

Me Me

151

Complexation of Porphyrins with Ions and Organic Molecules Me

Me

Me

Et

Me

A HO Me

O

CONH(CH2 )2 NHCOCH2

Me

NH

O

HN

Me N

N

R3 R1

R2

111-113

HO

Et

Me

H N

R

Et

O

Et

Et

Me

OH

111 : R=R3=Et, R1=Me, R2=CH2A 112 : R= A, R1=R2=Et, R3= CH2COPheOBu-t 113 : R= A, R1=R2=Et, R3=CH2COPheOH

mational state of the macrocycle on the nature of the metal’s central atom; the dependence is exhibited in the conversion of α,α,α,α-conformation to α,β,α,β-conformation at the substitution of Zn by Ni. Various porphyrin analogues possessing a chirality are also used for binding optically active compounds. Thus, in [142] complex lasalocid-sapphirine conjugates 111--113 were used to transfer aromatic amino acids from one water phase into another water phase via an organic layer. The proposed reaction mechanism includes the formation of a stable supramolecular ensemble between the conjugate and zwitterionic amino acid owing to the electrostatic coordination of carboxylate anion to protonated sapphirine and the formation of hydrogen bonds between the NH3+ group and polyester ionophore fragment of lasalocid. The above-considered examples of host--guest systems based on monomeric porphyrins clearly show the significance and broad potential of their applications. However, in many cases the sole monomeric porphyrin fragment can not efficiently funciutn without the synergic support of another porphyrin fragment, or several porphyrins. This aspect promoted the development of new chirogenic host--guest ensembles based on dimeric and oligomeric porphyrins, which shall be considered in the next section. 3.2

Host–guest systems based on dimeric and oligomeric porphyrins

Host--guest systems based on dimeric and oligomeric porphyrins are distinguished depending on the presence of chiral fragment, type of covalent bridge and number of porphyrin macrocycles. First, we shall consider examples of ensembles consisting of achiral and racemate bis-porphyrin molecules of the host and chiral molecules of the guest. This type of chirogenic systems was initially used to observe asymmetry in the interacting chiral molecule of the guest. On the whole, the achiral molecule of bis-porphyrin should possess at least two interacting sites and the sufficiently flexible covalent bridge to be capable of taking on a stereospecific three-dimensional conformation induced by the chiral guest molecule. In this case, a supramolecular ensemble possessing a chirality is formed.

152


R N.

R2

N

R

N

Fe

N

R

N

R1 Ce

O R

R R2

N

N N

N

.

R

N R

R

.

N.

N

N

Fe

N

R

R2

N N R1

R

114-115 114 : R = PhB(OH)2-m, 115 : R = PhB(OH)2 - p

116-121 116 : R =R1=R2= p-pyridyl 117 : R =R2= p-pyridyl, R1=Ph 118 : R =R2= p-pyridyl, R1=Ph(OMe)2-m,m 119 : R =R2 = p-(N-CH2PhB(OH)2-p)pyridyl, R1=PhOMe-p 120 : R = p-(N-CH2PhB(OH)2-p)pyridyl, R1=R2=PhOMe-p 121 : R = R2= p-(N-Me)pyridyl, R1=PhOMe-p

The signicance of these two factors is easily understood from the following examples. Thus, double-decker bis-porphyrins 114--121 linked centre-to-centre were developed based on µ-oxo-dimers and Ce complexes to observe the chirality of various guest molecules [143--148]. The association mechanism includes the interaction of the bidentate guest with two complexing groups of adjacent porphyrins, which makes these porphyrins rotate aroung the central axis and, respectively, to turn macrocycles left or right depending on the stereochemistry of the guest molecule. This asymmetric travel leads to the emergence of a significant optical activity (the consequence of interporphyrin exciton interaction) in the region where porphyrin absorbs. Though the electronic nature of induced CD has not been completely explained, these systems were successfully used to observe chirality of saccharides (for 114, 115, 119 and 120), dicarboxylic acids (for 116--118), dianions (for 121) and memorize chirality (for 118). The generated optical activity was excellently stored for three days at 0°C and even for one year at --37°C. Host--guest complexation occurs with a high degree of cooperativity, exhibiting a positive allosteric effect, and CD properties strongly depend on the guest [143, 144] and solvent [148] molecule structure and pH [147]. It is interesting to note that the effect of a change of optical activity controlled by the number of saccharide links was established for porphyrin 119 in the case of maltooligosaccharide guest molecules. An entropically unfavourable face-to-face arrangement of bis-porphyrins can be also preserved owing to the diametrically located covalent bridges as in the case of cryptand-containing porphyrins 82, 122 [86]. These host molecules are also capable of binding various saccharides predominantly by encapsulation. An observed very weak induction of CD in the region of Soret band is a consequence of the inability of considered bis-porphyrins to follow the stereochemistry of the guest molecules, probably, because of the structural features of the linking bridge. More flexible covalent bridges of various structures and lengths were used to ensure the face-to-face orientation in porphyrins 123--125 [149, 150]. bis-Porphyrins 123, 124 with a longer bridge showed a high correspondence to

Complexation of Porphyrins with Ions and Organic Molecules R2

R

R1

R1 NH

R2

R R2

R

N

(CH2)6

HN

N

R1

N

R

N

(CH2)6

(CH2)6

(H2C)6

R1

R1 NH

A

N

CO(CH2)4CO

N

R1 R

A

(CH2)6

HN

N

R1

A

R

N

153

R1 R2

R

R R1

R

122

R

N

122 : R=Pr, R1 = Me, R2=H

R

N

122 : R=Pr, R1 = Me, R2=H

N

R R

A

R1

R R

N N

R R

R R

N

M

R

N

M

R1

R

122

R

N

R R

R1

M

R1

A

R R

N N

R R

123 : M=Zn, R=H, R1=Mes, A = -CO(NHCMe2CO)9NH

-

123-125 124 : M=Zn, R=H, R1=Mes, A= -CO(NHCMe2CO)3NHC(CHPh)CONHCH2CONH(CHPh)CO(NHCMe2CO)3NH125 : M=2H, R=Me, R1=Ph(OMe)2-o,o, A = -OCH2PhCH2O

bis-pyridyl substituted chiral guest molecules of respective length, which leads to the formation of stable inclusion complexes of composition 1:1 with association constants of the order of 2.3·106 M-1. It was found that the helicity of the guest molecule plays a key role in transferring information of chirality and induction of a strong exciton interaction in the reigon of Soret band. bis-Porphyrin 125 with a shorter bridge was used for complexation with α-hydroxybenzeneacetic acid. Herewith, the activity of CD (A =260 cm – 1 M –1) was observed to increase significantly (more than sevenfold) as compared with respective monomer, owing to the conversion of chirality due to the nonplanar structure of the porphyrin ring to that due to the helicity of the ensemble as the whole. The stepwise increase of the concentration of the guest molecule leads to complex changes of CD signals, which reflect a multilevel equilibrium of the complexation process. Although the observed chiral responses have not yet been totally explained, these chiroptical systems are of certain interest as potential sensors of chirality in studies of the stereochemistry of interacting guest molecules. Another structural type of bis-porphyrins is based on combining two chromophore fragments via one covalent bridge. As the mutual arrangment of the porphyrin

154


R

R

R2

R1 N

R1

R1

N

N

R2

R2 R

M

N

R2

R2

N

N

A

M

R

R2

R2

N

N

R1

R

R2 R

R

R1

R

126-133 126 : M =Yb(acac), Gd(acac), R=R2=H, R1=Ph

A

CH2OCH2

128 : M =Zn, R = Me, R1=H, R2=n-C6H13

A

127 : M =Yb(acac), Gd(acac), R=R2=H, R1=Ph

A

CH2O(CH2)2OCH2

129 : M = Zn, R = R2 =H, R1 = Ph(t -Bu) -m,m MeO2C

CO2Me

A

macrocycles becomes less predictable in this case due to the reduced rigidity of the system as a whole, a number of additional factors, such as complexation with the bidentate guest molecule or shortening of the bridge, can bring two porphyrin fragments closer to each other and fixe bis-porphyrin in face-to-face orientation or linear conformation. Thus, series of lanthanide bis-porphyrins 126, 127 linked with ester bonds of various lengths were prepared as potential host molecules for binding polyions of cystine [151]. These receptors exhibit a significant selectivity to size in studies of the optical activity of guest molecules. Thus, the Yb complex of porphyrin 126 with a short bridge efficiently extracts cystine from an aqueous solution to organic phase, whereas cystathionine, homocystine and methionine are extracted weakly. In contrast, porphyrin 127 preferably extracts a longer homocystine. Complexation occurring via the formation of tweezer complexes of composition 1:1 leads to a complex CD signal in the Soret band region, which is significantly increased as compared with respective monomer. The authors of [152] report of a high selectivity to length in binding of chiral diamines by porphyrins 128, 129 with rigid bridges. The dimensional correspondence in the host--guest system is an important condition for efficient complexation with the association constant of the order of 2.4·106 M –1 for the best pair. Induced CD strongly depends on the volume and number of stereogenic sites, peripheral substituents of the porphyrin macrocycle and temperature. For instance, the largest value of A (1340 cm –1 M –1) was increased 1.5-fold by decreasing the temperature to --45°C. The binding link of the type of crown ester significantly enhances the potentialities of controlling the conformation of bis-porphyrin via specific complexation with ions of alkaline metals [153--155]. This effect is beautifully demonstrated by the example of an almost twofold increase of the binding constant (from 2.6·105 M –1 to 4.5·105 M –1) upon addition of Na+ ions to porphyrin (130)/trans-1,2-diaminocyclohexane system of the tweezer type due to the change of geometry of bis-porphyrin. Addition of Na+ ions significantly increases the value of A of induced CD. A similar effect of binding and CD increase was found in [154] for porphyrin 131 at the interaction with N-alkyl substituted 1,2-diaminocyclohexane. Howev-

Complexation of Porphyrins with Ions and Organic Molecules 130 : M =Zn, R = R2=H, R1=Ph A

155

131 : M = Zn, R = R2 = H, R1 =Ph

(OCH2CH2)4O OMe

MeO

A CONH

N

O

O

O

3

N O

NHCO

3

132 : M =Zn, Mg, R = R2 = H, R1 = Ph

A

CO2(CH2)5OCO

133 : M =Zn, Zn/2H, R=R2=Et, R1=H, A = -CH2CH2

er, a larger-size K+ ion is required in this case for a larger crown ester cavity. It is noted that porphyrin 131 is capable of recognizing the chirality of such potassium carboxylates as camphorates and mandelates. What is more, induced chirality of porphyrin 57 in the tweezer conformation was efficiently saved in the substitution of chiral diamine for achiral one in the presence of Ba ions. The authors of [155] note that in 24 h the intensity of CD in this system decreases only insignificantly. Although the discussed systems of the tweezer type exhibit a rather high optical activity due to the chiral spatial arrangement of two porphyrin chromophores, no detailed explanation of electronic and structural factors involved in this phenomenon has been given until now. One of the first attempts to solve this issue in the case of bis-porphyrin systems was made in [156--166] using a more flexible porphyrin 132, which is capable of taking on a tweezer conformation in the complexation with bidentate ligands. It has been found that the emergence of optical activity is based on the stereospecific differentiation of the relative volume of substituents at the asymmetric carbon atom due to the bis-porphyrin molecule taking on a spatially less hindered conformation. It should be also taken into account that other factors, such as formation of hydrogen bonds; presence of heteroatoms, solvent etc., can also affect the geometry of the ensemble as a whole. All this leads to an oriented arrangement of two tetrapyrrole fragments in porphyrin 132, which on the strength of exciton interactions induces a CD signal, whose sign can be used as a tool for determining the absolute configuration of various bidentate guests. In the cases when the relative size can not be determined directly, the conformation analysis is made using calculations according to the method of molecular mechanics with the view of explaining the observed contradiction between the predicted chirality and the one obtained in some cases. Though it is well known that metalloporphyrin chromophore contains two degenerated (or almost degenerated, due to the asymmetry determined by meso-substitution) electronic transitions along the meso5-15 and 10-20 axes, only the exciton interaction of the pair of 5-15 oscillators was chosen for explaining the induced optical activity in porphyrin 132. However, the contribution of 10-20 oscillators due to homo- and heteroconjugation can also play an important role in certain conformations. The works [156--166] also adduce some dependences of induced CD on the nature of the guest molecule, the solvent and temperature.

156


To investigate various aspects of optical activity of supramolecular systems based on bis-porphyrins and to explain the chirality induction mechanism, the works [98, 167--181] studied bis-porphyrin 133, in which porphyrin fragments are linked by an ethane bridge. In contrast with the above bis-porphyrin systems, this host molecule possesses an ability to sense the chirality of not only bidentate but also monodentate guest molecules due to the semi-flexibility / semi-rigidity of the bridge. This ability is provided for by the relatively short but sufficiently flexible C2 chain. As shown by the analysis of the absorption and CD spectra, porphyrin 133 forms with primary amines, by the cooperative mechanism, stable extended guest--host complexes of composition 1:2 with free Gibbs energy within the range of –6.7 to –8.4 kcal·mol –1. The specific chirogenic mechanism for porphyrin 133 in complexation with monodentate guest molecules includes competitive repulsive interactions between two most bulky substituents at the asymmetric centre of the ligand and ethyl groups of an adjacent porphyrin fragment. The most bulky group makes adjacent macrocycles move away one from another, thus generating a directed turn. The latter leads to the emergence of an (average to strong) signal of CD exciton interaction in the Soret band region. The chirality sign correlating with induced helicity makes possible the determination of the absolute configuration of monodentate guest molecules. The authors thoroughly discussed the mechanism of host-guest interactions, which enables a detailed investigation of various external and internal factors affecting the chirality induction processes. For instance, it was shown that the value of A linearly depended on the size of the largest substituent at the stereogenic centre for homologous ligands, thus making it possible to predict induced chirality [167, 168, 170]. The crucial role of solvent as an active part of the supramolecular system on the whole was clearly explained through the thorough analysis of solution--solvent selective interactions in near-boundary regions, where the distinctions between competing chiral interactions are small and the relative volume of a substituent can be formed owing to the interaction with the solvent [175, 176]. Temperature was found to be another important factor for controlling chirogenic processes. According to [171--173], a decrease of temperature, significanlty increasing the values of A, makes possible induction of chirality by binding alcohols, though they are known to possess a low affinity to Zn-porphyrins. However, the binding of alcohols and the ability to generate optical activity were significantly enhanced by the substitution of the central Zn ion in porphyrin by the Mg ion [174]. The authors of [181] note the importance of the phase transition as a factor of controlling the optical activity in porphyrin (133) / monodentate amine complexes, which leads to a significant increase of the anisotropy factors (16--9.1-fold) and solid-phase “switchover” of chirality owing to the formation of intermolecular aggregates of opposite helicity. Interaction of porphyrin 133 with bidentate guest molecules leads to the formation of extremely stable tweezer complexes of composition 1:1, with binding constants > 107 M –1. A further increase of the concentration of the bidentate guest shifts the supramolecular equilibrium towards the extended complex of composition 1:2, which makes it possible to study the respective stoichiometric effect [177--180]. In the case of enantiomeric 1,2-diphenylethylenediamine, an excellent chirality-switchover phenomenon was revealed, which is controlled by the stoichiometry of the supramolecular system as the result of the opposite mutual spatial arrangement of 1:1 and 1:2 complexes. A more detailed explanation of induced optical activity in supramolecular systems based on bis-porphyrins was first achieved using chiral 1,2-diaminocyclohexane. In particular, based on the obtained crystallographic structure of the (133) / (R,R)-1,2-diaminocyclohexane complex and the Кuhn--Kirkwood oscillatory interaction mechanism, CD was explained as a combination of two B||–B|| and B⊥ –B⊥ electronic homo interactions, which are oriented counterclockwise and are exhib-


157

ited as an intensive negative--positive CD signal (A = –590 cm –1 M –1), whereas the contribution of respective B|| –B⊥ hetero interactions is insignificant [180]. Thus, the high sensitivity, complete understanding of the mechanism of the process and broad applicability for various types of guest molecules show that porphyrin 133 can serve as a potent tool for studies of various aspects of supramolecular chirogenesis and for using as a universal sensor of chirality. Another bis-porphyrin with one bridge (134) was proposed for selective binding of oligosaccharides by the cooperative mechanism due to the use of advantages of the increased rigidity of the system considered [147, 182]. In particular, a fixed porphyrin--porphyrin distance matches well the size of maltotetraose, which leads to the formation of a stable 1:2 complex. This is reflected as a complex chiroptical response in the Soret band region, which response consists of three nonequivalent and asymmetric Cotton effects, probably, due to the superimposition of two or more CD signals. The further increase of the number of porphyrins in respective meso--meso linked oligomers (135) led to a significant increase of induced chirality [183]. These oligomers form heliciform structures owing to the fixation of meso-aryl substituents by way of forming hydrogen bonds between carboxyl groups of porphyrin and cyclic molecules of urea. The direction of the helix can be controlled by complexation with the chiral diamino derivative of 1,1′-binaphthyl, which leads to induced CD. The intensity of CD was increased by increasing the number of porphyrin links from 2 up to 8. An interesting effect of chirality memorizing was registered for porphyrin 135 (n = 8). Upon addition of the antipode to the helix formed by the opposite enantiomer, the sign of induced chirality decreased, but not disappeared completely. Another type of two-dimensional multiporphyrin arrangement was studied in [184, 185] for studies of their chirogenic properties. It was found that hexamers 136 induced a strong CD signal in the Soret band region in the presence of enantiopure (1-naphthyl)ethylamines, whereas other chiral amines lead only to weak Cotton effects. An even stronger chiroptical selectivity was found for a series of dendrimeric porphyrins 136--140 in complexation with bidentate chiral guest molecules. In particular, induced CD strongly depends on the number of porphyrin fragments, and the maximal value of A (2693 cm –1 M –1) is observed for 138. However, it should be noted that this effect has not be comprehensively studied to date. R

N

R

N

N

Zn N

R

N

N

N

N

Zn N

N

R

R

N Zn

n-2

N

R

134, 135 CO2H O 134 : n = 2, R =

B O

135 : n = 2, 3, 4, 8, R =

OC12H25 CO2H

158


R

A

R

N 136 : R =

N

N

N

R

R

A A = B = Ph(t-Bu)2-m,m or A = H, B = Ph(OC8H17)2-m,m

R 136-140

O2C 137 : R =

B

Zn

O2C

X

O2C O2C

X O2C

X

O2C

X

O2C

X 138 : R =

O2C O2C

O2C

X

Further modification of the geometry of porphyrin was done in [186] with the view of developing supermolecular systems for chiral recognition and enantioselective catalysis based on heterometallic trimer 141. At the first stage, various chiral carboxylic acids were totally linked to Sn porphyrins 141 by the type of hexacoordination to form a 1:4 ensemble possessing a chiral cavity. However, subsequent attempts of optical separation of the racemate mixture of chirally substituted pyridines and catalytic asymmetrical epoxidation were unsuccessful, apparently, due to the loss of rigidity inside the chiral cavity. A broader application in various fields proved to be chirogenic host--guest systems, in which both component parts possess a chiral activity due to a number of specific properties. In particular, chiral recognition is the most appropriate application of various derivatives of chiral bis-porphyrins. Thus, a series of bis-porphyrins 142--144 was prepared in [187] for recognition of dicarboxylate anions via a combination of Coulomb attraction and formation of hydrogen bonds. It was found that porphyrins 142 and 143 formed strong complexes with N-carbobenzoxyaspartic acid (N-Cbz-Asp) and N-carbobenzoxyglutamic acid (N-Cbz-Glu) with association constants within the limits of 104--105 M –1 and exhibit a preference towards glutamic acid as compared with aspartic acid. Porphyrin 143 shows an insignificant enantiomeric selectivity, while cyclic and more rigid porphyrin 144 exhibits a low affinity to guest molecules considered, but shows a significant chiral recognition. Apparently, the latter occurs from a good dimensional and geometric correspondence between the host and guest as the result of fixation of two porphyrin fragments by two covalent bridges, however, the detailed mechanism of recognition still requires additional studies.

159

Complexation of Porphyrins with Ions and Organic Molecules X

O2 C X

O2C 139 : R =

O2C

X

O2 C

X

O2C

X

O2 C

140 : R =

X

O2C

O2 C

O2 C O2 C O2 C O2 C

N X =

O2C

X

O2 C

X X

OCH2 Ph(OMe)2-m,m

N Zn

N

N

OCH2Ph(OMe)2-m,m

(CH2)2CO2Me

MeO2C(CH2)2 Me N

Me

N Ru(CO)(Py)

N

N

Me

Me (CH2)2CO2Me

MeO2C(CH2)2

Me

(CH2)2CO2Me Me

Me

MeO2C(CH2)2

Me

N N

MeO2C(CH2)2

(CH2)2CO2Me MeO2C(CH2)2

N

MeO2C(CH2)2

N

Sn(OH)2 N Me

N

N Sn (OH)2 N

Me

Me

Me

(CH2)2CO2Me

(CH2)2CO2Me

141

Spatially fixed porphyrin 145 was used to produce 80--86% and 48% enantiomeric excess in binding of esters of histidine and lysine by means of two-site interaction between two nitrogen atoms of the guest molecule and two zinc ions of the host molecule to form a tweezer structure [188]. Similarly, for chiral recognition of amino acid derivatives the work [152] used porphyrin 146 rigidly bound via 1,1′-binaphthyl fragment. This host molecule exhibits an extremely high affinity to lysine derivative due to the formation of tweezers.

160

N.Zh. Mamardashvili, V.V. Borovkov, G.M. Mamardashvili, Y. Inoue and O.I. Koifman Me

Et

Et N

Me

NH NH

Et

Me

HN Me

Me

Et

N

R

N

NH HN

Me

Et

Me

N

Et

Me

Me

HN

Et Et

Me

142,143 142 : R = A,

143 : R = B

-(H2C)2OCHN A =

B =

-(H2C)2OCHN

-(H2C)2OCHN -(H2C)2OCHN

Me R Et

Me N

Me

Et

NH HN NH

Me

N

Et Et Et Et

Et

N

Me

HN NH HN

Me

N Et

Me R Me -(H2 C)2OCHN 144 : R = -(H2 C)2OCHN

Optimization of the geometry of host--guest complexes showed that the selectivity mechanims is based on spatial repulsion between two methoxy groups of porphyrin 146 and the amide group of the amino acid derivative in the less preferable diastereomeric complex. After inclusion of the chiral residue of leucine into the covalent bridge of achiral dimeric porphyrin 124 the obtained chiral compound 147 was used by the authors of [189] for optical separation of a series of artificial bidentate ligands of different lengths, containing a leucine group in the middle of the molecule (the source of optical activity) and two end pyridine (providing for the two-site binding. A 80% value of enantiomeric excess with significant

Complexation of Porphyrins with Ions and Organic Molecules R

R

R

161

R

N

N N

Zn

N

N N

N

N

R

R

N

Zn

N

R

R

145 : R = Ph(t-Bu)-m,m

Me

Me Et

Me

Me

N

N

N

Zn N

A N

Et Me

Me

A

N

Me

Me

146 :

Et

N Zn

N

Et Me

Me

Me

Me

MeO MeO

R N N

R

A N N

N

Zn

N

A

R

Zn

N N

R 147 :

A = -CO(NHCMe2CO)3NHC(CHPh)CONHCH(i-Bu)CONH(CHPh)CO(NHCMe2CO)3NH-

enhancement of the CD signal in the Soret band region due to the formation of a preferable tweezer complex. Although the detailed mechanism of chirality enhancement is not presented in the work, comparative studies showed the observed effect to be stipulated by the stabilization of the twisted geometry of the host molecule due to the helicity of the guest molecule. An opposite tendency is described in [190]. In particular, in the presence of cit-

162

N.Zh. Mamardashvili, V.V. Borovkov, G.M. Mamardashvili, Y. Inoue and O.I. Koifman Ph

Ph N Co

Ph

Ph

Ph

N

N N

N

N

Co

N

Ph

N

O

O

O

O S

S

148

Et

Et

Et

Et N

N CH2

Zn N

*

N

Et

Et Et

Et

149

rene (1-methyl-4-isopropenylcyclohexene-1) a negative single CD signal of porphyrin 148 was significantly decreased and hypsochromically shifted with a different degree of modulation of chirality for two enantiomers. A similar CD signal decrease was found in the interaction of chiral bis-chlorine 149 with chiral ligands due to induced conformational changes; herewith, the chiroptical response of antipodal ligands differs significantly [191]. This makes it possible to use chlorine 149 for purposes of chiral recognition, using the new principle of enantioselectivity based on the model of two-site host--guest interaction in combination with the model of exciton interactions of chromophore fragments of dimer, the chiral arrangement of which is controlled by the stereochemistry of the guest molecule.

4

Conclusion

Material presented indicates that supramolecular systems based on porphyrins are of undoubted interest for establishing the absolute configuration of organic molecules of various nature and developing molecular receptors for a certain type of substrate. The occurrence of tetrapyrrole chromophores in such systems enables application of spectral methods traditionally used in porphyrin chemistry for studies of molecular recognition processes. Recent advances in this field are a promising prerequisite for the development of new molecular technologies, such as chiral nanotechnology, intelligent molecular devices and chiroptical memory.


163

References 1. B.D. Berezin and N.S. Enikolopian, Metalloporphyrins, Nauka: Moscow (1988) (in Russian). 2. B.D. Berezin, The Coordination Compounds of Porphyrins and Phthalocyanines, Nauka: Moscow (1978) (in Russian). 3. G.M. Trophymenko, B.D. Berezin and A.S. Semeikin, Rus. J. Inorg. Chem., 39 (9), 1493–1496 (1994). 4. N.Zh. Mamardashvili, L.V. Klopova and O.A. Golubchikov, Rus. J. Coord. Chem., 18 (1), 70–79 (1992). 5. N.Zh. Mamardashvili, L.V. Klopova and A.S. Semeikin, Abstracts of the XXII-th International Conference in Chemistry of Coordination Compounds, Minsk, 190 (1990). 6. N.Zh. Mamardashvili, L.V. Klopova, S.A. Zdanovich and O.A. Golubchikov, Abstracts of the III-rd Russian Conference in Chemistry and Application of Nonaqueous Solutions, Ivanovo, 128 (1993). 7. N.Zh. Mamardashvili, S.A. Zdanovich, L.V. Klopova and O.A. Golubchikov, Abstracts of the VIII-th International Conference in Spectroscopy and Chemistry of Porphyrins and Their Analogs, Minsk, 129 (1998). 8. B.D. Berezin and N.I. Sosnikova, Kinetika i Kataliz, 7 (2), 248–253 (1996). 9. O.A. Golubchikov and B.D. Beresin, Rus. J. Phys. Chem., 60 (9), 2113–2128 (1986). 10. V. Gutman, Chemistry of Coordination Compounds in Nonaqueous Solutions, Mir: Moscow (1971) (in Russian) 11. E. M. Kuvshinova, Rus. J. Phys. Chem., 57 (6), 1413–1422 (1987). 12. O.A. Golubchirov, E.M. Kuvshinova, A.S. Semeikin and C.U. Korovina, Rus. J. Phys. Chem., 63 (4), 912–917 (1989). 13. N.Zh. Mamardashvili, A.S. Semeikin and O.A. Golubchikov, Abstracts of the XV-th International Symposium in Macrocyclic Chemistry, Odessa, 94 (1990). 14. O.A. Golubchikov, E.M. Kuvshinova and A.S. Semeikin, Abstracts of the XV-th International Symposium in Macrocyclic Chemistry, Odessa, 93 (1990). 15. E.M. Kuvshinova, O.A. Golubchokov and B.D. Berezin, Rus. J. General Chem., 61 (8), 1799–1804 (1991). 16. B.D. Berezin, T.I. Potapova, R.A. Petrova and L.I. Kalenkova, Rus. J. Phys. Chem., 55 (1), 116–119 (1981). 17. N.Zh. Mamardashvili and O.I. Koifman, Rus. J. Org. Chem., 41 (6), 807–826 (2005). 18. G.M. Mamardashvili, N.Zh. Mamardashvili and O.I. Koifman, Rus. Chem. Rev., 74 (8), 839–855 (2005). 19. P.D. Beer, Chem.Commun., 689–690 (1996). 20. M. Dudic, P. Lhotak, H. Petrickova, I. Stibor and P. Proskova, Org. Lett., 5, 2129–2139 (2003). 21. J.K.M. Sanders, in: The Porphyrin Handbook, ed by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press: San Diego–San Francisco–New York–Boston–London–Sydney–Tokyo, 3, 347– 381 (2000). 22. H. Ogoshi, T. Mizutani, T. Hayashi and Y. Kuroda, in: The Porphyrin Handbook, ed by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press: San Diego–San Francisco–New York–Boston–London–Sydney–Tokyo, 6, 279–340 (2000). 23. E.V. Antina, V.P. Barannikov and A.I. Viugin, Rus. J. Inorg. Chem., 35 (2), 400–404 (1990). 24. B. Cheng and W.R. Scheidt, Inorg. Chim. Acta, 237, 5–14 (1995). 25. H. Imai, S. Nakagawa and E. Kyuno, J. Am. Chem. Soc., 114, 6719–6723 (1992). 26. M. Nappa and J.S. Valentine, J. Am. Chem. Soc., 100, 5075–5080 (1978). 27. B.D. Berezin and O.I. Koifman, Rus. Chem. Rev., 49, 2388–2417 (1980). 28. O.I. Koifman, T.A. Koroleva and B.D. Berezin, Rus. J. Coord. Chem., 4 (5), 1339–1342 (1978). 29. T.A. Koroleva, O.I. Koifman and B.D. Berezin, Rus. J. Phys. Chem., 7 (9), 2007–2012 (1981). 30. K. Belanger, M.H. Keefe, J.L.Welch and J.T. Hupp, Coord. Chem. Rev., 190, 29–35 (1999). 31. A.I. Viugin and G.A. Krestov, Solutions of Nonelectrolites in Liquids, Nauka: Moscow (1989) (in Russian). 32. V.I. Smirnov, A.I. Viugin and G.A. Krestov, Rus. J. Coord. Chem., 16 (7), 896–901 (1990). 33. E.F. Caldin and J.P. Field, J. Chem. Soc. Faraday Trans. 1, 78, 1923–1927 (1982). 34. O. Middel, W. Verboom and D.N. Reinhoudt, J. Org. Chem., 66, 3998–4005 (2001).

164


35. S.G. Puhovskaya, L.Zh. Guseva, E.M. Kuvshinova, N.Zh. Mamardashvili and O.A. Golubchikov Rus. J. Org. Chem., 11, 851–856 (1998). 36. S.J. Cole, G.C. Curthoys, E.A. Magnusson and J.N. Phillips, Inorg. Chem., 11, 1024–1025 (1972). 37. G.R. Miller and G.D. Dorough, J. Am. Chem. Soc., 74, 3977–3981 (1952). 38. N.Zh. Mamardashvili and O.A. Golubchikov, Abstracts of the XXII-th International Conference in Chemistry of Coordination Compounds, Minsk, 167 (1990). 39. R.P. Bonar-Law and J.K.M. Sanders, J. Chem. Soc. Perkin Trans. 1, 3085–3096 (1996). 40. C.A. Hunter, M.N. Meah and J.K.M. Sanders, J. Am. Chem. Soc., 112, 5773–5780 (1990). 41. O.E. Storonkina, G.M. Mamardashvili and N.Zh. Mamardashvili, Abstacts of the IX-th International Conference in Complexation of Macrocyclic Compounds, Ivanovo, 130, (2004). 42. R.S. Wylie, E.G. Levy and J.K.M. Sanders, Chem. Commun., 1611–1612 (1997). 43. H.L. Anderson, S. Anderson and J.K.M. Sanders, J. Chem. Soc. Perkin Trans. 1, 2231–2245 (1995). 44. I.Yu. Bazlova, Abstacts of the IX-th International Conference in Complexation of Macrocyclic Compounds, Ivanovo, 116, (2004). 45. W. Zielenkiewicz, N.Sh. Lebedeva, M. Kaminski, E.V. Antina and A.I. Vyugin, J. Therm. Anal. Cal., 58, 741–749 (1999). 46. H. Imai and E. Kyuno, Inorg. Chem., 29, 2416–2419 (1990). 47. H.L. Anderson, C.A. Hunter, M.N. Meah and J.K.M. Sanders, J. Am. Chem. Soc., 112, 5780–5789 (1990). 48. J. Froidevaux, P. Ochsenbein, M. Bonin, K. Scheck, P. Maltese, J.-P. Gisselbrecht and J. Weiss, J. Am. Chem. Soc., 119, 12362–12363 (1997). 49. H. Imai and Y. Uemori, J. Chem. Soc. Perkin Trans. 2, 1793–1794 (1994). 50. I.P. Danks, I.O. Sutherland and C.H. Yap, J. Chem. Soc. Perkin Trans. 1, 421–423 (1990). 51. N. Zh. Mamardashvili and O.A. Golubchikov, Rus. Chem. Rev., 69, 337–356 (2000). 52. L.Zh. Guseva, Abstacts of the IX-th International Conference in Complexation of Macrocyclic Compounds, Ivanovo, 97 (2004). 53. C.C. Mak, N. Bampos and J.K.M. Sanders, Chem. Commun., 1085–1086 (1999). 54. C.C. Mak, N. Bampos and J.K.M. Sanders, Angåw. Chem. Int. Ed., 37, 3020–3023 (1998). 55. Y. Kubo, Y. Murai, J. Yamanaka, S. Tokita and Y. Ishimavu, Tetrahedron Lett., 40, 6019–6021 (1999). 56. M. Dudic, P. Lhotak, H. Petrickova, I. Stibor, K. Lang and J. Sykora, Tetrahedron, 59, 2409–2415 (2000). 57. B.D. Berezin, O.I. Koifman and O.A. Golubchikov, Rus. J. Inorg. Chem., 18 (6), 1540–1544 (1986). 58. B.D. Berezin and A.N. Drobisheva, Rus. J. Phys. Chem., 41 (2), 402–408 (1967). 59. H.L. Anderson, S. Anderson and J.K.M. Sanders, J. Chem. Soc. Perkin Trans. 1, 2223–2229 (1995). 60. H.L. Anderson, C.L. Walter, A. Vidal-Ferran, R.A. Hay, P.A. Lowden and J.K.M. Sanders, J. Chem. Soc. Perkin Trans. 1, 2275–2279 (1995). 61. Z. Clyde-Watson, N. Bampos and J.K.M. Sanders, New J. Chem., 1135–1138 (1998). 62. A. Vidal-Ferran, N. Bampos and J.K.M. Sanders, J. Am. Chem. Soc., 36, 6117–6126 (1997). 63. A. Vidal-Ferran, Z. Clyde-Watson, N. Bampos and J.K.M. Sanders, J. Org. Chem., 62, 240–241 (1997). 64. S.L. Darling, C.C. Mak, N. Bampos, N. Feeder, S.J. Teat and J.K.M. Sanders, New J. Chem., 23, 359–363 (1999). 65. V. Marvaud, A. Vidal-Ferran, S.J. Webb and J.K.M. Sanders, J. Chem. Soc. Dalton Trans., 985–990 (1997). 66. H.-J. Kim, N. Bampos and J.K.M. Sanders, J. Am. Chem. Soc., 121, 8120–8121 (1999). 67. E. Mikros, A. Gaudemer and R. Pasternack, Inorg. Chim. Acta, 153, 199–210 (1988). 68. C. Verchere-Beaur, E. Mikros and M. Perree-Fauve, J. Inorg. Biochem., 40, 127–131 (1990). 69. C.A. Hunter and J.K.M. Sanders, J. Am. Chem. Soc., 112, 5525–5534 (1990). 70. Y. Kuroda and H. Ogoshi, Synlett., 319–324 (1994). 71. T. Mizutani and H. Ogoshi, Tetrahedron Lett., 30, 5369–5371 (1996). 72. T. Mizutani, T. Ema, T. Yoshida, Y. Kuroda and H. Ogoshi, Inorg. Chem., 32, 2072–2077 (1993).


165

73. Y. Aoyama, A.Yamagishi, M. Asagawa, H. Toi and H. Ogoshi, J. Am. Chem. Soc., 110, 4076–4077 (1988). 74. N.Zh. Mamardashvili, O.E. Storonkina and G.M. Mamardashvili, Rus. J. Coord. Chem., 30 (5), 416–420 (2004). 75. T. Mizutani, T. Murakami, T. Kurahashi and H. Ogoshi, J. Org. Chem., 61, 539–541 (1996). 76. H. Ogoshi, H. Hatakeyama, K. Yamamura and Y. Kuroda, Chem. Lett., 51–54 (1990). 77. H. Ogoshi, H. Hatakeyama, J. Kotani, A. Kawashima and Y. Kuroda, J. Am. Chem. Soc., 113, 8181–8183 (1991). 78. R.P. Bonar-Law and J.K.M. Sanders, J. Am. Chem. Soc., 117, 259–271 (1995). 79. T. Mizutani, T. Murakami, N. Matsumi,T. Kurahashi and H. Ogoshi, J. Chem. Soc. Chem. Commun., 1257–1263 (1995). 80. T. Mizutani, T. Kurahashi, T. Murakami, N. Matsumi and H. Ogoshi, J. Am. Chem. Soc., 119, 8991–9001 (1997). 81. M. Tichy, Adv. Org. Chem., 5, 115–118 (1965). 82. O.A. Golubchikov, Abstacts of the IX-th International Conference in Complexation of Macrocyclic Compounds, Ivanovo, 77 (2004). 83. F.A.J. Singelenberg and J.H. van der Maas, J. Mol. Struct., 243, 111–117 (1991). 84. V. Kral and J.L. Sessler, Tetrahedron, 51, 539–554 (1995). 85. O. Rusin and V. Kral, Chem. Commun., 2367–2368 (1999). 86. V. Kral, O. Rusin and F. P. Schmidtchen, Org. Lett., 3, 873–876 (2001). 87. G.M. Sanders, M. van Dijk, A. van Veldhuizen and H.C. van der Plas, J. Chem. Soc. Chem. Commun., 1311–1313 (1986). 88. T. Hayashi, T. Miyahara, N. Koide, Y. Kato, H. Masuda and H. Ogoshi, J. Am. Chem. Soc., 119, 7281–7290 (1997). 89. R. Fiammengo, P. Timmerman, F. de Jong and D.N. Reinhoudt, Chem. Commun., 2313–2317 (2000). 90. R. Purrello, R. Geremia, L. Randaccio, F.G. Gulino and V. Ravone, Angew. Chem. Int. Ed., 40, 4245–4248 (2001). 91. D. Voet and J.D. Voet, Biochemistry, 2nd edn, Wiley: New York–Chichester–Brisbane–Toronto– Singapore (1995). 92. J.W. Steed and J.L. Atwood, Supramolecular Chemistry, Wiley: Chichester–New York– Weinheim–Brisbane–Singapore–Toronto (2000). 93. J. Kyte, Structure in Protein Chemistry, Garland Publishing: New York–London (1995). 94. M.-C. Hsu and R.W. Woody, J. Am. Chem. Soc., 93, 3515–3525 (1971). 95. G. Blauer, N. Sreerama and R.W. Woody, Biochem., 32, 6674–6677 (1993). 96. S.C. Björling, R.A. Goldbeck, S.J. Paquette, S.J. Milder and D.S. Kliger, Biochem., 35, 8619–8627 (1996). 97. A. Boffi, J.B. Wittenberg and E. Chiancone, FEBS Lett., 411, 335–338 (1997). 98. V.V. Borovkov, G.A. Hembury and Y. Inoue, Acc. Chem. Res., 37, 449–459 (2004). 99. H. Ogoshi and T. Mizutani, J. Synth. Org. Chem. Jpn., 54, 906–917 (1996). 100. H. Ogoshi and T. Mizutani, Acc. Chem. Res., 31, 81–89 (1998). 101. H. Ogoshi and T. Mizutani, Curr. Opin. Chem. Biol., 3, 736–739 (1999). 102. J.-C. Marchon and R. Ramasseul, in: The Porphyrin Handbook, ed. by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press, An Imprint of Elsevier Science: Amsterdam–Boston– London–Oxford–Paris–San Diego–San Francisco–Singapore–Sydney–Tokyo, 11, 75–132 (2003). 103. X. Huang, K. Nakanishi and N. Berova, Chirality, 12, 237–255 (2000). 104. T.D. James, K.R.A.S. Sandanayake and S. Shinkai, Angew. Chem. Int. Ed. Engl., 35, 1911–1922 (1996). 105. A. Robertson and S. Shinkai, Coord. Chem. Rev., 205, 157–199 (2000). 106. S. Shinkai, M. Ikeda, A. Sugasaki and M. Takeuchi, Acc. Chem. Res., 34, 494–503 (2001). 107. S. Shinkai and M. Takeuchi, Biosens. Bioelectron., 20, 1250–1259 (2004). 108. J.H. Hartley, T.D. James and C.J. Ward, J. Chem. Soc. Perkin Trans. 1, 3155–3184 (2000). 109. H. Tsukube and S. Shinoda, Enantiomer, 5, 13–22 (2000). 110. H. Tsukube, S. Shinoda and H. Tamiaki, Coord. Chem. Rev., 226, 227–234 (2002). 111. H. Tsukube and S. Shinoda, Chem. Rev., 102, 2389–2403 (2002). 112. G. Simonneaux and P.L. Maux, Coord. Chem. Rev., 228, 43–60 (2002).

166


113. M.G. Finn, Chirality, 14, 534–540 (2002). 114. S. Allenmark, Chirality, 15, 409–422 (2003). 115. H.C. Aspinall, Chem. Rev., 102, 1807–1850 (2002). 116. M.C. Feiters, A.E. Rowan and R.J.M. Nolte, Chem. Soc. Rev., 29, 375–384 (2000). 117. Y. Furusho, T. Kimura, Y. Mizuno and T. Aida, J. Am. Chem. Soc., 119, 5267–5268 (1997). 118. Y. Mizuno, T. Aida and K. Yamaguchi, J. Am. Chem. Soc., 122, 5278–5285 (2000). 119. H. Kijima, M. Takeuchi and S. Shinkai, Chem. Lett., 781–782 (1998). 120. A. Synytsya, V. Kral, K. Volka, J. Copikova and J.L. Sessler, J. Chem. Soc. Perkin Trans. 2, 1876–1884 (2000). 121. G.R. Deviprasad and F. D’Souza, Chem. Commun., 1915–1916 (2000). 122. T. Mizutani, K. Wada and S. Kitagawa, J. Org. Chem., 65, 6097–6106 (2000). 123. H. Tamiaki, N. Matsumoto, S. Unno, S. Shinoda and H. Tsukube, Inorg. Chim. Acta, 300–302, 243–249 (2000). 124. H. Tsukube, M. Wada, S. Shinoda and H. Tamiaki, Chem. Commun., 1007–1008 (1999). 125. M. A. Awawdeh, J.A. Legako and H.J. Harmon, Sens. Actuator B, 91, 227–230 (2003). 126. C.M. Muzzi, C.J. Medforth, R.M. Smith, S.-L. Jia and J.A. Shelnutt, Chem. Commun., 131–132 (2000). 127. J.P. Collman, Z. Wang, C. Linde, L. Fu, L. Dang and J.I. Brauman, Chem. Commun., 1783–1784 (1999). 128. V.F. Slagt, M. Röder, P.C.J. Kamer, P.W.N.M. van Leeuwen and J.N.H. Reek, J. Am. Chem. Soc., 126, 4056–4057 (2004). 129. K. Ladomenou and R.P. Bonar-Law, Chem. Commun., 2108–2109 (2002). 130. C. Morice, P.L. Maux and G. Simonneaux, Tetrahedron Lett., 37, 6701–6704 (1995). 131. C. Morice, P.L. Maux, G. Simonneaux and L. Toupet, J. Chem. Soc. Dalton Trans., 4165–4171 (1998). 132. M. Inamo and I. Yoneda, Inorg. Chem. Commun., 2, 331–333 (1999). 133. Y. Kuroda, Y. Kato, T. Higashioji, J.-Y. Hasegawa, S. Kawanami, M. Takahashi, N. Shiraishi, K. Tanabe and H. Ogoshi, J. Am. Chem. Soc., 117, 10950–10958 (1995). 134. H. Imai, H. Munakata, Y. Uemori and N. Sakura, Inorg. Chem., 43, 1211–1213 (2004). 135. C.Z. Wang, Z.A. Zhu, Y. Li, Y.T. Chen, X. Wen, F.M. Miao, W.L. Chan and A.S.C. Chan, New J. Chem., 25, 801–806 (2001). 136. E. Galardon, M. Lukas, P.L. Maux and G. Simonneaux, Tetrahedron Lett., 40, 2753–2756 (1999). 137. S. Wolowiec, L. Latos-Grazynski, M. Mazzanti, and J.-C. Marchon, Inorg. Chem., 36, 5761–5771 (1997). 138. J.-P. Simonato, J. Pecaut and J.-C. Marchon, J. Am. Chem. Soc., 120, 7363–7364 (1998). 139. M. Claeys-Bruno, D. Toronto, J. Pecaut, M. Bardet and J.-C. Marchon, J. Am. Chem. Soc., 123, 11067–11068 (2001). 140. J.-P. Simonato, S. Chappellet, J. Pecaut and J.-C. Marchon, New J. Chem., 25, 714–720 (2001). 141. S. Gazeau, J. Pecaut and J.-C. Marchon, Chem. Commun., 1644–1645 (2001). 142. J.L. Sessler and A. Andrievsky, Chem. Eur. J., 4, 159–167 (1998). 143. M. Takeuchi, T. Imada and S. Shinkai, Bull. Chem. Soc. Jpn., 71, 1117–1123 (1998). 144. M. Takeuchi, T. Imada and S. Shinkai, Angew. Chem. Int. Ed., 37, 2096–2099 (1998). 145. A. Sugasaki, M. Ikeda, M. Takeuchi, A. Robertson and S. Shinkai, J. Chem. Soc. Perkin Trans. 1, 3259–3264 (1999). 146. A. Sugasaki, M. Ikeda, M. Takeuchi and S. Shinkai, Angew. Chem. Int. Ed., 39, 3839–3842 (2000). 147. A. Sugasaki, K. Sugiyasu , M. Ikeda, M. Takeuchi and S. Shinkai, J. Am. Chem. Soc., 123, 10239–10244 (2001). 148. M. Yamamoto, A. Sugasaki, M. Ikeda, M. Takeuchi, K. Frimat, T.D. James and S. Shinkai, Chem. Lett., 520–521 (2001). 149. Y.-M. Guo, H. Oike and T. Aida, J. Am. Chem. Soc., 126, 716–717 (2004). 150. Y. Mizuno and T. Aida, Chem. Commun., 20–21 (2003). 151. H. Tsukube, N. Tameshige, S. Shinoda, S. Unno and H. Tamiaki, Chem. Commun., 2574–2575 (2002). 152. T. Hayashi, T. Aya, M. Nonoguchi, T. Mizutani, Y. Hisaeda, S. Kitagawa and H. Ogoshi, Tetrahedron, 58, 2803–2811 (2002).


167

153. D. Monti, L.L. Monica, A. Scipioni and G. Mancini, New. J. Chem., 25, 780–782 (2001). 154. Y. Kubo, Y. Ishii, T. Yoshizawa and S. Tokita, Chem. Commun., 1394–1395 (2004). 155. Y. Kubo, T. Ohno, J.-I. Yamanaka, S. Tokita, T. Iida and Y. Ishimaru, J. Am. Chem. Soc., 123, 12700–12701 (1998). 156. X. Huang, B.H. Rickman, B. Borhan, N. Berova and K. Nakanishi, J. Am. Chem. Soc., 120, 6185–6186 (1998). 157. X. Huang, B. Borhan, B.H. Rickman, K. Nakanishi and N. Berova, Chem. Eur. J., 6, 216–224 (2000). 158. T. Kurtan, N. Nesnas, Y.-Q. Li, X. Huang, K. Nakanishi and N. Berova, J. Am. Chem. Soc., 123, 5962–5973 (2001). 159. T. Kurtan, N. Nesnas, F.E. Koehn, Y.-Q. Li., K. Nakanishi and N. Berova, J. Am. Chem. Soc., 123, 5974–5982 (2001). 160. Q. Yang, C. Olmsted and B. Borhan, Org. Lett., 4, 3423–3426 (2002). 161. G. Proni, G. Pescitelli, X. Huang, N.Q. Quraishi, K. Nakanishi and N. Berova, Chem. Commun., 1590–1591 (2002). 162. X. Huang, N. Fujioka, G. Pescitelli, F.E. Koehn, R.T. Williamson, K. Nakanishi and N. Berova, J. Am. Chem. Soc., 124, 10320–10335 (2002). 163. G. Proni, G. Pescitelli, X. Huang, K. Nakanishi and N. Berova, J. Am. Chem. Soc., 125, 12914–12927 (2003). 164. H. Ishii, Y. Chen, R.A. Miller, S. Karady, K. Nakanishi and N. Berova, Chirality, 17, 305–315 (2005). 165. A. Solladie-Cavallo, C. Marsol, G. Pescitelli, L.D. Bari, P. Salvadori, X. Huang, N. Fujioka, N. Berova, X. Cao, T.B. Freedman and R. Nafie, Eur. J. Org. Chem., 1788–1796 (2002). 166. H. Ishii, S. Krane, Y. Itagaki, N. Berova, K. Nakanishi and P.J. Weldon, J. Nat. Prod., 67, 1426–1430 (2004). 167. V.V. Borovkov, J.M. Lintuluoto and Y. Inoue, Org. Lett., 2, 1565–1568 (2000). 168. V.V. Borovkov, J.M. Lintuluoto and Y. Inoue, J. Am. Chem. Soc., 123, 2979–2989 (2001). 169. V.V. Borovkov, J.M. Lintuluoto and Y. Inoue, J. Phys. Chem. A, 104, 9213–9219 (2000). 170. V.V. Borovkov, N. Yamamoto, J.M. Lintuluoto, T. Tanaka and Y. Inoue, Chirality, 13, 329–335 (2001). 171. V.V. Borovkov, J.M. Lintuluoto, H. Sugeta, M. Fujiki, R. Arakawa and Y. Inoue, J. Am. Chem. Soc., 124, 2993–3006 (2002). 172. V.V. Borovkov, J.M. Lintuluoto, M. Fujiki and Y. Inoue, J. Am. Chem. Soc., 122, 4403–4407 (2000). 173. V.V. Borovkov, G.A. Hembury and Y. Inoue, J. Porphyrins Phthalocyanines, 7, 337–341 (2003). 174. J.M. Lintuluoto, V.V. Borovkov and Y. Inoue, J. Am. Chem. Soc., 124, 13676–13677 (2002). 175. V.V. Borovkov, G.A. Hembury, N. Yamamoto and Y. Inoue, J. Phys. Chem. A, 107, 8677–8686 (2003). 176. V.V. Borovkov, G.A. Hembury and Y. Inoue, Angew. Chem. Int. Ed., 42, 5310–5314 (2003). 177. V.V. Borovkov, J.M. Lintuluoto and Y. Inoue, Org. Lett., 4, 169–171 (2002). 178. V.V. Borovkov, J.M. Lintuluoto, M. Sugiura, Y. Inoue and R. Kuroda, J. Am. Chem. Soc., 124, 11282–11283 (2002). 179. V.V. Borovkov, J.M. Lintuluoto, G.A. Hembury, M. Sugiura, R. Arakawa and Y. Inoue, J. Org. Chem., 68, 7176–7192 (2003). 180. V.V. Borovkov, I. Fujii, A. Muranaka, G.A. Hembury, T. Tanaka, A. Ceulemans, N. Kobayashi and Y. Inoue, Angew. Chem. Int. Ed., 43, 5481–5485 (2004). 181. V.V. Borovkov, T. Harada, Y. Inoue and R. Kuroda, Angew. Chem. Int. Ed., 41, 1378–1381 (2002). 182. M. Ikeda, S. Shinkai and A. Osuka, Chem. Commun., 1047–1048 (2000). 183. C. Ikeda, Z.S. Yoon, M. Park, H. Inoue, D. Kim and A. Osuka, J. Am. Chem. Soc., 127, 534–535 (2005). 184. M. Takase, R. Ismael, R. Murakami, M. Ikeda, D. Kim, H. Shinmori, H. Furuta and A. Osuka, Tetrahedron Lett., 43, 5157–5159 (2002). 185. W.-S. Li, D.-L. Jiang, Y. Suna and T. Aida, J. Am. Chem. Soc., 127, 7700–7702 (2005). 186. S.J. Webb and J.K.M. Sanders, Inorg. Chem., 39, 5920–5929 (2000). 187. J.L. Sessler, A. Andrievsky, V. Kral and V. Lynch, J. Am. Chem. Soc., 119, 9385–9392 (1997).

168


188. M.J. Crossley, L.G. Mackay and A.C. Try, J. Chem. Soc. Chem. Commun., 1925–1927 (1995). 189. Y.-M. Guo, H. Oike, N. Saeki and T. Aida, Angew. Chem. Int. Ed., 43, 4915–4918 (2004). 190. R. Paolesse, D. Monti, L.I. Monica, M. Venanzi, A. Froiio, S. Nardis, C.D. Natale, E. Martinelli and A. D’Amico, Chem. Eur. J., 8, 2476–2483 (2002). 191. V.V. Borovkov, G.A. Hembury and Y. Inoue, J. Org. Chem., 70, 8743–8754 (2005).

5

Chemical Activation of Porphyrins in Coordination Core Reactions D.B. Berezin1,2 and B.D. Berezin1 1Institute

of Solution Chemistry, Russian Academy of Sciences, 153045 Ivanovo, Russia 2Ivanovo State University of Chemistry and Technology, 7 F. Engels Prospect, Ivanovo, 153000, Russia; email: [email protected]

Depending on the molecular structure and effects of the medium, porphyrin NH bonds of the coordination core can be considered as localized or partially delocalized. Delocalization can be caused by intramolecular effects such as molecule polarization by means of push-pull substituents or special nonplanar macrocycle conformations and by a series of external factors like strong solvation interactions in a solution containing the electron-donor component in a polymer state, sorption or other types of solid state effects. Activation of NH bonds in porphyrin molecules results in crucial changes of their coordination core reactivity. This chapter analyzes the causes and ways of activation in relation to the possibility of classifying porphyrins by their ability to activate or not to activate NH bonds. A number of reliable quantitative criteria allowing to talk about the degree of the chemical activity of NH bonds for porphyrins or their analogues of any structural group is presented. Activation of NH bonds can not only affect the H2P core reactivity but in some cases cause a reorganization of π-chromophore and reaction centres, which are different tauthomeric processes. Products of such a reorganization are good models for biological supramolecular systems. Chemical activation of porphyrin by itself is considered as a powerful tool for the reactivity control of these molecules, including biological aspects.

Keywords: porphyrin, coordination core, chemical activation, reactivity control

170

D.B. Berezin and B.D. Berezin

Introduction Coordination cores in molecules of porphyrin ligands (H2P) and their complexes with p, d and f metals (MP) are some of their main reaction centres [1–3]. These are complex multifunctional centres, as a rule, N4H2 and MN4, respectively. Their unique reactivity is determined by a combination of a number of factors. They include not only the chemical affinity of atoms and atomic groupings constituting these cores, with respect to electrophilic-nucleophilic reagents, their capability of synchronous donor-acceptor interactions, the possibility of redistributing electronic density from one atomic grouping to another for reasons of their involvement in the united system of π-conjugation [1, 4]. One of the most effective factors is a relatively rigid shielding of these reaction centres by an atomic environment and π-electron cloud of the aromatic macrocycle (Fig. 1). The extent of shielding, in turn, can change under the influence of external effects. This phenomenon was given the name of the macrocyclic effect (MCE) of porphyrins and metalloporphyrins [4–8]. MCE affects the physicochemical properties and reactivity of H2P and MP by means of π-electronic and structural (steric) components. Their effects on various physicochemical processes occurring in the molecule are, as a rule, different and nonsymbate one to another [8].

(a)

(b)

(c)

Figure 1 Atomic and π-electronic shielding of the reaction centre MN4 of metalloporphyrins MP (a), X(MP) (b), (X)2MP (c) [4–5].

Initially, research to achieve understanding of the mechanisms of complex enzymic reactions involving porphyrins and their analogues was aimed at studies of the electronic effects of substituents located mainly at the periphery, i.e., in β- and meso-positions of H2P and MP molecules, and their influence on the structure and properties of molecules [1, 9–11]. Namely the functional substitution was considered to be the most efficient means of controlling the properties of H2P and MP. Herewith, most porphyrins were traditionally seen as planar molecules and some (often significant) deviations from planarity, revealed using X-ray diffraction analysis (XDA) of crystals, were referred to crystal packing effects [1, 12, 13]. The last quarter of the 20th century witnessed the appearance of a large number of new porphyrin structures with unusual spectral and other physicochemical properties, drastically different from the earlier known compounds [14]. A large part of them are porphyrins with a strongly nonplanar structure [15–17]. As the result, an idea was proposed and took root in late 1980s that the key role in the exhibition of their unique properties by porphyrins and their complexes in the bioenvironment, the properties of the coordination core including, is played by the conformational mobility of the macrocycles [18, 19]. However, in recent years authors again come back increasingly often to the role of polarization effects interrelated with the planar structure of H2P [20–24]. Indeed,

171

Chemical Activation of Porphyrins in Coordination Core Reactions

conformational conversions of H2P molecules can cause secondary changes due to the asymmetric redistribution of π-electronic density in them. The emerging polarization of molecules on the whole and separate chemical bonds in particular, e.g., the chemical activation of NH bonds, can drastically change the reactivity of compounds. Of special significance is the role of the medium, in particular, the nature of solvent (if processes are run in a liquid-phase system), which is capable of significantly intensifying or inhibiting the polarization processes in H2P molecules. Thus, another important factor, which along with atomic electronic shielding (MCE), determines the reactivity of porphyrins in vivo, is polarization of the molecule and, as a rule, chemical activation of NH bonds in its coordination core. This conclusion has been evident earlier. It follows, e.g., from the reaction mechanism of complexation of porphyrins with d metal salts [1, 5, 25], the limiting stage of which is namely polarization by the solvent and dissociation of NH bonds of the H2P macrocyclic ligand. An important problem, which needs to be solved, is to elucidate the relation between the nonplanar structure of the porphyrin macrocycles and the chemical activity of NH bonds in the coordination core of these molecules. As a particular case of a jumpwise change of the properties under the influence of conformational regroupings, the phenomenon of NH activation is very important for modelling and studying the processes involving H2P in vivo, where in the course of acceptance of metal ions by bioporphyrins NH activation is performed also under the influence of the donor-acceptor protein-lipid environment.

1

Porphyrin Ligands with Localized and Delocalized NH bonds

In accordance with a classification proposed with account for the significance of the chemical activity of NH bonds, porphyrins can be divided into three groups [26]: • with the planar aromatic π-system and chemically low-active NH bond; • with the nonplanar π-system of aromatic chromophore and chemically active NH bond; • with the planar aromatic π-system and chemically active NH bond. The first group includes many porphyrin bioligands, e.g., chlorophylls, blood porphyrins, cytochromes [5, 27], as well as their synthetic β- and meso-substituted analogues, e.g., β-octaethylporphin {I, H2(β-Et)8P}, meso-tetraphenylporphin {II, H2TPP} and mesotetrapropylporphin {III, H2T(n-Pr)P}. These compounds are also called classical H2P [28],

R1

X

R

R1

R

N

X

R1

R1

meso-position

N

NH

R

R N

HN

β−position

R N

R1

R

R1

N X

R

N

N

HN

R1

R1

N

NH

N

NH

R

HN N X

I–III, V–VII IV, X VIII–IX I. R = H, R1 = Et; II. R = Ph, R1 = H; III. R = n-Pr, R1 = H; IV. R = Ph; V. R = Ph, R1 = Et; VI. R = R1 = Ph; VII. R = Ph, R1 = Br; VIII. X = H; IX. X = Br; X. R = H.

172


because the properties they exhibit totally correspond to those traditionally known in porphyrin chemistry. The second group are dodeca-substituted (peripherally highly substituted) H2P (compounds IV–VII) [28–33] and some other strongly nonplanar molecules. The third group comprises rigid highly aromatic porphyrazines (compounds VIII–IX) [34–37], as well as tetrabenzoporphyrins {e.g., H2TBP (X)}, having substituents in condensed rings but not in meso-positions [38]. Meso-substituted tetrabenzoporphyrins (e.g., IV) belong to the second group of compounds, dodeca-substituted porphyrins [31–32, 38–39]. H2P with nonclassical properties constitute the second and third groups of compounds [26, 28]. 1.1

Factors causing the delocalization of NH bonds in porphyrin molecules

The high chemical activity of NH bonds in H2P molecules, related to their partial or total delocalization [1, 4, 5, 35], can be achieved if several conditions are satisfied [28, 39]: • the structure of the macrocycle is changed to become rigid (aromatic) or, vice versa, strongly nonplanar; • electron-acceptor groups or groups of various electronic nature which lead to the push-pull-type polarization of π-chromophore [40–43], are present in the molecule; • the properties of the medium, contributing to the polarization of NH bonds of the H2P molecule, are changed by means of solvation and chemical interactions, e.g., with molecules of electron-donor solvents [4–5], or by adjusting the reaction NH centres of porphyrins to the reaction centres of the medium in the solid phase during the sorption, immobilization etc. [44–45]. At least two out of the three conditions, each of which is necessary but not sufficient, are required to be satisfied to achieve the NH activity of H2P. 1.1.1

Spatial structure and polarization of the porphyrin molecule

Analysis of vast literature data [4–5, 19, 22, 29–30, 34–37, 46, 56–57] and own studies [7–8, 24, 26, 28, 31–32] indicate that the porphyrin ligand molecule equally acquires a capability of delocalizing NH bonds both when it becomes more planar and highly aromatic and, vice versa, under the influence of significant extraplanar distortions of the π-chromophore. As a result, a situation, at first glance paradoxical, emerges when opposite structural changes lead to the same type of changes in some properties of the molecules [58]. Thus, for instance, with the advance of benzo- and aza-substitution of porphyrin molecules their

Saddle

Ruffled

Domed

Wave

Figure 2 Main types of nonplanar conformations of H2P and their complexes ( and assume the location of atoms over and under the initial plane of the macrocycle, respectively) [16, 19].


173

structural rigidity is enhanced in the sequence: porphyrins proper → benzoporphyrins → azaporphyrins, which is supported by numerous quantum chemical [35, 46, 48, 59], X-ray diffraction [1, 35] and fluorescence [48, 60] data. In the case when the periphery of the H2P molecule, i.e., its β- and meso-positions, is overloaded with bulky functional groups, porphyrin becomes more apt to extraplanar deformations and by the geometry of the most stable conformation belongs to one of the known types [16, 19]. Among them, the most frequent types are saddle, ruffled, domed and wave conformations (Fig. 2), as well as stepped conformation; their combinations are also widely known, which are especially characteristic of porphyrins with asymmetric substitution. Thus, a combination of the ruffled and domed conformations yields a gabled structure [57]. The largest distortion is achieved if even at free β-positions all four meso-positions are substituted by bulky groups (iso-Pr, cyclo-Hex, tert-Bu {compounds XI–XIII}). The distortion of the porphyrin macrocycle also increases to a maximum with the increase in the number of peripheral substituents of moderate voluminosity up to twelve (compounds IV–VII) at eight β-positions occupied. Each of the typical conformations can be characterized in one way or another by the degree of intensity of the nonplanar R structure. For instance, saddle conformations typical of dodeca-substituted H2P (compounds IV–VII), as well as of a significant number of Zn porphyrins (ZnP) (Table 1) are characNH N terized by the mean deviation of β-carbon atoms from the iniR R N HN tial plane of the macrocycle (∆Cβ, Å) obtained from XDA data. The value of ∆Cβ changes for the currently known strongly nonplanar, saddle-distorted porphyrins and their complexes R from 0.7 up to more than 1.3 Å [19]. A deviation of the order of 1 Å is considered to be very significant. The ruffled struc- XI. R == iso-Pr; iso-Pr; XI. R ture of nonplanar porphyrins, which include meso-substituted XII. XII. R R= = cyclo-Hex; cyclo-Hex; R= = tret-Bu. tert-Bu. molecules with bulky substituents (compounds XI–XIII), as XIII. XIII. R well as most nickel complexes (NiP) are characterized by the deviation from the averaged plane of the meso-carbon atoms (∆Cmeso, Å) (see Table 1). Table 1 Comparative X-ray analysis data for planar and nonplanar H2P as well as their complexes with Zn(II) and Ni(II) [19, 65]. Porphyrin

H2(β-Et)8P (I) H2TPP (II) H2T(n-Pr)P (III) H2T(iso-Pr)P (XI) H2T(cyclo-Hex)P (XII) H2T(tert-Bu)P (XIII) H2TPTBP (IV) H2(β-Et)8TPP (V) H2(β-Ph)8TPP (VI) H2(β-Br)8TPP (VII)

Configuration of ligand Pl Pl Pl Ruf Ruf Ruf Sad Sad Sad Sad

X-ray diffraction data(a) ∆Cβ, Å

∆Cmeso, Å

0.08[0.06] 0.23[0.18] 0.08 (0.27)[0.27] (0.26)[0.29] (0.33) (0.77) 1.17(1.09)[1.24] 1.28(0.36)[0.96] 1.26(0.92)

0.04[0.03] 0.14[0.45] 0.02 (0.29)[0.74] (0.06)[0.77] (0.90) (0.08) 0.03(0.04)[0.03] 0.07(0.01)[0.86] 0.32(0.25)

(a) ( ) X-ray analysis data for Zn complexes; [ ] for Ni complexes; ∆C and ∆C β meso, mean deviations of C atoms in β- and meso-positions of the initial plane of the macrocycle.

174 Property

D.B. Berezin and B.D. Berezin Ì

HOMO–LUMO energy gap

Ê H2P redox properties Ì stability of MP complexes Coordinating ability of H2P in electron-donor solvents

Ê

Ì

in proton-donor solvents Acid-base properties of H2P

Ê

Ì

N basicity

Relative H2P solvation heats by weakly solvating solvent by electron-donor solvent

Ê group 2 2 group 1 1 Ƚɪɭɩɩɚ Ƚɪɭɩɩɚ Rigidity of H2P macrocycle

NH acidity

Ƚɪɭɩɩɚ group 3 3

solubility of H2P

Figure 3 Porphyrin structure–property dependences: group 1, planar classical H2P; group 2, nonplanar nonclassical H2P; group 3, planar nonclassical H2P.

The properties related to the reactivity of the coordination core of the H2P molecule on the whole and the chemical activity of NH bonds in particular are especially dependent on the peculiarities of the macrocycle geometric structure [58]. Herewith, both porphyrins with the rigid π-macrocycle (compounds VIII–X) and strongly nonplanar compounds IV–VII acquire more pronounced acidic properties [35, 61–62], and in a some cases become capable of interaction with molecules of electron-donor solvents [30, 32, 55], which is exhibited both in their solvation characteristics [32,63–64] and in the change of the reactivity of these ligands in the complexation reaction (1) with salts of d metals in electron-donor media [36, 52, 55] as compared with predominantly planar porphyrins proper (compounds I–III). As the result, there arises a nonlinear dependence of the change of some properties of H2P on the change of structure of the molecules (Fig. 3). H2P + MX2(Solv)n–2

–2Solv

[H2P … MX2(Solv)n–4]#

MP + 2HX + (n–4)Solv. (1)

For porphyrins with nonplanar structure, an important role in activation of NH bonds is played by the type of nonplanar conformation of the ligand (see Section 1.4). The rise of rigidity of porphyrins with the advance of their benzo- (compound X) and aza-substitution (compounds VIII–IX) [35] leads not only to the flattening of the ligand, but also to its more pronounced π-electron deficiency. Therefore, the regular consequence of this is an enhancement of the acidic properties of NH groupings of molecules. Similar changes of reactivity are also observed in simpler five- or six-membered nitrogen-containing heterocycles with two or more heteroatoms or atomic groupings. Thus, in the sequence of compounds pyrrole – imidazole – 1,2,3-triazole – tetrazole the NH acidity in an aqueous solution increases 12.5-fold [66–67] with the rise of the intensity of π-electron deficiency of the aromatic heterocycle (Table 2). In a DMSO medium, the acidity of the heterocycles is 5–6 orders of magnitude lower, as in this solvent there is a problem with the solvation of anion [68]. The causes of the increase of the acidic properties and, in particular, of the chemical activity of NH bonds in nonplanar dodeca-substituted porphyrins remain a topical subject

175


of studies up to the present. As a rule, inclusion of acid-base centres into a unified system of conjugation leads to the symbateness in the change of the chemical activity of these centres [69]. It is observed as in the case of five-membered aromatic heterocycles, when an increase of acidity of NH groupings occurs against the background of a decrease of the main properties of nitrogen tertiary atoms (see the sequence pyrazole – triazole – tetrazole in Table 2). A similar situation is also observed in the case of complex macroheterocyclic Table 2 NH acidity (pKa) and N basicity (pKBH+) of five-membered nitrogen-containing heterocycles in DMSO and H2O [66–67].

N Heterocycle

N H

N H

Pyrrole pKa298(DMSO) pKa298(H2O) pKBH+298 (a)

N

N

N N

N N

N H

N H

Pyrazole

Imidazole

1,2,3-Triazole

Tetrazole

23.3

20.4

18.9

–

–

17.5

14.2

14.2

9.3

–

2.5

–

1.2

N H

4.9 –2.68(a)

measured in sulfuric acid

compounds, in particular, porphyrins with predominantly planar structure (I–III). Thus, introduction of one electron-acceptor chlorine atom each into each of the meso-phenyl rings of the H2TPP molecule (II), even into meta-positions, increases the strength of this NH acid in the elimination of the first proton ((pKa1, equation (2)) almost 1.5-fold in a DMSO medium. Simultaneously, the main properties (measured in acetonitrile) weaken in the total process of elimination of two protons from porphyrin dication (pK3,4, H4P2+ ' 2H+ + H2P) by two orders [61, 69]. Even the fact that the acidity and basicity of compounds was assessed in different solvents does not disturb the general pattern of acid-base interaction (equations (2)–(5)). H2P(solv) HP –(solv) H4P 2+(solv) H3P +(solv)

KK11

KK21

KK31

KK41

H+(solv) + HP –(solv) ,

(2)

H+(solv) + P2–(solv) ,

(3)

H+(solv) + H3P+(solv) ,

(4)

H+(solv) + H2P(solv) .

(5)

Interference of spatial factors related to the change of the geometric structure of the π-macrocycle very often results in the simultaneous increase of NH-acidic and N-basic properties of nonplanar porphyrins. For instance, already in passing from unsubstituted porphin (H2P) to slightly more nonplanar H2TPP (II) both the rise of acidity and basicity of the

176


molecule are observed simultaneously (Table 3). This phenomenon acquires a much more pronounced character in strongly nonplanar dodeca-substituted porphyrins IV–VII. Takeda and Sato [29] were among the first to explain the simultaneous enhancement of both acidity Table 3 Constants of acid-base interaction of porphyrins [61]. Cl

Porphyrin

Cl

NH

N

NH

N

NH

N

N

HN

N

HN

N

HN Cl

Cl

298 pKa1 (DMSO) 298 pK3,4 (AN)

H2P

H2TPP

H2T(m-Cl)PP

22.35

21.15

19.82

15.35

19.80

17.80

and basicity of these molecules as the result of spatial deshielding and increased accessibility of atoms and chemical bonds of the coordination core of the conformationally mobile macrocyclic ligand for chemical reagents. As it became clear later, this explanation is not sufficient, as far from any disturbance of the planar structure of the H2P molecule chemically activates NH bonds in it (see Section 3). On the contrary, according to the classical polarization views, a greater conformational mobility of the molecule should lead to a worsening of the conditions of the distribution of π-electrons on the conjugated system and, therefore, its localization on individual atoms and bonds, in particular, on coordination core atoms. This conclusion is consistent with the increase of the basicity of tertiary atoms of nitrogen in the molecule of strongly nonplanar H2P ligand; however, it does not explain but, vice versa, contradicts the enhancement of NH acidity. What is more, 1H NMR studies of a number of nonplanar porphyrins, whose conformations in the solid phase can be characterized as ruffled [70], have shown that rather significant extraplanar distortions of these molecules lead to a mere 5% decrease of the π-electron ring current. Herewith, the double-dipole model was used, which proved itself to be good in the assessment of ring currents of planar porphyrins, including chlorophyll ligand molecules [9, 10, 71–72]. Apparently, an insignificant decrease of the ring current is a specific feature of nonplanar porphyrins. Its presence makes it possible to assume that π-chromophores of such nonplanar molecules shall be as readily polarized under the influence of the electronic effects of substituents, light quanta and chemical reagents. Table 4 Dipole moments of nonplanar porphyrins [22]. Porphyrin

H2TPP (II)

cis-H2(β-Et)4TPP (XIV)

H2(β-Et)6TPP (XV)

H2(β-Et)8TPP (V)

∆Cβ, Å µ (D)

0.23 0

0.76 1.36

0.95 1.89

1.17 1.18


177

Indeed, a recent work [22] has found that dodeca-substituted porphyrins with an increased chemical activity of NH bonds possess, despite the highly symmetric structure, rather significant dipole moments (µ), while in planar H2P pertaining to a similar structural group this value is equal to zero (Table 4). Thus, practically planar tetraphenylporphin (II) has the zero dipole moment, whereas the symmetric octaβ-alkyl substitution of this molecule, leading to the nonR planar structure of the π-system in β-octaethyltetraphenylporphin {V, H2(β-Et)8TPP}, increases this R characteristic of H2P up to 1.2 D. In the case of porphyN NH rins (XIV–XV), which are more planar as compared HN N with compound V but are additionally polarized due to R1 R the asymmetric substitution, the dipole moments are even higher. R1 R The most nonplanar from asymmetrically substituted H2P (compound XV) possesses the maximal dipole moment (1.89 D). An unusual fact is that the dipole moXIV. R = Et, R1 = H; ment of saddle-distorted ligands is directed XV. R = R1 = Et; orthogonally or almost orthogonally to the plane of H2P (see Fig. 4). Ghosh et al. [23] acknowledge that the entire body of changes in the properties of H2P molecules in passing from the planar structure of the macrocycle to a strongly nonplanar conformation can not be explained by only the difference in the geometry of the molecule. Based on the above said, it can be concluded that the cause of changes in the physicochemical properties and reactivity of nonplanar porphyrin ligands, in particular, those caused by an increased chemical activity of coordination core’s NH bonds, is internal polarization of nonplanar molecules. It also determines many of their photophysical properties [22, 49–50] and reactivity [29–30, 52]. Polarization is a consequence of conformational conversions of H2P nonplanar molecules. Such structural changes are associated with the asymmetric change of bond lengths and valence angles, and lead to the redistribution of π-electronic density in it. Thus, they are the cause of secondary electronic changes in the molecule, which affect the reactivity of its reaction centres (N4H2, MN4 etc.). Polarization of H2P molecules under the influence of the electronic effects of substituents has an additional positive effect on the activation of NH bonds. In the case of highly aromatic molecules, such as tetrabenzoporphyrins (X) and porphyrazines (VIII–IX), the

Figure 4 Structural causes of the formation of the orthogonal dipole moment in saddle-distorted porphyrin ligands by the example of H2(β-Et)8TPP (V) [22, 33].

178


planar structure creates the most favourable conditions for their polarization by the light wave, which is exhibited in the change of the photophysical characteristics of the compounds. Besides, this is favourable for internal polarization of H2P by means of substituents of various electronic nature and for the emergence of the push-pull effect at the time when simultaneously donor and acceptor atomic molecular functions (compound XVI) occur [40–43, 71]. The influence of the push-pull effect proper on the reactivity of porphyrin coordination cores and on NH N NH bonds practically has C C NO C C (M e) N not been studied. Only the 2 N N H optical properties were investigated, including the nonlinear optics properties, as well as such parameters of “push-pull” porphyrins XVI as polarizability β and dipole moment µ [40–43]. The highest polarizabilities of the order of 5000·10 –30 esu and dipole moments of more than 10 D were found in “push-pull” porphyrins carrying electron-donor (e.g., NMe2) and electron-acceptor (e.g., NO2) groups in phenyl rings, which are linked with meso-positions of the macroring via an ethynyl spacer [41]. A later work [43] showed the most efficient among π-linking spacers (–CSC–, –N=N–, –CH=CH–, –CH=N–) to be exactly the ethynyl spacer. The consequences from introducing substituents of any electronic nature – donors or acceptors – into the H2P molecule (as a rule, with a chemically low active NH bond) have been studied well [1–3, 9–11]. A change of reactivity of the coordination core of such molecules is observed in complexation reactions of H2P with solvatosalts of d metals [MX2(Solv)n–2] (1) and acid-base interaction ((2)–(5)), as well as of dissociation (6) and metal exchange (7) in MP porphyrin complexes. Peripheral substituents of the same electronic nature in H2P molecules affect different processes differently. The cause of this is oooo MP + nH·Solv+ + nX – MP + M′Xm(Solv)p–m

H4P2+ 2X – + MX2(Solv)n + n–4HX,

(6)

M′P + MXm(Solv)p–m

(7)

the difference of the mechanisms of the above listed reactions and the different degree of involvement of NH bonds in these mechanisms. Thus, the complexation reaction (1) has a strong dependence on the donor-acceptor properties of the solvent, in which it is run, and for H2P with a chemically inactive NH bond (e.g., para-substituted derivatives {H2(p-R)4TPP} of tetraphenylporphin II) is accelerated by electron-donor substituents in phenyl rings, and is slowed down by electron-acceptor substituents [1–3, 73]. The rate of reaction (1) in this case is determined by the strength of reacting solvatosalts of the composition [MX2(Solv)n–2] [74–78] and is, therefore, maximal in carboxylic acids (AcOH, C2H5COOH) and other weakly coordinating solvents, such as esters, ketones, nitriles and alcohols [5]. Reaction (1) of porphyrins with a chemically active NH bond is, in contrast, accelerated by electron acceptors, which are simultaneously coordination-core activators,


179

and slowed down by electron donors [5, 31, 32, 35]. The highest catalytic effect for reaction (1) is achieved in a medium of polar electron-donor solvents, also activating NH bonds in the H2P molecule. A characteristic example are porphyrazines (VIII–IX, XVII–XVIII) substituted in β-positions by electron-acceptor halogen atoms [35, 54]. From the data of Table 5, it follows that the rate constants (kv) of reaction (1) of NH-active macrocycles with ZnAc2 increase with the rise of electronegativity of β-substituents, as well as at an increase Table 5 Kinetic parameters of the formation of zinc complexes with tetraazaporphin (VIII) and its halogen derivatives (IX, XVII–XVIIII) in AcOH–Py mixtures, CH2P = 2·10 –6 mol/l, CZnAc2 = 3.3·10–4 mol/l [54]. Porphyrin H2TAP (VIII) H2(β-Br)4TAP (IX)

Solvent

kv298 , · mol–1

s–1

E, kJ · mol –1

∆S #, J · mol–1 · K–1

AcOH

0.0087 0.0070

50 55

–190 –185

H2TAP (VIII) H2(β-Br)4TAP (IX) H2(β-Cl)4TAP (XVII) H2(β-F)4TAP (XVI)

AcOH – Py = 99–1

0.0054 0.0020 0.0011 0.00074

52 78 77 72

–180 –110 –128 –146


AcOH– Py = 3– 2

0.191 0.197 0.225 0.270

62 71 54 58

–106 –120 –161 –147


Py

5.79 very fast very fast very fast

44

–161

of concentration of the electron-donor component of the solution (Py). Thus, kv in reaction (1) of H2TAP in passing from pure acetic acid to pure pyridine increases more than 2.5-fold, and in the sequence of β-substituents H < Br < Cl < F from 1.5- to 7-fold [54]. The double-action effect of electron acceptor and solvent is exhibited both in the reactivity of the coordination core of rigid [34–37, 53–55] and strongly nonplanar [8, 19–20, 22–23, 29–32, 56–58,71] porphyrins with a chemically active NH bond. Dissociative processes in porphyrin complexes, such as dissociation of MP (6) or metal exchange (7), initiated by the atX tack of electrophilic reagent (H·Solv+, H3O+, HX, M2+) are, N as a rule, accelerated by electron-donor substituents, because X the latter increase the chemical affinity of the reaction centres N NH (nitrogen atoms) in reactions of the type of (6) and (7). Thus, N N for instance, when only the σ-effect of substituents affects the HN N MP dissociation reaction in a solvent–H2SO4 medium, as is X the case in metallotetraphenylporphyrins {MT(R)PP}, it is N accelerated, irrespective of the nature of metal (Fig. 5) {coX valent σ-complexes of Zn(II) and Fe(III), σπfwd, Mn(III) XVII. X = F: complex; σπinv, Cu(II), Ni(II), Pd(II) complexes; predomiXVII. X = F; XVIII. X = Cl XVIII. X = Cl nantly ionic Cd(II) complex}, in the following sequence of

180


N

-ɨɛɪ σ πinv . σ -ɨɛɪ

πinv

.

σ

M

-ɨɛɪ. πinv

N

-ɨɛɪ. πinv

σ

N

N

Figure 5 Electronic effects of coordination in molecules of metalloporphyrins [4–5].

the functional groups located in the phenyl rings of the macrocycles: –NO2, –NH3+, –Cl, –Br, –COOH < H < –CH3, –OH, –OCH3 [79–80]. If a substituent enters into direct contact with the π-macroring when being in β- or meso-positions of the MP molecule, reaction (6) proceeds with account of the electronic effects of substituents ±I and ±C, as well as the electronic effects of coordination – σ-donor–acceptor interaction (N → M), forward dative π-bond (πfwd, N → M) and inverse dative π-bond (πinv, N ← M) (Fig. 5) [1, 4–5]. A number of dissociation rates in such MP can change and even reverse depending on the combination of the above listed effects [80]. Studies of the effect of basicity of substituted meso-tetraphenylporphin (II) (Table 6) on the rates of metal exchange reaction (7) in the {CdP/Zn(ClO4)2/Py/0.5M H2O} system within the temperature range of 288–310 K has shown that electron-donor groups in the phenyl rings of H2TPP accelerate the metal substitution process [81–82]. Thus, the true reaction rate (kv298, l·mol–1 ·s–1) in passing from p-CN to p-OCH3 derivative increases from 1.3 up to 10.2. The metal exchange rates in the considered cases are 600–6000 times higher than the rate of complexation of zinc perchlorate with respective ligands (Table 6). Table 6 Kinetic parameters for the metal exchange reaction of {CdP/Zn(ClO4)2 /Py/0.5M H2O} of meso-substituted CdTPP [81]. Complex CdT(p-CH3)PP CdT(o-CH3)PP CdT(p-OCH3)PP CdT(p-CN)PP CdT(p-H)PP CdT(p-Cl)PP

kexch /kcompl

k298, l·mol–1 ·s–1

∆H #, kJ·mol –1

∆S #, J · mol–1

1640 6700 1000 1400 1530 600

8.3 8.5 10.2 1.2 5.7 1.3

70.2 61.0 58.1 54.7 48.1 45.1

8.7 –21.7 –30.5 –58.5 –68.5 –90.3

The rates of acid-base interaction processes involving the coordination core of the H2P molecule (equations (2)–(5)) react to the electronic nature of substituents in the porphyrin macrocycle in a way traditional for this type of reactions, i.e., the basicity increases and the acidity goes down at the drop of π-electronic density in the macrocycle [61]. Even the introduction of polarizing substituents into H2P molecules with inactive NH bonds (compound II) strongly changes the physicochemical properties and reactivity of the compounds. Thus, the substitution of one of β-positions in the H2TPP molecule (III) stabilizes two trans-tautomers [33, 72], increases the dipole moment of the molecule up to


181

approximately 7 D [33], due to the charge-transfer (CT) interactions cardinally changes the photophysical pattern of the process in its main S0 and excited (S1, T1) states [83]. 1.1.2

NH activation in the course of porphyrin–solution component interaction and porphyrin–solid phase interaction

The environment of the H2P macrocyclic ligand or its complexes in solid, liquid or gas phase plays an important and sometimes determining role for the successful execution of their useful functions [8, 68, 84]. Namely the rapid and significant response in the electronic structure and geometry, in the properties of compounds to the change of the nature of the medium is one of the conditions to recognize it as a biologically active or technically useful substance, e.g., in molecular recognition processes. For this reason, studies of the effect of the phase state of a substance and of the properties of the medium on the progress of various physicochemical processes involving biologically active substances or their models are still topical. Porphyrins belong to just the type of compounds, whose properties strongly depend on the change of environment of the molecules. This occurs, in particular, due to the complexity and multifunctionality of the coordination core of the H2P molecule (MP). As the greater part of chemical reactions proceed namely in the liquid phase, interactions of the “H2P·Solv” and “H2P·solution component” type have been studied better than solid-phase interactions [11, 64, 85]. Herewith, the most investigated interactions are those involving H2P with a chemically inactive NH bond. Their behaviour in solution is readily predictable, and changes of photophysical and coordination properties, solvation and stability of compounds to the action of light, temperature or chemical reagents obey the “classical” regularities, well known in porphyrin chemistry already by early 1980s [1, 9–10]. These compounds in the absence of active functional groups at the periphery of molecules are little soluble in organic solvents, both in weakly solvating (hexane, benzene) or weakly proton-donor ones (chloroform, nitromethane) and in electron-donor ones (e.g., pyridine, DMSO, acetonitrile). Their solubility under standard conditions changes within 1·10–5 –1·10 –3 mol/l [86]. The low solubility of H2P (Table 7) is predominantly due to the universal character of solvation of their molecules as the result of strong shielding by means of the macrocyclic effect of coordination-core atoms – in this case the only centres of specific solvation [6–8, 64]. This type of Table 7 Solubility of porphyrins in organic solvents [86]. S298 ·105, mol/l

Porphyrin

H2TAP (IX) H2TATBP (XXII) H2TBP (X) H(β-Et)8P (I) H2TPP (II) H(N-Me)TPP (XX) H(N-Ph)TPP (XXI) Measured in chlorobenzene(a) and pyridine(b).

Benzene

Ethanol

3.16(a) 39(a) (49)(b) 0.83 (560)(b) 290 590 1980 780

1.9 0.3 – 5.0 1.8 5.3 2.8

182


interaction assumes a weak sensitivity of compounds to the nature of the solvent, e.g., insignificant changes in their photophysical properties [48, 60, 71], high chemical and thermal stability of most H2P in solutions [1–5, 64]. At the same time, the nature of solvent plays a significant role in the change of reactivity of N4H2 coordination core, in particular, in reactions of acid-base interaction [61] and complexation [75]. Owing to the polarization of molecules of NH-active porphyrins [22, 71], their reactivity in solutions of electron-donor solvents significantly increases; this applies not only to nonplanar H2P possessing the intracyclic reaction centres more accessible for reagents [30, 32, 52], but also to planar porphyrins with rigid π-chromophore [53–55]. As we have already mentioned, the rise of aromaticity of the H2P molecule assumes not only the shielding of intracyclic reaction centres at the cost of the MCE structural component, but also the redistribution of the π-electronic density by means of the MCE electronic component, which entails the pronounced chemical activity of particular bonds [7–8]. In a number of chemical processes, the latter of the two factors prevails. At the same time, the solvation characteristics of planar and nonplanar H2P with a chemically active NH bond differ noticeably. Nonplanar H2P are well soluble in organic solvents irrespective of the type of bond, as, e.g., N-substituted porphyrins XIX–XXI with a chemically inert NH bond [85], whereas planar ones are still little soluble, especially in weakly proton-donor media (see Table 7) [86]. Thus, the solubility of predominantly planar compounds I–II with a chemically inactive NH bond in benzene decreases by 1–3 orders with the rise in the macrocycle rigidity (compounds X, IX, XXII). At the same time, for planar NH-active compounds the value of S298 is always higher in electron-donor solvents (e.g., in pyridine) as compared with universally solvating ones (Table 7) owing to the additional specific interaction of the Solv molecule with NH bonds of the coordination core of the H2P molecule. In a medium of weakly proton-donor solvents (EtOH), porphyrins, in contrast, are very weakly soluble, including at the expense of depolarization of NH bonds in them. Indeed, in some cases considered the solvent can regulate the extent of chemical activity of these bonds. Partially delocalized NH bonds of these compounds experience an additional activation (polarization) in a medium of electron-donor solvents [32] and, on the contrary, are completely localized already in a weakly proton-donor medium (carboxylic acids) [31]. These structural changes are reflected in a characteristic jumpwise change of the complexing properties of NH-active porphyrins in solvents of the types considered, and are also seen in the EAS [30, 55] (see Sections 1.2.1 and 1.3). Consider schematically the mechanism of action of electron-donor (Py) and electron-acceptor (AcOH) solvents on the state of the coordination core (N4H2) of H2P molecules of various structural groups: porphyrins proper (classical) and their rigid, or, conversely, strongly nonplanar nonclassical derivatives (Fig. 6) [31]. In molecules of classical H2P (compounds I–II) [28] the macrocycle is almost planar [13, 65], so the access of solvent molecules to –N= and –NH– solvation sites is hindered. Besides, NH bonds in such molecules are polarized very weakly. Both of these factors lead to a situation, when the interaction of NH groupings with molecules of electron-donor Py is almost absent in the main state. Thus, a change of enthalpy in the transfer of H2TPP (II) from inert solvent benzene to electron-donor Py, which characterizes the change of relative solvation, is very small (∆trH0 = –1.9 kJ/mol) [7, 64]. In this case the NH bond remains totally localized (is not activated), and the reaction proceeds with difficulty. Due to the low basicity, as well as owing to structural causes, the molecule of AcOH enters into only a very weak acid-base interaction with tertiary –N= atoms of the classical


183

meso

meso

meso

meso

Figure 6 State of the coordination centre of porphyrin molecules with a chemically inactive group (a) and a chemically active NH bond (b) in proton-donor (AcOH) and electron-donor (Py) media; b1, , transfer of π-electronic plane in the macrocycle [31, 39]. nonplanar H2P; b2, planar H2P;

H2P molecule. The H associate formed (Fig. 6, a) is, apparently, characterized by a low degree of proton transfer, because it is not registered in the EAS. Nevertheless, protons in this associate act as electron acceptors, decreasing the electronic density in the macrocycle and, thus, promoting the rather weak polarization of the NH bond, e.g., in reaction (1). This mechanism works only for predominantly planar porphyrins with a moderate rigidity of the macroring. In the case when the macrocycle acquires a nonplanar conformation under the influence of multiple substitution at the periphery of the molecule or in any other way [8, 16–17, 19, 87] (see also Section 1.4), the basicity of its tertiary nitrogen atoms sharply increases, and the action of AcOH leads to partial or total protonation of these N atoms (equations (4) and (5)) and their blocking as coordination centres in reaction (1) (Fig. 6, b1). Recently, it has been found that strongly distorted dodeca-substiR1 R1 R tuted H2P, e.g., compounds N R1 R1 V–VII, in their titration in toluene NH N N-X N solutions by weak organic bases N N R R form H complexes (or protonN HN N HN transfer complexes, PTC (equaN R1 tions (8)–(9)), which are characR1 terized by a rather profound R R1 R1 transfer of NH proton to the molecule of electron-donor solvent XIX. R = H; R1 = Et; X = Me XXI. and, for this reason, are clearly XX. R = Ph; R1 = H; X = Me XXI. R = X = Ph; R1 = H seen in the EAS [30, 55, 85]. This fact makes it possible to assert that the NH bond in them is noticeably polarized. The distortion of the macrocycle increases the accessibility of this bond for the attack by the Py molecule, and the formation of the H complex with proton transfer to pyridine additionally polarizes the bond, bringing it closer to the state characteristic of the transition state of reaction (1). Indeed, reaction (1) with nonplanar H2P proceeds very slowly in AcOH and easily in Py (see Sections 1.3 and 1.4). The

184


difference in complexation reaction rates in these solvents is up to more than seven orders of magnitude [8, 51, 57, 85]. Another group of nonclassical H2P – most often these are the aza derivatives of porphin [26, 28, 34–37], such as compounds VIII–IX, XVII, XVIII, XXII – is characterized (Fig. 6, b1) by a high aromaticity of the macrocycles and an enhancement of the macrocyclic effect (MCE) [6–8]. The MCE structural component [8] contributes to the shielding of –N= and –NH= centres from the penetration of reagents. This is one of the causes of a decrease of the main properties of tertiary nitrogen atoms of tetraazaporphin (H2TAP, VIII) and tetra-β-bromotetraazaporphin {H2(β-Br)4TAP, IX} [61]. For this reason, there is almost no interaction of them with AcOH. The bridge (meso-) aza atoms are involved in only a weak associative interaction with AcOH molecules, because their unshared electron pairs are involved to a significant degree in the conjugation with the π-system of the macroring [1, 35]. As for the state of the NH bonds, they are completely localized in a solution of AcOH, which does not contribute to the exchange of NH protons of H2P with the medium. Simultaneously, intramolecular effects, such as the enhancement of the MCE π-electronic component [8], evoke a significant polarization of NH bonds in molecules of rigid H2P. This polarization follows from an increase of their acidity [61], which for azaporphyrins in a DMSO medium is 10–12 orders higher as compared with classical H2P. Protonation of NH bonds facilitates the formation of the PTC of rigid macrocycles with molecules of electron-donor solvents, where these bonds are already partially delocalized [55] (Fig. 6, b2). Formation and stability of PTC are considered in more detail in Section 1.2. Apparently, the considerable potential of activating NH bonds and modifying the properties is in the effect of the solid-phase environment on porphyrins. Unfortunately, these issues are far from being understood at present, and there are only separate works on the topic in the literature [44–45, 88–89]. All these publications note the catalytic effect of the solid phase in H2P or MP formation reactions. Tsukahara and Suzuki [44] studied the complexation ability of H2TPP (II) sorbed on Kieselgel 60HR. They showed that its reaction with Zn2+ ions sorbed on SiO2 proceeds much faster in a sequence of weakly polar or nonpolar solvents (CCl4 > toluene > 1,2-dichloroethane > acetone) as compared with the reaction in solution. At room temperature, reaction (1) is complete in only 5 min in all solvents except weakly coordinating acetone at an H2TPP–salt concentration ratio of 1–18/260, whereas in solution it runs slowly and at increased temperatures [75]. This indicates the delocalization of NH bonds under the influence of sorbent’s active centres, as well as the specific adjustment of activated reaction centres of the solvatosalt to the reaction centres of porphyrin. Another example of solid-phase activation of NH bonds was observed when the chlorophyll ligand and its derivatives immobilized on water-soluble polymer material (polyvinyl alcohol) reacted in a glacial acetic acid and its aqueous solutions with Cu(II), Zn(II) and Co(II) acetates at rates not lower and even higher than in the absence of polymer carrier [45]. The above mentioned facts may come as a surprise because at first glance the solid-phase and, moreover, polymer environment strongly hinders sterically the attack of a reagent, e.g., the solvatosalt in reaction (1) on the coordination core of the H2P molecule. Letts and Mackay [88] studied the complexation reaction of H2TPP with Cu(II), Mg(II), Mn(II), Zn(II) and Co(II) ions in a water-emulsion system in the presence of cationic or anionic surfactants, noted both the facts of its inhibition and acceleration by various detergents. In the literature, there are facts of heterogeneous catalysis of not only MP formation reactions (1) but also of condensation processes to form H2P ligands. Thus, for instance, Onaka et al. [89] observed a more than twofold increase of the yield of tetra-meso-alkylporphyrins (35–45%) in a system containing porous montmorillonite K10.


1.2

185

Interaction of organic solvents and porphyrins with delocalized-type bonds

Porphyrins with chemically active NH bonds enter into quite a number of specific interactions involving these bonds. They can be the already mentioned weak acid-base interactions or tautomeric conversions, as well as their combinations. Depending on a particular process and type of ligand, such kinds of interactions can lead to absolutely different consequences – from the destruction of H2P chromophore to the formation of supramolecular structures on its basis. 1.2.1

Acid-base interactions

Traditionally porphyrins are assigned to compounds with weakly pronounced NH-acidic properties [1, 5, 9–10, 61]. The cause of the decrease of acidity of these aromatic tetrapyrrole compounds even as compared with individual pyrrole is, apparently, the spatial shielding of NH sites by atomic-electronic π-macrocyclic environment (the macrocyclic effect) [8]. However, due to the activation of NH bonds by means of the above considered structural polarization factors (the MCE electronic component [8]) they are not capable any more of not only donating a proton to a strong base, e.g., tetraalkyl ammonium hydroxide R4NOH, but of interacting by the acid-base type of interaction with molecules of electron-donor solvents (B) to form specific complexes with proton transfer (PTC) (equations (8)–(9); Fig. 7, a). H 2P + B H2P·B + B

Kk1 k –1 k21 K k –2

H2P·B, H2P·2B.

(8) (9)

Participation in such kind of processes is an individual feature of porphyrins with the delocalized type of chemical bond. It should be noted that planar (porphyrazines, R R N compounds VIII–IX, XVII–VXVIII, XXII– N NH XXV) and nonplanar (dodeca-substituted, comN N pounds IV–VII) porphyrins behave differently in HN N acid-base interaction reactions with electrondonor solvents. First, nonplanar H2P enter into reN R R actions of the type of (8)–(9) in an equilibrium manner [30], whereas for planar porphyrines PTC formation processes are realized in the kinetic reR R gime [55, 90]. Second, although proton-transfer XXIII. R = H; complexes are in both cases unstable in media XXIV. R = Br; with low permittivity (ε), the results of this instaXXV. R = NO2 bility are different. PTC with porphyrazines in low-polar media are, as a rule, subjected to destruction to break down the macrocycle, whereas in the case of dodeca-substituted H2P the decrease of permittivity of the medium only sharply shifts the equilibrium (8) of the formation of the H associate towards the initial compounds. For azaporphyrins, cases are known when the stability of PTC is high and, for structurally thermodynamic reasons, their formation inhibits the coordination reaction (1) R

R

186


λ, nm

λ, nm

Figure 7 Change of electronic absorption spectra in the course of formation in the C6H5Cl–DMSO system (a) and partial destruction in the C6H6 –tert-BuNH2 system (b) of the PTC of β-tetrabromotetraazaporphin (IX) [90].

as compared with media, where the formation of PTC with porphyrins is not registered spectrally, i.e., the extent of interaction of protons of NH groups with molecules of the solvent is lower. No such cases were observed for strongly nonplanar H2P. Interactions of planar derivatives of porphin, in particular, porphyrazines, have been studied in the literature in greatest detail [55, 90–99]. Proton-transfer complexes were first spectrally (EAS) observed in 1961 by Whalley in pyridine solutions of H2(β-Ph)8TAP (XXIII) and called pyridine salts [91], which is not quite correct as the PTC is not a salt but represents a product of incomplete acid-base interaction [92]. The scheme of complete transfer of proton from acid HA to base B has the form: HA + B

K1

AH…B

K1

A – ·HB+

Aδ– …H…Bδ+ K1

A – + HB+,

K1

A –…HB+

K1

(10)

where AH…B is a molecular complex with the H bond, Aδ– …H…Bδ+ is an intermediatetype complex with the delocalized H bond, A –…HB+ is an H-bound ion pair, A – ·HB+ is a contact ion pair, A – + HB+ are complete proton-transfer products. The H2P with partial proton-transfer to the base molecule (B or Solv) represents, depending on the strength of acid and base, permittivity and solvating power of the medium either an H associate (AH…B) or an ion pair (A –…HB+) or else an intermediate complex with the delocalized position of proton (Aδ– …H…Bδ+) (Fig. 8) [90]. Formation of such kinds of products is difficult even in solution and is possible only in the case of weak acids and bases with spatially shielded reaction centres [93]. The kinetics and mechanism of PTC formation with porphyrazines (VIII–IX, XXII–XXV) have been studied in detail in a series of works [55, 90, 94–99]. Based on the spectral and kinetic data, a probable structure of the PTC of the type of


187

Figure 8 Assumed structures of the porphyrin PTC with a weak base molecule B: (a) ionized, (b) H-bound [35, 55, 90].

H2P·2B (Fig. 8) has been proposed. It has also been shown that the formation of kinetically stable proton-transfer complexes is observed only if NH-acidic properties of H2P are sufficiently strongly pronounced, the electron-donor ability of the base is moderate and the permittivity of the medium is high, as, e.g., in DMSO or inert solvent–DMSO mixtures as well as in a DMSO–Py medium [90] (Table 8). Otherwise, e.g., in most media with low permittivity based on aliphatic amines {Et2NH, n-BuNH2, tert-BuNH2PhCH2NH2, Et3N, n-Bu3N, Pip}, PTC are subjected to gradual destruction to colourless products (Table 9, Fig. 7, b). Both the PTC formation and destruction rates are controlled by steric factors. The more bulky the amine molecule and the more shielded its reaction centre, the more complex the proceeding formation and destruction processes of PTC are (Fig. 9, Table 9). The observed phenomena are well described within the framework of the PTC formation and dissociation mechanisms [55, 90]. Table 8 Kinetic parameters of acid-base interaction of H2TAP derivatives with DMSO in C6H5Cl (CH2TAP = 0.5 · 105 mol/l; C 0DMSO = 3.88 mol/l). Porphyrin H2(β-Cl)4TAP (XVIII) H2(β-Br)4TAP (IX)

k298 ·105, l·mol–1 ·s–1

E, kJ · mol –1

∆S #, J · mol–1 · K–1

3.47 5.30

24 26

–250 –237

Table 9 Kinetic parameters of acid-base interaction of H2TAP derivatives with N bases in benzene {CH2TAP = 0.4–0.5·10–5 mol/l}. Porphyrin

N base

2 ·mol2 ·s–1

k298 ·102,

E, kJ · mol –1

l

∆S #, J · mol–1 · K–1

H2(β-Br)4TAP (IX)

BzNH2 n-BuNH2 tert-BuNH2 Et2NH

0.78±0.04 6.6±0.3 0.076 4.0±0.2

29±2 11±2 18 15±2

–226±5 –270±6 –283 –261±5

H2(β-Cl)4TAP (XVIII)

BzNH2 n-BuNH2 tert-BuNH2 Et2NH

0.65±0.03 3.8±0.1 0.09 1.10±0.05

31±3 23±2 28 30±3

–223±10 –235±7 –251 –218±10

188


Figure 9 Dependence of log k of the destruction of H2(β-Br)4TAP (IX, filled dots) and H2(β-Cl)4TAP (XVII, empty dots) in benzene in the presence of basic nitrogens at 298 K. eff

DMSO

Figure 10 Dependence of log keff on CDMSO for reaction (1) of formation of Mg(β-,4- NO2Ph)8TAP in benzene at CMgAc2 = 1.8·10 –4 mol/l; T = 308 K (1), 318 K (2), 328 K (3).

PTC formation processes are facilitated in the following sequence of compounds: H2TAP (VIII) < H2(β-Ph)8TAP (XXIII) < H2(β-,4-BrPh)8TAP (XXIV) < H2(β-,4NO2Ph)8TAP (XXV) < H2(β-Br)4TAP (IX) < H2(β-Cl)4TAP (XVIII), which is determined by the enhancement of the proton-donor properties of the ligands [35–36, 55, 90]. Thus, while compound IX enters into acid-base interaction with molecules of bases already under normal conditions, tetraazaporphin (VIII) and β-octaphenyltetraazaporphin (XXII) form PTC only in prolonged boiling in a solution of base B. Formation of PTC is a process of chemical activation of NH bonds in the H2P molecule by organic solvents.


189

λ, nm

Figure 11 Change of electronic absorption spectra for H2(β-Br)8TPP (VII) in toluene in the presence of piperidine, CH2P = 4.23·10 –4 mol/l; Cpip, mol/l: 0 (1), 0.008 (2), 0.016 (3), 0.025 (4), 0.033 (5), 0.041 (6), 0.058 (7), 0.083 (8), 0.083 (9), 0.166 (10).

lg CHP- / CH2P

Figure 12 Graphic determination of the number of base B molecules participating in the equilibrium with H2(β-Br)8TPP (VII) in toluene: 1, Pip; 2, DMF; 3, Py; 4, DMSO; n = tan α ≈ 1.

Indeed, in a number of cases H2P molecules activated in this way enter more readily into complexation reactions (1). At the same time, it should be noted that the complexing ability of PTC in reaction (1) is determined by its kinetic stability and can be both strongly increased and decreased with the change of the concentration of the base in the same reacting system (Fig. 10). The ability of strongly nonplanar dodeca-substituted porphyrins to interact with electron-donor solvents was found comparatively recently [29–30]. It was revealed that the substitution of a weakly solvating (inert) solvent (benzene, toluene) by an electron-donor solvent completely changes the structure and number of bands in the electronic spectrum of the molecule (Fig. 11) [30]. The fact that the number of bands decreases, and the EAS starts to resemble more the spectrum of a metal complex, indicates the formation of structures with a higher symmetry (D2h → D4h). Similar spectral changes were also observed in the interaction of porphyrazines with bases [55, 90]. In contrast with porphyrazines, PTC in dodeca-substituted porphyrins are formed in an equilibrium manner. Karmanova et al. [30] calculated the stability constants of forming delocalized acid-base forms from the results of titration of solutions of H2(β-Br)8TAP (VII). The stability of PTC proved to in-

190


crease with the rise of polarity of the medium, even if the electron-donor properties of the base weaken: Py < DMF < DMSO < Pip. In contrast with solutions of porphyrazines, the breakdown of PTC proceeded in DMSO without the destruction of ligands at the addition of N bases. Besides, it was found that in the course of titration of compound VII by the base only a PTC of the type of H2P·B is formed (equation (8)). In particular, this follows from the value of the slope of the dependence log CSolv = f (log CHP– /CH2P) (see Fig. 12). Table 10 Stability constants (Kst) of proton-transfer complexes (PTC) and thermodynamic parameters of acid-base interaction of H2 (β-Br)8TPP) (VII) with organic solvents [30]. Base solvent

Py DMF DMSO Pip

1.2.2

Characteristic of solvent DN

ε

33.1 26.6 29.8 –

12.3 36.7 46.7 4.28

Kst298

∆H, kJ · mol –1

∆S #, J · mol–1 · K–1

0.025±0.001 0.87±0.05 2.87±0.17 6.3±0.4

–18±1.1 –11±0.5 –17±1.6 –35±0.7

–90±4 –36±2 – 49±4 –102±3

Tautomeric processes

For porphyrins proper, not containing additional active functional fragments, e.g., mesotetraphenylporphin (H2TPP, II), tautomeric conversions are little characteristic. For them, as for any other H2P, the coordination core of the molecules of which contains pyrrole and pyrrolenine atoms of nitrogen, only the process of N–NH tautomerization is known (Fig. 13), whose conditions and mechanisms are considered in [35, 48, 60, 72,100–109]. This type of tautomerization weakly depends on the nature of solvent, and the energy of its activation is about 12 kcal/mol and little depends on the phase state of H2P. Quick-time N–NH exchange by protons is reduced to the conversion of two equivalent or practically equivalent trans-tautomers in each other. The synchronous (I) and asynchronous (II) mechanisms of ND–NH tautomerization is possible [100] (Fig. 13). A number of quantum chemical [48, 108–109, 101, 104], 1H NMR spectral [60, 72, 102–103] and photophysical [48, 100, 105–106] data are in favour of the asynchronous mechanism,. The two-step process II with thermally activated quantum mechanical tunnelling through the activation barrier is realized at T = 200–300 K [101] and in the case of predominantly planar, symmetrically substituted H2P is characterized by the rate constant ≈5·104 s –1 in CDCl3 [102]. In the course of process II the activation energy of the system is required only to achieve the cis-tautomer, which then is tunnelled by route II [107]. At temperatures ≈77 K the synchronous mechanism of tautomerization predominates, however, the rate of the process is much lower (≈10 –6 s –1 at 4.2 K) [101, 107], and the calculated activation barrier of the process is ≈18.5 kcal/mol [104]. The type II process (Fig. 13) can also run with participation of excited states of the H2P molecule as photoisomerization [105, 108]. A number of indirect proofs of its running with participation of the triplet T1 state have been found. Thus, Shushkevich et al. [106], by the example of tetrabenzoporphin (compound X), assessed the rate constant of NH rearrangement of the H2P molecule in triplet state to be ≈ 10 –1 s –1 (at 77 K). Introduction of an efficient electron-acceptor [107] or “heavy” [102] substituent into a pyrrole ring or meso-position of the macrocycle stabilizes NH tautomers. For instance, meso-substitution of the β-octaethylporphin (I) molecule by a bulky 2-(4-nitrobenzyl)-1-


NH

HN

N NH

N

N

HN

191

N II

N

HN

I

NH

N

Figure 13 Synchronous (I) and asynchronous (II) routes of N–NH tautomerization of porphyrins.

naphthyl fragment decreases the rate of the process to 0.004·104 s –1, i.e., by about three orders of magnitude. NH tautomerization does not change the properties of the H2P molecule significantly, because the rearrangement of the π-conjugation contour in the conversion of tautomers leads neither to the change of symmetry nor to the activation of the already available reaction centres or the emergence of new reaction centres. . Sophistication of the H2P structure can enrich the reactivity to a significant degree and often provides a possibility for the existence of the molecule in various molecular forms depending on the nature of the medium, on the presence of some or other reagents in solution. These manifestations determine the whole range of practically useful properties of the molecules [110]. One of the examples of such compounds are H2P carrying active (one or several) hydroxy or amine groups in meso-positions. In solutions, these compounds enter into tautomeric conversions of, respectively, keto-enol or amine-imine type. As a rule, being in a nonionized medium, these compounds exist in keto or imine form, which is reliably confirmed by spectral methods and follows from the data on their reactivity. Keto-enol equilibria of oxoflorins have been studied the most [111–112]. Existing in nonionized media in a keto form, oxoflorins (XXVI) are subjected to tautomerization during the action of strong acids (XXVIb), as well as in the course of coordination of these nonaromatic ligands by double-charged metal ions (XXVIc), and form meso-hydroxyporphyrins (enols) (Fig. 14) possessing all the features of aromatic molecules. For this reason, oxoflorins are stronger acids and bases as compared with H2P, they also enter more readily into complexation reactions [111]. The conversion mechanisms of oxoflorons (Fig. 14) have not been studied; however, it is evident that the process in this case involves chemically activated NH bonds. Chlorophyll and its structural analogues are also subject to keto-enol tautomerization involving the cyclo-pentane ring in accordance with scheme (11) [1]. As in the case of simpler molecules, the tautomerization in chlorophyll a (H2Cl a, XXVII) can be shifted towards ketone or enol, which is controlled by the nature of the peripheral substituents in the moloooecule, the oo

N

N M N

N

XXVc

MX2 -2HX

NH H N N

HN

XXV

OH

O

O

OH

HX -X

NH H N N H HN

XXVa

Figure 14 Tautomeric forms of oxoflorins.

HX -X

NH HN N H HN

XXVb

192

D.B. Berezin and B.D. Berezin H 2C

CH

H 3C

H H 3C H 2C

C H3

HN

N

C H3 H

H 2C C 20 H 39 O O C

C2H 5

N

NH

H

O C O O C H3

XXVII

molecule, the nature of the medium, by other factors. Thus, enol (compound XXVII) is formed not only in the medium of alcoholic alkaline [1, 113–114], but also in an ethanol–glycerol (1:1) system and also in solutions containing a large amount of the second component. The enol form of chlorophyll a (XXVI) is reactive. Thus, in an alkaline medium in the presence of air oxygen it is unstable and reacts to break down the cycle (11).

O2

NaOH, ROH H3COOC H

O

-

+

O Na

H3COOC H

C C + H3COOC O O O Na

(11)

Inversion of one of the pyrrole rings in the formation of the tetrapyrrole macrocycle of H2P leads to the formation of a new class of compounds – single-inverted analogues of porphyrins [115–116] (compounds XXVIII–XXIX). They, in particular, compound XXVIII, are characterized by an unusual type of tautomerization (12) associated with the intramolecular transfer of one of the intramolecular NH protons (the form XXVIIIa) to the external tertiary atom of the molecule (the form XXVIIIb: XXVIIIa + Solv

K1

XXVIIIb…Solv.

(12)

Ph The existence (in the porphyrin anaPh logue II) of individual tautomeric forms 7 5 3 RN 8 N 2 21 a (2-aza-21-carbatetraphenylporphin, 22 CH N N CH H 2 21 H2[ N, CH]TPP) and b (2-iminoPh 10 Ph Ph Ph 20 N HN 21-carbatetraphenylporphin, H[2NH, NH N 24 23 12 21 18 15 13 CH]TPP) in a medium of nonpolar 17 and polar solvents, respectively, was Ph Ph first found in [117–118]. As shown by further investigations, tautomeric proXXVIII a. XXVIII b. R = H cesses of this type take place in inverted XXVIII. R = CH3 H2P of any structural groups, including the analogue of unsubstituted porphin [119], β- and meso-substituted compounds [120], and play a role in the formation (by these ligands) of numerous structural types of complexes possessing the catalytic, receptor and other useful properties [121–123]. Berezin and coworkers [124], attracting a wide range of media with differing polarities


193

A 3.0

ɚ 2.5

b c

2.0

A/6

1.5

b

a

1.0

c 0.5 0.0 400

450

550

600

650

700

750

λ, nm

Figure 15 Electronic absorption spectra of compound XXVIII (c = 1.5·10 –5 mol/l) in organic solvents: a, C6H6; b, DMSO; c, Py.

and coordinating abilities, used electron absorption and fluorescence spectroscopies, as well as 1H NMR spectroscopy, to study the state of compounds XXVIII and XXIX in solution and to obtain stability constants of the tautomeric form XXVIIIb in benzene in the presence of electron-donor components of the solution – N,N-dimethylformamide (DMF), dimethylsulfoxide (DMSO), hexamethylphosphotriamide (HMPTA) and N,N-dimethylpropyleneurea (DMPU). Tautomeric conversion (12) is accompanied by a significant rearrangement of the main contour of π-conjugation in the molecule of compound XXVIII, which, in particular, follows from the EAS data (Fig. 15, a and b). The band structures in the EAS of H2TPP (II) proper and of the tautomers of its inverted analogue XXVIII are also strongly different. At the same time, Belair et al. [118] have shown that the EAS of compound XXVIII can be also described by the Platt–Gouterman four-orbital model used to calculate the absorption spectra of porphyrins proper [9–10, 33, 48, 59–60, 107–108]. Inversion of one of the pyrrole rings in macrocycle II leads to a strong batochrome shift of all EAS bands. For instance, the intensive Soret band (B band) in the region of 420–450 nm and the batochrome band I (Qx, 650–730 nm) are shifted by 19 and 81 nm in the case of tautomer a of compound XXVII and by 24 and 51 nm in tautomer b relative to the bands of compound II. The quantum chemical calculations also demonstrate a long wavelength shift of the EAS band corresponding to the electronic transition S0→S1 and an increase of the Qx band oscillator strength, occurring in the sequence of compounds II < XXVIIIb < XXVIIIa, which is explained by a decrease of the singlet excited state S1, especially for tautomer XXVIIIa [124]. Significant differences in the spectra of the tautomeric forms of compound XXVIII (Fig. 15) make it possible not only to reliably identify them by the spectral data, but also imply the individual character of the forms in solution. Thus, according to the EAS data, the individual form a exists in a medium of nonpolar solvents with weakly pronounced electron-donor properties, such as hexane (C6H14), benzene (C6H6), toluene, dichloromethane (CH2Cl2), chloroform (CHCl3). The type of the spectrum of tautomer a is given in Fig. 15, a. As a rule, the permittivity (ε) of such solvents is not higher than 5, and the donor numbers (DN) do not exceed 15 (Table 11) [68, 75].

194


Table 11 Properties of organic solvents and forms of existence of compound XXVIII in solutions. Solvent(a) C6H14 C6H6 CHCl3 AcOH TFA Diox (CH3)2CO NM PrOH-1 CH3CN THF Py DEA Pip DMF DMSO HMPTA DMPU

ε (µ)

DN

AN

Spectrally registered forms(b)

1.88 (0.08) 2.28 (0) 4.72 (1.15) 6.15 (1.70) 8.26 (2.28) 2.20 (0.45) 20.70 (2.10) 37.78 (3.56) 20.10 (1.65) 36.02 (3.44) 7.58 (1.75) 12.30 (2.37) 3.60 4.28 (3.87) 36.70 (3.80) 46.68 (3.96) 30.00 (5.54) 36.10 (4.23)

0 0.1 4.0 20.0 – 14.8 17.0 2.7 – 14.1 20.0 33.1 50.0 – 26.6 29.8 38.8 34.0

0 8.2 23.1 52.9 105.3 10.8 12.5 20.5 37.5 18.9 8.0 14.2 – – 16.0 19.3 10.6 –

a a a H4P2+ H4P2+ a a+b a + H4P+ a + H4P+ + H5P2+ a+b a+b a+b a+b a+b b b b b

(a)

Characteristics of the solvents were taken from [13, 14]; (b) H4P+ and H5P2+ are designations for single- and double-protonated forms of compound XXVIII.

Tautomer b is realized in a medium of polar electron-donor solvents, e.g., N,N-dimethylacetamide (DNAc) [117–118], DMF, DMSO (Fig. 15, b), HMPTA, DMPU. Tautomeric equilibrium (12) is shifted towards the form b only if a number of conditions are observed: the permittivity (ε), dipole moment (µ) and donor number (DN) of the solvent are sufficiently high, and the acceptor number (AN) does not exceed the donor number: ε ≥ 30, µ > 3.5 D; DN > 25: DN ≥ AN. Thus, in acetonitrile (CH3CN), a solvent with ε = 36 but low DN = 14.1 at AN = 18.9, the EAS of compound XXVII are already observed to register a mixture of tautomers a+b (Table 11). A similar situation is realized in pyridine (Py), despite its high value of DN (Fig. 15, c; Table 11). A mixture of forms a and b is also registered in other electron-donor solvents with low values of ε and µ, such as diethylamine (DEA, ε = 3.6; DN = 50) or piperidine (Pip, ε = 4.28) [68, 75]. Tautomeric conversion (12) can be registered by various spectral methods. Thus, the dependence of the position of short-wavelength maxima in the fluorescence spectra (754 nm in toluene and 721 nm in DMF), as well as a change of the structure of the emission bands depending on the nature of the solvent, is well consistent with the views of the existence of two stable tautomeric forms in compound XXVIII (Fig. 15). The fact that in both these solvents the fluorescence spectrum does not change in the variation of excitation and registration wavelength (Fig. 16, a and b) indicates the presence of only one of the two spectral forms of porphyrin XXVIII in each of the given media. Earlier, using a similar method, the individual character of forms XXVIIIa and XXVIIIb in CHCl3 and DMAc was confirmed [118]. In contrast, the fluorescence spectrum of compound XXVIII in Py strongly depends on the excitation wavelength (Fig. 16, c), thus dem-

195


λ, nm Figure 16 Steady state fluorescence spectra of compound XXVIII depending on the excitation wavelength (λexc = 360–460 nm) in toluene (a), DMF (b) and Py (c). T = 298 K.

onstrating the presence of a mixture of compounds – tautomers a and b in a pyridine solution. The state of inverted porphyrins in solutions is confirmed by the data of 1H NMR spectra (Table 12) obtained for 2-aza-21-carba-tetraphenylporphin (XXVIII) and its methylated analogue 2-(N-methylase)-21-carba-tetraphenylporphin (H[2NCH3,21CH]TPP, XXIX) in CDCl3 and DMSO d6. Transition of compound XXVIII from the tautomeric form a to form b is accompanied by the transfer of intracyclic proton NH, shielded by the ring current of the aromatic π-system of the molecule, to deshielded peripheral atom of nitrogen (2N). Indeed, the tautomerization in the medium of DMSO d6 leads to the emergence of a unit sigTable 12 1H NMR spectral characteristics of inverted H2P analogues in CDCl2 and DMSO d6 [124]. Compound

Chemical shift, δ (ppm)

Solvent

δ 21CH

(internal)

δ NH

(β + Ph)

δ CH

(external)

δ NH

δ NH3

XXVIII

CDCl3 DMSO d6

–5.07 (1H) 0.68 (1H)

–2.34 (2H) 2.22 (1H)

7.61–9.81 7.52–8.13

– 12.81 (1H)

– –

XXIX

CDCl3 DMSO d6

0.98 (1H) 0.87 (1H)

3.62 (1H) not determined

7.40–7.95 7.44–8.01

– –

3.32 3.52

nal (δ = +12.81 ppm) in compound XXVIII in a weak field, corresponding to NH proton in position 2 (Table 12), whereas in tautomer a in CDCl3 this spectral region is clean. Conversion (12) is accompanied by a very significant weak-field shift of signals of intracyclic CH and NH protons of molecules of XXVIIIa and XXVIIIb (by 5.75. and 2.56 ppm, respectively), and, the other way round, a shift of peripheral Cβ and C-phenyl protons into a strong field (by 1.78 ppm). This is indicative of a significant decrease of π-electronic ring current in the case of compound XXVIIIb as compared with XXVIIIa, i.e., of the decrease of aromaticity of tautomer b. The data of 1H NMR spectra for tautomers XXVIIIa and XXVIIIb are consistent with the results of [125]. As seen from the data of Table 12, inverted porphyrin XXIX, in which the extracyclic atom of nitrogen 2N is methylated, was not subjected to tautomerization (12), so the 1H NMR spectra of this compound in CDCl3 and DMSO d6 are close. Methylation of nitrogen

196


Figure 17 Correlation of the value of Kt of compound XXVIII in reaction (12) with characteristics of organic solvents: a) Kt = f(µ); b) Kt = f(DN); 1, DMF; 2, DMSO; 3, DMPU; 4, HMPTA.

atom 2N fixes in this molecule the conjugation contour, which corresponds to that of the tautomeric form b of nonmethylated analogue of XXVIII. As the result, by the extent of aromaticity and position of proton signals the macrocycle of XXIX is close to the tautomeric form XXVIIIb (Table 12). The increase of the acceptor number AN of the solvent relative to the donor number DN indicates an increase of its proton-donor properties, so equilibria (13) and (14) can be realized in a medium of solvents with DN < AN. As the result, a mixture of tautomer XXVIIIa with single- (H2[2NH,21CH]TPP+) and double- ({H3[2NH,21CH]TPP}2+) protonated forms of compound XXVIII is formed in solution (because the permittivities of these media are, as a rule, not high), if the acceptor number of the solvent AN < 50 as in weakly proton-donor propanol-1 (C3H7OH-1) and nitromethane (CH3NO2); or only the doubleprotonated form is formed, when AN > 50, e.g., in a medium of acetic acid (AcOH) or trifluoroacetic acid (TFA) (Table 11). According to the data of [115], the first proton in the molecule of XXVII is added to the external, more sterically accessible nitrogen atom 2N, and the second to the intracyclic atom 23N. H2[2N,21CH]TPP + HX

K1

{H2[2NH,21CH]TPP}+X – + HX

{H2[2NH,21CH]TPP}+X – , K1

{H3[2NH,21CH]TPP}2+2X –

(13) (14)

The tautomers of compound XXVIII (a and b) are converted into each other in an equilibrium manner during the change of the composition of the inert solvent–electron-donor solvent system (12). Berezin et al. [124] obtained tautomerization equilibrium constants (Kt = [b]/([a]·[Solv]2, where [a], [b] and [Solv] are the concentrations of tautomer a, tautomer b and titrant, respectively, for process (12) in C6H6 –DMF, C6H6 –DMSO, C6H6 –HMPTA and C6H6 –DMPU systems. As seen from the data of Table 13, the stability of the tautomeric form b increases with the increase of solvent’s polarity, which is characterized by the values of ε and µ, and its donor strength expressed by the donor number DN. The values of Kt were found to correlate the best with the values of µ and DN (Fig. 17, a and b). Apparently, the DN value of the solvent determines the strength of the bond of the donor site of compound XXVIIIb with the extracyclic NH proton, and the high dipole moment µ favours the shift of equilibrium (12) to the right owing to the better stabilization of ionic particles, formed in the conversion,


197

by the solvent. At the same time, this correlation is observed only for solvents, whose donor strength and polarity parameters are within a certain interval favourable for the formation of tautomer b. It should be noted that for bulky HMPTA the value of Kt is much lower (Fig. 17) than it could have been expected based on the values of its µ and DN, which are maximal in the sequence of solvents considered (Table 13). Evidently, this is explainable by the worsening of the steric conditions for the solvation of the 2NH grouping of inverted porphyrin by the bulky molecule of the solvent. Thus, the molar volumes (VM) for DMF, DMSO and DMPU are within the limits of 100 cm3/mol, whereas VM for HMPTA is almost twice as high (Table 13). Table 13 Tautomerization constants of porphyrin XXVIII in a C6H6 –donor solvent medium and characteristics of the solvents [124]. Solvent DMF DMSO HMPTA DMPU(b) (a)

Kt

ε(a)

µ, D(a)

0.23±0.01 0.39 ±0.01 0.47±0.02 0.67±0.01

36.7 46.7 30.0 36.1

3.8 3.96 5.54 4.23

DN(a) 26.6 29.8 38.8 34.0(c)

VM(a), cm3/mol 77.4 71.3 175.7 98.5

Characteristics of the solvents were taken from [68, 75]; (b) from [126]; (c) donor strength Ds.

The value of Kt in the temperature range of 298–328 K within the limits of experimental error does not depend on temperature [124]. From this, it follows that tautomerization process (12) proceeds with the activation energy E close to zero, which is consistent with the data of quantum chemical calculations of tautomers a and b [109, 124, 127]. Thus, the equilibrium state of (12) is largely controlled by entropy factors and, at an active participation of a solvent in the process, by solvation factors. Indeed, earlier studies showed tautomer b to be stabilized by the molecule of an electron-donor solvent even in a crystal [117]. Several ways are possible to realize the tautomerization mechanism of inverted analogues of porphyrins: within the involvement of a solvent by tunnelling or intermolecular transfer of NH proton, as well as owing to its intramolecular transfer with participation of a solvent [117, 128]. In the opinion of Furuta et al. [117], intramolecular proton transfer in the course of the tautomeric conversion of compound XXVIII is performed owing to the emergence of a strongly nonplanar conformation, in which the inverted pyrrole ring is included towards the coordination core, and the inner and outer nitrogen atoms are brought closer to each other. This treatment of the mechanism appears to be open to question, as the reaction centres of a tetrapyrrole molecule involved in proton transfer are too far away one from the other to enter into direct interaction, and the approximation of these centres to at least a distance of an effective H bond (≤3 Å) should strongly destabilize the molecule. According to the data of quantum chemical calculations and XDA data, both tautomers of compound XXVIII are almost planar and, besides, close energy-wise molecules. The difference in energy of the two tautomers is 3.4–5.7 kcal/mol according to the data of [109, 127]. The strong dependence of the value of Kt on the nature of solvent, the evident important role of the entropic contribution to the Gibbs energy of the tautomerization process, as well as the fact of the involvement of two solvent molecules in the proton transfer found by us (it follows from the values of tan α of the indicator dependence [log(Cb/Ca) = log (CSolv)

198


Figure 18 Dependence of the indicator relation (logCb/Ca) of the concentration of the tautomeric forms of compound (II) in benzene on the logarithm of the concentration of the donor solvent (log Csolv): 1, DMF; 2, DMSO; 3, DMPU; 4, HMPTA. tan α = 2.

N

N CH

HN

NH

N

CH

Solv - H Solv+

-

N

K1

K ɢɡɨ iso

N

a1

a

-N

Solv ...HN CH

N

N

HN

b1

Figure 19

HN

H Solv+

K2

CH

N

N

HN

b

Schematic tautomerization mechanism of inverted porphyrin analogues [124].

(Fig. 18)) enable us to propose an alternative mechanism of tautomerization involving a solvent (Fig. 19). According to the mechanism, conversion of tautomer a to tautomer b is realized via a number of consecutive equilibrium stages. At the first stage (a ↔ a1), there is a polarization under the influence of the molecule of electron-donor solvent (Solv), and then complete dissociation of one of the intracyclic NH bonds in the molecule of inverted porphyrin (Fig. 19). The proton can with equal probability be abstracted both from 22N and 24N – in both cases the second stage (a1 ↔ b1), of the isomerization proper of the conjugation contour, is accompanied by intracyclic cis-tautomerization (22NH→23N or 24NH→23N). Presumably, abstraction of the proton from the 22N grouping should be more probable due to its increased acidity. In this case, the proton is located slightly closer to the electronegative 2-aza atom. The last, third stage is accompanied by the protonation of the peripheral nitrogen atom (b1 ↔ b). Intermediate particles a1 and b1 are highly reactive and are not revealed in the EAS in any way. For this reason, the total tautomerization constant Kt (15) was calculated proceeding from the expression for the constants of the elementary stages of the process,


199

(16)–(18). Within the framework of this approach, the participation of two molecules of electron-donor solvent in the course of tautomeric process (12) is well explained. Kt = K1 ·Kiso ·K2 ,

(15)

K1 = [a1]/([a]·[Solv]),

(16)

Kiso = [b1]/[a1],

(17)

K2 = [b]/([b1]·[Solv].

(18)

Thus, inverted analogues of porphyrins, such are compound XXVIIIa, also pertain to tetrapyrrole macrocycles with chemically active NH bonds (see the mechanism). Tautomer XXVIIIb is less aromatic and, therefore, possesses no NH activity. 1.3

Quantitative assessment of the state of NH bonds in porphyrin molecules

The most evident quantitative assessment of the chemical activity of NH bonds in H2P molecules is the direct determination of their acidity by equations (2)–(3) (Section 1.1.1). However, the reliable and comparable data by the acidic dissociation constants of porphyrins are far from being readily accessible [61]. In this connection, in [26, 58] it was proposed to use a complex of criteria of state and reactivity of NH bonds in porphyrin molecules. They include the 1H NMR, kinetic and quantum chemical criteria. 1.3.1

Spectral criterion

1

The H NMR spectral criterion is based on the value of the chemical shift (19) of H2P NH proton signals in the spectrum during the substitution of inert, weakly solvating solvents (SI = C6H6, CHCl3, CH2Cl2), in which NH bonds of the porphyrin molecules are always localized, by strongly electron-donor solvents (SB = DMSO, Py), in the medium of which protons of these bonds can be delocalized. The weak-field shift of the value ∆δNH = δSI – δSB,

(19)

of δNH in porphyrins can be caused by two main causes – activation of the NH bond with abstraction of protons by a donor solvent from the region of their shielding by the ring current of the π-system and disturbance of the planar structure of the molecule [72]. The type and depth of nonplanar conformation varies in various H2P [19], so it would not be correct to compare the values of δNH of these compounds to make a judgement on the delocalization of the NH bond. When substituting the solvent, the conformational composition of H2P molecules does not in fact change [14], and the NH bond can only be activated, so to assess the latter, one should use namely the value of the relative chemical shift of the signals of NH protons, ∆δNH (19). It is positive, if porphyrin has chemically active NH bonds, and negative in the cases when the bonds, as in classical H2P, can not be activated by electron-donor solvent (Table 14). According to the 1H NMR spectral criterion, the sequence of chemical NH activity of H2P considered is as follows: IV kv(SA).

(20)

The values of the constants given in Table 15 make it possible to determine compounds I–II as classical H2P, and compounds IV–VII and VIII–X as nonclassical ones, and indicate an increase of the chemical activity of NH bonds in the following sequence of porphyrins pertaining to groups (2) and (3) (Fig. 3): X < VIII < IX and IV < V < VI < VII. From the data of Tables 14 and 15, it follows that polarization of the H2P molecule by peripheral substituents, in particular, those of opposite electronic nature (the push-pull effect) enhances the chemical activity of NH bonds. This is well seen by the example of H2(β-Br)8TPP (VII), carrying eight electron-acceptor functional groups in β-positions of the molecule, as well as compounds VIII–IX (Table 15).

201


Table 15 Dependence of the rate of reaction (1) of porphyrin with Zn(II) and Cu(II) acetates (kv, l ·mol–1 ·s–1) on the donor-acceptor nature of a solvent. Porphyrin

Concentration of salt, CMAc ·10–3, mol/l 2

Rate constant (kv, l ·mol–1 ·s–1) and direction of its change AcOH

H2(β-Br)4TAP (IX) H2TAP (VIII) H2TBP (X) H2(β-Et)8P (I) H2TPP (II) H2TPTBP (IV) H2(β-Et)8TPP (V) H2(β-Ph)8TPP (VI) H2(β-Br)8TPP (VII) (a)

M=Zn, (0.3) M=Zn, (0.3) M=Zn, (2.6) M=Zn, (5.0) M=Zn, (0.48) M=Zn, (2.6) M=Cu(II), (0.05) M=Cu(II), (0.005) M=Zn(II), (0.22)

Ref.

Py

0.007±0.0003 930±140 0.0087±0.0003 5.79±0.04 slow 0.0069±0.0018 3.67±0.05 slow 38.6±0.90 slow slow 0.100±0.001 0.00065±0.00004 1230±40(a) slow 754±20 slow fast

→ → → ← ← → → →

37,53–54 37,53–54 31–32,38 1,75 1,75 31–32 51–52 51–52 130

Csalt = 0.005 mol/l.

Chemical activation of NH bonds by molecules of an electron-donor solvents is exhibited not only in the change of the rate constants, but also of other kinetic parameters of indicator reaction (1), in particular, its entropic characteristics. A recent work [85] found significant differences in the kinetic parameters of reaction (1) of the complexation of N- (XIX–XX) and dodeca- (V–VI) substituted molecules of porphyrins taking in solution nonplanar conformations, which are predominantly saddled, but differ in symmetry and degree of deformation [15, 19]. It is not hard to notice (Table 16) that the change of entropy ∆S # in the course of activation of reaction (1) reagents for dodecasubstituted porphyrins is always more positive than for N-substituted porphyrins, but only an electron-donor solvent (DMF). The change of entropy of the reaction proceeding in solution without the change in the number of moles of substance is related to the change of solvation of particles in the formation of the transition state [1–8, 11, 21, 25, 31–32, 34–39, 52–54, 57, 74–77, 84–85]. From this perspective, the rise of the value of ∆S # in reaction (1) can be explained by either the desolvation of the transition state, which is little probable, or by an additional solvation of the initial reagents, achieved under certain conditions. In our opinion, the rise of ∆S # of the complexation reaction of NH-active dodeca-substituted porphyrins in an electron-donor solvent is directly related to the ability to form PTCs [30]. In the case of a nonpolar medium or NH-inactive compounds, there is no increase of ∆S # (Table 16). Formation of PTC leads to the increase of solvation of the initial H2P in reaction (1) and its acceleration owing to NH activation as compared with NH-inactive porphyrins. It is known [1, 5, 25] that one of the most energy-intensive contributions to the total energetics of reaction (1) is the dissociation of NH bonds in the H2P ligand. Formation of PTC is characteristic of dodeca-substituted H2P and not characteristic of N-substituted porphyrins (see also Section 1,4), which is reflected on the activation parameters of the reaction (Table 16). Earlier studies of the kinetics of coordination of tetrabenzoporphyrins with chemically active NH bonds by zinc acetate in a pyridine–diethylamine medium found [32] that, with the rise in the concentration of the base DEA, an increase of the rate of entropy is observed along with an increase of the reaction rate in the course of intermediate state formation. ooooo

202


Table 16 Effect of the nature of porphyrin on the entropic characteristics of complexation reaction (1) with Zn(II) and Cu(II) salts in C6H6 and DMF. Porphyrin

kv298 , l· mol –1 ·s–1

Ea , kJ · mol –1

∆S #, J · mol–1 · K–1

Salt

Solvent

H(N-Me)(β-Et)8P (XIX)

CoAc2 Co(Acac)2 Co(Acac)2

DMF DMF C6H6

30.94±1.39 18.13±0.69

53.6±1.3 59.7±1.8 very slow

–63±2 –29±1

H(N-Me)TPP (XX)

ZnAc2 Zn(Acac)2 Zn(Acac)2

DMF DMF C6H6

18.8±0.73 12.51±0.50

32.8±2.0 40.2±1.2 very slow

–119±9 –97±3

H2(β-Et)8TPP (V)

ZnAc2 Zn(Acac)2 CoAc2

DMF DMF DMF

9.66±0.11 4.01±0.21 17.55±0.47

62.6±0.7 88.9±4.0 60.0±2.2

–24±1 56±2 –26±1

H2P(β-Ph)8TPP (VI)

ZnAc2 Zn(Acac)2 Zn(Acac)2 CoAc2 Co(Acac)2 Co(Acac)2

DMF DMF C6H6 DMF DMF DMFC6H6

1.81±0.05 5.29±0.21 83.79±2.18 5.39±0.30 2.58±0.09 10.48±0.56

93.2±2.3 68.5±2.1 32.5±1.3 77.8±1.8 77.1±3.6 79.3±3.1

64±1 –9±1 –108±4 22±1 13±1 –32±1

Thus, for instance, during the change of the concentration of DEA in the H2TPTBP(IV)–ZnAc2 –Py–DEA system from 0 to 1 mol/l, the true reaction rate constant increases from 0.1 up to 3.93 l⋅mol –1 ⋅s –1, and the entropy change value increases from –143 up to –48 J·mol –1 ⋅s –1. Herewith, the value of kv was found to directly depend on the molar concentration of DEA (Fig. 20). Such changes can not be explained by a decrease of additional solvation in the formation of the transition state of reaction (1), but are caused by a decrease of the entropy of the initial state, as the result of the interaction of DEA molecules with chemically active NH bonds of H2P before entering into reaction (1).

kv, mol/(l·s)

CDEA, mol/l Figure 20 Dependence of the rate constant of reaction (1) on the concentration of DEA for H2TBP (X) in the Py–DEA system (a) and H2TPTBP (IV) in Py–DEA (b) and DMSO–DEA (c) systems.


1.3.3

203

Quantum chemical criterion

In the recent years, a very popular method has been the quantum chemical assessment of NH acidity of porphyrins [109]. Thus, Stuzhin [47] has shown that for NH-active H2P (e.g., IV–VII) the value of a relative stability of dianions P2– (δ∆Hf(0-2)) calculated by the AM1 method is more positive compared with classical H2P (compounds I–II) (Table 17). Porphyrins whose chemical activity of NH bonds is high are characterized by the value of δ∆Hf(0-2) > 0 (compounds V–VII, IX, XVII). This fact is consistent with the results of assessing the state of these bonds in accordance with the spectral and kinetic criteria and is supported by EAS data (Fig. 21). Table 17 Quantum chemical (AM1) and spectral (EAS) assessment of the chemical activity of NH bonds in H2P molecules (I–II, IV–X, XVII–XVIII).

δ ∆Hf (0-2), kcal/mol(a)

Porphyrin H2(β-F)4TAP (XVII) H2(β-Cl)4TAP (XVIII) H2(β-Br)4TAP (IX) H2TAP (VIII) H2TBP (X) H2(β-Et)8P (I) H2TPP (II) H2TPTBP (IV) H2(β-Et)8TPP (V) H2(β-Ph)8TPP (VI) H2(β-Br)8TPP (VII) (a)

+3.11 –0.30 +4.04 –23.18 –16.55 –39.01 –25.64 –2.34 +9.66 +13.28 +22.82

Parameters of Qx band in EAS [λQx, nm; (ε)] CHCl3(b)

DMSO(b)

– – 639 (4.71) 617 (4.75), C6H5Cl 663 (3.45) 619 (3.74) 646 (3.55) 694 (3.85) 696 (4.06) 709 (4.36) 738 (3.84)

– – 620 613 (4.71), Py 663 (3.53) 619 (3.74), DMF 646 (3.68), DMF 696 (3.57) 715 (3.97) 749 (4.64) 792 (4.16)

By the data of [47]; (b) using the data from [8, 35, 39, 55].

1.3.4

Insufficiency of absorption spectrum analysis of porphyrins

Nonclassical porphyrins with highly chemically active NH bonds are readily distinguishable from inert classical porphyrins by comparing their EAS in weakly solvating and electron-donor solvents (Fig. 21). With electron-donor solvents, these NH-active H2P form H associates, which are also called proton-transfer complexes (PTC) (21) [30, 32, 55, 93]. oooooooooooooooo

H2P + Solv

K1

HPδ–‘ …Hδ+…Solvδ–

K1 Solv

Solvδ–…Hδ+…Pδ+‘…Hδ+…Solvδ– (21)

However, electron spectroscopy data can not be a reliable criterion, as they make it possible to characterize the chemical NH activity of H2P only in the cases when this property is pronounced in compounds [32]. A large number of porphyrins, pertaining in accordance with the above three criteria to H2P with a chemically active NH bond, e.g., compounds IV, VIII and X, do not practically change their EAS in electron-donor solvents (Table 17) because their PTC formation equilibria (21) are strongly shifted to the left. An example of an EAS for a weakly NH-active porphyrin IV [32, 39] is given in Fig. 21, a. In contrast, compound V, which possesses a high chemical activity of NH bonds, forms in DMSO a PTC, which

204


(a)

(b)

λ, nm

λ, nm

Figure 21 Electronic absorption spectra of (a) H2TPTBP (IV) and (b) H2(β-Et)8TPP (V) in CHCl3 (1) and DMSO (2).

is accompanied by a decrease in the number and by a shift of EAS bands (Fig. 21, b). Herewith, the shift of the most batochromic Qx band (Table 17) is the greater, the easier the PTC is formed by porphyrin. Some porphyrins can not be tested for the chemical activity of NH bonds by all three criteria. It has been found [26, 58, 85] that H2P, for which at least two of the three assessment criteria of the state of NH bonds are observed, can be considered to be NH-active. A larger number of criteria of NH activity can be suggested using other methods. 1.4

Nonplanar structure of the macrocycle and chemical NH activity in its coordination core

Porphyrins can acquire a nonplanar structure as the result of a number of structural changes [19, 85, 87], such as: • Modification of the periphery of the H2P molecule: (a) multiple (nona-, deca-, undeca- or dodeca-) substitution of peripheral hydrogen atoms simultaneously in β- and meso-positions [19, 22, 85, 130–135] (b) substitution of meso-positions by bulky functional groups (iso-propyl, tert-butyl etc.) [19, 136–140] • Modification of the H2P coordination centre: (a) substitution of N atoms of intracyclic NH groups (N-substitution) [15, 85, 87, 141–144] (b) protonation of tertiary atoms of nitrogen {mono- (H3P+) or di- (H4P2+) cationic forms of H2P, equations (4)–(5)} [61, 145–147] (c) complexation with metals whose ionic radius is much smaller (Ni(II)) or, vice versa, exceeds (Zn(II)) the size of the coordination cavity (some metalloporphyrins) [16, 19, 65] • Changes in the H2P molecule leading to a decrease of aromaticity: (a) formation of radical forms of porphyrins [16, 65] (b) reduction of Cβ –Cβ bonds in pyrrole rings of H2P macrocycles (bacteriochlorins, corrins etc.) [9–10, 87] • Combined action of several factors (dodeca-substitution + N-substitution; dodecasubstitution + protonation etc.) [87, 143–145].


205

As the result of extensive X-ray diffraction studies [16, 19], all known H2P and MP with nonplanar structure were classified by the type of molecule distortion as saddled, ruffled, domed, waved, stepped, twisted, gabled etc. [19]. The porphyrin nonplanarity criteria can be not only XDA data [12–13,19, 65], which describe the structure of macrocycles in the solid phase, but also fluorescence data [22, 33, 49–50, 107, 135, 140, 142, 147], RR (resonance Raman) spectroscopy [134, 148–149], and in some cases NMR spectral results [72, 132, 150–151]. A crucially important issue is that of the relation between the type of nonplanar structure of the porphyrin macrocycle and the chemical activity of NH bonds in its coordination core. If such a relation does exist and, thus, a simple change of a nonplanar conformation of the H2P molecule may cardinally change its properties, the mechanism of the processes involving these molecules in vivo can be assumed to be similar [8, 17, 27]. The most pronounced distortion types of the planar structure of the H2P molecule include, besides dodeca substitution, the substitution of meso-positions by bulky, most often alkyl, groups, which leads to the ruffled conformation, as well as the substitution of one or a larger number of nitrogen atoms in the coordination core to form N-substituted porphyrin analogues [8, 19, 87]. In the latter case, the saddled conformation of an asymmetric type was found [142]. Section 1.3, which discussed the criteria of the chemical activity of NH bonds in H2P molecules, did not consider other types of strongly nonplanar macrocycles, except dodecasubstituted.

Figure 22 Absorption spectra (solid line) and fluorescence spectra (dashed line) of H(N-CH3)TPP (XX) in toluene, T = 298 K [142].

At the same time, both meso-substituted H2P [19, 49–50, 70, 72, 136–140] and N-substituted porphyrin analogues [15, 24, 61, 71, 85–87, 141–144] by their spectral characteristics, physicochemical properties and reactivity behave as compounds with a nonplanar structure. Thus, all bands in the visible range of their EAS are shifted batochromically relative to the predominantly planar H2P [15, 139], which, however, is not a reliable criterion of nonplanarity [23]. The photophysical characteristics describing the properties of excited states are more reliable [107]. Significant Stokes shifts of adjacent bands in the fluorescence spectra relative to absorption bands (22), low quantum yields of fluorescence and triplet states’ yields are considered to be features of nonplanar porphyrins [33, 49]. Besides, oooooooo

∆ν1 = ν1fl – ν1abs

(22)

206


characteristic modes appear in vibrational spectra of nonplanar porphyrins in the region of 1300–1700 cm–1, the so called marker lines ν2–ν4, sensitive to the change of H2P spatial structure [149]. The 1H NMR spectra are observed to have weak-field shifts of NH proton signals owing to the deshielding of the coordination core in the distortion of the molecule planar structure [72]. As the result of activation of extraplanar vibrations, porphyrins of the structural groups considered acquire an increased solubility in organic solvents (Table 7) [86, 152], redox properties [153] unusual for predominantly planar H2P. The stability of the ligands and their complexes to the action of elevated temperatures [64, 154] and chemical reagents [136, 154] decreases. For instance, strongly nonplanar H2T(tert-Bu)P (XIII), which exists predominantly in ruffled conformation (∆Cmeso = 0.9 Å), proves so reactive that in the presence of acids or d metal salts nucleophilically attaches solvent molecules (CH3OH), thus forming porphodimethene structures [136]. N-methylation of the H2TPP (II) molecule decreases the stability of its zinc complex in a medium of glacial AcOH 100-fold [154]. Apparently, as the result of partial localization of π-electronic density the basicity of tertiary nitrogen atoms in pyrrole rings of nonplanar macrocycles (XI–XIII, XIX–XXI) increases, as well as the ability of the nitrogen atoms to be protonated even in the presence of mean-strength acids [61,85, 154]. Their complexation ability in a medium of solvents with moderate donor-acceptor characteristics, e.g., in acetonitrile, increases too [155]. The issue of the appurtenance of these nonplanar macrocycles (XI–XIII, XIX–XXI) to porphyrins with chemically active NH bonds has not been discussed earlier. To the best of our knowledge, there are no data at all at present on the acidity of NH bonds in porphyrin molecules with bulky meso-substituents. Sufficiently contradictory information is available on the effect of intracyclic N-substitution into H2P molecule on their acidic properties [15, 61]. In such a case, use can be made of the above criteria of the chemical activity of NH bonds in porphyrin molecules [26]. Unfortunately, there is no information in the literature on the dependence of these 1H NMR spectra of meso-substituted H2P (XI–XIII) and their N-substituted analogues (XIX–XXI) on the nature of solvent. Besides, in most cases the proton signal of an NH group is not exhibited at all in 1H NMR spectra of N-substituted porphyrins [15]. For this reason, we shall use two other NH activity criteria – quantum chemical and kinetic [26]. Table 18 compares the results of a quantum chemical calculation of enthalpies of forming H2P ligands, their mono- (HP – , ∆Hf(–1), kcal/mol) and dianions (P 2– , ∆Hf(–2), kcal/mol), as well as heats of deprotonating the ligands to mono- and dianions in the gas phase for predominantly planar NH-inactive porphyrins (I–III), as well as nonplanar mesosubstituted porphyrins (XI–XIII) and N-substituted H2P analogues (XIX–XXI). The values of ∆Hf(–2) taken as estimates show that, in contrast with dodeca-substitution of H2P (compounds IV–VII), meso- or N-substitution in the macrocycle does not lead to chemical activation of NH bonds. Thus, the values of ∆Hf(–2) show only very weak tendencies to the shift towards more positive values (a mere 5–8 units), whereas even in compounds with weakly pronounced NH activity, such as H2TBP (X) and H2TPTBP (IV), these shifts reach 26.5 and 23.3 units, respectively (Table 18). As N-substituted porphyrins are monobasic acids, analysis of quantum chemical data for them can be done using the value of ∆Hf(–1). As it follows from the data of Table 18, this value does not practically increase relative to unsubstituted porphin, either, as compared with NH-active compounds (IV and X). Thus, quantum chemical data indicate the absence of a noticeable chemical activity of ooooo

207


Table 18 Enthalpic characteristics of deprotonation of {∆H f(–1) and ∆H f(–2) (kcal/mol)} and formation of mono- and dianionic forms {∆H (–1) and ∆H (–2) (kcal/mol)} of porphyrins with planar and nonplanar structure (calculated by the AM1 method). Macrocycle

∆Hf(0)

∆Hf(–1)

∆Hf(–2) δ ∆Hf(0–1) δ ∆Hf(0–2) ∆H(–1)

∆H(–2)

H2P H2(β-Et)8P (I)(a) H2TPP (II)(a) H2T(n-Pr)P (III) H2T(iso-Pr)P (XI) H2T(cyclo-Hex)P (XII) H2T(tert-Bu)P (XIII) H(N-Me)(β-Et)8P (XIX) H(N-Me)TPP (XX) H(N-Ph)TPP (XXI) H2TBP (X) H2TPTBP (IV)

221.00

200.81

264.09

20.19

347.01

777.49

151.91 178.54 131.62 195.96 129.24 352.27 392.90 265.25 401.29

132.54 160.12 111.92 176.05 113.26 329.70 372.72 237.53 371.91

192.37 217.27 167.04 231.19 – – – 281.80 403.63

19.37 18.42 19.70 19.91 15.98 22.57 20.18 27.72 29.38

347.83 348.78 347.50 347.29 351.22 344.63 347.02 339.48 337.82

774.86 773.13 769.82 769.63 – – – 750.95 736.74

(a)

–43.09 –39.01 –25.64 –40.46 –38.73 –35.42 –35.23 – – – –16.55 –2.34

Data of [47].

NH bonds in nonplanar meso-substituted H2P and N-substituted porphyrin analogues. Let us turn to the data on the kinetics of complexation reaction (1) of these ligands, as they make the basis of the kinetic criterion of H2P NH activity. It follows from the obtained experimental data (Table 19) that the complexation rates of N-substituted compounds decrease with the increase of the electron-donor value of the solvent in the sequence: PrOH-1 > DMF > DMSO > Py [31, 85]. This corresponds to the behaviour of porphyrins with a localized NH bond. Moreover, in a medium of pyridine, which activates NH bonds, in the reaction of nonplanar H(N-Me)TPP (XX) with zinc acetate the rate of the process not only fails to increase but, upon reaching a certain concentration of the product, an equilibrium is settled [31]. An additional argument in favour of the absence of NH activity in N-substituted porphyrins are the negative values of activation entropy change in reactions (1) of their coorTable 19 Kinetic parameters of reaction (1) of N-substituted porphyrins with ZnAc2 in organic solvents [31, 39, 85]. Porphyrin

H(N-Me)(β-Et)P (XIX)

H(N-Me)TPP (XX)

Csalt ·103, mol/l

Solv

kv298 , l· mol –1 ·s–1

Ea , kJ · mol –1

∆S #, J · mol–1 · K–1

2.6 2.0 2.0 0.2 0.2

Py DMSO DMF DMF PrOH-1

0.014 5.20 very fast 131.4 very fast

77.0±1.4 52.8±2.8 – 7.1±0.1 –

–30±2 –56±3 – –189±15 –

2.6 2.0 2.0 0.2 0.2

Py DMSO DMF DMF PrOH-1

very slow 6.14 very fast 18.8 very fast

– 34.1±2.8 – 32.8±2.0 –

– –124±7 – –119±9 –

208


dination by d metal salts (Table 16). In contrast, the positive values of ∆S # = S # – S init in NH-active H2P are associated with a significant solvation of the initial state on the coordinate of reaction (1). There are practically no data on the complexation kinetics of nonplanar porphyrins in ruffled conformation. Our studies, using H2T(iso-Pr)P (XI) and H2T(cyclo-Hex)P (XII) (∆Cmeso for their complexes with Ni(II) are equal to 0.74 and 0.77 Å [19], respectively), have shown that the rates of this reaction with copper, zinc or cadmium acetates in electrondonor solvents (DMF, DMSO), assuming NH activation, are very low, and respective complexes within the temperature range of 298–338 K are not formed in practice. This fact is a good proof of a localized NH bond in the molecules of meso-substituted porphyrins. It can also be noted that H2P with bulky meso-substituents, as a rule, do not in practice enter into complexation reactions with d metal salts in acetic acid and other solvents with the pronounced proton-donor function. The disturbance of the planar structure of the macrocycle leads to partial localization of the π-electronic density in pyrrole rings and an increase of basicity of the molecule. As protonated forms of porphyrins produced in this case in proton-donor media in reaction (1) are inactive [1, 7], this rule is also valid for other nonplanar H2P, such as N- or dodeca-substituted molecules. Some dodeca-substituted compounds also enter into reaction (1) in a medium of AcOH, but at very low rates [51–52]. The best medium for MP synthesis in the case of nonplanar meso-substituted porphyrins is a polar solvent with moderate donor-acceptor properties, e.g., CH3CN, in which reaction (1) with more nonplanar H2T(cyclo-Hex)P (XII) proceeds at 298 K very rapidly, and with H2T(iso-Pr)P (XI) with kv ~ ⋅10 l⋅mol –1 ⋅s –1. Thus, the criteria of the chemical activity of NH bands in the coordination core are a universal tool for the assessment of the reactivity of porphyrins and are applicable for both H2P belonging to various structural groups and their analogues. The second important conclusion is that a disturbance of the planar structure of the H2P molecule does not automatically assume the presence of chemically active NH bonds. The chemical activity of these bonds depends in a specific way on the type of conformation, in which a nonplanar macrocycle is. Thus, only the occurrence of the symmetrically distorted pronounced saddled conformation, possessing a high orthogonal dipole moment [22], leads to the emergence of an NH activity in nonplanar H2P. Nonplanar porphyrins in ruffled conformation are low-polar and exhibit no chemical activity of NH bonds.

2

Reactivity of the Coordination Core in Molecules of Porphyrin Ligands

As we have already mentioned, the chemical activity of NH bonds is vividly exhibited in the character of the progress of those chemical processes, which affect these bonds in one way or another. For H2P ligands and their analogues, these are primarily complexation reactions (1), acid-base interaction reactions ((2)–(5)), as well as coordination-core reactions of nucleophilic substitution. 2.1

Complexation reactions

The features of reaction (1), which follow from its mechanism (Fig. 23), have already been considered in Section 1. Here we would only list them once again, bearing in mind the significance of the state of NH bonds in the H2P molecule on the rate of this reaction: 1. The reactivity of porphyrins as aromatic macrocyclic ligands in complexation

Chemical Activation of Porphyrins in Coordination Core Reactions N... H . solv N

+

N

solv solv

solv . H ... N

solv . H

X. solv solv Ɇ solv X. solv

N

1

N solv . H N

Salt ɋɨɥɶ

ɉɨɪɮɢɪɢɧ Porphyrin

N solv . H ......N

N

X. solv solv Ɇ solv X. solv

ɂɧɬɟɪɦɟɞɢɚɬ Intermediate

2

solv .H......N

N

209

× solv Ɇ

X. solv X. solv

Metalloporphyrin Ɇɟɬɚɥɥɨɩɨɪɮ

solv

Transition state Figure 23 A schematic mechanism of the complexation reaction of porphyrins with d metal salts [1, 5, 11, 25].

reaction (1) is controlled by the intensity of the macrocyclic effect, i.e., is determined by the rate of the electronic and structural MCE components affecting the process [6–8]. 1.1. The structural component always hinders reaction (1) to a greater or smaller extent due to the spatial shielding of the coordination core by the surrounding atoms and π-electronic cloud. 1.2. The π-electronic MCE component is determined by the extent of aromaticity of the H2P macrocycle, and in the case of high delocalization of π-electrons in the macroring induces the polarization of the molecule and some chemical bonds in it. Thus, the character of the macrocyclic effect depends on the type of H2P structure – planar or nonplanar [8]. 2. The progress of reaction (1) (Fig. 23) for porphyrins with a different degree of rigidity strongly depends on the nature of a solvent. 2.1. Reaction (1) of H2P with a chemically inactive NH bond proceeds the most readily in a medium of solvents – moderate-strength ligands (carboxylic acids, alcohols), because it is determined by the strength of the solvatosalt [MX2Solv(n-2)] [75]. The strength of the solvatosalt depends on the coordinating ability of the metal ion and the nature of ligands – molecules of a solvent (Solv) and anions of a salt (X – ). If the bonds M–X and M–Solv in the solvatosalt are sufficiently strong, an intermediate is formed, a complex with H2P of amine type (Fig. 20), which is easily broken by electron-donor additives [1, 25]. 2.2. Complexation reaction (1) for H2P with a chemically active NH bond is accelerated in solutions of electron-donor solvents. In this case, the state of the NH bond is the major factor determining the rate of reaction (1) [1, 32, 129]. Chemical activation of NH bonds not only leads to accelerate reaction (1), e.g., owing to the interaction of these bonds with molecules of an electron-donor solvent, but also to a possibility of controlling more complex and fine processes, e.g., in the course of tautomeric conversions. Thus, it has been found that, depending on the state of tautomeric equilibrium

210

D.B. Berezin and B.D. Berezin F

F

F

F

F

F

F

F

F

F

F

F

FN

F

Cu III N N

F F

F DDQ

F

F F F

F p-TolSO2NHNH2

N

C

F

F

HN C

II

F

N

Cu

F F

F N

N

F

F

F

F

F

F

F

F

F

Figure 24

F

F

F

Redox processes involving the fluorinated analogue of compound CuXXVIII [121].

of compound XXVIII, the oxidation of metal occurring in the ligand can be controlled; herewith, the metal is in a bidentate state (XXVIIIb) or a tridentate state (XXVIIIa) [117, 122–124]. Thus, Maeda and Furuta [122] propose a simple method of oxidation of Cu(II)P DDQ to Cu(II)P, which proceeds without the change of the geometry of the coordination node (plane-square complexes) (Fig. 24). The inverse process is performed in the presence of p-toluene sulfonyl hydrazide. Both processes proceed quantitatively and almost instantaneously. They are realized as easily electrochemically {Cu3+ \Cu2+ = 0.15V} [121]. 2.2

Proton ionization of NH bonds in H2P molecules

The rise of chemical activity of NH bonds leads to the formation of proton-transfer complexes with weak organic bases [30, 55, 90, 93]. In a PTC, the uncompleted acid-base interaction is realized (equations (8)–(10)). Herewith, an obligate event is enhancement of NH-acidic properties of H2P with a chemically active bond in reactions accompanied by the total proton transfer (equations (2)–(3)), i.e., when the interaction of H2P with a strong base is realized [61]. What is more, the assessment of the values of proton dissociation constants of NH-active porphyrins (pK1 and pK2, equations (2)–(3)) is a good proof for the validity of chemical activity criteria of NH bonds. These criteria can be used on other objects in the absence of such constants [26]. Table 20 presents the values of dissociation constants for a number of porphyrins with chemically inactive (I–III, XX) and active (IV–X, XXII–XXIII) bonds. As it follows from the table data, the NH-acidic properties are indeed enhanced in sequences of compounds, in which an increase of NH activity is observed (Section 1.3). But H2P nonplanar macrocycles, which possess no NH activity, drop out of this sequence and behave similar to planar compounds I–III. Thus, for nonplanar [142] H(N-Me)TPP (XX), the other way round, the NH activity is observed to be decreased almost 10-fold as compared with unsubstituted tetraphenylporphin [61] (Table 20). Thus, the rise of chemical activity of NH bonds in H2P molecules increases the NH acidity of compounds in a cardinal way. For instance, the acidity of nonplanar NH-active compounds can be enhanced by up to 13 orders, and of planar ones, up to 15 orders as compared with classical unsubstituted porphin according to the values of pK1 (equation (2), Table 20).


211

Table 20 Constants of acid dissociation of porphyrins in DMSO–KOH [222] and DMSO–NR4OH systems. Porphyrin H2(β-Br)4TAP (IX) H2Pc (XXII) H2(β–Ph)8TAP (XXIII) H2TAP (VIII) H2TBP (X) H2T(n-Pr)P (III) H2P H2TPP (II) H(N-Me)TPP (XX) H2TPTBP (IV) H2(β-Et)8TPP (V) H2(β-Ph)8TPP (VI) H2(β-Br)8TPP (VII) (a)

2.3

pK1 7.26 ±0.02 10.73± 0.03 13.73 ±0.03(a) 12.36 ±0.18 9.68 ±0.02(b) 18.53±0.05 23.91±0.02 22.35±0.02 21.15±0.03 22.07±0.04 18.42±0.05 17.93 12.52±0.02 9.14±0.06

pK2

Reference

7.96±0.02 – – – – – – – – – – 17.49 15.39±0.06 12.90±0.03

35, 53 35, 61, 155 156 61 156 61, 155, 157 61, 155 61, 155, 157 61, 155 61, 155 155 62 62 62

Sulfo derivative of H2Pc(SO3– ); (b) solvent DMF.

Nucleophilic substitution reactions in the coordination core

Porphyrins can be subjected to numerous nucleophilic substitution reactions in the coordination core to form a number of simple or bridge N-substituted porphyrin analogues [15]. A scheme of one of the simplest substitution processes – at the saturated carbon atom, namely, alkylation, benzylation etc. – is given in Fig. 25.

regrouping

Figure 25 Schematic reaction of the nucleophilic substitution by H2P molecule coordination core atoms in saturated carbon atom [15, 85].

Although the substitution process is nucleophilic, by one of the tertiary nitrogen atoms of the coordination core, progress of the reaction at the regrouping stage depends to a great extent on the chemical activity of the NH bond.

3

Biosignificance of the Phenomenon of NH Activation

Studies of the phenomenon and conditions of the chemical activation of NH bonds in porphyrin molecules are of great importance for understanding the processes occurring in porphyrin-containing biosystems. Although the major part of bioporphyrins are attributed to

212


the group of planar classical porphyrins with a chemically low-active NH bond [28, 58], selective activation of bonds in these molecules in vivo, nevertheless, can occur. It can proceed under the influence of the protein-lipid environment, leading to conformational changes of a certain character [17, 19] and, probably, to the polarization of the macrocycle [22, 26]. As we showed above, depending on the type of H2P conformation emerging in solution, the polarization of the macrocycle and individual bonds in it can occur or not occur at all [26, 85]. It can be expected that processes in biosystems proceed in accordance with similar mechanisms. Incorporation of metals (Fe, Mg, Co, Ni and others) into molecules of heme precursors, chlorophylls or cytochromes is performed in vivo on the enzymic level [10–11, 14, 27, 258]. Nevertheless, the effect of conformationally initiated activation of NH bonds can also accelerate metal inclusion processes by several orders, as it occurs in solution [26, 35, 52, 85]. Unfortunately, until now it has not been established in detail in conformations of which types precursors of bioporphyrins exist and how they structurally change in vivo, as well as which factors regulate the process of metal inclusion into H2P ligand [14]. It can only be assumed that their main conformation in this process would be saddled [16, 19], as the result of the formation of which the synthesis of MP proceeds more efficiently owing to the accessibility of the reaction centre on the whole and the activation of NH sites in particular [26]. Thus, studies of the peculiar features of conformational conversions of the porphyrin macrocycles and related changes in the physics, chemistry and reactivity of compounds has wide prospects. Thus, for instance, knowledge of the requirements imposed on porphyrins – molecular-recognition receptors or photosensitizers in photodynamic therapy [110] – from the point of view of the geometry of conformations, from the positions of the best correspondence of reaction centres of the participants in the process will undoubtedly make it possible to intensify it significantly.

References 1. B.D. Berezin, Coordination Compounds of Porphyrins and Phthanocyanines, Nauka: Moscow, 280 pp. (1978) (in Russian); Coordination Compounds of Porphyrins and Phthanocyanines, Wiley: New York–Toronto, 286 pp. (1981). 2. P. Hambright, Dynamic Coordination Chemistry of Porphyrins, in: Porphyrins and Metalloporphyrins, ed. by K.M. Smith, Elsevier: Amsterdam–Oxford–New York, pp. 233–278 (1975). 3. J.W. Buchler, Synthesis and Properties of Metalloporphyrins, in: The Porphyrins, ed. by D. Dolphin, Acad. Press: New York, San Francisco, London, vol. 1, pp. 389–483 (1978). 4. B.D. Berezin, Application of Porphyrins for Studies of Electronic, Steric and Solvation Effects of Coordination, in: Porphyrins: Spectroscopy, Electrochemistry, Applications, ed. by N.S. Enikolopyan, Nauka: Moscow, pp. 182–213 (1987) (in Russian). 5. B.D. Berezin and N.S. Enikolopyan, Metalloporphyrins, Nauka: Moscow, 159 pp. (1988) (in Russian). 6. B.D. Berezin and M.B. Berezin, Zhurn. Fiz. Khim., 63 (12), 3166–3181 (1989). 7. B.D. Berezin, M.B. Berezin and D.B. Berezin, Ross. Khim. Zhurn., 41 (3), 105–123 (1997). 8. D.B. Berezin and T.N. Lomova, The Macrocyclic Effect and Biological Activity of Metalloporphyrins, in: Problems of Solution Chemistry: Structure, Thermodynamics, Reactivity, ed. by A.M. Kutepov, Nauka: Moscow, 326–362 (2001). 9. Porphyrins and Metalloporphyrins, ed. by K.M. Smith, Elsevier: Amsterdam–Oxford–New York, 890 pp. (1975). 10. The Porphyrins, ed. by D. Dolphin, Acad. Press: New York–San Francisco–London, vol. I–VII (1978–1979).


213

11. Porphyrins: Structure, Properties, Synthesis, ed. by N.S. Enikolopyan, Nauka: Moscow, 333 pp. (1985) (in Russian). 12. J.L. Hoard, Stereochemistry of Porphyrins and Metalloporphyrins, in: Porphyrins and Metalloporphyrins, ed. by K.M. Smith, Elsevier: Amsterdam–Oxford–New York, pp. 317–380 (1975). 13. W.R. Scheidt, Porphyrin Stereochemistry, in: The Porphyrins, ed. by D. Dolphin, Acad. Press: New York–San Francisco–London, vol. 3, pp. 463–512 (1978). 14. The Porphyrin Handbook, ed. by K.M. Kadish, K.M. Smith and R. Guliard, Academic Press: New York, vols. I–X (2000). 15. D.K. Lavallee, The Chemistry and Biochemistry of N-substituted Porphyrins, VCH Publishers: New York, 313 pp. (1987). 16. W.R. Scheidt and Y.-J. Lee, Struct. Bonding (Berlin), 64, 1 (1987). 17. M.O. Senge, J. Photochem. Photobiol. B: Biol., 16, 3 (1992). 18. K.M. Barkigia, L. Chantranpong, K.M. Smith and J. Fayer, JACS, 110, 7566–7567 (1988). 19. M.O. Senge, Highly Substituted Porphyrins, in: The Porphyrin Handbook, ed. by K.M. Kadish, K.M. Smith and R. Guliard, Academic Press: New York, vol. 1, pp. 239–347 (2000). 20. S.G. Di Magno, A.K. Wertching and C.R. Ross, JACS, 117 (31), 8279–8280 (1995). 21. B.D. Berezin, Koord. Khim., 22 (4), 285–289 (1996). 22. V.S. Chirvony, I.V. Sazanovich, V.A. Galievsky, A. van Hoek, T.J. Schaafsma, V.L. Malinovskii and D. Holten, J. Phys. Chem., 105 (32), 7818–7829 (2001). 23. A. Ghosh, I.H. Wasbotten and J. Conradie, J. Phys. Chem., 107 (15), 3613–3623 (2003). 24. D.B. Berezin, Zhurn. Obshch. Khim., 74 (3), 506–511 (2004). 25. B.D. Berezin and O.I. Koifman, Usp. Khim., 42 (11), 2007–2037 (1973) (in Russian). 26. D.B. Berezin, Zhurn. Fiz. Khim., 2006 (in press) (in Russian). 27. Inorganic Chemistry, ed. by G. Eichhorn, Mir: Moscow, 1437 pp. (1978) (Russian translation). 28. B.D. Berezin and D.B. Berezin, State of Art of the Chemistry of Nonclassical Porphyrins and their Complexes, in: Advances of Porphyrin Chemistry, ed. by O.A. Golubchikov, Chemistry Research Institute, St. Petersburg University: St. Petersburg, vol. 2, pp. 128–141 (1999) (in Russian). 29. J. Takeda and M. Sato, Chem. Lett., 971–972 (1995). 30. T.V. Karmanova, T.V. Gromova, B.D. Berezin, A.S. Semeykin and S.A. Syrbu, Zhurn. Obshch. Khim., 71 (5), 856–861 (2001) (in Russian). 31. D.B. Berezin and O.V. Toldina, Zhurn. Inorg. Khim., 47 (12), 2075–2081 (2002) (in Russian). 32. D.B. Berezin, O.V. Toldina and R.S. Kumeyev, Zhurn. Fiz. Khim., 78 (8), 1427–1432 (2004) (in Russian). 33. V.S. Chirvonyi, Picosecond Photoinduced Processes in Porphyrin Molecules and their Complexes with Nucleic Acids, DSc (Physics and Mathematics) Thesis, Inst. Mol. and Atomic Physics, Nat. Acad. Sci. of Belarus: Minsk, 315 pp. (2005) (in Belorussian). 34. B.D. Berezin and O.G. Khelevina, Tetrasubstitution and Physico-chemical Properties of Porphyrins, in: Porphyrins: Structure, Properties, Synthesis, ed. by N.S. Enikolopyan, Nauka: Moscow, 83–113 (1985) (in Russian). 35. P.A. Stuzhin, G. Khelevina and B.D. Berezin, Azaporphyrins: Acid-base and Coordination Properties, in: Phthalocyanines: Properties and Applications, ed. by C.C. Leznoff and A.B.P. Lever, VCH Publ.: New York, vol. 4, pp. 19–77 (1996). 36. P.A. Stuzhin and O.G. Khelevina, Coord. Chem. Rev., 147, 41–86 (1996). 37. P.A. Stuzhin and O.G. Khelevina, Structure and Coordination Properties of Azaporphyrins, in: Advances of Porphyrin Chemistry, ed. by O.A. Golubchikov, Chemistry Research Institute, St. Petersburg University: St. Petersburg, vol. 1, pp. 150–202 (1997) (in Russian). 38. B.D. Berezin, E.B. Karavaeva and T.I. Potapova, Formation, Stability and Properties of Tetrabenzoporphyrin and its Complex Compounds, in: Problems of Solvation and Complexation, Ivanovo Inst. of Chemistry and Technology: Ivanovo, issue 6, pp. 25–31 (1979) (in Russian). 39. O.V. Toldina, Complexation of meso-Substituted Tetrabenzoporphyrins with Metal Salts in Organic Solvents, PhD (Chemistry) Thesis, Ivanovo University of Chemistry and Technology: Ivanovo, 168 pp. (2004) (in Russian). 40. K.S. Suslick, Ch.-T. Chen, G.R. Meredith and L.-T. Cheng, JACS, 114 (17), 6928–6930 (1992). 41. S.M. LeCours, H.-W. Guan, S.G. DiMagno, C.H. Wang and M.J. Therien, JACS, 118 (6),

214


1497–1503 (1996). 42. M. Yeung, A.C.H. Ng, M.G.B. Drew, E. Vorpagel, E.M. Breitung, R.J. McMahon and D.K.P. Ng, J. Org. Chem., 63 (21), 7143–7150 (1998). 43. T.E.O. Screen, I.M. Blake, L.H. Rees, W. Clegg, S.J. Borwick and H.L. Anderson, J. Chem. Soc. Perkin Trans., 1, 320–329 (2002). 44. S. Tsukahara and N. Suzuki, Inorg. Chim. Acta, 245, 105–108 (1996). 45. A.G. Govorov, A.B. Korzhenevsky and O.I. Koifman, Russ. J. Phys. Chem., 70 (4), 582 (1996). 46. V.M. Mamayev, I.P. Gloriozov and V.V. Orlov, Izv. Vuz. Khim. Khim. Tekhnol., 25 (11), 1317–1332 (1982) (in Russian). 47. P.A. Stuzhin, J. Porphirins Phthalocyanins, 7 (12), 796-815 (2003). 48. V.A. Kuzmitsky, K.N. Solovyev and M.P. Tsvirko, Spectroscopy and Quantum Chemistry of Porphyrins, in: Porphyrins: Spectroscopy, Electrochemistry, Applications, ed. by N.S. Enikolopyan, Nauka: Moscow, 7–126 (1987) (in Russian). 49. S. Gentemann, C.J. Medforth, T.P. Forsyth, D.J. Nurco, K.M. Smith, J. Fajer and D. Holten, JACS, 116 (16), 7363–7368 (1994). 50. S. Gentemann, N.J. Nelson, L. Jaquinod, D.J. Nurco, S.H. Leung, C.J. Medforth, K.M. Smith, J. Fajer and D. Holten, J. Phys. Chem. B, 101 (7), 1247–1254 (1997). 51. N.S. Dudkina, P.A. Shatunov, E.M. Kuvshinova, A.S. Semeykin, S.G. Pukhovskaya, O.A. Golubchikov, Zhurn. Obshch. Khim., 68 (1), 2042–2047 (1998(in Russian). 52. O.A. Golubchikov, S.G. Pukhovskaya and E.M. Kuvshinova, Spatially Distorted Porphyrins: Structure and Properties, in: Advances of Porphyrin Chemistry, ed. by O.A. Golubchikov, Chemistry Research Institute, St. Petersburg University: St. Petersburg, vol. 4, pp. 45–75 (2004) (in Russian). 53. O.G. Khelevina, N.V. Chizhova and B.D. Berezin, Koord. Khim., 17 (3), 400–404 (1991). 54. O.G. Khelevina, S.V. Timofeeva, B.D. Berezin and S.I. Vagin, Zhurn. Fiz. Khim., 68 (8), 1423–1426 (1994) (in Russian). 55. O.A. Petrov, Koord. Khim., 27 (7), 483–492 (2001). 56. P. Bhyrappa, M. Nethaji and V. Krishnan, Chem. Lett., 869–872 (1993). 57. X.-Zh. Song, W. Jentzen, S.-L. Jia, L. Jaquinod, D.J. Nurco, C.J. Medforth, K.M. Smith and L. Shelnutt, JACS, 118 (51), 12975–12988 (1996). 58. D.B. Berezin and B.D. Berezin, J. Porphyrins Phthalocyanines, 8 (4–6), 516 (2004). 59. A. Rosa, G. Ricciardi, E.J. Baerends and S.J.A. Gisbergen, J. Phys. Chem. A, 105 (13), 3311–3327 (2001). 60. G.P. Gurinovich, A.A. Sevchenko and K.N. Solovyev, Spectroscopy of Chlorophyll and Related Compounds, Nauka i Tekhnika: Minsk, 520 pp. (1968) (in Russian). 61. V.G. Andrianov, O.V. Malkova and D.B. Berezin, Acid-basic Properties of Porphyrins, in: Advances of Porphyrin Chemistry, ed. by O.A. Golubchikov, Chemistry Research Institute, St. Petersburg University: St. Petersburg, vol. 3, pp. 107–129 (2001) (in Russian). 62. D.B. Berezin, Yu.B. Ivanova and V.B. Sheinin, Zhurn. Fiz. Khim., 2006 (in Russian) (in press). 63. A.I. Vyugin, E.V. Antina and G.A. Krestov, Dokl. Akad. Nauk SSSR, 315 (5), 1149–1151 (1991) (in Russian). 64. A.I. Vyugin, E.V. Antina and M.B. Berezin, Solvation Effects and Coordination Properties of Porphyrins and Metalloporphyrins in Solutions, in: Advances and Problems of Solvation Theory: Structural and Thermodynamic Aspects, ed. by A.M. Kutepov, Nauka: Moscow, pp. 208–241 (1998) (in Russian). 65. W.R. Scheidt, Systematics of the Stereochemistry of Porphyrins and Metalloporphyrins, in: The Porphyrin Handbook, ed. by K.M. Kadish, K.M. Smith and R. Guliard, Academic Press: New York, vol. 3, pp. 49–112 (2000). 66. M.I. Terekhova, E.S. Petrov and E.M. Rokhmina, Khim. Geterotsykl. Soed., 8, 1104–1108 (1979) (in Russian). 67. G.I. Koldobsky and V.A. Ostrovsky, Khim. Geterotsykl. Soed., 5, 579–592 (1988) (in Russian). 68. Yu.A. Fialkov, Solvent as a Means of Controlling a Chemical Process, Khimiya: Leningrad, 240 pp. (1990) (in Russian). 69. B.D. Berezin, V.G. Andrianov and O.V. Malkova, Dokl. Akad. Nauk SSSR, 314 (6), 1432–1435 (1990) (in Russian). 70. C.J. Medforth, C.M. Muzzi, K.M. Shea, K.M. Smith, R.J. Abraham, S. Jta and J.A. Shelnutt, J. Chem. Soc. Perkin Trans., 2, 839–844 (1997).


215

71. D.B. Berezin, Effect of Electronic, Solvation and Steric Factors on the Chromophore Properties of Porphyrins and their Protonated Forms, PhD (Chemistry) Thesis, Ivanovo State University of Chemistry and Technology: Ivanovo, 187 pp. (1997) (in Russian). 72. C.J. Medforth, NMR Spectroscopy of Diamagnetic Porphyrins, in: The Porphyrin Handbook, ed. by K.M. Kadish, K.M. Smith and R. Guliard, Academic Press: New York, vol. 5, pp. 1–80 (2000). 73. F. Longo, E. Brown, D. Quimby, A. Adler and M. Meot-Ner, Ann. New York Acad. Sci., 206, 420–424 (1973). 74. B.D. Berezin and O.A. Golubchikov, Zhurn. Fiz. Khim., 60, 2113–2128 (1986) (in Russian). 75. B.D. Berezin and O.A. Golubchikov, Coordination Chemistry of Solvation Complexes of Transition Metal Salts, Nauka: Moscow, 240 pp. (1992) (in Russian). 76. B.D. Berezin, Koord. Khim., 17 (5), 597–606 (1991) (in Russian). 77. B.D. Berezin, Usp. Khim., 60 (9), 1946–1968 (1991) (in Russian). 78. O.A. Golubchikov, Izv. Vuz. Khim. Khim. Tekhnol., 47 (5), 10–26 (2004) (in Russian). 79. B.D. Berezin, G.A. Tsvetkov and L.P. Shormanova, Zhurn. Inorg. Khim., 25 (10), 1645–2652 (1980) (in Russian). 80. T.N. Lomova, L.P. Shormanova and M.E. Klyueva, Electronic and Steric Effects in the Functional Substitution of Metalloporphyrins, in: Advances of Porphyrin Chemistry, ed. by O.A. Golubchikov, Chemistry Research Institute, St. Petersburg University: St. Petersburg, vol. 1, pp. 129–149 (1997) (in Russian). 81. J. Reid and P. Hambright, Inorg. Chim. Acta, 33, L135–L136 (1979). 82. P. Hambright, Chemistry of Water Soluble Porphyrins, in: The Porphyrin Handbook, ed. by K.M. Kadish, K.M. Smith and R. Guliard, Wiley: San Diego, vol. 3, pp. 129–210 (2000). 83. V.S. Chirvony, A. van Hoek, T.J. Schaafsma, P.P. Pershukevich, I.V. Filatov, I.V. Avilov, S.I. Shishporenok, S.N. Terekhov and V.L. Malinovskii, J. Phys. Chem. B, 102 (48), 9714–9724 (1998). 84. B.D. Berezin, Solvation Processes in Porphyrin Chemistry, in: Problems of the Chemistry of Solutions and Liquid-phase Materials, Ivanovo Institute of Solution Chemistry RAS: Ivanovo, pp. 238–248 (2001) (in Russian). 85. E.N. Misko, Spectral, Kinetic, Thermochemical Characteristics of Solvation and the Coordination Properties of Some N- and Dodeca-substituted Porphyrins, PhD (Chemistry) Thesis, Ivanovo Institute of Solution Chemistry RAS: Ivanovo, 198 pp. (2005) (in Russian). 86. G.M. Mamardashvili and B.D. Berezin, Thermodynamics of Porphyrin Dissolution, in: Advances of Porphyrin Chemistry, ed. by O.A. Golubchikov, Chemistry Research Institute, St. Petersburg University: St. Petersburg, vol. 3, pp. 130–149 (2001) (in Russian). 87. A.M. Stolzenberg, S.W. Simerly, B.D. Steffey and G. Scott Hatmond, JACS, 119 (49), 11843–11854 (1997). 88. K. Letts and R.A. Mackay, Inorg. Chem., 14 (12), 2990–2993 (1975). 89. M. Onaka, T. Shinoda, Yu. Izumi and E. Nolen, Chem. Lett., 117–120 (1993). 90. O.A. Petrov, The Reactivity of Tetraazaporphyrins in Processes of Acid-base Interaction and Formation of Molecular Complexes, DSc (Chemistry) Thesis, Ivanovo State University of Chemistry and Technology: Ivanovo, 264 pp. (2004) (in Russian). 91. M. Whalley, J. Chem. Soc., 866–869 (1961). 92. B.D. Berezin, V.G. Andrianov and O.V. Malkova, Dokl. Akad. Nauk SSSR, 314 (6), 432–1435 (1990) (in Russian). 93. B.D. Berezin, O.G. Khelevina, N.D. Gerasimova and P.A. Stuzhin, Zhurn. Fiz. Khim., 56 (11), 2768–2772 (1982) (in Russian). 94. O.A. Petrov, O.G. Khelevina, and B.D. Berezin, Zhurn. Fiz. Khim., 69 (5), 811–815 (1995) (in Russian). 95. O.A. Petrov, O.G. Khelevina and S.V. Timofeeva, Zhurn. Fiz. Khim., 69 (10), 1771–1775 (1995) (in Russian). 96. O.A. Petrov, N.V. Chizhova and B.D. Berezin, Koord. Khim., 25 (3), 235–240 (1999) (in Russian). 97. O.A. Petrov and N.V. Chizhova, Koord. Khim., 25 (5), 393-400 (1999) (in Russian). 98. O.A. Petrov, Zhurn. Fiz. Khim., 76 (9), 1577–1582 (2002) (in Russian). 99. O.A. Petrov, G.V. Osipova, A.S. Semeykin and B.D. Berezin, Koord. Khim., 31 (12), 941–945 (2005) (in Russian).

216


100. V.A. Kuzmitsky and K.N. Solovjev, J. Molec. Struct., 65, 219–230 (1980). 101. V.M. Mamaev, S.Ya. Ishchenko and I.P. Gloriozov, Izv. Vuz. Khim. Khim. Tekhnol., 32 (1), 3–21 (1989) (in Russian). 102. M. Asakawa, H. Toi, Y. Aoyama and H. Ogoshi, J. Org. Chem., 57 (21), 5796–5798 (1992). 103. J. Braun, M. Schlabach, B. Wehrle, M. Köcher, E. Vogel and H.-H. Limbach, JACS, 116, 6593 (1994). 104. J.R. Reimers, T.X. Lü, M.J. Crossley and N.S. Hush, JACS, 117 (10), 2855–2861 (1995). 105. V.A. Kuzmitsky, Khim. Fiz., 15 (12), 61–74 (1996) (in Russian). (in Russian). 106. I.K. Shushkevich, V.N. Knyukshto, V.A. Kuzmitsky, V.N. Kopranenkov, A.M. Vorotnikov and K.N. Solovyev, Zhurn. Prikl. Spektrosk., 57 (1–2), 92–99 (1992) (in Russian). 107. V.A. Kuzmitsky, V.I. Gael and K.N. Solovyev, Quantum-chemical Calculations of the Electronic Structure and Spectroscopic Properties of Tetrapyrrole Molecular Systems, in: Spectroscopy and Luminescence of Molecular Systems, Belorussian State University: Minsk, pp. 96–120 (2002) (in Russian). 108. K.N. Solovyev, A.L. Gladkov, A.S. Starukhin and S.F. Shkirman, Spectroscopy of Porphyrins: Vibrational States, Nauka i Tekhnika: Minsk, 415 pp. (1985) (in Russian). 109. A. Ghosh, Quantum Chemical Studies of Molecular Structures and Potential Energy Surfaces of Porphyrins and Hemes, in: The Porphyrin Handbook, ed. by K.M. Kadish, K.M. Smith and R. Guliard, Wiley: San Diego, vol. 7, pp. 1–38 (2000). 110. The Porphyrin Handbook, ed. by K.M. Kadish, K.M. Smith and R. Guliard, Wiley: San Diego, vol. 6, 350 pp. (2000). 111. P.S. Clezy, Oxophlorins (Oxyporphyrins), in: The Porphyrins, ed. by D. Dolphin, Acad. Press: New York, vol. 2, pp. 103–130 (1978). 112. M.G.H. Vicente, Reactivity and Functionalization of β-Substituted Porphyrins and Chlorins, in: The Porphyrin Handbook, ed. by K.M. Kadish, K.M. Smith and R. Guliard, San Diego: Wiley, vol. 1, pp. 149–200 (2000). 113. B.D. Berezin and O.B. Lapshina, Zhurn. Fiz. Khim., 48 (2), 303–307 (1974) (in Russian). 114. B.D. Berezin and O.B. Lapshina, Zhurn. Fiz. Khim., 50 (8), 2007–2011 (1976) (in Russian). 115. P.J. Chmielewski, L. Latos-Grazynski, K. Rachlewicz and T. Glowiak, Angew. Chem. Int. Ed. Engl., 33 (7), 779–781 (1994). 116. H. Furuta, T. Asano and T. Ogawa, JАCS, 116 (2), 767–768 (1994). 117. H. Furuta, T. Ishizuka, A. Osuka, H. Dehima, H. Nakagawa and Y. Ishikawa, JACS, 123 (25), 6207–6208 (2001). 118. J.P. Belair, Ch.J. Ziegler, Ch.S. Rayesh and D.A. Modarelly, J. Phys. Chem. A, 106 (27), 6445–6451 (2002). 119. T. Morimoto, Sh. Tanaguchi, A. Osuka and H. Furuta, Eur. J. Org. Chem. 3887–3890 (2005). 120. J.L. Shaw, Sh.A. Garrison, E.A. Aleman and C.J. Ziegler, J. Org. Chem. 69 (22), 7423–7427 (2004). 121. H. Maeda, Y. Ishikawa, T. Matsuda, A. Osuka and H. Furuta, JACS, 125 (39), 11822–11823 (2003). 122. H. Maeda and H. Furuta, J. Porphyrins Phthalocyanines, 8 (1), 67 (2004). 123. A. Srinavasan and H. Furuta, Acc. Chem. Res., 38 (1), 10–20 (2005). 124. D.B. Berezin, I.A. Maltsev, A.S. Semeykin and V.L. Bolotin, Zhurn. Fiz. Khim., 79 (12), 2220–2226 (2005) (in Russian). 125. P.J. Chmielewski and L. Latos-Grazynski, J. Chem. Soc. Perkin Trans., 2, 503–505 (1995). 126. P. Smirnov, L. Weng and I. Persson, Phys. Chem. Chem. Phys., 3, 5248 (2001). 127. P.J. Chmielewski and L. Latos-Grazynski, Inorg. Chem., 36 (5), 840 (1997). 128. H. Furuta, T. Morimoto and A. Osuka, Org. Lett., 3 (9), 1427 (2003). 129. E.M. Kuvshinova, A.S. Semeikin and O.A. Golubchikov, Russ. J. Phys. Chem., 71 (3), 371–373 (1997). 130. B.D. Berezin, T.V. Karmanova and T.V. Gromova, Zhurn. Inorg. Khim., 47 (12), 890–895 (2002) (in Russian). 131. V. Bhyrappa and V. Krishnan, Inorg. Chem., 30 (2), 239–245 (1991). 132. C.J. Medforth, M.O. Senge, K.M. Smith, L.D. Sparks and J.A. Shelnutt, JACS, 114 (25), 9859–9869 (1992). 133. M.O. Senge and W.W. Kalish, Inorg. Chem., 36 (26), 6103–6116 (1997). 134. J.A. Shelnutt, X.-Z. Song, J.-G. Ma et al., Chem. Soc. Rev., 27, 31–41 (1998).


217

135. J.L. Retsek, S. Gentemann, C.J. Medforth, K.M. Smith, V.S. Chirvony, J. Fajer and D. Holten, J. Phys. Chem., 104 (29), 6690–6693 (2000). 136. T. Ema, M.O. Senge, M.I. Nelson, H. Ogoshi, K.M. Smith, Angew. Chem. Int. Ed. Engl., 33, 1879–188 (1994). 137. M.O. Senge, T. Ema, K.M. Smith, J. Chem. Soc. Chem. Comm., 733 (1995). 138. M. Veyrat, R. Ramasseul, J.-C. Marchon, I. Turowska and W.R. Scheidt, New J. Chem., 19, 1199–1202 (1995). 139. M.O. Senge, I. Bischoff, N.T. Nelson and K.M. Smith, J. Porphyrins Phthalocyanines, 3, 99–116 (1999). 140. S. Gentemann, C.J. Medforth, T. Ema and N.J. Nelson, Chem. Phys. Lett., 245, 441–447 (1995). 141. S.-I. Aizawa, Y. Tsuda, Y. Ito, K. Hatano and S. Funahashi, Inorg. Chem., 32 (7), 1119–1123 (1993). 142. I.V. Sazanovich, A. Van Hoek, A.Yu. Panarin, V.L. Bolotin, A.S. Semeykin, D.B. Berezin and V.S. Chirvony, J. Porphyrins Phthalocyanines, 9 (1), 59–67 (2005). 143. M.O. Senge, W.W. Kalisch and S. Runge, Liebigs Ann./Recueil, 1345–1352 (1997). 144. T.E. Clement, L.T. Nguyen, K.M. Smith et al., Heterocycles, 45 (4), 651–658 (1997). 145. M.O. Senge, T.P. Forsyth, L.M. Nguyem and K.M. Smith, Angew. Chem. Int. Ed. Engl., 33 (23/24), 2485–2487 (1994). 146. B. Cheng, O.Q. Munro, H.M. Marques and W.R. Scheidt, JACS, 119 (44), 10732–10742 (1997). 147. V.S. Chirvony, A. Van Hoek, V.A. Galievsky, I.V. Sazanovich, T.J. Schaafsma and D. Holten, J. Phys. Chem., 104 (42), 9909–9917 (2000). 148. C.J. Medforth, J.A. Shelnutt, M.D. Berber and K.M. Barkigia, JACS, 113, 4077–4087 (1991). 149. L.D. Sparks, C.J. Medforth, M.S. Park et al., JACS, 115 (2), 581–592 (1991). 150. J.W. Dirks, G. Underwood, J.C. Matheson and D. Gust, J. Org. Chem., 44 (14), 2551–2555 (1979). 151. C.J. Medforth, R.E. Haddad, C.M. Muzzi, N.R. Dooley, L. Jaqwinod, D.C. Shyr, D.J. Nurco, M.M. Olmstead, K.M. Smith, J.-G. Ma and J.A. Shelnutt, Inorg. Chem., 42 (7), 2227–2241 (2003). 152. D.B. Berezin, Zhurn. Obshch. Khim., 75 (5), 807–810 (2005) (in Russian). 153. K.M. Kadish, Electrochemistry of Metalloporphyrins in Non-aqueous Media, in: The Porphyrin Handbook, ed. by K.M. Kadish, K.M. Smith and R. Guliard, San Diego: Wiley, vol. 8, pp. 1–114 (2000). 154. D.B. Berezin, E.N. Misko, E.V. Antina and M.B. Berezin, Zhurn. Obshch. Khim., 2006 (in Russian) (in press). 155. I.V. Tsvetkova, V.G. Andrianov and B.D. Berezin, Izv. Vuz. Khim. Khim. Tekhnol., 37 (1), 73–76 (1994) (in Russian). 156. V.B. Sheinin, V.G. Andrianov, B.D. Berezin and T.A. Koroleva, Zhurn. Org. Khim., 21 (7), 1564–1570 (1985) (in Russian). 157. V.B. Sheinin, Studies of the Acid-base Ionization of Some Natural and Synthetic Porphyrins, PhD (Chemistry) Thesis, Ivanovo Inst. of Chemistry and Technology: Ivanovo, 157 pp. (1981) (in Russian). 158. V.Ya. Bykhovsky, Tetrapyrrols: Diversity, Biosynthesis, Biotechnology, in: Advances of Porphyrin Chemistry, ed. by O.A. Golubchikov, Chemistry Research Institute, St. Petersburg University: St. Petersburg, vol. 1, pp. 27–51 (1997) (in Russian).

218


6

Synthesis, Structure Peculiarities and Biological Properties of Macroheterocyclic Compounds M.K. Islyaikin1, E.A. Danilova1, Yu.V. Romanenko1, O.G. Khelevina1 and T.N. Lomova2 1Ivanovo

State University of Chemistry and Technology, 7 F. Engels Prospect, Ivanovo, 153000, Russia; email: [email protected] 2Institute

of Solution Chemistry, Russian Academy of Sciences, Ivanovo, 153045, Russia; email: [email protected]

Recent advances in the chemistry of macroheterocyclic compounds – structural analogues of porphyrin and hexaphyrins – are reviewed. Aromaticity of various macrocyclic molecules and their fragments was studied using the geometric (EN, GEO and HOMA) and magnetic (NICS) criteria based on experimental data and results of quantum chemistry calculations at the DFT level. The coordinating and biological properties of macroheterocyclic compounds as well as of their metal complexes are under consideration.

Introduction A conspicuous group of natural macrocyclic pigments among the great diversity of macroheterocyclic compounds are hemoglobins, chlorophylls, cytochromes and other species. They play an exceptional role in vital processes of photosynthesis and respiration, are involved in regulation of fine metabolic processes in living organisms. These substances are tetrapyrrole macroheterocyclic compounds (porphyrins), the basis of which is porphin 1. It

CRC but not checked

220

M.K. Islyaikin, E.A. Danilova, Yu.V. Romanenko, O.G. Khelevina and T.N. Lomova

has a planar structure, and their inner macroring consisting of 16 carbon and nitrogen atoms contains 18 π-electrons and is aromatic. Owing to their unique structure, porphyrins and their metal complexes exhibit quite a number of interesting properties and find the most broad application in science, technology and medicine. In particular, they are used as photosensitizers for photodynamic therapy of cancer [1], as catalysts [2] and electrochemical sensors [3]; are promising materials for nonlinear optics, optoelectronics, liquid-crystal systems [4] etc.

N

X HN

N

NH

N

N

N

NH

HN

N

X

X

N

X

N

1 X = CH; 1a X = N

2

Numerous synthetic derivatives of porphyrins are known, e.g., tetraazaporphin (porphyrazine) 1a and tetraazatetrabenzoporphin (phthalocyanine) 2, which are of great practical significance. In particular, phthalocyanines are deep-blue colour pigments and dyes unsurpassed for their purity and brightness [5, 6]. The endeavour to expand the colour gamut of phthlanocyanine dyes led to the discovery of a new generation of macroheterocyclic compounds, structural analogues of phthalocyanine, in whose molecule one (3) or two (4) isoindole fragments are substituted by residues of aromatic diamines,

N

N

R

N

N

N N

NH

R

NH

N

N NH

R N

N

3

4

where R are aromatic or heteroaromatic cycles. Perfection of the methods of synthesis, use of various diamines made it possible to obtained macroheterocyclic compounds (Mc), which differ in the number and composition of small cycles combined into a macrocyclic system. As the result, Mc were synthesized, which differ in the size of the coordination cavity, composition and nature of atoms in it, which is of undoubted interest for coordination chemistry. The broadest potentialities of a

CRC but not checked

Synthesis, Structure Peculiarities and Biological Properties of Macroheterocyclic Compounds

221

structural modification of the macrocycle skeleton, which do not rule out the incorporation of pharmacophore groupings, introduction of substituents by the periphery, as well as metal atoms into the inner coordination cavity, make this class of compounds rather promising and attractive for the search of substances with practically valuable properties, in particular, with potential biological activity.

1

Synthesis, Structure and Properties of the Initial Compounds

The common method of synthesis of symmetric-structure macroheterocyclic compounds 4 is interaction of diamines (A) with phthalodinitrile or its functional derivatives 1,3-diiminoisoindoline (B) or 1,1-dialkoxy-3-iminoisoindolines (Scheme 1).

A

NH

2H N

A

2

NH2

+

2

B

B

B

NH

A

Scheme 1

The reaction of three-link BAB products with the functional derivatives of phthalodinitrile – 1,3-diiminoisoindoline or 1,1-dialkoxy-3-iminoisoindolines – is used to produce nonsymmetric Mc 3 (Scheme 2).

A

B NH

A

NH B HN

+

B NH

B

B

B

Scheme 2

As initial diamines, use is usually made of aromatic five- or six-membered carbo- and heterocyclic compounds containing amino groups in 1,3-positions. The use of 1,4-, bi- and polynuclear diamines leads to the formation of Mc with an increased internal cavity [7, 8]. Being included into the macrosystem, cyclic fragments of diamines can support or interrupt macroring conjugation. In this connection, of special interest are five-membered heterocyclic diamines, e.g., based on 1,2,4-triazole 5–8 and 1,3,4-thiadiazole 9, as they are heteroanalogues of pyrrole and their presence in the macrosystem can lead to the formation of an intracyclic conjugation system similar to the porphyrazine system. Substituted diaminotriazoles – 1(H)-3,5-diamino-1,2,4-triazole 5 (guanazole), 3,5-diamino-1-phenyl-1,2,4-triazole 6, as well as 3,5-diamino-1-(α-naphthyl)-1,2,4-triazole 7 – are obtained by condensation of hydrazine hydrate, phenyl hydrazine chloride or α-naphthyl hydrazine, respectively, with dicyandiamide [9, 10].

222

M.K. Islyaikin, E.A. Danilova, Yu.V. Romanenko, O.G. Khelevina and T.N. Lomova N N H2N

R NH2

N

5 R = H; 6 R = ɋ6ɇ5; 7 R= ɋ10ɇ7 8 R = C12H25

N N H2N

S

5-8

NH2

9

1-Dodecyl-3,5-diamino-1,2,4-triazole 8 was produced by direct alkylation of initial diamine 5 by dodecyl bromide in MeOH in the presence of MeONa [11]. 2,5-Diamino1,3,4-thiadiazole 9 is formed in the interaction of dithiourea with H2O2 [12]. Key substances for the synthesis of substituted macroheterocyclic compounds are substituted o-dinitriles [13]. The main methods of producing substituted phthalodinitriles are dehydration of diamides of substituted phthalic acids, substitution of halogen atoms in o-dihalogen derivatives for nitrile groups by the Rosemund–von Braun reaction, as well as substitution of mobile atoms and groups by the reaction of nucleophilic substitution and conversion of substituents of dinitriles already available in molecules. Methods for producing substituted phthalodinitriles are reviewed in rather many works. Some of them are presented in [13–16]. Maleodinitrile and its derivatives are also of interest for the synthesis of Mc. However, in the literature there is almost no information on the synthesis of Mc with pyrrole fragments. At the same time, already the first attempts of using substituted maleodinitriles and their functional derivatives led to the discovery of new macroheterocyclic compounds [17, 18], which were obtained based on 3,4-di(4-tert-butylphenyl)-2,5-diiminopyrroline [19] 10 and 1,2-dialkylthiomaleodinitriles [20] 11–13. Alk

NH

AlkS

CN

NH NH

AlkS

C4H9 C8H17 C12H25

CN

10

11, 12, 13

Phthalodinitrile and its substituted are rarely used for Mc synthesis directly. Usually, they are transferred into more reactive alkoxy- or diiminoisoindole derivatives. The reaction of phthalodinitrile with alcoholates of alkaline metals in an alcoholic medium (Alk = Me, Et, Pr) smoothly proceeds at room temperature [21] in accordance with Scheme 3.

CN

AlkONa

CN

14

OAlk

OAlk N

AlkOH

AlkO

NH

N

-AlkONa

NH

NH

NNa

15

16

OAlk

16a

Scheme 3

Introduction of a substituent into a phthalodinitrile molecule is accompanied by a decrease in symmetry, which predetermines the possibility of forming regioisomer alkoxy compounds: 4(7)- in the case of 3-substituted and 5(6)- for 4-substituted phthalodinitriles.

223


Substituents have a significant effect on the reactivity of phthalodinitrile. Thus, substituted phthalodinitriles, containing electron6 acceptor or weak electron-donor substituents in positions 3 or 4, R pass rather easily into respective alkoxy compounds already at 5 3 4 room temperature (e.g., 17–20). In contrast, 3-amino-, 3,6-diNH hydroxy-, 3,6-dimethoxyphthalodinitriles form no alkoxy comR pounds under these conditions. 3(6)NO217 There are several points of view on the mechanism of this reaction 3(6)Cl18 in the literature. The dominating one is the hypothesis [21] formu3(6)Br19 lated well back in 1956, in accordance with which phthalodinitrile 4(5)tBu20 14 in a strongly polarizing alcoholic medium passes into bipolar ion 14a (Scheme 4). The latter, by attaching a molecule of alkaline metal alcoholate, transforms into an isoindolenine derivative 15: OAlk 1 N2

7

OAlk

OAlk

CN N

N Na

CN N 14

14a

NNa 15

Scheme 4

It should be emphasized that the formation of bipolar ion 14a is postulated, and the role of the “polarizing effect of the medium” is not disclosed. Herewith, no proofs supporting this mechanism are given in the literature. Borodkin [22] studied the interaction of phthalodinitrile with sodium methanolate in benzene and methanol. He notes that, in the former case, a product of sodium methanolate addition to one of the nitrile groups is formed, whereas in the reaction in methanol the reaction product is an isoindolenine derivative. Therefore, the mechanism given in Scheme 4 is, apparently, not realized. It is also not ruled out that the isoindolenine derivative is formed as the result of the intramolecular cyclization of the monoaddition product. The results of kinetic studies of the reaction of phthalodinitrile with phenol given in a review [23] are in favour of such a mechanism. As phthalodinitrile can be recrystallized from alcohol and other organic solvents, while addition of alcoholate to its alcoholic solution already at room temperature leads to the formation of alkoxy compounds, the activating role of alkaline metal becomes evident. Interaction of phthalodinitrile and its substituted derivatives with nucleophilic reagents in the presence of alkaline metal cations was studied using the MNDO quantum chemical method [24]. The action of cation was found to cause a polarization of the nitrile group, as the result of which the carbon atom of this grouping acquires a significant positive charge and becomes a strong electrophilic site. As the result of the nucleophilic attack, a product of sodium methanolate molecule addition for this group is formed (Fig. 1); after an intramolecular regrouping, this product turns into an isoindole derivative. Thus, based on the generalization of the literature data and performed theoretical studies, the following mechanism of isoindolenine cycle formation appears to be the most probable: formation of the product of alkaline metal alcoholate addition for one of the nitrile

224

M.K. Islyaikin, E.A. Danilova, Yu.V. Romanenko, O.G. Khelevina and T.N. Lomova H16

O15

C7 C1

N8

2.7

6A

126.4

Li

C2 C9

N10

Figure 1 A model of the product of addition of LiOH to the phthalodinitrile molecule.

groups of phthalodinitrile followed by the intramolecular cyclization of the addition product (Scheme 5): OAlk

OAlk CN

AlkONa

NNa

N

CN

CN

NNa 21

14

15

Scheme 5

The effect of substituents on the direction and rate of this conversion was explained within the framework of this two-step mechanism [25]. 1,3-Diiminoisoindoline and its substituted species are obtained by the interaction of respective phthalodinitriles in methanol in the presence of sodium methanolate with ammonia [21]. Apparently, this compound is formed via 1-methoxy-3-iminoisoindolenine, which in the interaction with ammonia is converted into the end product. This reaction mechanism was considered in [26] using the semiempirical AM1 method. Conformational and tautomeric conversions of 1,3-diiminoisoindoline, as well as the influence of substituents and solvents on the energetics of isomeric transitions was studied in [27]. Analysis of the critical points of potential energy surface showed the cis-form 22 to be the most stable (Scheme 6).

N

H

H

NH N

H N

N

NH N

H

H

N N

H

H

22

22a

23

∆Hf = 81.24 kcal mol-1

∆Hf = 83.19 kcal mol-1

∆Hf = 88.14 lcal mol-1

Scheme 6


225

Figure 2 Structure of 2-(1-aminoisoindolenine-3-ylidenamino)-5-thioxo-1,3,4-thiadiazole-4-in 24.

Figure 3 Packing of the solvate of 2-(1-aminoisoindolenine-3-ylidenamino)-5-thioxo-1,3,4-thiadiazole-4-in 24 in a crystal.

Planar inversion (53 → 53a) proceeds with overcoming a relatively low energy barrier (~20 kcal·mol –1). For realizing tautomeric conversions (53a–54) as the result of intramolecular transfer, this value is ~70 kcal·mol –1. Calculations performed in “supramolecular” approximation showed that the prototropic rearrangement with participation of hydroxylcontaining solvents (MeOH, H2O) was made with lower energy expenses by about 17 kcal·mol –1. The data obtained are qualitatively consistent with ab initio calculations (6-31G basis) for intermolecular proton transfer in formamidine hydrate [28], where the activation energy is also observed to decrease by 20.8 kcal·mol –1 as compared with the intramolecular transfer in formamidine. Of special interest is the structure of products of condensation of 1,3-diiminoisoindoline with amines at one of the imino groups, as they are intermediate substances in Mc synthesis. The structure of 2-(1-aminoisoindolenine-3-ylideneamino)-5-thioxo-1,3,4-thiadiazole-4-in 24 was studied using X-ray diffraction analysis [29]. It has been shown that compound 24 is crystallized as the solvate with one molecule of DMFA and water. In the crystal, the molecule is planar, despite the much shorter intramolecular contact N(1)...S(1) 2.74 Å (the sum of van der Waals radii 3.34 Å [30]) (Figs. 2, 3).

226

M.K. Islyaikin, E.A. Danilova, Yu.V. Romanenko, O.G. Khelevina and T.N. Lomova H

166

∆Hf, kcal/mol

S

N N S

B

N

C

N

164

NH2

162

0

50

100 ϕ, deg

150

200

Figure 4 Internal-rotation energy profiles for tautomer 24a. The structural formula is given for ϕ = 180°.

The energy profiles of inner rotation of the thiadiazoline (thiadiazole) fragment around the ordinary bond N(3)-C(9) in molecule 24a were studied using the semiempirical quantum chemical AM1 method. The results of the analysis are given in Fig. 4. In the configuration space studied, the internal rotation proceeds with overcoming the low energy barrier ∆∆G = 4.02 kcal·mol –1, which makes it possible to assign this molecule to structurally nonrigid molecules. The low activation barrier of internal rotation suggests that not internal but external factors shall be of determining significance in the stabilization of this or that configuration, which is what is observed in the crystal (Figs. 2, 3), where form C is stabilized owing to intermolecular interactions. bis(1-Imino-3-isoindolinylideneamino)arylenes and azoles (TZP) are important intermediate products in the synthesis of both symmetric- and nonsymmetric-structure Mc [31–33]. bis(1-Imino-3-isoindolinilydeneamino)arylenes 25, 26 are the most studied of them.

N

X

NH NH 25 X = CH;

N HN HN 26 X = N

Despite the similarity in structure, they exhibit different reactivities. Thus, in the interaction with diiminoisoindoline, compound 25 forms a triisoindolebenzene macrocycle, but from compound 26 a similar product can not be obtained under the same conditions. A structural feature of three-link products 25 and 26 is the occurrence of two sufficiently “heavy” isoindole fragments connected with the diamine residue by aza bridges, which appears to determine the aptitude of the molecules to the formation of nonplanar

227


Figure 5 Model of complex 25 with DMFA.

Figure 6 Model of complex 26 with AlOH.

structures as the result of internal rotation relative to single N–C bonds. Planar inversion of hydrogen atoms of terminal imino groups, the tautomerism and internal rotation in molecules of bis(1-imino-3-isoindolinilydeneamino)arylenes (aryl = 1,3-phenylene, 2,6-pyridindiyl) were studied using quantum chemical methods [34]. Analysis of the critical potential energy surface points showed the internal rotation to proceed with low energy barriers and contribute to the structural nonrigidity of molecules. Solvation by aprotonic solvents, such as DMFA (Fig. 5) and, to an even greater degree, complexation with metals (Fig. 6) were found to stabilize the molecule, without rendering any significant effect on the electronic structure of terminal imino groups. Besides, the distance N(5)–N(6) decreases as compared with the initial three-link product. These factors together make it possible to consider such metal complexes as convenient templates for Mc synthesis. The most widespread method of producing TZP is by heating aromatic diamine with 1,3-diiminoisoindoline or 1,1-dialkoxy-3-iminoisoindoline in methanol at a temperature of about 40°C for several hours [31]. Compounds 26, 27–29 are produced in accordance with Scheme 7: NH NH

NH

NH H3CO

R

N H2N-R-NH2

HN

NH

HN

MXn

NH Cl

N

NH . HCl

N NH

NH H N N

R= N 26

26, 27-29

OCH3 NH

Cl

N

N 27

R

L(X)

N

26a,b; 27a-e; 28a,b,d; 29b

M N HN N N S 28

M = Zn (a), Al (b); Cu (c); Co (d); Ni (e)

Ph N N N 29

Scheme 7

228


Information on products of TZP interaction with metal salts is rather scarce and of contradictory character. Thus, the work [35] reports the synthesis of complexes of 2,6-bis(1-imino-3-isoindolinilydeneamino)-pyridine with Cu, Ni, Co and Au of 1:1 composition. A later paper [36] presents data on the synthesis of copper complexes of TZP based on 1,3-phenylene, 1,2,4-triazolidene-3,5 and 4-chloro-1,3,5-triazinylidene-2,6, which represent 2:1 complexes. The work [37] reports the synthesis of metal complexes 28a,b of 1:1 composition (Scheme 7), which were further used for template synthesis of ABBB-type Mc.

2

ABBB-type Macroheterocyclic Compounds

Recent years have witnessed an increased interest in noncentrosymmetric analogues of porphyrin [38] and phthalocyanine [39, 40]. These compounds exhibit practically valuable properties: nonlinear optical properties, ability to form ordered monomolecular Langmuir–Blodgett layers, etc. A rather elegant approach is when one of the pyrrole or isoindole fragments is substituted by a residue of aromatic diamine, as the result of which an ABBB-type structure is formed. The first representative of this class of Mc was the triisoindolebenzene macrocycle 30 obtained by Elvidge and Golden [31] in 1957 by the inN N X teraction of 1,3-bis(1-imino-3-isoindolinilydeneamino)phenylene with 1,3-diiminoisoindoline in N N boiling ethanol. H Similar to phthalocyanine, Mc are capable of N forming complexes with metals. Thus, respective N N metal complexes were first obtained by the interaction of a solution of 30 with copper(II) or nickel(II) acetates in boiling pyridine or with cobalt(II) in boiling benzyl alcohol [31]. However, these compounds were characterized only by the ele30 X=CH; 31 X=N mental analysis data. The same complexes were studied in [41] using electron and IR spectroscopy. In particular, it has been shown that complexation is accompanied by a batochrome shift of the long-wavelength absorption band, lying at 510 nm in the spectrum of a solution of metal-free compound 30 in α-chloronaphthalene. It has also been found that the nature of the metal complex former renders a significant effect on the value of this shift. Thus, the batochrome shift increases in the sequence Ni < Co < Cu (λmax = 520, 530, 550 nm, respectively). The work also notes that metal complexes are cations; however, the nature of counterion (L) is left without discussion. Later publications [42, 43] take acetate anion in the cases when metal acetates were used for complexation. The formation and destruction kinetics of nonsymmetric-structure Mc metal complexes is considered in [44, 45]. It has been shown that metal complexes 30a–c are rather stable compounds and possess a high thermostabilizing activity [46] with respect to polycapramide, and complex 30a (L =AcO) found application in practice as a polymer thermal and light stabilizer [47] and is manufactured under the name of Stabilin-9. Finding practically valuable properties inspired investigators to study in greater detail oooo

229


N

N

L

N

N Cu

N M

N

N

N

N

30a,

M=Cu; 30b,

30c,

M=Co,

NH

N

L

N N

N

M=Ni;

32

L=OAc, Cl, OH

the structure of nonsymmetric Mc. Thus, for a complex nonsymmetric-structure Mc with copper, the work [48] presents structural formula 32, which assumes the absence of counterion. However, in performing the complexation with anhydrous copper(II) chloride under conditions excluding the presence of moisture, complex 30a was synthesized, which contains chloride anion as a counterion [49]. X-ray electron data [50] are also in favour of structure 30. Attempts to use this approach for the synthesis of triisoindolepyridine macrocycle 31 failed to lead to the desired result [31]. Nevertheless, using respective complexes of 2,6-bis(1-imino-3-isoindolinylideneamino)pyridine as templates for cyclization with 1,3-diiminoisoindoline, Bamfield and Mac [35] succeeded in producing complexes of compound 31 with copper(II), nickel(II) and gold(III) (structures 31a–c).

N

N

NH

M

N

31a, M = Cu;

N

N

N

N N

31b, M = Ni

N N

N

Au N

N N N

31c

These compounds proved to be unstable to the light and to alkaline media, so further studies in this direction were stopped.

230

3


Azolophthalocyanines

Of special interest among macroheterocyclic compounds of ABBB type are Mc with azole fragments because, as compared with tetrapyrrole precursors, they have no centre of symmetry during the preservation of the internal macroring of the structure close to that of the porphyrazine one. Analysis of the literature [31] as well our own experimental studies have shown that attempts to obtain Mc with three isoindole fragments and heteroaromatic diamine residue by the method similar to the above method for the benzenetriisoindole macrocycle fail. Possible causes are the discrepancy of geometric sizes between the reaction centres of bis(1-imino-3-isoindolinilydeneamino)heteroarylenes and 1,3-diiminoisoindoline, as well as the low stability of TZP under reaction conditions [32]. As shown by quantum chemical calculations presented above, TZP-based metal complexes can be of interest as templates for the synthesis of nonsymmetric-structure Mc. Synthesis of metal complexes of azole-containing macroheterocyclic compounds was done by template condensation of respective TZP metal complexes with phthalodinitrile in a phenol medium [37] (Scheme 8).

N N

R M

CN

N Ln

N

CN

N

N

PhOH NH

HN

N

M = Zn (a), Al (b) N N

N N

N

S

133a,b

34a,b

R M N

N Ln N N

33a,b; 34a,b; 35b; 36a,b Ph N N N 35b

N 36a,b

Scheme 8

The choice of the condensation medium is explained by that initial TZP metal complexes are well-soluble in phenol. Besides, phthalodinitrile in the interaction with phenol passes into reactive phenoxy compounds [51]. Phenol, being a weak acid, slows down side reactions, in particular, formation of phthalocyanines, which, in the case of weakly-soluble reaction products, are extremely difficult to get rid of. The EAS of Mc metal complexes 33a,b–36a,b are characterized by sufficiently intensive absorption bands in the region of 500–600 nm; the intensity and position of these bands are largely dependent on the nature of solvent used for spectral measurements. For instance, the zinc and aluminium complexes of Mc with a thiazole fragment 34a,b feature three absorption bands in the region of 500–620 nm, which indicates the pronounced aromatic character of the macrocycles. Triazole-containing Mc of ABBB type, called triazolephthalocyanines, are the most extensively studied to date [52]. The methods of synthesis and the properties of triazolephthalocyanine metal complexes are given in the literature [53].


231

The major methods of their production are statistical condensation and interaction of TZP with respective nitrile or its functional derivatives in the presence of nickel acetate in organic solvents. In the former case, a mixture of substituted 1,3-diiminoisoindoline, 3,5-diamino-1,2,4-triazole and nickel acetate in a molar ratio 3:1:1 is heated in organic solvents (usually in 2-ethoxyethanol, butylonitrile, butanol) to form Mc containing the same substituents in all three isoindole fragments. In the latter case, one succeeds in obtaining ABB′B-type Mc with different substituents. Nickel proved to be the most convenient in the synthesis of this type of Mc; the N N successful use of copper salts is also reN N ported. Other metals do not lead to the N formation of required compounds. Recently, Torres et al. succeeded in proN N Ni X X ducing metal-free triazolephthalocyaN nine [54], however the purification of N N this substance proved very difficult. There are reports in the literature [53] on the production of nickel complexes with substituted triazolenaphthalocyanines of ABBB and ABB′B types. Y Triazolephthalocyanines have a con37a, b stant dipole moment, are capable of 37a X = OC8H17, Y = NO2, forming ordered Langmuir–Blodgett 37b X = OC8H17, Y = tBu layers [55, 56] possessing semiconductor properties [57], exhibit interesting nonlinear optical properties [58, 59], liquid-crystalline properties [60] and possess a high thermal stability [61]. The use of substituted pyrrolines instead of substituted isoindoles led to the discovery of a new class of noncentrosymmetric Mc of ABBB type – derivatives of porphyrazine, called by the authors [18] triazoleporphyrazines. Thus, hexa(4-tert-butylphenyl)triazoleporphyrazine 38 and its N-butyl-substituted analogue 39 were produced with good yields by the interaction of 3,4-di(4-tert-butylphenyl)-2,5-diiminopyrroline with 3,5-diamino1,2,4-triazole or 1-dodecyl-3,5-diamino-1,2,4-triazole, taken in a molar ratio 3:1, in dried butyl alcohol (Scheme 9). N N N NH 3

NH +

N N H 2N

N

N

R

N

N2H

N H N

R N N N

NH

38 R = H, 39 R = C12H25.

Scheme 9

232


Respective complexes 38a–c are synthesized by the interaction of triazoleporphyrazine 38 with nickel, copper and cobalt acetates in DMFA at a temperature of 100°C (Scheme 10). R R N R

N

N N N

H N

R

N N

R N H N

R

M(OAc)2 nH2O

N R

R

R

R

1

N

N N N

N

M

N

N

N N R

R

38a-c

R = 4-tBuPh

M = Ni (a), M = Cu (b),

M = Co (c)

Scheme 10

The structure of synthesized compounds was established based on the results of elemental analysis, NMR, IR and electron spectroscopy, as well as mass spectrometry. Comparative analysis of the spectral characteristics of Mc 38, its alkyl-containing analogue 39, as well as metal complexes 38a–c has shown that triazoleporphyrazine 38 exists in a configuration where the hydrogen atom is localized in position 1 of the triazole cycle. Similar to porphyrazine 1a, triazoleporphyrazine 40 is the molecule with a complex multicontour conjugation system. The structure of 40 can be presented by means of three tautomeric species 40a–c (Scheme 11). N N N N N

N H H N

N N N

1a

N N N

N H N

H N N N

40a

N N N N N

Scheme 11

N H H N

N N N

N

N

N

N

N HH N

N

N

40b

N

N

40c

The features of the geometric and electronic structure of molecules of porphyrazine 1a, tautomeric species 40a–c, as well as the Ni complex of triazoleporphyrazine 41 were considered in [62] using the quantum chemical method (DFT B3LYP/6-31G d,p). The results of the calculations are given in Table 1 and Fig. 7. Table 1 Total (E(RB+HF-LYP)) and relative (Erel) energies, ionization potential (IP) and dipole moment (µ) of structures 1a, 40a–c and 41 optimized at the level of DFT B3L YP/6-31G**. Structure 1a 40a 40b 40c 41 (a)For

structures 40a–c.

E(RB +HF-LYP), a.u. –1053.72663826 –1085.75771329 –1085.77012851 –1085.77770003 –2592.8681607

Erel,(a) kcal·mol –1

IP, eV

µ,

– 12.54 4.75 0.00 –

5.82 6.45 6.48 6.36 6.41

0.00 2.66 7.21 6.13 6.67

D


3

1.38

1.35 9

1.010

1.32 5

1.350

1 .4 67

1.360

8.2

10 8 .2

6

10

1.467

1 8. 10

1.45

N

1.010

1. 32 2

8 0. 12

130.4

130 .5

1.026

1. 35 0

11 1.4

127 .2

1.3 14

8

127.4 110.7 2 . 06 N N105.4 11 06. 1 111.6 H 127.4 12 2. 3 N 4 7 3 N 128.1 . N 111.0 1 1.31 106.4 7

H

1.39 3

1.860

1.869

1. 36 9

1.365

1. 36 7

1.45 4

N

127.5 12 110.0 9 7. . 4 06 1.878 Ni N N105.6 110 6.9 110.7 127.4 12 0. 4 6 N 7 3 N 127.5 . N 105.8 1 1.31 110.3 9 89

10 8 6.

1 6.

1.457

10

1.473

40c

4 1.45

1.328

1.326

1.349

1.294

5 7. 10 111.0 1.326 1 .3 103.0 128.0 N N 61 2 N 9. 11 1.3

127.3 106.4 0 . 1.012 08 N H H N110.9 110 8.0 106.7 125.9 12 2. 4 4 N 6 3 N 127.9 . N 105.3 1 1.33 111.2 7 12 4. 3

1.31 5

0 7. 10 5 111.6 3 .3 1 1. 34 102.9 129.0 N N 9 9 N 0. 12

77

1.45 0

N

N

1.3

9 1.44

1.279

40b

1.353

1.31 9

1.42 5

40a N

N

5 8. 10 2 107.8 3 1.3 1 .3 107.4 128.3 N N 58 1 N 1. 12

106.7 106.9 1.313

1.349

1.309

77 1.3

1.47 2

48 125.3 1.

.3 25

11 87N .38 1.3109.9 1

1

1.348

83 1 .3

H

1.47

1.347

95 1.3

1.366

N

1 1. 12

1.307

1.307

1.302

1.8 1.381 13

126 .8 1.48 6 110.1 1 6. 0 N 105.1 110 5 .7 112.2

.7 127

10 109.7 6. 3 105.8N .6 5 10112.6 1.4 84 12 5.0

1.347

N

1 1.369 113.8 108.1 .362 104.6 N 1 2 20 .35 N .2 32 .6 . 9 5 1 11

N1

1.485

2.1 10

1.36 3

H 11 1.328 1.0127.7 N N

233

1.352

41

Figure 7 Lengths of bonds (Å) and valence angles (°) in molecules of tautomeric species 40a–c and Ni complex 41 optimized on the level of DFT B3LYP/6-31G**.

The difference in the values of energy in the sequence 40a > 40b > 40c (Table 1) is relatively small. Tautomer 40c, the structure of whose internal ring is the most close to that of the porphyrazine ring, is the most advantageous from the energy point of view. In the sequence 40a, 40b, 40c the length of the N–N bond of the triazole cycle decreases and, in the case of structure 40c, approaches the length of the double bond N=N (Fig. 7). At the same time, the N–C bonds linking the N–N grouping with the internal macrocycle become longer to make 1.425 Å for 40c. This indicates that the N–N bond in this structure is significantly isolated from the conjugation system of the macroring. In the case of complex 41, the length of this bond is 1.294 Å and it occupies an intermediate position between respective bonds in structures 40b and 40c. Based on the geometric characteristics (Fig. 7) we [62] calculated the HOMA [63] and

234


NICS [64, 65] criteria and analyzed the aromaticity of 1a, 40a-c and 41 (Table 2, Fig. 8). Table 2 EN, GEO and HOMA indices of tautomers 40a –c and Ni complex 41. N N

N N

H H

N

d

N

1a

NH

H N

c

40a Structure

N N

N N

N

a

b

N N

d

N

N

a

N

N

N H

N

H N

N

c

40b

N

b

N N

d

N

N N

N N

N

a

H H N

c

40c

N

b

N N

d

N N

N N

N

a Ni N

c

b

N N

41

EN

GEO

HOMA

Porphyrazine 1a fragment a, c fragment b, d internal cycle

0.115 0.211 0.265 0.022

0.322 0.230 0.514 0.036

0.562 0.559 0.221 0.942

40a triazole cycle a fragment b fragment c fragment d internal cycle

0.122 0.044 0.332 0.332 0.316 0.015

0.464 0.034 0.812 0.484 0.801 0.094

0.414 0.922 –0.145 0.184 –0.117 0.892

40b triazole cycle a fragment b, d fragment c internal cycle

0.103 0.081 0.280 0.252 0.018

0.322 0.032 0.154 0.325 0.039

0.567 0.887 0.566 0.422 0.943

40c triazole cycle a fragment b, d fragment c internal cycle

0.098 0.118 0.220 0.289 0.016

0.287 0.225 0.253 0.568 0.036

0.615 0.657 0.527 0.143 0.948

41 triazole cycle a fragment b, d fragment c internal cycle

0.080 0.090 0.237 0.241 0.019

0.277 0.074 0.356 0.383 0.069

0.643 0.835 0.408 0.376 0.913

As it follows from the data presented in Table 2, the total aromaticity increases in the sequence of triazoleporphyrazines 40a < 40b < 40c < 41, which corresponds to the tendency of bond length equalization in these molecules. Transition from 40a to tautomers 40b and 40c is observed to be accompanied by an increase of the total aromaticity against the background of a decrease of the local aromaticity of the triazole cycle. The latter is even due to the fact that part of the triazole cycle including atom N-4 and adjacent carbon atoms are involved in the conjugation system of the

235

Synthesis, Structure Peculiarities and Biological Properties of Macroheterocyclic Compounds H

-11.6

N N N

N H

-15.5

H N

1a

N N

N -3.84

N N

N N N

N -6.90

H N

40a

N N

-13.2

N N N

N N N

N H

-14.1

H N

40b

N N

-12.6

N N N

N

N

N

N

NH

-14.8

-5.60

HN

N

N

40c

N N N N

N N Ni N

-5.87

N N N

41

Figure 8 Calculated NICS values (ppm) for structures 1a, 40a–c, 41.

internal macroring. This leads to the seclusion of the N–N bond, which in the case of structure 40c acquires a pronounced double character. As the result, the alternation of single and double bonds in the triazole cycle increases, which is manifested in an increase of index GEO from 0.034 up to 0.225 in passing from structure 40a to 40c. It should be noted that the difference in the HOMA values of total and local aromaticity of the triazole nucleus decreases in the sequence 40a > 40b > 40c (0.508, 0.320 and 0.042, respectively). Aromaticity of the internal cycle in this case increases (40c, HOMA = 0.948). Formation of nickel complex 41 is accompanied with a greater equalization of bonds and an increase of the HOMA value for the entire molecule up to 0.643. The bonds at the bridge nitrogen atoms become shorter and equalize with the simultaneous elongation of C–N bonds, nitrogen atoms of which directly interact with nickel. As the result of these changes, the aromaticity of the internal macroring (HOMA = 0.913) decreases as compared with similar characteristics of metal-free structures 40b and 40c. Analysis of the aromaticity criteria for [18]heteroannulenes designated on respective structural formulas 40b, 40c and 41 by bold lines (Table 2) shows that the largest value of HOMA = 0.803 is in structure 40b, in which the conjugation contour passes through the azo group of the triazole cycle. The respective contour of structure 40c is characterized by a smaller value, HOMA = 0.753. In the case of metal complex 41 the contour including the N–N group of triazole appears to be more aromatic (HOMA = 0.756) than the alternative contour (HOMA = 0.634). The magnetic criterion NICS (Fig. 8) confirms the tendencies of aromaticity change in the sequence 1a, 40a–c, 41, revealed using the HOMA criterion. Thus, 40a is the least aromatic, NICS = –6.90 ppm, whereas 40c appears to be the most aromatic of the tautomers (NICS = –14.8 ppm). A rather interesting fact was also that the triazole nucleus successively loses local aromaticity, which is also consistent with HOMA criteria for this sequence of tautomers. Analysis of aromaticity using the NICS criterion supports the conclusion that for the efficient participation in the formation of the aromatic macrocycle, its constituent heterocycles should lose part of their aromaticity, which should be compensated for by the gain in energy owing to the established conjugation in the macrosystem. The spectrum of compound 38 given in Fig. 9 is characterized by four intensive absorption bands. Their intensity decreases with the advance to the long-wavelength region. The band at 244 nm, which appears to be the result of electronic transitions in the phenyl nuclei of substituents, is the most intensive. The band at 324 nm, by analogy with tetraazaporphin, can be interpreted as a Soret band. The bands in the visible spectrum are the result of electronic transitions involving highest occupied (HUMO) and lowest unoccupied (LUMO) molecular orbitals. An intensive band at 417 nm, exhibited in the spectrum of

236


D 1.2

A 1.5 328

1.0

324

421

391

417

0.8

535 527 1

0.5

2

0.0 300

374

1.0 325338

400 Fig. 9

500

600 700 λ, nm

0.6 328 0.4

376

472 518 497

629 606

1

2 3

0.2 0.0 300

400

500 Fig. 10

600

700 λ, nm

Figure 9 Electronic absorption spectra: 1, 38; CHCl3, c = 2.16·10 –5; 2, 39, CHCl3, c = 1.71·10 –5 (mol·l –1). Figure 10 Electronic absorption spectra: 1, M=Ni, 38a, CH2Cl2, c = 1.71·10 –5; 2, M=Cu, 38b, CHCl3, c = 2.05·10 –5; 3, M=Co, 38c, CHCl3, c = 1.03·10 –5 (mol·l –1).

substituted porphyrazine [66] as a weak band at 450 nm, apparently, is due by its origin to the disturbance of symmetry of the molecule owing to the substitution of the pyrrole fragment by the triazole fragment. The long-wavelength band at 527 nm in substituted triazoleporphyrazine 2 is shifted hypsochromicly by 139 nm as compared with the band of respective porphyrazine [66]. The shape of absorption spectral bands of 38 and of dodecyl-substituted compound 39 is practically the same (Fig. 9), which indicates the identity of their chromophore systems. At the same time, in compound 39 the 1H-triazole form is fixed by the presence of the alkyl substituent in position 1. This fact supports the presence of a hydrogen atom in position 1 of the triazole cycle in the molecule of triazoleporphyrazine 38. The absorption bands in the case of 39 are slightly shifted towards the long-wavelength side, which is a consequence of the exhibition of a weak +I effect by the substituent. Three groups of bands can be singled out in the EAS of complexes 38a–c (Fig. 10): the long-wavelength absorption band at 602–629 nm and a broadened band at 570 nm, which is exhibited in the case of compounds 38a and 38b as an inflection; a broad band in the mid part of the spectrum at 472–518 nm; and two bands in the ultraviolet part of the spectrum at 325–391 nm. The arrangement and sufficiently high intensity of the long-wavelength absorption band indicate the aromatic character of metal complexes. The nature of metal has an influence on the position of the absorption bands; the batochromic shift of the long-wavelength absorption band increases in the sequence Ni < Co < Cu.

4

State of Triazoleporphyrazines in Proton-donor Media

Study of the state and stability of the macrocycles and their complexes in proton-donor media is an important issue in the chemistry of porphyrazine, the solution of which makes it possible to determine the limits of their possible applications, to develop efficient methods of synthesis and isolation of pure ligands and complexes. The stability of porphyrazines and their metal complexes in a proton-donor medium is determined by the form of existence of compounds in this medium (neutral or acidic). Porphyrazines (H2Pz) are weak multicentre conjugated bases. The number of por-


237

phyrazine donor sites involved in acid-base interaction, the character of the interaction, as well as the values of stability constants of acidic species produced depend on the structure of porphyrazine and the properties of the proton-donor medium. Porphyrazine ligands can produce acidic species both by the intracyclic and extracyclic (meso-) atoms of nitrogen. In metal complexes, ϕn orbitals of the intracyclic nitrogen atoms participate in the formation of bonds with complex-forming metal, therefore, only meso-atoms of nitrogen are involved in acid-based interaction [67]. Triazoleporphyrazines, which differ from porphyrazines by the substitution of one of the pyrrole cycles by a triazole residue, have two additional nucleophilic centres in the triazole cycle, which can be considered as possible protonation sites. Besides, the proximity of these atoms to the internal macroring makes it possible to assume a significant effect of protonation on the properties of triazoleporphyrazines. The process of acid-base interaction with participation of porphyrazines is of a complex character [68]. Proton transfer from acid HA to base B proceeds via the stages of forming acidic associate, H associate, ion-ionic associate and totally ionized protonated species: HA + B

B…HA

acidic iassot acidic associate H associate protonated associate

1

B…H…A

BH+…A –

BH+ + A– (1)

H associate

ion-ionic associate protonated associate species

The forming acidic species differ one from another by the degree of proton transfer from the molecule of acid to the donor site and should be spectrally distinguishable. The final result of acid-base interaction depends on the electronic and geometric structure of acid or base, as well as on the solvation features of anion A– and cation BH+, which is determined by the nature of solvent. In media with low ionizing ability, the electron-donor centres of porphyrazine, by participating in weak acid-base interaction, form H associates and ionionic associates. The total proton transfer is possible only in a strongly ionizing medium. Interaction of porphyrazines with acids is accompanied with characteristic changes in the visible range of the electronic absorption spectrum (EAS), which correspond to the formation of various acidic species. A quantitative measure of basicity is the stability constant of the acidic species (Ks). Porphyrazines are Hammett indicators [67], and the value of Ks can be determined using the Hammett equation: pKs = nH0 + log Ii ,

(2)

where H0 is the Hammett acidity function, Ii = Ci /Ci–1 is the ratio of an ith and (i–1)th acidbase species being in equilibrium (indicator ratio), n is the number of donor sites involved in the acid-base interaction at a given stage. The values of acidity functions are known from the literature [69]. In mixtures of carboxylic acids with organic solvents, for which the H0 functions are not known, one can determine the concentration constants of stability of acidic species, using a modified Hammett equation: pKs = –x logCHA + log Ii ,

(3)

where x is the number of acid molecules involved in acid-base interaction at a given stage. The values of n and x can be determined as the slope of lines log Ii = f(H0) or log Ii = f(logCHA). We investigated the stage and stability of 1H-hexakis(4-tert-butylphenyl)triazoleporphyrazine 38 (H2TrPz) in a medium of benzene and 100% acetic acid.

238

.


N N N

N

N N

H N N

H N

N

38

Introduction of a small amount of acetic acid (0.4 mol·l –1, H0 = 7.5) into a benzene solution of triazoleporphyrazine leads to a hypsochromic shift of the Q band in the EAS (Fig. 11). Formation of the acidic species is completed in an acetic acid–H2SO4 –antipyrin medium (H0 = 3.95). Herewith, the absorption maximum of the Q band is shifted hypsochromicly to 473 cm –1. The stability constant of acidic species 38, as determined by the Hammett equation, is equal to 4.48±0.18. The number of donor sites involved in acid-base interaction is 0.97, i.e., ~ 1 (Fig. 12). Taking into account that acid-base interactions of triazoleporphyrazines occur in a medium with low ionizing ability, it can be concluded that an ion-ionic associate is formed in this case. A 1.6

531 1 2 3 4

1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

450

500

550

600

650 λ, nm

Figure 11 Change of electronic absorption spectra of triazoleporphyrin 38 in the process of acid-base interaction in time (benzene, 1; benzene–acetic acid system: 2, H0 = 6.08; 3, H0 = 5.33; 4, H0 = 4.5).


239

log I 0.5 0.0

4.0

4.5

5.0

5.5

6.0 H0

6.5

-0.5 -1.0 -1.5 -2.0

Figure 12 Dependence of log I on H0 for the acid-base interaction of triazoleporphyrin 38 in a benzene–acetic acid medium. Α 1.6

531

1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

450

500

550

600

λ, nm

Figure 13 Change of acid-base interaction of triazoleporphyrin 38 in the process of destruction in a benzene–acetic acid medium at a temperature of 50°C and H0 = 5.95.

Comparison of the values of pKs for triazoleporphyrazine, tetraazaporphin (1.0) and octaphenyltetraazaporphin (–1.33) has shown that substitution of the pyrrole cycle by a triazole fragment increases the basicity of the macrocycle. An earlier work [68] has shown that the first stage of the acid-base interaction of porphyrazines proceeds via one of four meso-atoms of nitrogen and leads to a batochromic shift of the Q band in the EAS. The high basicity of the macrocycle of triazoleporphyrazine, the hypsochromic shift of the Q band of its EAS suggest that the protonation centre is the nitrogen atom of the triazole cycle in position 4 (or N-3, in accordance with the numeration of Table 2). Quantum chemical studies we carried out at the DFT level [70] confirm this assumption. Triazoleporphyrazine proved to be less stable than porphyrazines. In the course of time at an elevated temperature, it breaks down already in solutions of acetic acid in benzene. Figure 13 shows a change of the EAS of triazoleporphyrazine in the process of destruction.

240

M.K. Islyaikin, E.A. Danilova, Yu.V. Romanenko, O.G. Khelevina and T.N. Lomova 313 K

log C0H2TrPz / CH2TrPz

0.07 0.06 0.05

323 K

0.04 0.03

333 K

0.02 0.01 0.00

0

10

20

30

40

50 τ, min

Figure 14 Dependence of log (C 0H2TrPz / CH2TrPz) on time τ for the destruction reaction of triazoleporphyrin in a benzene–acetic acid medium at temperatures 313 K, 323 K, 333 K and H0 = 4.87. log ke

333

-4.0 -4.1

323

-4.2 -4.3 313

-4.4 -4.5 -4.6 1.10

Figure 15 rin 38.

1.12

1.14

1.16

1.18

1.20

1.22 log CCH3COOH

Dependence of log keff on log C CH3COOH for the destruction reaction of triazoleporphy-

The stability of the macrocycle in solutions is assessed by kinetic stability, which is characterized by the rate constant of its destruction. We studied the kinetic stability of triazoleporphyrazine in protonated form in solutions of acetic acid in benzene with acidity functions H0 equal to 5.32, 5.17, 4.87, 4.80 and 4.60, within the temperature range of 313–333 K. Kinetic measurements were carried out at a large excess of acetic acid, i.e., under conditions of the pseudofirst-order reaction. The first order for triazoleporphyrazine is confirmed by linear dependences of log(C 0H2TrPz /CH2TrPz) on reaction time (Fig. 14). The effective reaction rate constants were calculated by equation (4): k = 1/τ ln(C 0H2TrPz /CH2TrPz).

(4)

Table 3 presents the kinetic parameters of the triazoleporphyrazine destruction reaction. The experimental data show that the rate of destruction increases with the increase of the concentration of acetic acid. The dependences logkeff = f(logCAcOH) are of linear


241

character (Fig. 15) with the slope close to 2, i.e., the reaction is of second order with respect to the concentration of acetic acid. Table 3 Kinetic parameters of destruction of triazoleporphyrin 38 (CH02TrPz = l·10–6 mol/l). CCH3COOH, mol·l–1

H0

T, K

keff ·105, s–1

E, kJ·mol–1

∆S #, J · mol–1 · K–1

13.10

5.32

313 323 333

2.00±0.10 3.15±0.15 4.90±0.18

37±2

–224±2

13.98

5.17

313 323 333

2.44±0.10 3.85±0.11 5.95±0.14

37±2

–223±3

14.85

4.88

313 323 333

3.21±0.12 5.20±0.20 8.20±0.26

39±3

–215±2

15.72

4.80

313 323 333

3.50±0.10 5.71±0.45 9.00±0.14

39±2

–215±4

16.60

4.60

313 323 333

4.25±0.20 6.80±0.20 10.60±0.40

39±4

–213±2

Destruction of porphyrazines can be protolytic and acidolytic under the action of acid molecules, or a more complex solvoprotolytic process under the action of solvated proton. Destruction of porphyrazines in aqueous solutions of H2SO4 is hydrolytic [71], i.e., proceeds under the action of hydroxonium ions H3O+. To split the macrocycle, the system must in all cases be destabilized by two H3O+ ions. Destruction of triazoleporphyrazine in a benzene-acetic acid medium is acidolytic, i.e., proceeds under the action of AcOH molecules. The following scheme of destruction of the triazoleporphyrazine macrocycle can be proposed. First there occurs the protonation of nitrogen atom, which was localized in position 4 of the triazole cycle (centre N-3 in Table 4); herewith, the aromaticity of the macrocycle decreases. In the further successive interaction with two acetic acid molecules, there occurs the protonation of nitrogen meso-atom. Further, acetate ion attacks the α-carbon atom of the pyrrole fragment, and the bond C–N in the macrocycle breaks down. Quantum chemical studies were carried out by the AM1 method to elucidate the most reactive nucleophilic centres (nitrogen atoms) in the molecule of the monoprotonated species of triazoleporphyrazine. Table 4 presents the main thermodynamic characteristics of diprotonated species, which show that the most probable sites of attack by acetic acid molecules are nitrogen atoms N-8 and N-10 of the macrocycle (Table 4). We considered the effect of metal on the basicity of the triazoleporphyrazine macrocycle by the example of acid-base interaction of Cu(II) and Ni(II) complexes with 1H-hexakis(4-tert-butyl)triazoleporphyrazine 1 and 2, of a nickel(II) complex with 3,4-di(4-tert-butylphenyl)dibenzotriazoleporphyrazine 3 in benzene–AcOH and dichloromethane–AcOH media, as well as of a Ni(II) complex with [4-(n-triphenylmethylphenoxy)] 7,8:12,13:17, 18-tribenzotriazoleporphyrazine 4 in a dichloromethane–AcOH medim.

242


Table 4 Major thermodynamic characteristics of diprotonated species (AM1 method, 298 K). 2

10

N

1

H

N N a

4

N

N 3

d N 8

N

9

H N7 c

5

N b N

. 2H

6

Position of protons

∆Hf, kcal·mol –1

∆S, kcal·mol –1 ·deg – 1

∆G, kcal·mol – 1

∆∆Hf, kcal·mol –1

3,2 3,4 3,6 3,8 3,10 3,5 3,7 3,9

730.64 720.08 712.93 711.58 712.27 715.61 733.39 712.68

126.1 125.5 124.4 124.5 125.1 123.6 123.7 123.9

693.06 682.68 675.86 674.48 674.99 678.78 696.53 675.76

–72.99 –83.55 –90.70 –92.05 –91.36 –88.02 –70.24 –90.95

∆G – calculated by the formula ∆G = ∆Hf – ∆ST ∆∆Hf – thermal effect calculated for diprotonation: ∆∆Hf = ∆Hf (H2TrPzH22 +) – [Hf (H2TrPzH+) + ∆Hf(H+)], where ∆Hf(H+) = 314.9 kcal·mol –1.

N N

N N N N N

N

N Ni

N

N

N N

Ni N

N

38b

N N N

N

N

N

N

38a

N

Cu

N

N

N

N

N N N

N

N

N

N

N

42

43

N

N Ni

N

N

N

O

CPh3


243

A 0.5

628

1 2 3 4 5 6 7

0.4 0.3 0.2 0.1

0.0 400

450

500

550

600

650

700

750 λ, nm

Figure 16 Change of electronic absorption spectra of compound 42 in the process of acid-base interaction in a dichloromethane–acetic acid medium: 1, in CH2Cl2, 2–7, in CH2Cl2 –AcOH (C AcOH = 0.176–17.64 mol·l –1).

In the acidification of solutions of complexes in benzene and dichloromethane by a small amount of 100% acetic acid, as in the case of the triazoleporphyrazine ligand, there occurs a hypsochromic shift of the Q bands in the EAS of the complexes. Formation of acidic species in a benzene–AcOH medium for compound 38b occurs within the interval of H0 = 6.55–4.61 (CAcOH = 0.40–16.48 mol·l –1; for compound 38a, H0 = 6.60–4.60 (CAcOH = 0.87–16.59 mol·l –1); compound 42, H0 = 6.55–4.60 (CAcOH = 1.74–16.59 mol·l –1) (Fig. 16). The thermodynamic constants of stability (pKs) of acidic species of compounds 38a,b, 42 in a benzene–AcOH medium and the concentration stability constants of acidic species of compounds 42 and 43 in a dichloromethane–AcOH medium are presented in Table 5. Table 5 Stability constants (pKs) of acidic species of triazoleporphyrins. Compound

Medium

pKs

H2TrPz 38 38a 38b 42 42 43

benzene-ACOH benzene-ACOH benzene-ACOH benzene-ACOH dichloromethane-ACOH dichloromethane-ACOH

4.48±0.18 6.22±0.03 6.50±0.04 6.45±0.05 0.43±0.04 0.27±0.07

Based on the dependences logIi = f(H0) and logIi = f(logCAcOH), it has been found that one proton is involved in acid-base interaction (the slope of these lines in a benzene–AcOH medium for compounds 38a is 0.89; 38b, 1.14; 42, 1.07; and in a dichloromethane–AcOH medium for compound 42, 0.80; for 43, 0.82) (Figs. 17, 18). The high basicity of the macrocycle of triazoleporphyrazine complexes, a hypsochromic shift of the Q band in the EAS suggest that the protonation centre is a nitrogen atom in the pyrrole cycle in position 1. Quantum chemical studies of protonated species of triazoleporphyrazine we recently carried out [70] support this suggestion.

244

M.K. Islyaikin, E.A. Danilova, Yu.V. Romanenko, O.G. Khelevina and T.N. Lomova log I 0.9 0.6 0.3 0.0

5.4

5.7

6.0

6.3

6.6

6.9 H0

-0.3

Figure 17 Dependence of log I on H0 for the acid-base interaction of compounds 38a,b and 42 in a benzene–acetic acid medium: °, compound 38a; •, compound 38a; S , compound 42. log I 0.9 0.6 0.3 0.0 -1.0

-0.5

,0.0

0.5

1.0 log CCH3COOH

-0.3 -0.6

Figure 18 Dependence of log I on log C AcOH for the acid-base interaction of compounds 42 and 43 in a dichloromethane–acetic acid medium: S , compound 42; •, compound 43.

Experimental data show that nickel and copper complexes are more basic than the ligand; and the copper complex is more basic than the nickel complex. The presence of the triphenylmethylphenoxy substituent in the molecule of triazoleporphyrazine decreases the basicity of the macrocycle.

5

ABAB-type Macroheterocyclic Compounds

The first representative of this vast class of Mc (2+2 or ABAB type) 4 was the symmetric pyridine macrocycle [72, 73] 44, which Campbel [73] called hemiporphyrazine. Subsequently, this term was extended to the entire class of Mc 4. Torres [33] proposed to add the name of residue R as a prefix to the word hemiporphyrazine. For instance, compound 120 can be called pyridinohemiporphyrazine (HpH2). Methods of synthesis and properties of ABAB-type Mc and their metal complexes are considered in detail in a review [33].


245

N N

NH N

N NH

N N 44

The structure of compound 44, as well as of its complexes with Ni 44a, Cu, Co, Mn, Zn and Ge, was studied using the X-ray diffraction analysis [74–79].

Figure 19

Structure of 44a.

It has been found that the molecule of symmetric macroheterocyclic compound 44, constructed from isoindole and pyridine nuclei, is in a solid state capable of acquiring both the planar and nonplanar configuration depending on the nature of complex-forming metal, as well as on the presence or absence of solvated molecules of solvent. Hecht and Luger [75] present information on the studies of symmetric benzene macroheterocyclic compound 45 by the XDA method. It has been shown that in the crystalline state the macrocycle exists as a solvate, which includes molecules of ethanol and water and has a nonplanar structure. N NH N

N

NH N 45

Spectral studies of a symmetric-structure tert-butyl-substituted benzene Mc we have carried out in 1981 using 1H NMR spectroscopy [80] show that this substance in solution can form stable complexes with a solvent, in particular, with DMFA. However, their structure remained unknown.

246


N(1s) O(1s) N(3) N(5) O(2s) N(2s)

45a

45b

Figure 20 Structure of complexes 45a,b according to the X-ray diffraction analysis data. (Hydrogen atoms at carbon atoms are omitted for clarity.)

In this connection, we [81] carried out an X-ray study of complexes of dibenzenehemiporphyrazine with DMFA. Monocrystals suitable for X-ray studies were obtained by slow diffusion of water vapours into a solution of 45 in DMFA. It was shown that 45 formed complexes with DMFA of composition 1:1 45a and 1:2 45b, in which the macrocycle has a saddled shape (Fig. 20). In the case of 1:1 complexes, exocyclic atoms of nitrogen N(2), N(3), N(5), N(6) lie in one plane to an accuracy of 0.04 Å. Benzene rings and isoindole fragments deviate from the midplane of the molecule to the opposite sides. This conformation of the macrocycle is, probably, due to abridged intramolecular contacts H(1N)…C(28) 2.72 Å (the sum of van der Waals radii [71] 2.87 Å), H(1N)…C(10) 2.75 Å, H(1N)…C(9) 2.77 Å, H(1N)…C(27) 2.77 Å, H(4N)…C(10) 2.68 Å, H(4N)…C(11) 2.74 Å, H(4N)…C(23) 2.79 Å, H(4N)…C(28) 2.69 Å. The macrocycle forms an oval cavity, whose size is determined by the distances H(10)…H(28) 3.33 Å and H(1N)…H(4N) 2.64 Å. Owing to this, the DMFA molecule is above the macrocycle, thus forming three-centre hydrogen bonds N(1)-H(1N)…O(1S) 2.14 Å (angle N-H…O 160°) and N(4)-H(4N)…O(1S) 2.17 Å (angle N-H…O 152°). Addition of one more DMFA molecule (complex 45b of composition 1:2) leads to a significant change of conformation of the macrocycle (Fig. 20). Exocyclic atoms do not lie in one plane. The angle between the lines N(2)...N(6) and N(3)...N(5) is 5.8°. In compound 45b, one molecule of DMFA occupies the position similar to that found for a complex of composition 1:1, thus forming three-centre hydrogen bonds N(1)-H(1N)…O(1S) 1.98 Å (angle N-H…O 172°) and N(4)-H(4N)…O(1S) 2.12 Å (angle N-H…O 166°). The second molecule of the solvent is on the opposite side of the macrocycle and is linked with it by the weak intermolecular hydrogen bond C(10)-H(10)...O(2S) 2.24 Å (angle C-H…O 158°). In crystalline state, intermolecular interactions have a strong influence on the structure of macromolecules, thus distorting the manifestation of internal factors. An idea of the structure of compounds in isolated state is given by quantum chemical calculations. In this connection, we carried out a theoretical study of the structural features of Mc 45 using the semiempirical method AM1 for the planar 45PL (D2h) and nonplanar 45A (C2v), 45B (C2h) configurations and crystalline solvate 45*DMF, as well as the transitory


45PL, D2h, ∆Hf = 259.8 kcal·mol –1

45B, C2v, ∆Hf = 242.5 kcal·mol –1

247

45A, C2h, ∆Hf = 250.4 kcal·mol –1

45*DMF, C1, ∆Hf = 200.0 kcal·mol –1

Figure 21 Spatial structure of symmetric benzene Mc 45 according to the data of quantum chemical calculations by the AM1 method.

state 45TS between configurations 45A–45B. Models of molecules, as well as the calculated values of the heats of formation of planar 45PL (D2h) configuration and nonplanar configurations 45A,B and solvate 45*DMF are given in Fig. 21. In the case of planar configuration 45PL (D2h), hydrogen atoms in the internal coordination sphere are spatially close. In particular, the distance between hydrogen atoms of benzene nuclei is 2.17 Å, and between hydrogen atoms of imino groups 2.59 Å. The spatial closeness of atoms in the internal sphere leads to their strong mutual repulsion. The molecule is stabilized as the result of its removal from the plane of the benzene and isoindole fragments. Two stereoisomers were found which conformed to the potential energy surface minima: chair 45A (C2h) and saddle 45B (C2v). The formation energy gain is 9.4 and 17.3 kcal·mol –1, respectively, as compared with the planar configuration 45PL, symmetry D2h. Thus, the most energetically advantageous is isomer 45A with the saddled configuration, in which benzene nuclei are removed from the plane, determined by the exocyclic nitrogen atoms, by 42.8°. The isoindole fragments are inclined to the opposite side by an angle of –21.2°. The same configuration is fixed in solvate 45*DMF, which is consistent with the XDA data for 45a. The calculated value of the energy barrier between conformers 45A => 45B is equal

248


to 0.6 kcal·mol –1, and for the inverse transition this characteristic corresponds to 8.5 kcal·mol –1. The so small conformation-transition energies commensurable with the energy of the hydrogen bond suggest that similar conversions can occur in solvents. Herewith, depending on the nature of the solvent, one or another conformation can be stabilized. This conclusion drew confirmation from a recently published work [82], which, using tertbutyl-substituted Mc as an example, considered the effect of solvents by means of the 1H NMR method. Thus, symmetric-structure Mc are structurally nonrigid molecules capable of existing as various conformers separated by low activation barriers. The predominant content of this or that species in crystalline state and in solution is determined by the character of intermolecular interactions, the significant of which is specific solvation. It should be noted that unsubstituted Mc are hard to dissolve in organic solvents. However, the solubility of Mc can be significantly increased by introduction of bulky substituents, e.g., tert-butyl groups. Symmetric-structure tert-butyl-substituted Mc 46–50 were synthesized in accordance with Scheme 19 by the interaction of 4-tert-butylphthalodinitrile or respective alkoxy- or three-link compounds with aromatic diamines in ethylene glycol or butanol. It proved the most expedient to synthesize compounds 46–50 from 4-tert-butylphthalodinitrile via respective alkoxy compounds, without isolating the latter from the stock (Scheme 12). N

OAlk

CN AlkONa, AlkOH

N

R(NH2)2

R

N

N

N M(CH3COO)2

NH HN

R

N

Ln

N R N 46a-c – 50a-c N NH

N 47

HN

N M N

R N 46 – 50

R= 46

NH

NH R(NH2)2 N

R(NH2)2

NH HN

CN R(NH2)2

N

R

N N Ph

N 48

N 49

N N

M = Cu (a); Co (b); Ni (c) (1-Nph )

N 50

Scheme 12

At the introduction of tert-butyl groups, the character of Mc electronic spectra remains almost the same. Usually, a small long-wavelength shift of the bands is observed, with absorption in the region of 300–450 nm preserved. The experimental values obtained by means of NMR spectroscopy are consistent with these data. Proton signals of imino groups are in the region of 10.2–12.6 ppm. The presence of signals of intracyclic protons in a weak field is experimental proof of the absence of a unified macrocyclic conjugation system in compounds 46–50. The conclusion of the absence of a unified aromatic system in the macrocycles is consistent with the results obtained in studies of a germanium complex of the symmetric pyridine macroheterocycle by 1H NMR spectroscopy [84]. A later work [85] registered signals of intracyclic protons in a weak field at 15.7 and 15.2 ppm (CDCl3) in an 1H NMR spectrum of symmetric-structure tert-butyl-substituted Mc with fragments of 1-dodecyl-1,2,4-triazole containing two nitrile groups in one of the isoindole fragments.

249


Respective metal complexes were obtained by the interaction of compounds 46–50 with excess anhydrous acetates of respective metals in butanol (Scheme 12) [83]. The thermal stability of the compounds increases in complexation. The complex of macroheterocyclic compound with fragments of 1-phenyl-1,2,4-triazole 49a with copper [83] proved to be the most thermally stable; the maximal exo effect (depletion of mass, 86%) in air for it is observed at 602°C. Introduction of tert-butyl groups leads to a decrease of stability of Mc in proton-donor media [86]. This, apparently, can be explained by that the introduction of two tert-butyl groups possessing a pronounced +I effect into the Mc molecule is accompanied by an increase of electronic density on exocyclic atoms of nitrogen, which facilitates protonation and subsequent breakdown of molecules. In complexation, stability of tert-butyl-substituted compounds increases by approximately an order of magnitude. While for compound 49 the efficient rate constant of destruction is equal to 0.11×10 –3 s –1, for its complex with copper this value is 0.47×10 –4 s –1 [86]. Complexation is accompanied by a decrease of solubility of tert-butyl-substituted Mc in organic solvents.

6

ABABAB-type Macroheterocyclic Compounds

A macroheterocyclic compound based on 2,5-diamino-1,3,4-thiadiazole was first synthesized in 1970s [87, 88]. By analogy with macrocycles known at the time, that compound was presented as a symmetric ABAB-type compound. However, studies performed simultaneously and independently by two research groups [17, 89] showed it to be a new class of ABABAB-type compounds, the basis of the structure of which is a macrocyclic system consisting of six alternating substituted pyrrole and 1,3,4-thiadiazole fragments linked by aza bridges. These systems have a formally conjugated internal macroring containing 30 π-electrons, which conforms to the Hückel’s rule. tert-Butyl-substituted macroheterocyclic compound 51 was synthesized by the interaction of equimolecular amounts of 2,5-diamino-1,3,4-thiadiazole and 4-tert-butylphthalodinitrile in phenol, or 5-tert-butyl-1,3-diiminoisoindoline in 2-ethoxyethanol (Scheme 13). The presence of bulky tert-butyl groups in the molecule was a cause of a sufficiently good solubility of 51 in organic solvents, which enabled its chromatographic purification.

CN 3

3 H2N

CN

N N

PhOH

S

N

S

NH2 N

N

N

N

N

N H NH 3

NH NH

N 3 H 2N

N N S

EtOEtOH NH2

Scheme 13

N S

N

H N H

N

N N

S N

250


Condensation of equimolecular amounts of 3,4-bis(4-tert-butylphenyl)pyrroline2,5-diimine with 2,5-diamino-1,3,4-thiadiazole during the boiling of the stock for 24 h yielded compound 52 with an increased coordination cavity; the compound incorporates substituted pyrrole fragments into the macrosystem [90] (Scheme 14).

N NH 3

NH + 3 H2N NH

S i

N N S

NH2

N N

N N

N H

N S

N N

H

H N N

N

S

N

N

N

52 Scheme 14

The structure of Mc was established using the methods of electron, IR, 1H and 13C NMR spectroscopy, as well as mass spectrometry data. Thus, the PMR spectrum of compound 51 measured in deuterated chloroform, along with signals in the region of 1.40–1.25 ppm, which characterize absorption of protons of tert-butyl groups, and signals in the region of 7.94–7.61 ppm, which are evoked by absorption of protons of benzene nuclei of isoindole fragments, yielded a singlet in the region of 12.35 ppm, which vanishes upon addition of deuterated water and can be assigned to the absorption of protons of imino groups. The location of the latter absorption in the weak field indicates the nonaromatic character of the macrocycle. It should be emphasized that under similar conditions, an ABAB-type macrocycle is formed from 3,5-diamino-1,2,4-triazole. The cause of so significant differences in the structure of the condensation products is the nature of the thiadiazole fragment. As shown by the XDA data of 3,5-diamino-1(H)-1,2,4-triazole [91] and 2,5-diamino-1,3,4-thiadiazole [92], the lengths of S-C bonds in the thiadiazole cycle are much larger than the respective N-C lengths in the triazole cycle. As the result, the angle between the C-NH2 bonds in the molecule of 2,5-diamino-1,3,4-thiadiazole proves to be 12.5° larger than that in the molecule of 3,5-diamino-1(H)-1,2,4-triazole, which predetermines the possibility of forming macroheterocyclic compounds of larger size in the case of 2,5-diamino-1,3,4-thiadiazole. The spectral curves obtained for solutions of Mc 51, 52 (Fig. 22) in chloroform have an unusual appearance and are practically the same. Two most intensive absorption bands are observed in the region of 392 and 413 nm for 51 and at 428 and 452 nm for 52, as well as two bands of average intensity respectively at 463 and 501 nm and at 515 and 552 nm. The presence of the main absorption in a so short-wavelength region (390–450 nm) indicates the nonaromatic character of the macrocycle. The features of the electronic and geometric structure of ABABAB-type macroheterocyclic compounds were studied using semiempirical quantum chemical methods [90]. Analysis of the potential rotation energy surface of thiadiazole nuclei has shown that these


D

251

1.5 S N

N

N H

1

N

N S

392

N N

N

N N H N S N H N N

413 428

452

N

S N

N

N

N H

N

N

N

S

N

N N H N S N H N N

N

0.5

0 250

350

450

550

650

Wavelength, nm

Figure 22 Electronic absorption spectra (solvent, CHCl3) of 51, c = 2.63 g-mol·l –1; of 52, c = 1.65·10 –5 g-mol·l –1.

Figure 23 A model of Mc 51 (the geometric parameters optimized by the AM1 method).

molecules are structurally nonrigid. The most preferable configurations are nonplanar ones; in them, sulfur atoms of thiadiazole fragments are oriented outside. Six bridge atoms of nitrogen lie in one plane, whereas thiadiazole nuclei as well as pyrrole or isoindole nuclei are out of this plane. The result is an umbrella-like configuration (Fig. 23). The so unusual structure of Mc, containing three thiadiazole fragments and three pyrrole or isoindole fragments, assumes the occurrence of unusual coordination properties as compared with porphyrazines and phthalocyanines. These Mc proved to be capable of forming stable complexes of composition 3:1. What is more, these metal complexes could be synthesized both by the interaction of metal-free compounds with salts of respective metals (stepwise synthesis) and by heating a mixture of initial substances in the presence of a metal salt (template synthesis). The complexes were synthesized by the interaction of metal-free Mc with metal salts,

252


taken in a molar ratio of 1:3–5, in DMFA at 100°C. As complex formers, metals were taken which have relatively small covalent radii (Ni, Cu, Co) and are widely used for the synthesis of metal complexes based on macroheterocyclic compounds. 3 R

3

S R

N S N

N N N

R

R=tBu,

N M

M

S N N M

N N N

S

R

N

N

3X

N N

N N

M N

N S

R

N M M

N N

R

N

N N N

N R

N

S 3X N R

N R

51a M=Ni, X=Cl (OH) 51b M=Cu, X=OAc (OH) 51c M=Co, X=OAc (OH)

52a M=Ni, X=OAc (OH) 52b M=Cu, X=OAc (OH) 52c M=Co, X=OAC (OH)

Metal complexes 51a–c and 52a–c are formed with good yields. Their structure is confirmed using mass spectrometry, electron and IR spectroscopy, as well as flame ionization spectroscopy. Mass spectra of compound 52a are observed to have peaks corresponding to a complex with three metal atoms (Fig. 24). The isotopic composition of the signals for the complex obtained completely coincides with that calculated theoretically. Rather unexpected was also the fact that, besides clusters of signals corresponding to the structure of compounds with three metal atoms [3+3+3M]+, the mass spectra of the complexes obtained have intensive peaks corresponding to structures [3+3+3M+O]+, in which one atom of oxygen is present instead of acetate or chloride anions.

a

b

Figure 24 MALDI-TOF. Isotopic distribution of peaks in molecular ions of compound 51a; a, experimental; b, theoretical.

253


Complexation leads to a significant change of the spectral pattern – to the disappearance of a characteristic splitting of the most intensive bands in the region of 400 nm, observed in metal-free precursors, and to the appearance of broadened absorption bands. H2N

OR N N

CN H2N

CN OR Ni(OAs)2*4H2O

S

BuOH

NC

NH2

SAlk

NC SAlk Ni(OAs)2*4H2O

EtOEtOH

RO S RO RO

N N

N Ni

N

N X N Ni X X Ni N N N N N S RO

RS

OR N N N N

S RS

S OR

53 R = C5H11 54 R = C10H21 X = OAc(OH)

RS

N N

N

N

Ni X N Ni X X Ni N N N N N S N

SR N N N N

S SR

55 R = C4H9 56 R = C18H17 57 R = C12H25 X = OAc(OH) Scheme 15

The method of stepwise synthesis failed to produce ABABAB-type macroheterocyclic compounds based on 3,6-dialkyloxyphthalodinitriles. Nickel complexes 53, 54 were obtained by the interaction of 2,5-diamino-1,3,4-thiadiazole with 3,6-dipentoxyphthalodinitrile or 3,6-didecyloxyphthalodinitrile in the presence of nickel acetate in boiling butanol or amyl alcohol for 20 h. This approach also proved efficient for the synthesis of metal complexes of Mc with pyrrole nuclei 55–57 (Scheme 15).

7

Coordination Properties

Macroheterocyclic compounds of ABABAB type – structural analogues of hexaphyrin [93] – are interesting objects for studies by coordination chemistry methods. As per the classification proposed in [94] for monomacrocyclic ligands, compound 58 in accordance with the type of heteroatoms should be assigned to the group of mixed polythiaazamacrocycles with a set of donor atoms SnNm. However, the specific structure, namely the “extracyclic” arrangement of atoms of S and the formation of the internal “cavity” by three pyrrole and six thiadiazole atoms of N, provides macroheterocyclic compound 52 with properties of an N-donor ligand (“polyazamacrocycles with donor atoms Nm”). Lomova et al. [95] were the first to determine the acid-base characteristics and stability of the ABABAB-type macrocycle to the action of acids by the example of macroheterocyclic compound 52 constructed from six successively alternating fragments of thiadiazole and 3,4-bis(4-tert-butylphenyl)pyrrole linked by six aza bridges (McH3). Owing to the presence of six 4-tert-butylphenyl substituents, compound (McH3) 52 dissolves in nonpolar organic solvents, thus forming true solutions with a concentration of (0.1–10)·10 –5 mol·l –1. In transition from neutral organic solvents to individual and mixed protolytic solvents, the EAS of compound 52 are observed to have noticeable changes, ooooooo

254


N

N N

S

N H

N S

N

N N

N S

N

H N

H S

N

N

N N

N

H

H N

N

S

N

N

N

N

H

N

N

N

N

N

N

N

S

58

N

52

which depend on the concentration of acid (Fig. 25). Experimentally, by dilution of solutions with respect to acid, these changes in a benzene–AcOH medium were shown to be reversible. The protonation process of McH3 in a benzene–AcOH proton-donor medium at different temperature and acidities (H0 = 4.44–6.61) was studied by the method of spectrophotometric titration (Fig. 26). It was found experimentally that protonation is a two-stage process (equations (5) and (6)). Respective protonation constants K1 and K2 were found by the method of least squares using Microsoft Excel by the mass action law equation (3) made up for a three-component system with account for the proportionality of concentrations and optical densities (A0, A∞ and Ap), the neutral and protonated species of McH3 and their equilibrium mixtures at the working wavelength of 449 nm. McH3 + AcOH ⇔ (McH3)H+ ·AcO – ,

(5)

(McH3)H+ + AcOH ⇔ (McH3)H22+ ·2AcO – ,

(6)

A

1 2 3

400

500

600

λ, nm

Figure 25 Electronic absorption spectra of McH3 in a benzene–AcOH medium, CAcOH, %: 1, 5; 2, 95; 3, 100. CMcH3 = 1.47·10 –6 mol·l –1.


255

A 0.08

0.06

0.04

0.02 4.2

5.2

6.2

7.2

H0

Figure 26 A curve of spectrophotometric titration of McH3 in benzene by acetic acid at 298 K, the first stage; working wavelength, 449 nm. log I 1.2 tan α = 1 0.6 0 5.5

6

6.5

7

-0.6

H0

-1.2

Figure 27 Dependence of log I on H0 for McH3 in a benzene–acetic acid medium, the first stage (I = (Ap – A0)/(A∞ – Ap), correlation coefficient 0.99, tan α = 1.15.

Ap − A0 K=

A∞ − A0 1 × Ap − A0 ⎛ Ap − A0 0 1− A∞ − A0 ⎜⎜ H 0 − CMcH3 × A − A ∞ 0 ⎝

⎞ ⎟⎟ ⎠

n

.

(7)

Here n is the number of protons attached in one stage. The numerical values of K (Table 6) and n, equal to 1 at both stages, were found by optimizing the dependence in coordinates of equation (7) reduced to a linear form: log(Ap – A0)/(A∞ – Ap) – H0 (Fig. 27). The Table 6 Equilibrium constants K1 and K2 of the McH3 protonation reaction in the benzene–acetic acid system at various temperatures. T, K

K1, l·mol –1

K2, l·mol – 1

298 308 321

0.14 0.23 0.36

1.3

256


-ln K1 3 2 1 0 3.1

3.2

3.3

3.4 1/T 10-3

Figure 28

Dependence –lnK1 on 1/T for McH3 (benzene–acetic acid).

521 nm

Α

1 2 3

λ, nm

Figure 29 Electronic absorption spectra of McH3 in 18.1 M H2SO4: 1, initial; 2, in 48 h; 3, after heating up to 353 K.

thermodynamic characteristics of protonation were determined from the temperature dependence of K (Fig. 28). The thermodynamic parameters of reaction (1) are equal to: ∆H° = 31.6 kJ·mol –1, ∆S° = –90.2 J·mol –1 ·K–1. Analysis of the values of equilibrium constants (Table 6) and thermodynamic parameters shows that compound 52 is a weak base, exceeding amphoteric H2O by only an order and being by 10 orders weaker than NH3 (pK1 at 298 K are, respectively, equal to 0.85, 1.7 and –9.5 [96]). The protonation reaction of McH3 by the first stage in a benzene medium is an endothermic process characterized by a not too high negative value of ∆S°. The latter corresponds to a reaction of formation of charged associated particles more solvated as compared with initial particles (equation (5)). In sulfuric acid with concentrations higher than 17 mol ·l–1 at 298 K, we observe a new protonated species (McH3)Hnn+ (n > 2) with an additional blurred band in the EAS with λmax 521 nm (Fig. 29), which is confirmed by the shape of the EAS in chloroform after the reprecipitation of McH3 from sulfuric acid. At concentrations of the acid smaller than 17 mol ·l–1, there occurs the destruction of McH in the course of its dissolution, which is also confirmed by the analysis of the recrystallization products from the medium of H2SO4. The low stability of (McH3)Hnn+ does not make it possible to study quantitatively the


1/A 120

257

1

90 60 2

30

3 0

40

80

120

160 τ, min

Figure 30 Dependence of 1/A on time (τ) for McH3. C 0H2SO4, mol/l: 17.1 (2); 17.4 (1, 3). T, K: 343 (1); 333 (2); 324.2 (3).

protonation equilibrium and to determine how many protons are attached additionally in changing from AcOH to H2SO4. However, it can be assumed that in a sulfurous solution for (McH3)Hnn+ n = 3 (for a solution in AcOH, n = 2), because up to 100% H2SO4 the EAS with the band of 521 nm does not undergo changes, i.e., no new protonated species are observed. The destruction kinetics of (McH3)Hnn+ was studied quantitatively in H2SO4 with a concentration of 17.1–18.1 mol ·l–1 at temperatures of 324–343 K. The kinetic parameters of the destruction reaction are presented in Table 7. The order of the reaction with respect to the concentrations of McH3 (Fig. 30) and H3O+ were found experimentally (Fig. 31) and Table 7 Efficient rate constants (kef), energy (E) and entropy (∆S #) of activation of the McH3 destruction reaction in concentrated H2SO4. CH2SO4, mol · l – 1 17.1 17.4 18.1

keff ×103, s–1 ·l·mol –1 324.2 K

333 K

343 K

0.50±0.05 0.37±0.02 0.080±0.007

3.7±0.4 3.1±0.3 0.52±0.05

19.0±0.2 13±1 2.4±0.2

E, kJ·mol–1

∆S #, J · mol–1 · K–1

179(a) 175(a) 146(a)

234(a) 224(a) 121(a)

168

157

k×105, s–1 ·l4 ·mol –4 17.1–18.1 0.189 (a) Efficient

1.32

5.74

values of E and ∆S # at a fixed concentration of H2SO4.

are equal to 2 and 3, respectively. Kinetic equation (8) is treated with account for the run (in the reaction system) of equilibrium protonation reaction of (McH3)H22+ and water and the irreversible destruction of the macrocycle (equations (9)–(11)). –dCMcH3 /dτ = k × C 2McH3 × C 3H3O+,

(8)

258


log [H3O+] 0.7

-1

0.9 1 2

-3

3

-5 tan α = 3 log ke -7

Figure 31 324.2 (3).

Dependence of log keff of the destruction of McH3 on log[H3O+]. T, K: 343 (1); 333 (2);

(McH3)H22+ + H3O+ H2O + H2SO4

KK31

KK41

(McH3)H33+ + H2O, H3O+ + HSO4– ,

(McH3)H33+ + H3O+ + H2So4 → thiadiazole and pyrrole derivatives.

(9) (10) (11)

The experimental rate constant k will be expressed as k = k1 × K3 × K4 –1. With account for the results of studies of McH3 acid-base properties, it was suggested that in the transitory state of limiting stage (11) McH3, protonated most probably in two opposite, the most basic, meso-atoms of N and in equilibrium with its triple-protonated species (McH3)H33+, is further attacked by proton-donor particles, which are not only H3O+ but also the uncharged H2SO4 molecule. Interaction with the latter is electrostatically more advantageous for protonated McH3, and its equilibrium concentration is several times higher than that of H3O+. The difficulty of splitting the macrocycle conforming to the aromaticity rule is expressed in a very high activation energy (Table 7), which is two times higher than that for phthalocyanine [97]. Solvation effects are not observed to promote the destruction process, as far as it can be judged from the positive ∆S # (Table 7). The destruction rate of MH3 at standard temperature is several orders lower as compared with aromatic phthalocyanine. Thus, in 17.1 M H2SO4 keff is, respectively, equal to 0.17·10 –5 s –1 ·l·mol –1 (found by extrapolation by the data of Table 7) and 0.122·10 –3 s –1 ·l·mol –1 [97]. The kinetics of the reaction of MCH3 with nickel cations was studied by the spectrophotometric method in a DMFA medium at temperatures higher than 348 K, using the change of the optical density of the solution at the working wavelength of 420 nm. The effective (experimental) rate constants keff (Table 8) of the reaction carried out at an excess of the nickel salt Ni(OAc)2 with respect to McH3 were calculated from the first-order equation (12), in which A0, Aτ, A∞ are the optical densities of the solutions at a working wavelength at times 0, τ and at the end of the reaction. The choice of parameters of the coordination reaction is determined by the possibilities of the spectrophotometric determination of the rate with a satisfactory accuracy (Table 8).


keff =

1

τ

ln

A0 − A∞ . Aτ − A∞

259

(12)

Table 8 Efficient rate constants of the complexation reaction of nickel(II) with macrocyclic ligand 52 in DMFA(a). CNiAc2 ×104, mol · l – 1

T, K

kef ×105, s–1

1.47

348 353 358 353 358 353 358 348 353 358 348 353 358

4.5±0.1 2.2±0.1 0.76±0.08 2.5±0.1 3.7±0.1 3.4±0.2 4.6±0.2 4.1±0.1 4.3±0.2 5.2±0.3 3.7±0.1 7.4±0.4 12.0±0.6

1.58 1.70 1.87

2.06

(a)

Concentration of McH3 did not exceed (2±5) · 10–6 mol · l –1.

The polydentate ligand with an extended coordination cavity and a large amount of electron-donor heteroatoms – potential coordination cores – features a rather complex kinetics of complexation reactions. In the course of conversion, the character of the McH3 absorption spectral curve changes significantly and, according to the data of Fig. 32 and [98], coincides in the end of the reaction with the spectrum of a complex compound of the 3:1 metal–macrocyclic ligand composition. The logarithm log keff linearly correlates with the value of log[Ni(OAc)2] (equation A 0.4

1

0.3 2

0.2 0.1

3

0.0 400

450

500

550

600 λ, nm

Figure 32 Electronic absorption spectra of McH3 in DMFA (1), of the complexation product of McH3 with Ni(OAc)2 in DMFA (2) and, after isolation from the reaction mixture, in CHCl3 (3).

260


log ke

-4.0

2 1 -4.5

-3.81

3.78

3.75

log CNi(OAc)2

Figure 33 Dependence of log keff on logCNi(OAc)2 for the reaction of macrocyclic compound McH3 with Ni(OAc)2 in DMFA. T, K: 353 (1), 358 (2) (ρ equals 0.93 and 0.99, respectively).

(13), Fig. 33) only within a narrow range of concentration of the salt (1.58–1.87)·10 –4 mol·l –1, where the order close to two is observed with respect to [Ni(OAc)2]. Respectively, only for this interval of reaction mixture concentrations with respect to the nickel salt we can write down kinetic equation (14). At Csalt close to 1.47·10 –4 and 2.06·10 –4 mol·l –1 the order is, respectively much lower and higher than two. logkeff = logkeff + nlog[Ni(OAc)2],

(13)

–dCMcH3 /dτ = k·CMcH3·Csalt 2.

(14)

Experimental kinetic equation (14) is treated with account for the reversible and irreversible reactions (equations (15)–(18)), where acetate ion is dropped in writing the formula of the salt): (McH3)2 McH3 + Ni2+

KK11

kK2 1

[NiMcH2]+ + Ni2+ [Ni2McH]2+ + Ni2+

2 McH3,

[NiMcH2]+…H+, slow, KK31

kK41

(15) (16)

[Ni2McH]2+…H+,

(17)

[Ni3Mc]3+ + H+, slow.

(18)

The rate of the limiting stage (14) is equal to k4 ·С[Ni2McH]2+…H+ ·СNi2+ or, after expressing the concentration [Ni2McH]2+…H+ with account of equilibrium (17): k4 ·K3 · С[NiMcH2]+…H+ ·(СNi2+)2. The latter coincides with the right-hand side of experimental equation (18) owing to the equilibrium (19) and a fast irreversible transition of McH3 into [NiMcH2]+…H+ (equation (16)). Thus, the experimental rate constant is k = k4 ·K3.


261

The value of K1 is first determined experimentally in this work and calculated by equation (19): K1 = (1.6±0.18)·105 mol –1 ·l. The equilibrium with constant K1 (equation (15)) proves kinetically significant, because it provides for the presence of McH3 molecules in DMFA solutions during the complexation in predominant concentrations as compared with the dimer (McH3)2. Ap − A0 K=

A∞ − A0 1 Ap − A0 Ap − A0 0 1− Cs − CMcH 3 A − A A∞ − A0 0 ∞

(19)

(in the equation, C0McH3 and Cs are the initial concentrations of, respectively, McH3 and DMFA; A0, Ap, A∞ are the optical densities at a working wavelength of 420 nm for solutions of (McH3)2, equilibrium mixture at a definite concentration of DMFA and McH3). It is characteristic that the second order with respect to the salt is disturbed at the broadening of the salt concentration interval. The assumed scheme of stepwise reactions explains this regularity. When the consentation of the salt is increased, stages (16) and (17) merge into one equilibrium stage, which in the expression of the concentration [Ni2McH]2+…H+ leads to the third order with respect to the salt. In contrast, at low CNi2+ the coordination reaction does not proceed to the end, and stage (16) becomes slow and limits the entire process. Respectively, the order with respect to the salt becomes equal to unity. The high sensitivity of the reaction to the concentration of the salt enables an assumption on the detailed mechanism of limiting stage (18) as a dissociative one according to the Langford–Gray classification [98]. The temperature dependence of the rate constants is nonlinear, which does not make it possible to obtain the value of activation energy and, respectively, entropy for the reaction. If the narrow temperature range (353–358 K) is used, one can obtain the activation energy, regularly decreasing with the salt concentration rise from 82 down to 25 kJ·mol –1. The value of ∆S # in this case changes from –109 down to –266 J·(mol·K) –1. Large negative values of activation entropy (–109 to –266 J·(mol·K) –1) indirectly confirm the statement of the dissociative mechanism of reaction (14), thus indicating a more ordered transitory state of the reacting system as compared with the initial state. Indeed, the polar transitory state [[Ni2McH]2+ ·[Ni(OAc)2(DMFA)3]] #, as the state formed from the pentacoordinated nickel complex (the dissociative mechanism of reaction (14)) and from the macrocyclic ligand polarized by the N–H bond, is, probably, more solvated in the medium of the main solvent, DMFA. Thus, most probably, the nickel coordination reaction of the macrocyclic ligand with an extended coordination cavity McH3 proceeds as follows: the macrocycle is readily coordinated with one nickel ion, the second ion is coordinated with greater difficulty, and the reaction does not proceed to the end (equilibrium (17) is established), the third nickel ion is coordinated with even greater difficulty, also, apparently, reversibly, however, due to the low rate of the direct reaction, no equilibrium is achieved under the conditions of the experiment. The stepwise entry of metal cation into the coordinating space of McH3 opens prospects for finding conditions of synthesizing heteronuclear complexes with ligands of similar structure and for their practical applications.

262

8


Studies of the Practically Valuable Properties of Macroheterocyclic Compounds and their Metal Complexes

By the time macroheterocyclic compounds were discovered [72, 73], phthalocyanines had already been sufficiently well studied and found wide use as pigments and dyes [5]. For this reason, the first studies mainly concerned the coloristic properties of Mc [99, 100]. It is not hard to note that synthesized Mc were primarily tested for the same practically valuable properties revealed in phthalocyanines. This is due both to similarities in their structure and to the fact that macroheterocyclic compounds were, as a rule, studied by the same investigators who studied phthalocyanines. And, though in most cases the presence of such properties was confirmed, Mc, as a rule, proved less efficient than phthalocyanines [101–106]. A true qualitative breakthrough in the approach to the assessment of the properties of Mc was the discovery of their ability to catch and deactivate free radicals. A groups of Russian scientists [46] has found that metal-containing macrocycles possess a thermo- and light-stabilizing activity, whereas respective phthalocyanines catalyze the destruction of polycapramide. At present, the copper complex of the triisoindolebenzene macrocycle is manufactured on a industrial scale under the trade mark of Stabilin-9 [47]. Apparently, the same quality is the basis of stabilization of siloxane rubbers (SKT) [107], as well as of flame retardation of polyorganosiloxane polymers [108] by metal complex Mc. The development of new approaches to the synthesis enabled a structural modification of Mc, which led to the discovery of new practically valuable properties, in particular, liquid crystalline [60], nonlinear-optical [109], as well as the ability to form ordered Langmuir–Blodgett layers [55]. Analysis performed has shown that macroheterocyclic compounds and their complexes with metals are substances possessing a whole range of practically valuable properties, the study of which is far from completion. 8.1

Biological properties

One of the factors that impede studies of their properties is the low solubility of unsubstituted Mc in water and in organic solvents. As we reported above, introduction of bulky substituents led to a significant increase of solubility and, because of this, opened broad prospects for studies of the properties, biological properties including, of this class of compounds. At the first stage, we studied the antimicrobial activity of a number of tert-butyl-substituted Mc and their metal complexes. The antimicrobial activity was assessed by the germination of test cultures (Escherichia coli str. 676,0 and Staphylococcus aureus Losmanov) and the diameter of the lysis zone. tert-Butyl-substituted Mc and their complexes with copper, which include 1,2,4-triazole fragments, proved to possess moderate antimicrobial properties. A qualitatively new step in studies of the biological properties of Mc was a systemic approach, which consisted in predicting the range of biological activity with the view to reveal promising structures and perform laboratory tests of chosen substances. Prediction of the kinds of activity for a number of macroheterocyclic compounds, their metal complexes, as well as initial substances for their synthesis was performed using the PASS program [110]. A fragment of the calculated spectrum of biological activity is presented in Table 9.

263


Table 9 Prediction of the range of biological activity of Mc, their metal complexes and initial substances for their synthesis. Compound

Coef.

Type of activity Anti- Nucleo- Carcino- Antitumour phil.reag. genic bacter. exchange

N NH H2N

NH2

N

Antiviral

Kconf

14.8

20.8

–

16.9

21.3

Keff

2.0

6.1

–

1.8

5.4

Kconf

30.2

–

1.3

20.7

11.8

Keff

4.2

–

3.2

2.2

3.0

Kconf

15.7

18.9

–

–

17.6

Keff

4.0

2.4

–

–

5.2

Kconf

12.1

–

–

25.8

12.3

Keff

1.7

–

–

9.2

3.1

Kconf

14.8

28.9

–

15.4

6.6

Keff

2.0

8.5

–

5.5

43.7

Kconf

14.8

28.9

–

15.4

2.8

Keff

2.0

8.5

–

5.5

3.3

I(a) N N

H2N

NH2

S II N NH

N

N

N

NH

HN

NH

III

HN

N N N

N

S

HN

NH

HN

NH

IV

N NH N

N N

N

Cu

N

N N

N

V NH

N

N NH N

N N

N

Co

N

N

N N

N

NH

VI

264


Table 9 Prediction (continued)of the range of biological activity of Mc, their metal complexes and initial substances for their synthesis. Compound

Coef.

Type of activity Anti- Nucleo- Carcino- Antitumour phil.reag. genic bacter. exchange

Antiviral

N NH N

N

N

Ni

N N

Kconf

14.8

28.9

–

15.4

6.6

Keff

2.0

8.5

–

5.5

43.7

Kconf

19.7

21.1

–

25.3

11.8

Keff

1.3

6.2

–

9

4

N N

N N

NH

N

N

VII

N N N

S Zn N

N N N

VIII Note: For reading convenience, in this section we use different numeration of compounds.

As it follows from the data presented in Table 9, for all tested compounds one should expect the manifestation of an antitumour activity, which should decrease in the following sequence: II > III > I = V = VI = VII > IV > VIII. In this connection, chemical compounds presented in Table 10 were tested for the antitumour activity on a model of transplantable tumour L-1210. Table 10 Indices of antitumour activity of compounds at repeated intraperitoneal injection of DBA with L-1210 to female mice. Compound No

Injection protocol, days

Daily dose, mg/kg

Total dose, mg

ALE, days

ALEI, %

1

3

2

4

5

6

Control I Control I

2–3 2–3 2–6 2–6

– 150 – 50

– 300 – 250

5.3±0.2 5.4±0.4 10.0±0.4 9.8±0.3

– 1.8 – –2.5

265


(continued) Table 10 Indices of antitumour activity of compounds at repeated intraperitoneal injection of DBA with L-1210 to female mice. Compound No


Daily dose, mg/kg

Total dose, mg

ALE, days

ALEI, %

1

3

2

4

5

6

Control II Control II Control III Control IV Control IV Control V Control VI Control VI Control VII Control VIII Control IX(a) Control IX

2–3 2–3 2–6 2–6 2–6 2–6 2–6 2–6 2–3 2–3 2–6 2–6 2–3 2–3 2–6 2–6 2–3 2–3 2–3 2–6 2–6 2–6 2–3 2–3

– 150 – 50 – 250 – 250 – 40 – 125 – 250 – 50 – 250 – 250 – 250 – 100

– 300 – 250 – 1250 – 1250 – 80 – 625 – 500 – 250 – 500 – 1250 – 1250 – 200

5.3±0.2 6.0±0.1 10.0±0.4 10.5±0.17 8.3±0.5 8.0±0.2 8.3±0.5 4.0±0.1.1 5.3±0.2 4.8±0.1 8.3±0.5 7.0±0.6 5.3±0.2 3.6±0.2 10.0±0.4 9.6±0.4 5.3±0.2 5.8±0.2 5.3±0.2 6.6±0.6 8.3±0.5 5.4±0.1.3 5.3±0.2 4.8±0.9

– 13.2 – 5.0 – –3.6 – –51.8 – –10.3 – –15.6 – –32.1 – –4.0 – 9.4 – –20.5 – –34.9 – –9.4

(a) IX, macroheterogenic compound of symmetrical structure with fragments of 1,2,4-triazole. ALE, average life expectancy; ALEI, average life expectancy increase.

Preliminarily, maximum tolerated doses (MTD) were determined for each compound, based on which daily therapeutic doses were chosen. The values of MTD made it possible to assign all tested compounds to toxicity class IV (low-toxic compounds) [111]. Experimental studies of the antitumour activity were carried out under conditions of intraperitoneal introduction of preparations. The results of this part of work are presented in Table 10. As it follows from the table, compounds 3,5-diamino-1,2,4-triazole (II) and the nickel complex of symmetric-structure macroheterocyclic compound with fragments of 3,5-diamino-1,2,4-triazole (VII) exhibit a moderate antitumour activity with respect to lymphoid leukemia L-1210. In the sequence of triazole-containing compounds studied, two of them (3,5-diamino1,2,4-triazole and the nickel complex of Mc with fragments of 1,2,4-triazole) possess a moderate antitumour activity. By the value of the increase of the average life expectancy,

266


these compounds can be arranged in the following sequence: I > VII > III > V > VI > IX. Thus, the antitumour activity of triazole derivatives studied depends both on the structure of the organic moiety of the molecule and on the nature of metal complex former. Introduction of metal into the Mc molecule significantly increases the value of ALEI, thus weakening the negative effect of descriptors present in the Mc molecule. The nickel complex of Mc VII proved the most efficient in the sequence of all macrocycles and their complexes studied, while the cobalt and copper complexes exhibited no activity with respect to leukemia L-1210. In the sequence of 1,3,4-thiadiazole derivatives, no manifestation of antitumour activity was found for any of the compounds studied, and the TZP with a fragment of 1,3,4-thiadiazole IV was even capable of significantly initiating the development of the disease (Table 10). 2,5-Diamino-1,3,4-thiadiazole I itself exhibited a weak activity with respect to leukemia L-1210, which is consistent with the literature data [112]. The results obtained served as the basis for continuation of works on the assessment of the antitumour and antimetastatic activities of compounds II and VII on a solid LLC (Lewis-lung carcinoma) tumour resistant to all currently known antitumour preparations. The results of studies are presented in Table 11. Table 11 Indices of antitumour and antimetastatic activity of compounds in female mice C57BL/6 in repeated injection. Group

Control II VII

Daily dose, mg/kg – 50 50

Tumour weight, g

Number of metastases


Mtm

% inhibition

Mtm

MII, %

– 6–14 6–14

9.16±0.57 9.80±0.43 9.91±0.58

0.0 –7.0 –8.2

5.0±0.2.2 9.3±0.4.9 5.8±0.3.6

– –59.4 –16.0

As indicated by the data presented, the compounds studied exhibited no specific activity with respect to LLC tumour. Thus, of the compounds we studied, only 3,5-diamino-1,2,4-triazole and the nickel complex of Mc with fragments of 3,5-diamino-1,2,4-triazole exhibited a moderate antitumour activity on an L-1210 lymphoid leukemia model, which gives grounds for directed synthesis and search for more active compounds among their analogues (1,2,4-triazole derivatives) for subsequent works on the development of antitumour means. In conclusion, it should be noted that accumulated material is too little to judge on concrete structure–property dependences. At the same time, the already available results enable a statement to be made that Mc are promising compounds for the search of substances with interesting biological properties. The work was supported by the grant 05-03-33003a from the Russian Foundation for Basic Research.

References 1. R.K. Pandey and G. Zheng, Porphyrins as Photosensitizers in Photodynamic Therapy, in: Porphyrin Handbook, vol. 6, ed. by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press: New York, Chapter 43, pp. 157–230 (2000).


267

2. T. Aida and S. Inoue, Metalloporphyrins as Catalysts for Precision Macromolecular Synthesis, in: Porphyrin Handbook, vol. 6, ed. by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press: New York, Chapter 42, pp. 133–156 (2000). 3. T. Malinski, Porphyrin-Based Electrochemical Sensors, in: Porphyrin Handbook, vol. 6, ed. by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press: New York, Chapter. Vol. 44, 231–256 (2000). 4. J.-H. Chou, H.S. Nalwa, M.E. Kosal, N.A. Rakow and K.N. Suslick, Applications of Porphyrins and Metalloporphyrins to Materials Chemistry, in: Porphyrin Handbook, vol. 6, ed. by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press: New York, Chapter 41, pp. 43–132 (2000). 5. P. Erk and H. Hengelsberg, Phthalocyanine Dyes and Pigments, in: Porphyrin Handbook, vol. 19, ed. by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press: New York, Chapter 119, 105–150 (2003). 6. G. Booth, in: The Chemistry of Synthetic Dyes, ed. by K. Venkataraman, Academic Press: New York, pp. 241–282 (1971). 7. G.P. Shaposhnikov, V.E. Maizlish, V.P. Kulinich, Yu.G. Vorobyev and M.K. Islyaikin, Izv. Vuzov, Khim. Khim. Tekhnol., 48 (7), 22–31 (2005) (in Russian). 8. S.V. Lindeman, V.E. Shklover, Yu.T. Struchkov, I.I. Ponomarev, S.A. Siling, S.V. Vinogradova and V.V. Korshak, Izv. Akad. Nauk SSSR, Ser. Khim., 2015–2023 (1984) (in Russian). 9. G. Pellizzari and C. Roncagliolo, Gaz. Chem. Ital., 31 (1), 477–513 (1901). 10. G. Pellizzari, Ber., 649–651 (1891). 11. F. Fernandez-Lazaro, J. de Mendoza, O. Mo, S. Rodríguez-Morgade, T. Torres, M. Yanez and J. Elguero, J. Chem. Soc. Perkin Trans., 2 (7), 797–803 (1989). 12. E. Fromm, Justus Liebigs’ Ann. Chem., 433, 1–9 (1923). 13. W.M. Sharman and J.E. Van Lier, Synthesis of Phthalocyanine Precursors, in: Porphyrin Handbook, ed. by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press: Amsterdam– Boston–London–New York–Oxford–Paris–San Diego–San Francisco–Singapore–Sydney–Tokyo, vol. 15, pp. 1–60 (2003). 14. S.A. Mikhalenko and E.A. Lukyanets, Zhurn. Org. Khim., 6, 171–174 (1970) (in Russian). 15. I.G. Abramov, M.V. Dorogov, S.A. Ivanovskii, A.V. Smirnov and M.V. Abramova, Mendeleev Commun., 2, 78–80 (2000). 16. G.P. Shaposhnikov and V.E. Maizlish, Izv.Vuzov, Khim. Khim. Tekhnol., 47 (5), 26–35 (2005) (in Russian). 17. M.K. Islyaikin, E.A. Danilova, L.D. Yagodarova, M.S. Rodríguez-Morgade and T. Torres, Org. Lett., 3 (14), 2153–2156 (2001). 18. M.K. Islyaikin, M.S. Rodríguez-Morgade and T. Torres, Eur. J. Org. Chem., 15, 2460–2464 (2002). 19. T.F. Bauman, A.G.M. Barrett and B.M. Hoffman, Inorg. Chem., 36 (24), 5662–5665 (1997). 20. A. Davison, R.H. Holm, R.E. Benson and W. Mahler, Inorg. Synthesis, 10, 8–26 (1967). 21. F. Bauman, B. Binert, G. Rosch, H. Vollmann and W. Wolf, Angew. Chem., 68 (4), 133–150 (1956). 22. V.F. Borodkin, Zhurn. Prikl. Khim., 31, 813–816 (1958) (in Russian). 23. S.A. Siling and S.V. Vinogradova, Usp. Khim., 63 (9), 810–824 (1994) (in Russian). 24. M.K. Islyaikin, A.V. Lyubimtsev, R.P. Smirnov and A. Baranski, Izv. Vuzov, Khim. Khim. Tekhnol., 38 (4–5), 81–87 (1995) (in Russian). 25. A.V. Lyubimtsev and A. Baranski, Zhurn. Org. Khim., 34 (10), 1535–1541 (1998) (in Russian). 26. A. Baranski and A. Lyubimtsev, Zhurn. Org. Khim., 34 (10), 1542–1546 (1998) (in Russian). 27. A.V. Lyubimtsev, A. Baranski, M.K. Islyaikin, R.P. Smirnov, Khim. Geterotsykl. Soed., 8, 1074–1079 (1997) (in Russian). 28. V.I. Minkin, B.Ya. Simkin and R.M. Minyaev, The Quantum Chemistry of Organic Compounds: Reaction Mechanisms, Khimiya: Moscow, 248 pp. (1986) (in Russian). 29. O.V. Shishkin, E.V. Kudrik, M.K. Islyaikin, A.V. Lyubimtsev and S.A. Siling, Izv. Akad. Nauk, Ser. Khim., 2, 334–338 (1998) (in Russian). 30. Yu.V. Zefirov and P.M. Zorky, Usp. Khim., 64 (5), 446–460 (1995) (in Russian). 31. J.A. Elvidge and J.H. Golden, J. Chem. Soc., 700–709 (1957).

268


32. M.E. Baguley and J.A. Elvidge, J. Chem. Soc., 709–719 (1957). 33. F. Fernández-Lazáro, T. Torres, B. Hauschel and M. Hanack, Chem. Rev., 98 (2), 563–576 (1998). 34. M.K. Islyaikin and A. Baranski, Khim. Geterotsykl. Soed., 8, 1047–1055 (2001). 35. P. Bamfield and P.A. Mack, J. Chem. Soc. C, 1961–1964 (1968). 36. V.A. Gnedina, R.P. Smirnov and G.A. Matyushin, Izv. Vuzov, Khim. Khim. Tekhnol., 16 (4), 589–592 (1973) (in Russian). 37. E.V. Kudrik, M.K. Islyaikin and R.P. Smirnov, Zhurn. Org. Khim., 33 (7), 1107–1110 (1997) (in Russian). 38. L. Latos-Grazynski, Core-Modified Heteroanalogues of Porphyrins and Metalloporphyrins, in: Porphyrin Handbook, vol. 2, ed. by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press: New York, Chapter 14, pp. 361–416 (2003). 39. M.S. Rodríguez-Morgade, Gema de la Torre and T. Torres, Design and Synthesis of LowSymmetry Phthalocyanines and Related Systems, in: Porphyrin Handbook, vol. 15, ed. by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press: New York, Chapter 99, pp. 125–160 (2003). 40. N. Kobayashi, Synthesis and Spectroscopic Properties of Phthalocyanine Analogs, in: Porphyrin Handbook, vol. 15, ed. by K.M. Kadish, K.M. Smith and R. Guilard, Academic Press: New York, Chapter 100, pp. 161–262 (2003). 41. R.P. Smirnov, V.F. Borodkin and G.I. Lukyanova, Izv. Vuzov, Khim. Khim. Tekhnol., 1, 118–121 (1964) (in Russian). 42. L.M. Fedorov, A.I. Mikhailov, R.P. Smirnov, N.A. Kolesnikov and M.I. Alyanov, Izv. Vuzov, Khim. Khim. Tekhnol., 15 (12), 1817–1820 (1972) (in Russian). 43. L.M. Fedorov, R.P. Smirnov and G.P. Shaposhnikov, Chemistry and Technology of Dyeing, Synthesis of Dyes and Polymer Materials, Interinstitute Collected Volume, No 1, pp. 57–61 (1973) (in Russian). 44. R.P. Smirnov, B.D. Berezin and L.A. Skvortsova, Zhurn. Fiz. Khim., 41 (7), 1630–1634 (1967) (in Russian). 45. R.P. Smirnov and B.D. Berezin, Zhurn. Fiz. Khim., 43 (10), 2494–2498 (1969) (in Russian). 46. R.P. Smirnov, V.N. Kharitonov and L.N. Smirnov, Proc. Ivanovo Institute of Chemistry and Technology, 4, 111–116 (1972) (in Russian). 47. Chemical Additives to Polymers (reference book), A.S. Baranova, G.S. Baryshnikova, N.S. Glazunova, V.M. Delyustro, K.A. Zolotareva, I.P. Maslova, L.A. Pugacheva and L.A. Skripko, Khimiya: Moscow, 264 pp. (1981) (in Russian). 48. N.A. Kolesnikov, L.M. Fedorov and E.E. Kolesnikova, Izv. Vuzov, Khim. Khim. Tekhnol., 19 (9), 1352–1354 (1976) (in Russian). 49. V.F. Borodkin, T.I. Chesnokova and M.K. Islyaikin, Izv. Vuzov, Khim. Khim. Tekhnol., 23 (8), 1044–1046 (1980) (in Russian). 50. T.M. Ivanova, M.Yu. Bazanov, A.V. Petrov and E.S. Yurina, Koord. Khim., 32 (1), 75–78 (2006) (in Russian). 51. R.K. Bartlett, L.V. Renny and K.K. Chan, J. Chem. Soc. C, 1, 129–133 (1969). 52. F. Fernández-Lázaro, A. Sastre and T. Torres, J. Chem. Soc. Chem. Com., 1525–1526 (1994). 53. M. Nicolau, B. Cabezón and T. Torres, Coord. Chem. Rev., 190–192, 231–243 (1999). 54. S. Esperanza, M. Nicolau and T. Torres, J. Org. Chem., 67, 1392–1395 (2002). 55. F. Armand, M.V. Martínez-Díaz, B. Cabezón, P.-A. Albouy, A. Ruaudel-Teixier and T. Torres, J. Chem. Soc. Chem. Commun., 1673–1674 (1995). 56. F. Armand, B. Cabezón, M.V. Martínez-Díaz, A. Ruaudel-Teixier and T. Torres, J. Mater. Chem., 7 (9), 1741–1746 (1997). 57. F. Armand, B. Cabezón, O. Araspin, A. Barraud and T. Torres, Synth. Metals, 84, 879–880 (1997). 58. B. Cabezón, F. Fernández-Lázaro, M.V. Martínez-Díaz, S. Rodríguez-Morgade, A. Sastre and T. Torres, Synth. Metals, 71, 2289–2290 (1995). 59. G. Rojo, F. Agulló-López, B. Cabezón, T. Torres, S. Brasselet, I. Ledoux and J. Zyss, J. Phys. Chem. B, 104, 4295–4299 (2000). 60. B. Cabezón, M. Nicolau, J. Barberá and T. Torres, Chem. Mater., 12, 776–781 (2000). 61. V. Stefani, B. Cabezón, E.L.G. Denardin, D. Samios and T. Torres, J. Mater. Chem., 10, 2187–2192 (2000).


269

62. M.K. Islyaikin, V.R. Ferro and J.M. García de la Vega, J. Chem. Soc. Perkin Trans., 2 (12), 2104–2109 (2002). 63. T.M. Krygovski and M.K. Cyranski, Chem. Rev., 101, 1385–1419 (2001). 64. P.v.R. Scheyer, C. Maerker, A. Dransfeld, H. Jiao and N.J.R.v.E.J. Hommes, Am. Chem. Soc., 118 (26), 6317–6318 (1996). 65. M.K. Cyranski, T.M. Krygovski, M. Wiciorowski, N.J.R.v.E. Hommes and P.v.R. Shleyer, Angew. Chem. Int. Ed., 37 (1/2), 177–180 (1998). 66. L.E. Marinina, S.A. Mikhalenko and E.A. Lukyanets, Zhurn. Org. Khim., 43 (9), 2025–2029 (1973) (in Russian). 67. P.A. Stuzhin, O.G. Khelevina and B.D. Berezin, Phthalocyanines: Properties and Applications, ed. by C.C. Leznoff and A.B.P. Lever, VCH: New York, vol. 4, pp. 19–77 (1996). 68. P.A. Stuzhin and O.G. Khelevina, Koord. Khim., 24 (10), 783–793 (1998) (in Russian). 69. P.A. Stuzhin, A. Ul-Hak, N.V. Chizhova, A.S. Semeykin and O.G. Khelevina, Zhurn. Fiz. Khim., 72 (9), 1585 (1998) (in Russian). 70. O.G. Khelevina, V.R. Ferro, M.K. Islyaikin, E.A. Veselkova, M.G. Stryapan, J.M. García de la Vega, J. Phys. Org. Chem., 18 (4), 329–335 (2005). 71. B.D. Berezin and O.G. Khelevina, Porphyrins: Structure, Properties, Synthesis, ed. by N.S. Enikolopyan, Nauka: Moscow, p. 83 (1985) (in Russian). 72. J.A. Evidge and R.P. Linstead, J. Chem. Soc., 5008–5012 (1952). 73. J.B. Campbel [E.I. du Pont de Nemours and Co.], US Patent 2765308, Applied for 15.08.1952, Issued 02.10.1956. 74. J.C. Speacman, Acta Cryst., 6 (10), 784–791 (1953). 75. H.-J. Hecht and P. Luger, Acta Crystal., 30 (12), 2843–2848 (1974). 76. W. Hiller, J. Sträle, K. Mitulla and M. Hanack, Liebigs Ann. Chem., 1946–1951 (1980). 77. S.-M. Peng, Y. Wang, T.-F. Ho, I.-C. Chang, C.-P. Tang and C.-J. Wang, J. Chem. Soc. (Taipei), 33, 13–21 (1986). 78. S.-M. Peng, Y. Wang, C.-K. Chen, J.-Y. Lee and D.-S. Liaw, J. Chem. Soc. (Taipei), 33, 23–33 (1986). 79. S.P. Konovalov and M.K. Islyaikin, Izv. Vuzov, Khim. Khim. Tekhnol., 36 (4), 76–81 (1993) (in Russian). 80. V.F. Borodkin, V.A. Burmistrov and M.K. Islyaikin, Khim. Geterotsykl. Soed., 1, 62–64 (1981) (in Russian). 81. O.V. Shishkin, A.Yu. Kovalevsky, M.V. Shcherbakov, M.K. Islyaikin, E.V. Kudrik and A. Baran- ski, Kristallografiya, 46 (3), 461–464 (2001) (in Russian). 82. V.V. Aleksandriiskii, M.K. Islyaikin, I.G. Shutov, G.A. Zhurko and V.A. Burmistrov, Russ. J. Phys. Chem., 79 (1), S130–S134 (2005). 83. E.A. Danilova and M.K. Islyaikin, Synthesis and Properties of tert-Butyl-substituted Macroheterocyclic Compounds and their Complexes with Metals, in: Advances of Porphyrin Chemistry, Chemistry Research Institute, St.-Petersburg University: St.-Petersburg, vol. 4, pp. 356–375 (2004) (in Russian). 84. J.N. Esposito, L.E. Sutton and M.E. Kenney, Inorg. Chem., 6 (6), 1116–1120 (1967). 85. G. de la Torre and T. Torres, J. Org.Chem., 61, 6446–6449 (1996). 86. E.A. Danilova, M.K. Islyaikin and R.P. Smirnov, Zhurn. Org. Khim., 65 (11), 1882–1884 (1995) (in Russian). 87. V.F. Borodkin and N.A. Kolesnikov, Khim. Geterotsykl. Soed., 2, 194–195 (1971) (in Russian). 88. N.A. Kolesnikov and V.F. Borodkin, Izv. Vuzov, Khim. Khim. Tekhnol., 15 (6), 880–882 (1972) (in Russian). 89. N. Kobayashi, S. Inagaki, V.N. Nemykin and T. Nonomura, Angew. Chem. Int. Ed., 40 (14), 2710–2717 (2001). 90. M.K. Islyaikin, E.A. Danilova and L.D. Yagodarova, Izv. Vuzov, Khim. Khim. Tekhnol., 46 (2), 3–7 (2003) (in Russian). 91. G.L. Starova, O.V. Frank-Kamenetskaya, V.V. Makarsky and V.A. Lopyrev, Kristallografiya, 25 (6), 1292–1294 (1980) (in Russian). 92. Hitoshi Senda and Juro Maruha, Acta Cryst., 43, 347–349 (1987). 93. J.L. Sessler and D. Seidel, Angew. Chem. Int. Ed., 42, 5134–5175 (2003). 94. K.B. Yatsimirsky and Ya. D. Lampeka, The Physical Chemistry of Complexes of Metals with Macrocyclic Ligands, Naukova Dumka: Kiev, 256 pp. (1985) (in Russian).

270


95. T.N. Lomova, E.E. Suslova, E.A. Danilova and M.K. Islyaikin, Zhurn. Fiz. Khim., 79 (2), 263–269 (2005) (in Russian); T.N. Lomova, E.E. Suslova, E.A. Danilova and M.K. Islyaikin, Russ. J. Phys. Chem., 79 (2), 201–206 (2005). 96. R. Bell, The Proton in Chemistry, Mir: Moscow, 381 pp. (1977) (Russian translation). 97. B.D. Berezin, A Study of the Physico-chemical Properties of Phthalocyanine Complex Compounds, DSc (Chemistry) Thesis, Inst. of General and Inorganic Chemistry: Kiev, 56 pp. (1966) (in Russian). 98. C. Langford and H. Gray, Ligand Substitution Processes, Mir: Moscow, 157 pp. (1984) (Russian translation). 99. B.I. Stepanov, Introduction to the Chemistry and Technology of Organic Dyes, 3rd edn, Khimiya: Moscow, p. 541 (1984) (in Russian). 100. L.G. Krolik and B.D. Vitkina, Zhurn. VKhO im. D.I. Mendeleeva, 11 (1), 60–69 (1966) (in Russian). 101. L.M. Fedorov, N.A. Kolesnikov, R.P. Smirnov and M.I. Alyanov, Izv. Vuzov, Khim. Khim. Tekhnol., 15 (4), 537–540 (1972) (in Russian). 102. R.P. Smirnov, V.V. Andreyanov, Yu.G. Vorobyev, V.A. Shorin and L.M. Fedorov, Izv. Vuzov, Khim. Khim. Tekhnol., 27 (10), 1239–1241 (1984) (in Russian). 103. C.G. Birch and R.T. Iwamoto, Inorg. Chem., 12 (1), 66–73 (1973). 104. C.G. Birch and R.T. Iwamoto, Inorg. Chem. Acta, 6 (4), 680–682 (1972). 105. M.I. Bazanov, L.V. Kokhova, V.A. Bogdanovskaya, M.R. Tarasevich, R.P. Smirnov and N.A. Kolesnikov, Izv. Vuzov, Khim. Khim. Tekhnol., 29 (7), 43–46 (1986) (in Russian). 106. M.I. Bazanov, V.V. Kudrinsky, N.A. Kolesnikov, R.P. Smirnov,. Izv. Vuzov, Khim. Khim. Tekhnol., 36 (10), 59–64 (1993) (in Russian). 107. R.P. Smirnov, A.L. Smirnov, D.G. Snegirev and G.A. Zdorikova, Izv. Vuzov, Khim. Khim. Tekhnol., 35 (6), 66–71 (1992) (in Russian). 108. L.N. Smirnov, G.A. Zdorikova, A.L. Smirnov and D.G. Snegirev, Izv. Vuzov, Khim. Khim. Tekhnol., 35 (6), 71–76 (1992) (in Russian). 109. G. Rojo, F. Agulló-López, B. Cabezón, T. Torres, S. Brasselet, I. Ledoux and J. Zyss, J. Phys. Chem. B, 104, 4295–4299 (2000). 110. Yu.V. Burov, L.V. Korolchenko and V.V. Poroykov, Byul. All-Union Scientific Centre on Safety of Biologically Active Compounds, Moscow, pp.4–25 (1990) (in Russian). 111. N.F. Izomerov, I.V. Sakotsky and N.N. Sidorov, Toxicity Parameters of Industrial Poisons in Single Exposure, Meditsina: Moscow, 239 pp. (1977) (in Russian). 112. D.L. Hill, Cancer Rev., 4, 215–220 (1980).

7

The Photochemical Aspect of Reactions of Flavonols with Molecular Oxygen E.A. Venedictov Institute of Solution Chemistry, Russian Academy of Sciences, 1 Akademicheskaya Street, Ivanovo, 153045, Russia email: [email protected]

This chapter presents the results of research into the kinetics of photochemical reactions of biological pigments and related compounds with molecular oxygen.

Introduction Flavonoids belong to the most widespread natural compounds [1–10]. Their sources are mainly higher plants. They also occur in algae, lichens, mosses, fungi, ferns and some representatives of insects and microorganisms. They are found in roots, wood, leaves, flowers, fruits. These compounds are present in various cell organelles of plants. They are revealed in a broad colour palette of the flora and are represented in a whole range of products of plant raw material processing (juices, wine, oil and extracts). Owing to the ability of flavonoids to pigment plant tissues, protection of plants from the damaging action of solar UV radiation was historically considered to be their most significant function [6–12]. Subsequently, this point of view found its experimental proof. The rise of flavonoid biosynthesis in response to an increase of the level of UV radiation, and the localization of these compounds mainly in the upper cell layer of plant tissues are considered at present as an adaptation mechanism of plants to their new habitat conditions [13–15]. In reality, the physiological functions of flavonoids are more diverse. They are believed to be involved in electron transfer processes, stimulate plant growth and protection from harmful biological effects occurring with participation of pathogenic organisms and insects [1–10, 16].

272

E.A. Venedictov

Making up sometimes up to 5% plant biomass, flavonoids are an essential part of our diet. For this reason, much attention is given to the studies of their pharmacological properties [1–10, 17]. In particular, these compounds have been known since long time ago to have antiischemic, antiinflammatory, fungicidal and anticarcinogenic activities. At present, they are considered as a broadly accessible and potent means capable of not only preventing but also curing many kinds of illnesses. There is a strongly pronounced antioxidant activity of their action behind all the diversity of biological and pharmacological properties of flavonoids [17–20]. It has been no secret since long time ago that active species of oxygen and its metabolites are responsible for virtually all negative processes occurring in the living organism on the cell level [21–28]. They are also the main cause of spoilage of many foodstuffs. This field of biochemistry has been given special attention; a large body of information has been accumulated over two recent decades in favour of the ability of flavonoids to inhibit oxidation processes and to reduce the risk of damaging essential biomolecules in stress situations [26, 34, 35], as well as to react with many forms of active oxygen species [18–20, 26, 29–33]. Along with the antioxidant properties, in recent years flavonoids were noted to possess a prooxidant activity [20, 36, 37], though the ability of their particular representatives to be included into photodynamic processes has been known since long ago [9, 38]. Flavonols are of special interest, as according to modern views they are the main components of the photoprotector system of plants and the most potent antioxidants. This review deals with some aspects of the relation between the structure of flavonols and their photochemical reactions with oxygen. It should be noted that ideas and assumptions presented here are far from being those, which could be taken for granted. 1

Structure of Flavonols

Flavonoids form a large family of natural O-heterocyclic colorants, the main structural unit of which is 2-phenyl-4-benzopyrone (flavan). For this reason, flavonoids can be considered as its derivatives. One of the classifications [10] divides these compounds into ten main classes: catechins, leucoanthocyanidins, flavanols, dihydrochalcones, anthocyanidins, flavanonols, flavones, flavonols and aurons. The structural features of some of them are shown in Fig. 1. The structural unit of flavonols is flavone. For this reason, these compounds are often considered as its derivatives. The structural types of some of them are given in Table 1. A feature of flavones is the presence of a C2C3 double bond conjugated with the π-electronic system of cycle B. Emerging cross conjugation between the carbonyl group and cycles A and B [7] determines the character of their electronic absorption spectra. Flavonols can contain from one up to six free OH groups [1–4, 7, 9]. Compounds with etherified OH groups (OR) are known to exist, as a rule, at C-3, C-5 and C-7 carbon atoms. Methyl and acetyl radicals often occur as substituting R groups, as well as residues of carbohydrates (glucoses, rhamnoses, galactoses and others). Flavonols containing 3-, 7- and 4′-OH groups are weak acids and show three stages of titration within the range of pKa 6.7–11.6 [8, 32, 39, 40]. Compounds with the 7-OH group possess the maximum acidity, whereas the acidity of other OH groups is on the level of phenol and is to some extent related to the structure of molecules. For derivatives of flavonol with ortho-substitution of hydrogen atom in cycle B the acidity of the 3-OH group decreases. The cause of this is noncoplanarity of cycles B and C, which emerges due to steric hindrances and prevents the

The Photochemical Aspect of Reactions of Flavonols with Molecular Oxygen

273

3' 8 7 6

2'

1

2

O

A

C

1' 3

4'

B

H

5'

O

6'

H

4

5

H

H

O I

OH

II

O

O OH

OH O IV

III

Figure 1 Structural formulas of some representatives of the family of flavonoids: flavone (I), catechine (II), flavonol (III) and anthocyan (IV).

delocalization of excess electron charge on the C2C3 double bond. The 5-OH group is the most hard to ionize due to the strong intramolecular hydrogen bond formed with its participation [8, 39]. Owing to the structural features of flavonols, their 3- and 5-OH groups and the carbonyl atom of oxygen mutually affect one another, as the result of which cyclic unstressed intramolecular hydrogen bonds can emerge between them (an H bond) [1, 40–43], as shown schematically in Fig. 2. These bonds are formed in noncoordinating solvents and Table 1 Structural types of some essential flavones. Flavone

Galangine Fisetine Chrysine Apigenine Luteoline Kampferol Morine Quercetine Rhamnetine Myricetine Rutine

Substituents 3

5

6

7

8

1'

2'

3'

4'

5'

OH OH H H H OH OH OH OH OH P

OH H OH OH OH OH OH OH OH OH OH

H H H H H H H H H OH OH

OH OH OH OH OH OH OH OH M H H

H H H H H H H H H H H

H H H H H H H H H H H

H H H H H H OH H H H H

H OH H H OH H H OH OH OH OH

H OH H OH OH OH OH OH OH OH OH

H H H H H H H H H OH H

M, methoxy group; P, rutinose residue.

274

E.A. Venedictov

O

O O O

O

H I

Figure 2

O H II

Structure of 3-hydroxyflavone (I) and 5-hydroxyflavone (II) with intramolecular H bonds.

O

hν

O

O O

H

N

O OH

T

Figure 3 Structure of normal (N) and tautomeric (T) forms of 3-hydroxyflavone.

open partially or totally in proton-donor and proton-acceptor media capable of forming thermally more stable intermolecular H bonds, as well as, apparently, in self-association of these compounds in highly concentrated liquid and solid solutions [44]. Formation of intramolecular H bonds explains many properties of flavonols, including photochemical properties. Under the action of light, such compounds undergo isomerization [40, 42, 47–69]. Formation of the phototautomeric configuration (Fig. 3) is achieved owing to the total intramolecular transfer of proton due to the increase of acidity of 3- or 5-OH groups in excited state [39]. Theoretical studies predict the formation of an intramolecular H bond between oxygen atom of the 3-OH group and hydrogen ortho-atom of benzene cycle B of flavonols [45]. A greater extent of coplanarity of cycles B and C is assumed to be achieved owing to this and, as a consequence, an enhancement of intramolecular H bond of the 3-OH group and carbonyl atom of oxygen according to the feedback principle [40]. Another important structural element determining the physicochemical properties of flavonols is cycle B. It is capable of undergoing a turn around the C1′C bond, thus contributing to a change of the extent of conjugation of π-electron systems of cycles B and C. There are no reliable data yet on its equilibrium position relative to the plane of the rest of the molecule. At the same time, theoretical studies show cycle B to be removed from the plane of cycles A and C [45, 46]. The angle of turn of cycle B is from 16 to 43° [46]. Nevertheless, this geometry [45] does not completely restrict the conjugation of π-electron systems of cycles B and C. For compounds containing the ortho-substituted cycle B, the advantageous

275


OH

OH OH

HO

OH

O M+n

HO

O

OH OH

O

O

OH

O

M+[n-1] II

I

M+n O

M O

HO

O O OH

O

M+[n-1] III

Figure 4 Structural images of the coordination compounds of quercetine (I): complexes of the composition 1 : 1 (II) and 1 : 2 (III).

conformation due to steric reasons is the strongly nonplanar conformation with all ensuing consequences. According to theoretical studies [57], the close-to-coplanar conformation of cycles B and C seems to be achieved only in excited state. Cycle B plays the role of an electron-density donor with respect to γ-pyronic cycle C. This is indicated by the results of theoretical studies, which show a high density of electrons at C-2 carbon atom of flavonols [45]. This is also testified by the data of PMR studies showing an increase of aromaticity of cycle C [57]. Intramolecular charge transfer is enhanced in excited state, if electron donors are substituents in cycle B [68]. One of the consequences of the structural features of hydroxyflavones is their capability of coordination interactions with metal ions [4, 70–72]. Complexation can occur with participation of carbonyl oxygen and 3-OH or 5-OH group, as well as involving ortho-OH groups of cycle B. This property is the most pronounced in natural flavonols, which can have up to three coordination sites (Fig. 4), the filling of which depends on the nature of metal and specific conditions of the medium. Complexation of flavonols can occur not only in the main state but also in excited state [68], both from normal and tautomeric forms, thus enhancing intramolecular charge transfer from cycle B to cycle C, which, in turn, would lead to an increase of the stability of the complex. The ability of flavonols to bind heavy metal ions is believed to be one of the key mechanisms of their antioxidant action [16, 18–20, 34].

276

2

E.A. Venedictov

Spectral Luminescent Properties of Flavonols

Flavonols are weakly-coloured compounds, which absorb light in the UV range of the solar spectrum. Owing to the presence of two cross-conjugated systems, the electronic spectra of these compounds consist of two groups of close-intensity bands lying in the range of 240–270 and 320–380 nm and related to the electronic transitions of ππ*type in the “benzoyl” (cycle A) and “cinnamoyl” (cycles B and C) fragments of molecules, respectively [2, 4, 7, 8, 17]. The first group of bands coincides with the mixed range of action of the lethal effect of UV radiation on living organisms, which is due to the absorption of proteins and nucleic acids [23]. The second lies in the region of near UV light considered to have an indirect damaging effect [23, 73] associated with the photochemical properties of natural and artificial photosensitizers, which are either products of cell metabolism or get into the organism with food or drugs. Electronic spectra bear an imprint of the structural features of flavonols. Here we shall give only the general ideas of the structure–property and medium–property relationships. More profound views can be found in [2, 4, 7, 8, 17, 38, 39, 52, 66, 68]. Available data indicate that an increase of the number of OH groups in cycle B induces a batochromic shift of the long-wavelength band and an increase of its absorption coefficient. This fact can be explained within the framework of an extended system of conjugated bonds, including those due to superconjugation with participation of OH groups. As for the short-wavelength band, the rise of the number of OH or OCH3 groups in this cycle usually does not lead to a significant change of its position. The successive increase of the number of these groups in cycle A leads to a similar effect. However, it is less pronounced. The spectra of many flavones change at high pH due to the ionization of “acidic” OH groups. Herewith, a batochromic shift of the maxima of the bands and the redistribution of their intensities are observed. The strongest changes in the absorption spectra of flavonols are related to the 3-OH group. Its elimination is accompanied by a hypsochromic shift of the absorption bands. A similar change is observed at the substitution of hydrogen atom of this group by methyl radical. The effect of methylation of OH groups on the absorption spectrum of quercetine is demonstrated by the data of Table 2. The effect of a solvent on the spectra of flavonols is expressed to a lower degree [52]. No unequivocal relation between the parameters of the spectra and the nature of a solvent Table 2 Parameters of the first band in the absorption spectra of quercetine and its derivatives in different solvents [86].

λ max, nm (ε, cm –1 ·l·mol –1)

Flavonoid

Quercetine 3,5,7,3′,4′-Pentamethoxyflavone 5-Hydroxy-3,7,3′,4′-tetramethoxyflavone 3-Hydroxy-5,7,3′,4′-tetramethoxyflavone Mg-Quercetine

pyridine

alcohol

acetone

benzene

379 (21700) 339 (18200) 360 (19100)

374 (20900) 340 (21900) 352 (21900)

369 (21000) 332 (19000) 350 (18900)

330 (19600) 351 (19200)

360 (22300)

367 (20300)

351 (19800)

356 (22000)

474 (21230


277

has been revealed. Nevertheless, it is believed that the effect of the medium reflects effects of general and specific solvation. The strongest changes in the absorption spectra of hydroxyflavones are caused by complexation with metal ions, which leads to a batochromic shift of both bands. Herewith, the complexation involving the 3-OH group is characterized by a maximal effect (Fig. 6). Additional complexation via the ortho-site of cycle B (Fig. 4) enhances this effect. Thus, it is seen that the diversity of flavone structures provides for the maximum effect of absorption and reflection of solar UV radiation at the most harmful directions of its action. In contrast with flavone [42, 67, 69, 74], flavonols possess fluorescence. This is explained by the fact that in substitution of hydrogen atoms of flavone at C-3 carbon atom by an OH group the excited triplet nπ* level of the molecular proves much higher than the excited singlet level of ππ* type and renders no effect on its radiative ability [67, 74]. Fluorescence of flavonols is of dual character. It is considered as a radiation from two excited singlet states with energies of 69–71 and 53–55 kcal/mol, respectively [39, 40, 50, 56]. Short-wavelength luminescence is determined by the properties of normal molecules, and long-wavelength luminescence is due to the properties of their tautomers formed as the result of reversible intramolecular phototransfer of proton from the 3-OH group to carbonyl oxygen. Intramolecular proton phototransfer is an essential feature of flavonols. It occurs during a time shorter than 8×10–12 s and is controlled by structural factors and, in particular, specific properties of the medium [40, 51, 52, 56, 59, 64, 66]. Proton phototransfer can be due both to the properties of singlet and triplet stages depending on the structure of a flavonol. As a rule, it takes place in compounds containing a 5-OH group [57, 63, 69]. Intramolecular proton phototransfer determines the short lifetime of an excited singlet state of flavonols, thus limiting its chemical potentialities. However, it does not totally suppress the interconversion of excited singlet molecules into triplet ones. The latter are characterized by a lifetime of 7.2–7.5 µs and energy of 61–53. 4 kcal/mol [60, 65]. Fluorescence of tautomers of flavonols is characterized by a relatively large lifetime, which has some dependence on the nature of substituents, their number and substitution site in cycle B and is from 1.3 up to 4.4 ns [39, 48, 50, 56, 65]. The lifetime of the triplet state of tautomers depending on the structural features of flavonols is from 9 up to 28 µs [49, 56, 60, 65]. The energy level of this state is within the limits of 31–33 kcal/mol [65]. In contrast to fluorescence [49], oxygen efficiently quenches triplet states of flavonols [57, 60, 62]. Its removal from solution is an obligatory condition for studies of their properties. Studies have shown that the rate constants of quenching of triplet molecules by oxygen are close to diffusion rates. In n-heptane, for instance, for the main and tautomeric forms of unsubstituted flavonols they are 2.1×109 M–1 s–1 and 2.7×109 –3.2×10 9 M–1 s–1 [60, 62], respectively. Unlike flavonols containing no 5-OH group, quercetine does not luminesce in practice and gives no long-lived triplet states in direct excitation [57]. One of the causes is a high probability of internal radiativeless conversion of absorbed energy. However, formation of a long-lived triplet state of quercetine can be caused by the triplet–triplet transfer of energy [57]. In this case, sensitizers having a triplet level no less than 57 kcal/mol transfer their energy to the triplet level of the main form of quercetine with a rate constant exceeding 1.2×10 9 M–1 s–1. Then there occurs rapid tautomerization of triplet molecules as the result of intramolecular proton phototransfer from the 5-OH group to carbonyl oxygen. This is one more significant fact, which indicates that quercetine is

278

E.A. Venedictov

capable of participating in quenching of excited triplet states of proteins and nucleic acids, whose energy exceeds 57 kcal/mol [75]. The work [57] obtained the characteristics of the triplet state of the tautomer of quercetine. It has an energy within the range of 17.1–24 kcal/mol and lifetime from 12.9 up to 18 µs depending on the solvent. The triplet tautomer of quercetine is quenched by oxygen. In isopropanol, the quenching rate constant is 1.7×109 M–1 s–1. Finally, it is necessary to note the luminescent properties of coordination compounds of flavonols. Compounds with Mg2+, Zn2+, Cd2+, Al3+ and a number of other elements possess a high yield of fluorescence [68, 70, 71]. These results are interpreted from positions of limitations imposed by complexation on intramolecular proton phototransfer [68]. Electron-donor substituents in para-position of cycle B increase the yield and lifetime of fluorescence [68]. The latter is associated with the increase of the probability of transition from an excited singlet state into the basic one with illumination and a reduction of the probability of internal radiativeless conversion. Introduction of a heavy atom leads to a reverse phenomenon of the decrease of the quantum yield and a reduction of the fluorescence lifetime. As the rates of these transitions are relatively small, one should take into account the possibility of formation of a long-lived triplet state of coordination compounds of flavonols in absorption of light by them. As will be shown below, these conditions are realized also in the case of quercetine complexes. Nevertheless, unlike other flavonols, quercetine attracts attention by its capability of highly efficient utilization of high-energy solar radiation. 3

Photoproduction of 1O2 (1∆g)

By their photochemical activity, flavonols can be conditionally divided into two groups, one of which contains compounds only with a 3-OH group, and the other, compounds with 3- and 5-OH groups simultaneously. Flavonols of the first group exhibit a photochemical activity [76] and enter into photochemical interaction with oxygen [49, 50, 60], undergoing destruction, whose mechanism was discussed from various mechanistic positions, including with attraction of 1O2 [60]. Natural representatives of the second group are photochemically inactive. Photoformation of 1O2 by compounds of the first group is supported by the results of experiments on the sensitization of luminescence from 1∆g state of O2 in solution of unsubstituted flavonol [60]. In this case, the quantum yield of 1O2 is 0.18. As dual fluorescence of this compound is not quenched by oxygen [49], generation of 1O2 is associated with the quenching by oxygen of its triplet states. Unlike unsubstituted flavonol, quercetine does not sensitize photoformation of 1O2 [57]. These data are an additional argument in favour of the absence of photochemical activity in this group of compounds [76]. A similar pattern was observed in studies of the photochemical properties of 3hydroxy-5,7,3′,4′-tetramethoxy- and 5-hydroxy-3,7,3′,4′-tetramethoxyflavones [77]. At the same time, 3,5,7,3′,4′-pentamethoxyflavone was shown to be capable of photosensitizing the formation of 1O2. In this case, the quantum yield of 1O2 is 0.49. This is no wonder, as it is known to exhibit a high photochemical activity [11, 12] and possess intensive phosphorescence, which indicates the formation of an excited triplet state [12]. These data show that in such molecules as quercetine the photochemical inertness is related to both 3- and 5-OH groups. Of special interest are the properties of triplet states and their role in the formation of


279

1

O2. This issue becomes especially topical when we speak of the triplet state of quercetine, accounting for the special significance of this compound as a photoprotector. Though the answer to the posed question requires special investigation, some conclusions can be made based on the consideration of the quenching of triplet states by oxygen within the framework of the formalism of the equilibrium formation of exciplex intermediate [78, 79]. 3

S + O2 (3Σg– ) ↔ 1(S … O2) → S + 1O2 (1Σg+ and/or 1∆g), ↔ 3(S … O2) → S + O2, ↔ 5(S … O2).

Here 3S is the triplet molecule of a sensitizer; 1, 3, 5{S … O2 } are the singlet, triplet and quintet states of the exciplex. In this scheme, the processes of conversion of singlet and triplet exciplexes are considered to be the most probable, as they are allowed by the spin selection rules. The theoretical values of statistical balances of these processes are 1/9 and 1/3, respectively. Their real values can be assessed from the ratio kQ /k diff, where kQ and k diff are the rate constant of quenching of triplet molecules by oxygen and the diffusion rate, respectively. For triplet states of the normal form of unsubstituted flavonol and its tautomer, these values in n-heptane are 1/7 and 1/5, respectively. Comparing them with theoretical values, it can be concluded that the probability of 1O2 formation at the expense of the triplet state of the normal form is expected to be higher. For a triplet state of tautomeric quercetine, the ratio kQ /kdiff is equal to only ~1/2 [57]. Evidently, in this case all triplet molecules are quenched by oxygen according to the mechanism of internal interconversion. Therefore, this process can be considered within the framework of the photoprotector action of quercetine as an important link for transforming the energy of the tautomer’s excited state to thermal energy. Thus, from the data presented it can be concluded that natural flavonols containing a 5-OH group provide for highly efficient radiativeless dissipation of absorbed solar energy without forming 1O2. 4

1

O2 Reactions

The first steps in this direction were made in [80], which showed the selective character of the reaction of flavonols with 1O2, aimed to open the C2C3 double bond of cycle C. The physical character of quenching 1O2 by quercetine was first demonstrated in [57]. The same work postulated the interaction mechanism determined by the formation of exciplex intermediate with reversible partial charge transfer, as it is proposed for other phenols [25, 81, 82, 93–96]. Herewith, the possibility of not only the above-considered process was shown, but also of oxidation of quercetine and electron-vibration transfer of energy with excitation of the vibratory levels of its molecules. Therefore, the total rate of quenching in this case can be presented by a sum of the rates of three elementary processes: k t = k q + k r + k E-V, where k t is the overall quenching rate constant; k q is the physical rate constant determined by the formation of the exciplex intermediate; k r is the oxidation rate constant of flavonoids; k E-V, the physical rate constant determined by electron-vibration energy exchange. The results of measurements and assessments of these constants are given in Tables 3 and 4.

280

E.A. Venedictov

Though for quercetine the value of k E-V is relatively small, it is still comparable with k r in acetone. Therefore, it is natural that the expediency and reality of such a scheme for other representatives of this family requires additional consideration and experimental proof. Electron-vibration energy exchange should be expected exactly for molecules containing C–H and O–H bonds, as they have the highest vibratory states [57, 90–92]. The rate constant of this process could not be measured, but yields to numerical assessment on the basis of the additivity principle [90–92] by the formula Table 3 Rate constants of interaction of flavonoids with 1O2 in different solvents. Solvent

kt × 10–5, M –1 ·s–1

3-Hydroxyflavone

CCl4 Benzene-h6 Acetonitrile Methanol-h4 Pyridine D2O (pD = 3) D2O (1.0 M NaOH)

0.172 0.187 1.05 2.3 750 2.5 1900

83 83 83 83 83 83 83

3-Methoxyflavone

CCl4 Benzene-h6

0.19 0.15

83 83

Quercetine

Acetone Acetone Diethyl ester Methanol-d4 Dimethylformamide Pyridine

3.1

24 53 156

Galangine Kampferol Fisetine Rutine Luteoline Chrysine Catechine

Methanol-d4 Methanol-d4 Methanol-d4 Methanol-d4 Methanol-d4 Methanol-d4 Methanol-d4

3,5,7,3',4'-Pentamethoxyflavone

Flavone

kr × 10 –5, M–1 ·s–1

Reference

0.09 0.26(a) 1(a) 8.9(a) 49(a) 117(a)

57 85–87

12 7.1 31 16 13 2.4 58

7.4 4.8 11 1.1 0.18 0.06

84 84 84 84 84 84 84

Benzene-d6 Pyridine

35

1.4(a) 0.5(a)

77, 87, 88 87

3-Hydroxy-5,7,3',4'-tetramethoxyflavone

Benzene-d6 Acetone Dimethylformamide Pyridine

25

1.4(a) 15(a) 110(a) 250(a)

87, 88 87 87 87

5-Hydroxy-5,7,3',4'-tetramethoxyflavone

Benzene-d6 Pyridine

20

1.1(a) 0.9(a)

87, 88 87

84 86, 87 86, 87

(a)The values of k were obtained relative to the reactivity of tetracene and were refined using its r values of kr [89] equal to 5.2×107 M–1 ·s–1 (pyridine), 7.7×107 M–1 ·s–1(DMFA), 2.2×107 M–1 ·s–1 (acetone) and 1.4×107 M–1 ·s–1 (benzene).


281

k E-V = mk C–H, arom + nk C–H, aliph + pk O–H . Here k C–H, arom = 480 M –1 s–1, kC–H, aliph = 390 and kO–H = 2145 M –1 s–1 are rate constants of 1O2 quenching by particular C–H (aromatic), C–H (aliphatic) and O–H bonds, respectively [92]; m, n and p are the number of each of these bonds in the molecule. The values of k E-V for some flavonoids are presented in Table 4. The same table gives the values of k q. For compounds with a large number of OH groups, k E-V is only a small part of k q, not exceeding 5%. At the same time, the values of k E-V and k q are commensurable for a whole range of simplest hydroxyflavones. For this reason, the role of this process in 1O2 quenching for them should not be underestimated. The values of k q for some flavonoids are given in Table 4 together with their oxidation potentials. The donor-acceptor nature of 1O2 quenching is confirmed by the dependence between k q and the oxidation potential, in which the rate constant increases when the oxidation potential decreases. Therefore, this process can be believed to be controlled by charge transfer and can be considered within the framework of exciplex intermediate formation. Table 4 Parameters of the elementary reactions of 1O2 physical quenching by some flavonoids and their oxidation potentials. Flavone Quercetine Rutine Kampferol Catechine (a) from

kE-V ×10–4 M–1 ·s–1

kq×10 –5(a), M –1 ·s–1

Eox(b) , V

1.5 3 1 1.5

15 15 2.3 ~58

0.60 0.60 ~0.95 0.57

[84]; (b) from [32].

If we turn to the results of [84], it can be found that in the sequence of flavonols between the values of k q and k r a correlation takes place, which in methanol is described by the equation ln k r = (9.01 ± 0.93) + (0.34 ± 0.07) ln k q , r = 0.961. Based on this, the following conclusion can be made [97]: exciplex formation is the common stage of physical and chemical quenching of 1O2. A study performed within the structure–property framework [84] made it possible to single out a number of structural features affecting the quenching of 1O2 by flavonols. It has been found that compounds with the catechol structure of cycle B are stronger quenchers (Table 3). The effect of the 3-OH group on the process occurring in methanol-d6 proves to be relatively weak. This conclusion is supported by the data of an earlier work, which studied the quenching of 1O2 by unsubstituted flavonol and 3-methoxyflavone in relatively inert solvents, such as benzene and CCl4 [83]. The works [77, 86, 88] have shown that complete substitution of hydrogen atoms of OH groups of quercetine by methyl radicals leads to a rise of efficiency of 1O2 quenching. The presence of a free 3- or 5-OH group leads to an insignificant decrease of the quenching rate (Table 3). An amplification of 1O2 quenching in these cases as compared with quercetine should be, apparently, ascribed to a decrease of energy expenses for exciplex formation owing to an increase of the density of π-electrons in molecules of these compounds due to

282

E.A. Venedictov

ln kr

20

1

10

2

0 0

20

40

DN Figure 5 Dependence of the logarithm of the rate constant of oxidation of quercetine (1) and 3-hydroxy-5,7,3′,4′-tetramethoxyflavone (2) by singlet oxygen on the donor strength of the solvent.

the electron-donor effect of the methoxy groups. Herewith, the greatest contribution is, apparently, made by methylation of ortho-OH groups of cycle B. According to the data of [83, 86, 87], the efficiency of 1O2 quenching by flavonols rises in highly specific media. Due to the absence of data in the case of elementary reactions with unsubstituted flavonol, we carried out a simplified consideration of this phenomenon. In the order of increasing k t, the solvents are in the following sequences: CCl4 ≤ benzene < acetonitrile < methanol ≤ heavy water (pD = 3) < pyridine < heavy water (1 M NaOH) for unsubstituted flavonol and acetone < methanol-d4 < N,N-dimethylformamide < pyridine for quercetine, which give no grounds to relate the quenching rate to the polarity of solvent. At the same time, it is evident that the efficiency of the process increases with the rise of the ability of the solvent to form a strong intermolecular H bond with the OH group. This is illustrated by the dependence of k t on its donor ability of its molecules (DN). For unsubstituted flavonol, this type of dependence has the form: ln k t = (9.04 ± 0.35) + (0.19 ± 0.02) DN,

r = 0.988.

It is also confirmed for quercetine: ln k t = (9.95 ± 1.60) + (0.21 ± 0.7)DN,

r = 0.912.

Noteworthy is the qualitative and quantitative similarity of the correlation equations. The coefficients at DN, which determine the slopes of the dependences, do not practically differ from each other, though structurally flavonols differ significantly. This indicates that 1 O2 quenching in both cases experiences practically the same effect of the medium. Despite the fact that k t is a complex value, for a simplified explanation of this fact it is necessary to admit that the effect of the solvent is related to the presence, mainly, of the 3-OH group. To understand the nature of this phenomenon better, we considered the features of the behavior of quercetine in particular elementary reactions. The rate constant of physical quenching changes in different media within one order of magnitude. The largest value of k q was found in pyridine; the smallest, in acetone. Despite the fact that the change of k q depends in a complex way on the solvent, nevertheless, there is a hint to a relation to its donor properties. This makes us ascribe the observed effect to the solvation stabilization of

283


exciplex intermediate, the result of which is additional delocalization of the fractional positive charge emerging during its formation. In contrast with k q, the value of k r in different media changes within broader limits. The value of k r is related to the solvent in the same manner (Fig. 5). Unlike quercetine, oxidation of compounds without the 3-OH group experiences practically no effect of the medium (Table 3). Therefore, the observed difference in the effect of the solvent is quite naturally related to the solvation of the 3-OH group. Undoubted is a symbateness in the change of k q and k r in these media, which is similar to that considered above, though is less pronounced. This behaviour of k q and k r, evidently corresponds to the case when the oxidation rate is controlled by the formation of exciplex intermediate [98]. Another example is oxidation of 3-hydroxy-5,7,3′,4′-tetramethoxyflavone (Fig. 5). The observed tendency for flavonols is apparently of general character. Still, it is not difficult to notice some difference. It is seen that the oxidation of 3-hydroxy-5,7,3′,4′tetramethoxyflavone possesses the least pronounced sensitivity to the solvent. To explain this fact, it is necessary to admit that in this case the emerging cationic site in the formation of exciplex intermediate is capable of more efficient intramolecular delocalization. In [80], oxidation of flavonols by 1O2 molecules was demonstrated, unlike flavones, to affect the C2C3 double bond. Herewith, the main product of oxidation is depside (Fig. 6, VI). In the literature, two most possible mechanisms of this reaction are discussed (Fig. 6): one assumes oxidation to proceed as an ene reaction (Fig. 6, III) with concerted addition of 1 O2 [80]; the other postulates the formation of a dioxetane intermediate (Fig. 6, IV) [84]. Based on the above analysis, it can be concluded that the primary intermediate of oxidation

O

1

O O O

O2

OH

O O OH

O II

O I

O III

O

H

O

O IV

O

OH

O I O

O O V

O CO

OO

O COOH VI

Figure 6 Hypothetic mechanisms of addition of 1O2 to flavonols.

O H

284

E.A. Venedictov

k, s-1 100000

50000

0 0

0.00002

0.00004

c, mol/l Figure 7 Effect of the concentration of Mg-quercetine on the rate constant of 1O2 luminescence decay in pyridine (pulse fluorimetry method [88]; 1O2 photosensitizer, anthracene, λexc = 337 nm).

of flavonols is exciplex (Fig. 6, II). What is more, the difference in the kinetic behaviour of flavonols and flavone suggests that its formation is related to the involvement of π-electrons of the C2C3 double bond of cycle C. Assessing on the whole the ability of flavonols to quench 1O2, it can be concluded that the nature of the molecular environment is one of the most important factors making it possible to achieve the high efficiency of the process. 5

Photochemical Properties of Coordination Compounds of Quercetine

With many metal ions in stable degrees of oxidation (Mg 2+, Zn 2+, Cd 2+), quercetine forms compounds of various compositions stable in the dark. The structures of these compositions are shown in Fig. 4 (II). Unlike quercetine, its coordination compounds efficiently quench luminescence of 1 O2 [85, 86]. For instance, Mg-quercetine presumably having the structure II (Fig. 4), quenches 1O2 in pyridine with a rate constant of (9.3 ± 0.5)×10 8 M–1 s–1 (Fig. 7). Close values were obtained for compounds with Zn2+ and Cd2+ [86]. In the presence of foreign sensitizers, photoexcitation leads to oxidation of coordination compounds [85, 86], which is related to the properties of 1O2. Figure 8 presents the data showing the extent of difference in the rates of photosensitized oxidation of Mg-quercetine and a typical 1O2 acceptor tetracene [99–101]. This result is explained by a higher reactivity of the former. Comparing the oxidation of these compounds, we can obtain k r for Mg-quercetine. In pyridine, the ratio of their observed rate constants is 12. According to the principle of competing reactions [102], this value reflects the ratio of their true rate constants (k r /k r,t). Hence, knowing k r,t for tetracene, it is easy to find kr. Taking kr,t = 5.2×107 M–1 s–1 [89], we get that k r = 6.2×108 M–1 s–1. Based on k t and k r, we can find k q. Its value is 3.1×0 8 M –1 s–1, which is significantly lower than k r. These data are in favour of the chemical mechanism of 1O2 quenching by Mg-quercetine. As seen from the data presented, the rate constants of the elementary reactions of 1O2


285

ln (A0/At) 1.2

1 2

0.6

0

0

30 Time, s

60

Figure 8 Kinetics of photosensitized oxidation of Mg-quercetine (1) and tetracene (2) in pyridine during their joint presence in solution (1O2 photosensitizer: Pd-tetra(t-butyl)phthanocyanin, λex > 640 nm, A0 and At are the initial and current optical densities of the solution at the measurement wavelengths, respectively).

quenching found for Mg-quercetine are by approximately two orders of magnitude higher than for quercetine (Table 3). We have here a rather interesting example of activation of quercetine with respect to the reaction of 1O2, similar, apparently, to the activation of unsubstituted flavonol due to the ionization of the 3-OH group [83]. To date, there is sufficient proof of accelerated destruction of coordination compounds of quercetine in the light [44, 86]. Consider the photochemical properties of these compounds in greater detail by the example of Mg-quercetine. Experiments yield the following results [86]. The range of action of photodestruction coincides with the absorption spectrum of this compound. For this reason, it is natural to consider the photodestruction as a self-sensitized photoreaction. Its yield depends on oxygen. The involvement of 1O2 in it is indicated by the fact that para-hydroquinone inhibits the photoreaction with the efficiency, with which it quenches the luminescence of 1O2 (Fig. 9). Hence, an important conclusion can be made that Mg-quercetine is a photosensitizer of 1O2. As the lifetime of its fluorescence in pyridine does not exceed 2×10–9 s, and the fluorescence is not quenched by oxygen, it can be concluded that 1O2 is formed as the result of quenching of excited triplet molecules of Mg-quercetine by oxygen. We can not but mention another significant fact, namely, that the photooxidation mechanism of coordination compounds of quercetine depends on the nature of a metal. For instance, inhibition of Cd-quercetine photooxidation by para-hydroquinone is much less efficient [86]. In this case, the contribution of the reaction with participation of 1O2 is reduced to 60–70%. It is evident that in the case of coordination compounds with heavy elements other oxidation routes are possible in principle; oxidation of excited triplet molecules by oxygen can be assigned to these routes [25, 99]. The determination of 1O2 quantum yield is more complicated. A low spectral resolution of the setup used in [86] did not make possible the direct proof of luminescence photosensitization from the 1∆g condition of O2 in solutions of these compounds. To assess the

286

E.A. Venedictov

k, rel. units 1 1 2 3

0.5

0 0

0.0006

0.0012

c, mol/l Figure 9 Effect of p-hydroquinone on the rate of self-sensitized photooxidation of Mg2+-quercetine (1), tetracene (3) and 1O2 luminescence (2) in pyridine [86].

quantum yield of 1O2, we used the kinetic data of the direct photooxidation of Mg-quercetine and tetracene, obtained under conditions of monochromic excitation by the light with λ = 447 nm (Fig. 10). Consider this issue in more detail. As shown in Fig. 9, the photooxidation reactions of Mg-quercetine and tetracene are inhibited by para-hydroquinone with the same efficiency. These data indicate that oxidation is related exclusively to 1O2. Therefore, photoconversion of these compounds can be presented as the following scheme [97–101]: hν 1

3

S → S→ S –

3

S + O2 (3 Σ g ) → S + 1O2

1

O2 → O2(3 Σ g ) + hνlum,

1

O2 + Solv → O2 (3 Σ g ) + Solv + Q

1

O2 + S → O2 (3 Σ g ) + S

–

–

1O

–

2

+ S → oxidation products

(krad) (ks) (kq) (kr),

where S, 1S and 3S is the photosensitizer in the basic and excited singlet and triplet states, respectively; Q is the energy of 1O2 dissipated as heat; Solv is solvent; krad, ks, kq, kr are rate constants of respective processes. Hence, we can write the following expression for the oxidation rates for each of these compounds:


287

0.49

0.48

sequence 1 sequence 2

A

0.47

0.46

0.45 0.44

0.43 0

100

200 Time, min

300

400

Figure 10 Optical density changes in solutions of Mg2+-quercetine (1) and tetracene (2) at a wavelength of 447 nm during the excitation with a monochromatic light (λ = 447 nm) in air-saturated pyridine (initial concentrations of Mg2+-quercetine and tetracene are, respectively, 6.5·10 –5 and 1.1·10 –4 mol/l) [86].

k r γ I [ S ]τ -. –d[S] ⁄ ( dt ) = -----------------------1 + k t [ S ]τ Here γ is the quantum yield of 1O2; I, the intensity of the absorbed light flux; [S], concentration of sensitizer; τ = 1/(krad + ks [Solv]), the lifetime of 1O2 in the solvent; kt = (kq + kr). For tetracene, the term (1 + kt [S]t) in most common solvents can be accepted to be unity [86, 89, 98]. If solutions of these compounds are illuminated under identical conditions, then for the extent of their conversion not exceeding 15%, the relative value of γrel can be assessed from the ratio

α = β γrel / (1 + kt [S]τ), where α, β and γrel are the ratios of the observed rate constant, kr and γ for Mg-quercetine to analogous values for tetracene; [S] is the initial concentration of Mg-quercetine. It is seen from the data presented in Fig. 10 that in pyridine α = 1.8. The value of β, according to the above-presented data, in the same solvent is equal to 12. Then, using τ = 18×10–6 s, for the concentration of Mg-quercetine equal to 6.5×10–5 mol/l, we obtain γrel = 0.32. Taking for tetracene γ ~ 0.71 [103], we find that γ ~ 0.22. As is seen, this value is sufficiently high and comparable with the value of γ for unsubstituted flavone, but is lower that the value of γ for 3,5,7,3′,4′-pentamethoxyflavone. Therefore, complexation, by eliminating the effect of the hydrogen atom of the 3-OH group and restricting the effect of the 5-OH group on the photophysical properties of quercetine, contributes to the photoproduction of the triplet state of quercetine capable of exciting O2.

288

E.A. Venedictov

Summing up these results, it can be inferred that an unequivocal conclusion on the photochemical inertness of quercetine with respect to oxygen can not be correct without accounting for the effect of complexation with a whole range of metal ions on this property. 6

Conclusion

At present, there is no doubt that flavonols, depending on conditions, can play the role of both photoprotectors and photosensitizers. This confirms the reality of information mentioned at the beginning of the chapter on the photodynamic properties of some flavonoids. Realization of photochemical reactions is determined, apparently, by whether conditions and structural changes favour these processes.

References 1. Biochemistry of Phenolic Compounds, ed. by J. Harborne, Mir: Moscow, 451 pp. (1968) (Russian translation). 2. T.Y. Mabry, K.R. Mackham and M.B. Thomas, The Systematic Identification of Flavonoid Compounds, N.Y.–Heidelberg–Berlin, 353 pp. (1970). 3. V.L. Kretovich, Fundamentals of Plant Biochemistry, Vyssh. Shkola: Moscow, 464 pp. (1971) (in Russian). 4. A. Blazhei and L. Shutyi, Phenolic Compounds of Plant Origin, Mir: Moscow, 239 pp. (1977) (Russian translation). 5. V.A. Baraboi, Plant Phenols and Human Health, Nauka: Moscow, 160 pp. (1984) (in Russian). 6. T.S. Lebedeva and K.M. Sytnik, Pigments of the Plant World, Naukova Dumka, 86 pp. (1986) (in Russian). 7. G. Britton, The Biochemistry of Natural Pigments, Mir: Moscow, 422 pp. (1986) (Russian translation). 8. V.P. Georgievsky, A.I. Rybachenko and A.L. Kazakov, The Physico-chemical and Analytical Characteristics of Flavonoid Compounds, Rostov State University: Rostov-on-Don, 143 pp. (1988) (in Russian). 9. P. Atkins, Molecules, Mir: Moscow, 216 pp. (1991) (in Russian). 10. M.N. Zaprometov, Phenolic Compounds, Nauka: Moscow, 272 pp. (1993) (in Russian). 11. A.C. Waiss (Jr) and J. Corse, J. Am. Chem. Soc., 87, 2068 (1967). 12. A.C. Waiss (Jr), R.E. Lundin, A. Lee and J. Corse, J. Am. Chem. Soc., 89, 6213 (1967). 13. M. Tevini, Photochem. Photobiol., 2, 401 (1988). 14. M.A.K. Jansen, V. Gaba and B.M. Greenberg, Trends in Plant Sci., 3, 131 (1998). 15. P. Carletti, A. Masi, A. Wonisch, D. Grill, M. Tausz and M. Ferretti, Environ. Experim. Botany, 50, 149 (2003). 16. M.S. Simmonds, J. Phytochem., 64, 21 (2003). 17. B. Havsteen, Biochem. Pharm., 32, 1141 (1983). 18. C.A. Rice-Evans, N.J. Miller and G. Paganga, Free Radic. Biol. Med., 20, 933 (1996). 19. C.A. Rice-Evans, N.J. Miller and G. Paganga, Trends in Plant Sci., 2, 152 (1997). 20. G. Cao, E. Somc and R.L. Prior, Free Radic. Biol. Med., 21, 749 (1997). 21. B. Halliwell and O.I. Aruoma, FEBS Lett., 281, 9 (1991). 22. A.C. Bowling and M.F. Beal, Life Sci., 56, 1151 (1995). 23. J.G. Scandalios, Trends Biochem. Sci., 27, 483 (2002). 24. S.V. Konev and I.D. Volotovsky, Photobiology, Belorussian State University: Minsk, 383 pp. (1979) (in Russian). 25. C. Foote, in: Free Radicals in Biology, Mir: Moscow, vol. 2, p. 96 (1979) (Russian translation). 26. Y. Sorata, U. Takajama and M. Kimura, Biochim. Biophys. Acta, 799, 313 (1984). 27. E.R. Blazek, J.G. Peak and M.J. Peak, Photochem. Photobiol., 49, 607 (1989). 28. K. Fukuzawa, K. Matsuura, A. Tokumura, A. Suzuki and J. Terao, Free Radic. Biol. Med., 22, 923 (1997). 29. M. Thompson, C.R. Williams, G.E.P. Elliot, Anal. Chim. Acta, 85, 375 (1976).


289

30. L. Ramanathan, N.P. Das, J. Agric. Food Chem., 40, 17 (1992). 31. S. Nieto, A. Carrido, J. Sanhueza, L.A. Loyola, G. Morales, F. Lieghton and A. Valenzuela, J. Am. Oil Chem. Soc., 70, 773 (1993). 32. S.V. Jovanovic, S. Steenken, M. Tosic, B. Marjanovic and M.G. Simic, J. Am. Chem. Soc., 116, 4846 (1994). 33. U.N. Wanasundara and F. Shahidi, Food Chem., 50, 393 (1998). 34. P. Sestili, A. Guidarelli, M. Dacha and O. Cantoni, Free Rad. Biol. Med., 25, 196 (1998). 35. C. Zhao, Y. Shi, W. Wang, Z. Jia, S. Yao, B. Fan and R. Zheng, Biochem. Pharm., 65, 1967 (2003). 36. Z. Cai, X. Li and Y. Katsumura, Free Rad. Biol. Med., 27, 822 (1999). 37. Y.H. Miura, I. Tomita, T. Watanabe, T. Hirayama and S. Fokui, Biol. Pharm. Bull., 21, 93 (1998). 38. K. Nishie, A.S. Waiss (Jr) and A.C. Keyl, Photochem. Photobiol., 8, 223 (1968). 39. O.S. Wolfbeis, M. Leiner, P. Hochmuth and H. Geiger, Ber. Bunsenges. Phys. Chem., 88, 759 (1984). 40. A.J. Strandjord, D.E. Smith and P.F. Barbara, J. Phys. Chem., 89, 2362 (1985). 41. T. Hayashi, S. Kawai and T. Ohno, Chem. Pharm. Bull., 19, 792 (1971). 42. Yu.L. Frolov, Yu.M. Sapozhnikov, S.S. Barer and N.A. Tyukavkina, Izv. AN SSSR, Ser. Khim., No 10, 2364 (1974) (in Russian). 43. M.C. Etter, Z. Urbanchyk-Lipkowska, S. Baer and P.F. Barbara, J. Mol. Struct., 144, 155 (1986). 44. G.J. Smith, S.J. Thomsen, K.R. Markham, C. Andary and D. Cardon, Photochem. Photobiol., 136, 87 (2000). 45. N. Russo, M. Toscano and N. Uccela, J. Agric. Food Chem., 48, 3232 (2000). 46.A.Yu. Nazarenko, Zhurn. Strukt. Khim., 19, 525 (1978). 47. P.K. Sengupta and M. Kasha, Chem. Phys. Lett., 68, 382 (1979). 48. O.A. Salman and H.G. Drickamer, J. Chem. Phys., 77, 3329 (1982). 49. M. Itoh and Y. Fujiwara, J. Phys. Chem., 87, 4558 (1983). 50. P. Chou, D. McMorrow, T.J. Aartsma and M. Kasha, J. Phys. Chem., 88, 4596 (1984). 51. D. McMorrow and M. Kasha, J. Phys. Chem., 88, 2235 (1984). 52. A.J.G. Strandjord and P.F. Barbara, J. Phys. Chem., 89, 2355 (1985). 53. P. Chou and T.J. Aartsma, J. Phys. Chem., 90, 721 (1986). 54. M. Kasha, J. Chem. Soc. Faraday Trans., 2, 82, 2379 (1986). 55. P. Chou and T.J. Aartsma, J. Phys. Chem., 90, 721 (1986). 56. M. Itoh, Y. Fujiwara, M. Sumitani and K. Yoshihara, J. Phys. Chem., 90, 5672 (1986). 57. A.P. Darmanyan, O.T. Kasaikina and N.P. Khrameyeva, Khim. Fizika, 6, 1083 (1987) (in Russian). 58. B. Dick and N.P. Ernsting, J. Phys. Chem., 91, 4261 (1987). 59. G.A. Brucker and D.F. Kelley, J. Phys. Chem., 92, 3805 (1988). 60. W.E. Brewer, S.L. Studer, M. Standiford, P.-T. Chou, J. Phys. Chem., 93, 6088 (1989). 61. P.F. Barbara, P.K. Walsh and L.E. Brus, J. Phys. Chem., 93, 29 (1989). 62. M.L. Martinez, S.L. Studer and P.-T. Chju, J. Am. Chem. Soc., 112, 2427 (1990). 63. M.L. Martinez S.L. Studer and P.-T. Chju, J. Am. Chem. Soc., 113, 5881 (1991). 64. G.A. Brucker, T.C. Swinney and D.F. Kelley, J. Phys. Chem., 95, 3190 (1991). 65. K. Tokumura, N. Yagata, Y. Fujiwara and M. Itoh, J. Phys. Chem., 97, 6656 (1993). 66. T.S. Swinney and D.F. Kelley, J. Chem. Phys., 99, 211 (1993). 67. C.M. Christoff, V.G. Toscano and W.J. Baader, Photochem. Photobiol., 101, 11 (1996). 68. A.D. Roshal, A.V. Grigorovich, A.O. Doroshenko, V.G. Pivovarenko and A.P. Demchenko, Photochem. Photobiol., 127, 89 (1999). 69. Y. Norikane, H. Itoh and T. Arai, Photochem. Photobiol., 161, 163 (2004). 70. K.P. Stolyarov and N.N. Grigoryev, Introduction to the Luminescent Analysis of Inorganic Substances, Khimiya: Leningrad, 363 pp. (1967) (in Russian). 71. T. Hayashi, S. Kawai and T. Ohno, Chem. Pharm. Bull., 19, 792 (1971). 72. M. Kopach, S. Kopach and E. Skuba, Zhurn. Org. Khim., 74, 1035 (2004) (in Russian). 73. A. Paretzoglou, C. Stockenhuber, S.H. Kirk and S.I. Ahmad, Photochem. Photobiol., 43, 101 (1998). 74. A.A. Efimov, R.N. Nurmukhametov, I.L. Belaits and A.I. Tolmachev, Opt. Spektrosk., 30, 622 (1971) (in Russian). 75. L.P. Kayushin, Z.P. Gribova and O.A. Azizova, Electronic Paramagnetic Resonance of

290

E.A. Venedictov

Photoprocesses in Biological Compounds, Nauka: Moscow, 304 pp. (1973) (in Russian). 76. T. Matsuura, T. Takemoto and R. Nakashima, Tetrahedron, 29, 3337 (1973). 77. E.A. Venedictov and O.G. Tokareva, Mendeleev Commun., 2, 84 (1997). 78. A.P. Darmanyan and C.S. Foote, J. Phys. Chem., 96, 3723 (1992). 79. F. Wilkinson, D.J. McGarvey and A.F. Olea, J. Phys. Chem., 98, 3762 (1994). 80. T. Matsuura, H. Matsushima and R. Nakashima, Tetrahedron, 26, 435 (1970). 81. V.Ya. Shlyapintokh and V.B. Ivanov, Usp. Khim., 45, 202 (1976) (in Russian). 82. E.A. Lissi, M.V. Encinas, E. Lemp and M.A. Rubio, Chem. Rev., 93, 699 (1993). 83. S.L. Studer, W.E. Brewer, M.L. Martinez and P.-T. Chou, J. Am. Chem. Soc., 111, 7643 (1989). 84. C. Tournaire, S. Croux, M.-T. Maurette, I. Beck, M. Hocquaux, A.M. Braun and E. Oliveros, Photochem. Photobiol., 19, 205 (1993). 85. E.A. Venedictov, O.G. Tokareva and N.O. Konstantinova, Zhurn. Fiz. Khim., 70, 1284 (1996) (in Russian). 86. O.G. Tokareva, PhD Thesis, ISC RAS: Ivanovo (1997) (in Russian). 87. E.A. Venedictov and O.G. Tokareva, Zhurn. Fiz. Khim., 72, 2185 (1998) (in Russian). 88. E.A. Venedictov and O.G. Tokareva, Kinet. Kataliz, 41, 186 (2000) (in Russian). 89. E.A. Venedictov and E.J. Tulikova, Tetrahedron Lett., 44, 3215 (2003). 90. M.A.J. Rodgers, J. Am. Chem. Soc., 105, 6201 (1983). 91. R. Schmidt and H.-D. Brauer, J. Am. Chem. Soc., 109, 6976 (1987). 92. R. Schmidt, J. Am. Chem. Soc., 111, 6983 (1989). 93. M.J. Thomas and C.S. Foote, Photochem. Photobiol., 27, 683 (1978). 94. A.A. Gorman, L.R. Gould, I. Hamblett and M.C. Standen, J. Am. Chem. Soc.,106, 6956 (1984). 95. A. Pajares, J. Gianotti, E. Haggi, G. Stettler, F. Amat-Guerri, S. Criado, S. Miskoski and N.A. Garcia, Photochem. Photobiol., 119, 9 (1998). 96. M.I. Gutierrez, A.T. Soltermann, F. Amat-Guerri and N.A. Garcia, Photochem. Photobiol., 136, 67 (2000). 97. A.A. Krasnovsky (Jr), E.A. Venedictov and O.M. Chernenko, Biofizika, 27, 966 (1982) (in Russian). 98. J.-M. Aubry, B. Mandard-Cazin, M. Rougee and R.V. Bensasson, J. Am. Chem. Soc., 117, 9159 (1995) 99. A.N. Terenin, The Photonics of Dye Molecules, Nauka: Leningrad, 616 pp. (1967) (in Russian). 100. B. Stevens, S.R. Perez and J.A. Ors, J. Am. Chem. Soc., 96, 6846 (1974). 101. A.S. Cherkasov, in: Molecular Photonics, Nauka: Leningrad, 244 pp. (1970) (in Russian). 102. L. Gammet, Fundamentals of Physical Organic Chemistry, Mir: Moscow, 534 pp. (1972) (in Russian). 103. A.P. Darmanyan, Khim. Fizika, 6, 1192 (1987) (in Russian).

8

Solvation of Drugs as a Key for Understanding Partitioning and Passive Transport Exemplified by NSAIDs German L. Perlovich1,2 and Annette Bauer-Brandl2 1Institute

of Solution Chemistry, Russian Academy of Sciences, 153045 Ivanovo, Russia [email protected] 2University of Tromsø, Institute of Pharmacy, Breivika, N-9037 Tromsø, Norway [email protected]

Passive transport properties of drug molecules are of utmost importance for their pharmacological and biopharmaceutical effectiveness. Diffusion in different media and through lipid bilayers are in many cases the rate-determining steps for the distribution in the body. In the present review, an attempt is made to demonstrate the importance of solvation of drug molecules for the diffusion and partition/distribution in phases of different lipophilicity. Different approaches known in the literature to describe solvation of compounds with flexible conformation are discussed as well as the experimental methods to directly measure the energy of solvation. NSAIDs are chosen as an example of a class of drugs of different molecular structures that have already been studied thoroughly in many aspects. Thermodynamic characteristics of solvation of the drug molecules yielded by independent classical experimental methods (Gibbs energy, enthalpic and entropic terms of Gibbs energy) are used in order to better understand the diffusion and distribution properties. Correlations between in vitro data (the partition coefficient, enthalpy of solvation) with biopharmaceutically relevant characteristics (plasma half-life) are also discussed.

292

German L. Perlovich and Annette Bauer-Brandl

Keywords: NSAID, sublimation, solubility, solvation, calorimetry, thermodynamic functions, driving forces, crystal lattice packing architecture, hydrogen bond networks topology, plasma half-life, passive transport, distribution / partitioning.

Introduction Combinatorial chemistry, high throughput screening and in vitro receptor affinity studies nowadays confront the developer of new drug formulations with a huge number of potential new drug substances. Although receptor affinity is in many cases undoubtedly a key issue for potent drugs, other factors may be equally important for the usefulness in vivo, such as solubility, partitioning behavior, absorption properties, active and passive transport properties and biodegradation. In many cases, unfortunately, these important aspects are studied only late in the drug discovery / drug development process. Therefore, as most new drug candidates are in the first place barely tested in vitro, one faces this enormous amount of promising new drug substances of high receptor affinity with horrible physicochemical material properties. In many cases the substances are of extremely low solubility in any of the physiological media resulting in tremendous absorption and distribution problems. This should be a serious drawback for a drug candidate to become a useful drug preparation, because it is hard to compensate such bad properties even by using the most advanced drug delivery systems. It would be much more efficient and economical to pick those candidates with less difficult properties right in the beginning of the drug development process. Therefore, adding studies of those physicochemical and biophysical characteristics of the compounds, which may be biopharmaceutically relevant to the selection procedure and the optimization process of new drug candidates at an early phase of development, is an important issue today. It is the purpose of the present review to draw attention to physicochemical properties of importance in this context, and to conclusions that may be drawn from them.

1

Partitioning and diffusion

In the past, relative measures of the lipophilic–hydrophilic properties of drug molecules, like the logarithm of the partition coefficient between water and octanol, log P, have been used extensively [1]. The log P is a relative measure of the lipophilic–hydrophilic balance of a compound and characterizes the equilibrium of the partitioning process in a water–octanol system resulting from respective driving forces. It is a thermodynamic measure for processes that, however, are kinetically controlled in vivo: most of the transport and delivery processes in a biological environment take place under essentially non-equilibrium conditions and in non-homogeneous media. Water/octanol partitioning is frequently used as a simple model to mimic the distribution of drug substances in biological membranes, which is important because the pharmacokinetic properties of drug molecules are strongly dependent on their interaction with biological membranes. The conditions of absorption and distribution in biological systems are various and complicated, just to name the difference between trans- and paracellular pathways from the point of view of their lipophilic–hydrophilic composition [2–4]. Moreover, questions connected with the mechanism of diffusion into biological membranes (e.g., what controls the size of activation volume for passive transport? what is the nature and dimension of the units that are transferred by the diffusion process: is it the drug molecule or the drug molecule + its solvation shell?) have widely been outside the focus of discussion up to now [2, 4, 5, 6–8].

Solvation of Drugs as a Key for Understanding Partitioning and Passive Transport

293

boundary of phases phase 2

phase 1

a

b

c

partitioning Figure 1

1.1

Schematic illustration of the partition process.

The partitioning (distribution) process

The partitioning between water and octanol is the phase transition of the solute molecule across the partition boundary of the two separated phases (water and octanol) due to the thermodynamic driving force, which is defined by the difference of the chemical potentials of solute molecules dissolved, respectively, in water and in octanol. However, if one sticks to the thermodynamic formalism, a lot of questions appear about a model to describe the elementary act of the partitioning process. First of all, it is obvious that the solute molecules in the solution interact with their nearest neighbor solvent molecules in a stronger way compared to molecules further out in the bulk. Therefore, a so-called solvation shell is created around the solute molecule. Strength and nature of the drug–solvent interactions within this solvation shell are non-homogeneous (non-uniform): there are weak van-der-Waals interactions as well as stronger donor–acceptor interactions, hydrogen bonds (as the extreme case of a donor–acceptor interaction) and electrostatic interactions. Since most of the drugs are compounds with the ability to create hydrogen bonds, the molecules are solvated in solutions, particularly in aqueous solutions, by strong hydrogen bonds. During the partitioning/distribution process, the molecule solvated in water must rebuild this solvation shell into the octanol solvation shell (resolvation). One can imagine several variants of the resolvation mechanism which differ by the height of the activation energy for the partitioning (Fig. 1): (a)

Complete destruction of the water solvation shell and creation of the new shell in the octanol phase. This case would be preferred with weak drug–water hydrogen bonds; the fraction of the hydrogen bond energy in this case would be much less compared

294


to the van-der-Waals energy, in the case of big and topologically complicated molecules. (b)

Partly destroying the solvation shell during the phase transition step. For example, those water molecules forming strong hydrogen bonds with the drug molecule are kept within the solvation shell, and others are destroyed and replaced by interaction with octanol molecules.

(c)

The complete solvation shell transits without changes from the one phase into the other.

The last two variants ((b) and (c)) will be realized if the energy fraction of the hydrogen bonds is comparable with the van-der-Waals energy. Using the proposed classification, the effective size of molecules taking part in the partitioning process varies in size of the aggregate (solvation shell). As noted above, the main driving force of the partitioning process is the difference of the chemical potentials of the solute molecules in the considered phases, whereas the height of the activation barrier is determined both by thermodynamic and kinetic factors [9]. There is mutual solubility of water and octanol (which is important in the considered water–octanol systems), which will influence the height of the barrier of activation, and the size of the solvation shell as well. It may therefore be assumed that the height of the activation barrier of partitioning widely depends on the solvation characteristics of the solute (drug molecule). 1.2

Influence of the solution pH on the partition/distribution coefficients

Numerous drugs contain ionogenic functions and, dependent on the pH value of the solution and the strength of the acid, may exist in various degrees of ionization. In the literature a whole set of studies has been devoted to the investigation of the correlation between the pH value and the distribution coefficients (D = P7.4) and/or the partition coefficients (P) [10, 11]. The equation relating the partition coefficient, P, (for the unionized form of the compound) to the distribution coefficient, D, (for the ionized form) is well known: log D = log P – log (1 + 10pH–pK),

(1)

where pK is the acid strength. The D values are always smaller than the P values. This means that the activation barrier of partitioning is always higher for the ionic form of a drug compared with its nonionic form. This can be explained by large energy expenses for the resolvation of molecules in the ionic form during the transfer from the water into the octanol phase [2, 6, 7]. However, some authors [12, 13] describe the partitioning process as the transfer of the solvated form of the drug molecules from one phase into the other. This model concept seems reasonable, particularly taking into consideration that the saturated octanol phase (as in the partitioning experiments) contains more than 26 mol % of water [14]. Therefore, a close look on the solvation of molecules is useful to understand the nature of partitioning and passive transport. 1.3

Diffusion of drugs

Since the lipid membrane is in the liquid crystalline state [15, 16], the mechanism of diffusion of drugs into lipid membranes should be similar to diffusion processes into crystals (amorphous state). It is well known that the diffusion of molecules/atoms is controlled by

Lateral diffusion coefficient, cm2/sec


295

1

2

Molecular weight, Da Figure 2 Relationship between lateral diffusion coefficients and molecular weight of drugs. (Data from Johnson et al. [5].)

defects of the medium (i.e. the lattice of the solid/liquid crystal). A better model concept for the characterization of self-diffusion processes in the considered case could be the “void volume” (which is often used for the description of diffusion processes in liquids) or vacancies (bi-vacancy etc., which is used for solid/liquid crystals) [17–19]. The elementary diffusion step can be made if the following conditions are fulfilled: a) availability of a “void volume” of the size being comparable to the size of the diffusing particle (molecule, aggregate etc.); b) the activation energy of defects (“void volume”) should be enough in order to remove it; c) the activation energy of the diffused molecule should be high enough in order to partially destroy the interaction with the nearest neighbors (lipid matrix or octanol, for example) and to remove the molecule from the region where the defect of the structure is situated. Therefore, the activation energy of the diffusion process is a function of the Gibbs energy of solvation: ∆G* = f (∆Gsolv).

(2)

Consider the lateral diffusion coefficients of drug molecules into the lipid bilayers of the stratum corneum (SC), which have been calculated from permeability values [5]. The lateral diffusion coefficients versus the respective molecular weights are presented in Fig. 2 (data are obtained from the work by Johnson et al. [5]). These authors paid special attention to the bifunctional size dependence of the transport characteristics (diffusion coefficient) within SC. There is an overall dependence of the lateral diffusion coefficient on the molecular weight of the compound. However, if one picks compounds with a molecular weight of ≈ 120 Da (the dashed line in Fig. 2) as an example, the ratio of the maximum and minimum diffusion coefficients (points 1 and 2) may be approximately a factor of 200. This huge variation may be explained as follows. Firstly, probably a number of different diffusion mechanisms take part in the overall diffusion (i.e. different defects of the structure are present in the elementary diffusion step). Secondly,

296


aggregates (solvated drug molecules) of different size may take part in the elementary diffusion act. Thirdly, the studied drugs (at pH = 7.4) are ionized to a different degree and therefore build hydrated aggregates (solvates) of different sizes. The noted reasons are considered to be closely interrelated. Nevertheless, the conception of solvated drugs appears useful for explanation of the transport parameters.

2

Solvation of drugs: relevance and theoretical approaches

The main qualification of drug molecules to passively penetrate through membranes by diffusion is still acknowledged to be their lipophilic–hydrophilic balance. However, the diffusion rate depends strongly both on the concentration gradient and the energetic parameters of drug–membrane interaction. The diffusion process consists of a number of elementary activation steps, each having its own defined energy barrier. These energy barriers may be split up into two general forms: a) nonspecific drug–membrane interactions (van-der-Waals interactions) and b) specific interactions (hydrogen bonding, donor–acceptor interactions). The overall diffusion characteristics are a function of both the absolute strength of the mentioned interactions, and of the balance between them. In other words, the solvation properties of drug molecules, i.e. an understanding of the nature of interactions and the estimation of relative and absolute energetic terms thereof, are the key to understand the mechanism of not only passive transport, but also the mechanism of drug–receptor interactions. 2.1

The main definitions

The solvation of one mole of solute molecules in the solvent can be defined as the total change of the standard thermodynamic functions (∆G0, ∆H0, ∆S0) of the compound when transferring it from the gas phase (ideal gas, single molecules without interaction) into the solvent. The thermodynamic cycle of solvation is illustrated by Fig. 3, from which it follows that 0 0 0 ∆Ysolv = ∆Ysol − ∆Ysub ,

(3)

where ∆Y 0 is the standard change of any of the thermodynamic functions of the solvation (∆Y 0solv), dissolution (∆Y 0sol) and sublimation (∆Y 0sub) process. Therefore, the following equations may be defined: ∆G0solv = ∆G0sol – ∆G0sub ,

(4)

0 ∆H 0solv = ∆H 0sol – ∆H sub ,

(5)

T∆S 0solv = T∆S 0sol – T∆S 0sub .

(6)

In order to study the solvation process, which is experimentally not accessible, we need to investigate the other two processes: sublimation and dissolution.


297

Gas

Solid state

Solution

Dissolution

Figure 3 Thermodynamic cycle of the solvation process. Accessible surface ........ ........... ............ ............. .............. ............... ............... .............. .............. ........... .........

....... .......... ........... ............. .............. ............... ............... ............... .............. ............. ........... .........

Solvent probe

Excluded (Molecular) surface

Figure 4 Schematic picture illustrating an accessible surface (volume) and an excluded (molecular) surface (volume) introduced by Lee and Richards [20, 21].

2.2

Models describing the solvation of molecules

More than 30 years ago, Lee and Richards [20, 21] introduced the definitions of accessible volume and excluded volume (surface), which turned out to be very useful for describing various solvation effects. Pharmaceutics and medical chemistry are not exceptions to the broad use this approach has found, and therefore in the following paragraph we pay more attention to definitions and applications of this approach. The van der Waals molecular surface (SW) is the envelope surface of a set of intersecting spheres with given atomic radii centered on the nuclei of selected atoms of the molecule. The van der Waals volume (V W) is the volume enclosed by the van der Waals surface. The accessible molecular surface (S acc) is the surface defined by the center of the solvent, considered as a rigid sphere (probe sphere), when it rolls around the van der Waals surface. The accessible molecular volume (V acc) is the volume enclosed of the accessible molecular surface. The excluded (molecular) surface (S exc) is the surface envelope of the volume excluded to the solvent, considered as a rigid sphere, when it rolls around the van der Waals surface. The excluded (molecular) volume (V exc) is the volume enclosed by the excluded (molecular) surface (Fig. 4). As a rule, the presented approach has been used for calculation of hydration energy, where a water molecule is used as a “probe” solvent. Since the size of the water molecule is much smaller compared to a drug molecule, the assumption that it is spherical in shape

298


is reasonable. In some cases the noted approach is also applied to big solvent molecules (benzene, n-hexane), the size of which is comparable to the size of the solute molecule. In such cases various modifications to the model are being made: an ellipsoid (or cylinder, or any other geometric body) is used to describe the solvent molecule. It should be kept in mind that a lot of substances are topologically complicated and their conformation is flexible. The calculation of excluded (molecular) and accessible volumes (surfaces) of such molecules is very sensitive to the conformational state they are in. Therefore, different ways to calculate the mean statistical molecular conformational states have been suggested. For example, let us consider some modifications of the Lee’s and Richards’s approach: a)

The hydration shell model

Kang et al. [22–25] developed a model providing the opportunity to estimate Gibbs energies of hydration of conformationally flexible solute molecules. The main basic points of this approach are the following: the free energy of conformation i, ∆Gt(i), in solution can be represented by the equation: ∆Gt(i) = ∆E (i) + ∆Gh(i) ,

(7)

where ∆E (i) is the conformational energy, and ∆Gh(i) = Σk (Vwa,k(i ) ·∆gh,k)

(8)

is the free energy of hydration of the solute molecule in conformation i, expressed as the sum of the free energies of hydration of all its constituent atoms or groups k. Vwa,k(i) is the water accessible volume of group k in a given conformation i of the compound, and ∆gh,k is the free energy density of hydration of group k. In polyfunctional molecules, a correction term is added to account for the polarization of the hydration shell around polar groups by the presence of neighboring polar groups. The free energy of hydration of a compound is obtained as a Boltzmann weighted average over all conformations. In order to estimate ∆gh,k, free hydration energies of various classes of compounds have been used, which have been averaged for definite atoms (functional groups). b)

The dynamic polar molecular surface area (PSAd ) model

Out of the numerous useful models used to describe and predict various pharmacokinetic and pharmacodynamic properties of drugs, the model with the PSAd predictor should be mentioned, which has been developed by Palm et al. [7, 8] during the past years. The molecular surface area is divided into a polar fraction (nitrogen atoms, oxygen atoms and the hydrogen atoms attached to these heteroatoms) and a nonpolar fraction. The dynamic average of the surface properties is then calculated from all low-energy conformations according to a Boltzmann distribution at 37°C. In a first step, the conformational analysis of the drug has been carried out in vacuum, chloroform and water by Monte Carlo simulations. The low-energy conformations were exclusively chosen for further analysis. The authors noted that for conformational flexible molecules in vacuum and chloroform there is a wide spectrum of low-energy conformational states, whereas in water (hydrogen bonding solvent) the number of these variations is restricted. Therefore, water renders an essential influence on the stabilization of the conformation state of the drug molecule. As a next step,


299

-3.5 -4.0

logPc [cm⋅c-1]

-4.5 -5.0 -5.5 -6.0 -6.5 -7.0 30

40

50

60

70

80

90

100

110

120

PSAd, water [Å2]

Figure 5 Relationship between log Pc, the Caco-2 cell monolayer permeability coefficients and PSAd,water, dynamic polar molecular surface area of β-adrenoreceptor antagonists. (Data from Palm et al. [8].)

different low-energy conformations are compared. It appeared that the dispersion of these values was from 6 to 54% of the dynamic average (the last value corresponds to conformationally flexible drugs). The correlation analysis shows that there are linear dependencies between the logarithm of the Caco-2 cell monolayer permeability and PSAd values of water (Fig. 5) for a number of β-adrenoreceptor-blocking agents that have been chosen as a subject of investigations. The models presented above demonstrate attempts of estimating the thermodynamic functions of solvation of solute compounds. There may be experimental data on the solvation of some organic substances; however, for drug substances such data are practically absent. This is an essential obstacle for the development and improvement of models describing the transport characteristics of drugs from the point of view of the solvation phenomenon. In recent publications [26–31], in contrast to the equilibrium partitioning approach in water–octanol systems, we have tried to define an absolute lipophilicity scale for drug molecules of the group of non-steroid anti-inflammatory drugs, NSAIDs. The work is based on the analysis of drug solvation characteristics, where the experimental data were obtained using classical thermoanalytical methods: sublimation, solution calorimetry and the solubility method. This enables to us compare the solvation characteristics of a wide range of compounds using quantitative parameters derived from experimental data. The main results of the noted studies will be presented below.

3

Experimental methods to measure solvation characteristics and choice of subjects

3.1

Sublimation experiment

In order to study thermodynamic characteristics of the sublimation of molecular crystals, usually the Knudsen effusion method is used. In principle, this technique monitors the rate

300


of vapor loss through a small orifice under conditions of free molecular diffusion by gravimetry. However, during the last ten years, a transpiration (inert gas flow) method to investigate the sublimation process of various molecular crystals [32, 33] and NSAIDs [27–31] has been developed. The method has a number of advantages in comparison to effusion method. Firstly, the transpiration method works over a wider temperature interval, which increases the number of experimental data points and promotes a statistically better interpretation of the results. Secondly, the transpiration method works at lower temperatures (in comparison to the effusion method), which decreases the risk of chemical decomposition of the substance. It should be noted that both sublimation methods have practically not been used for drug substances except for the above mentioned and the following works: investigation of (±)-ibuprofen sublimation by Ertel et al. [34] applying the effusion method; and studies on the polymorphism of caffeine, theophylline and carbamazepine by Griesser et al. [35] using the transpiration method. In brief, the transpiration method can be described as follows. A stream of an inert gas passes above the sample at a constant temperature and at a known slow constant flow rate in order to achieve saturation of the carrier gas with the vapor of the substance under investigation. The vapor is condensed at some point downstream, and the mass of sublimate and its purity determined. The velocity of the carrier gas stream through the sublimation chamber should be chosen very carefully in order to establish and maintain the thermodynamic equilibrium between the solid and the gaseous state of the substance. The apparatus needs to be tested before starting the exact measurements by determining the relation between P and ν and choosing the gas flow velocity value adequate to the appearance of a plateau on P = f(ν) curve. The velocity of carrier gas flow for the considered compounds was 1.8 dm3/h. The equipment is calibrated using the benzoic acid standard substance. The standard value of sublimation enthalpy obtained should be in good agreement with the value recommended by IUPAC of ∆H 0sub = 89.7±0.5 J·mol –1 [36]. In order to gain valid results, the saturated vapor pressures should be measured 5 times at each temperature with the standard deviation being within 3–5%. The experimentally determined vapor pressure data can be described in (ln P; 1/T) coordinates by eq. (9): ln P = A + B/T.

(9)

The value of the enthalpy of sublimation is calculated by the Clausius–Clapeyron equation: ∆H Tsub = –RT2 ·∂(ln P)/∂(T),

(10)

whereas the entropy of sublimation at a given temperature T was calculated from the following relation: ∆S Tsub = (∆H Tsub – ∆G Tsub)/T,

(11)

where ∆G Tsub = –RT·ln(P/P0) and P0 = 1.013·105 Pa. In order to improve the extrapolation to room conditions, heat capacity (Cpc298 value) of the crystals was measured using a calibrated sapphire standard (Perkin Elmer). Heat capacity was introduced as a correction for the recalculation of the ∆H Tsub value at 298 K (∆H 298 sub value), according to eq. (12) [64]:

3.2


301

298 = ∆H T + ∆H T 298 ∆Ηsub sub cor = ∆H sub + (0.75 + 0.15·Cpc )·(T – 298.15)

(12)

Method of isothermal saturated solubility

In order to investigate solubility, usually the method of isothermal saturation is used. This is a routine method and has extensively been described in the literature (for example [37]). The Gibb’s energy of solution is derived from equilibrium conditions, i.e. solubility: ∆G 0sol = – RT·ln (X2),

(13)

where X2 is the mole fraction of solute molecules. The standard values of solution enthalpies for slowly dissolving compounds can be derived from the temperature dependencies of solubility by the equation: ∆H 0sol = –RT2 ·∂(ln X2)/∂(T),

(14)

However, in this case the thermodynamic functions of ∆G 0sol and ∆H 0sol have selfconsistent experimental errors and their respective dependence is difficult to interpret. In order to make discussion of the data more correct, one needs to measure ∆H 0sol values by a direct method (for example, calorimetry). 3.3

Isothermal calorimetry

Isothermal calorimetry provides the opportunity to measure heat effects of dissolution in various solvents with an accuracy of approximately 0.1% (depending on the total heat effect). As already has been mentioned above, it is important that this method and the isothermal saturated solubility are independent experiments. Therefore, also the experimental statistical errors obtained by the two methods are independent and not correlated. This fact enables us to analyze enthalpy entropy-compensation effects [38] both with respect to the solubility of definite compounds in a series of solvents, as well as in the form of a series of solutes in a particular solvent. The noted techniques have been discussed in more detail in [26, 27]. 3.4

Choice of drugs

Non-steroidal anti-inflammatory compounds (Table 1) are destined for solvation studies due to the following reasons. First of all, for the named substances, a lot of solubility, partitioning, distribution, X-ray structure, pharmacokinetic and pharmacodynamic experimental data are available in the literature. This fact essentially facilitates analysis and interpretation of the obtained results within a logical line. Secondly, the drugs have different basic structures: phenyl derivatives (BA, ASA, IBP); biphenyl derivatives (DIF, FBP); benzophenone substituted (KETO) and naphthalene substituted (NAP). Moreover, all the considered drug substances are carboxylic acids. In conclusion, comparison of solvation characteristics of a wide spectrum of compounds using quantitative thermodynamic parameters derived from experimental data is possible.

302


Table 1 Structural formulas of NSAIDs studied. Compound

Structural formula

Compound

Phenyl derivatives

Structural formula

Biphenyl derivatives

Benzoic acid (BA)

Diflunisal (DIF)

COOH

COOH OH

F F

Acetylsalicylic acid (ASA)

COOH

Flurbiprofen (FBP)

O O C

CH3

CH3

C

H

COOH F

Ibuprofen (IBP)

CH 3

H CH 3

C

CH 2

C

CH 3

COOH

Benzophenone derivative Ketoprofen (KETO)

H

Naphthalene derivative

O

CH3

C

C

H

Naproxen (NAP)

CH3

H

C COOH

COOH

H3CO

4

Crystal structures of NSAIDs

4.1

Description of hydrogen bond networks topology by graph set assignment

Hydrogen bond networks play an important role both in the creation of architecture and in the energetic parameters of crystal lattices. Therefore, before we analyze packing architectures of the drugs under investigation, let us consider an approach describing the topology of hydrogen bond networks by means of graph set assignment introduced by Etter [58]. Graph sets describe the topological nature of the hydrogen-bond pattern and the numbers of donors and acceptors involved while highlighting the common features of molecular aggregates that are not addressed by empirical formulae or by symmetry considerations of the crystallographic space group. A graph set is specified as Gda (r), where G is the pattern designator, a and d are the numbers of acceptors and donors, respectively, that are used in forming the hydrogen-bond pattern, and the degree r is the total number of atoms involved in the pattern. The pattern designator G describes the pattern of hydrogen bonding and can be one of four types: S, C, R and D. As illustrated in Scheme 1, S (or self) denotes an intramolecular hydrogen bond, C refers to infinite intermolecular hydrogen-bonded chains, R refers to intermolecular rings, and D refers to non-cyclic intermolecular diads and other finite hydrogen-bonded sets.

303


R O

R

H

O

O

O

H

H

O

R S(6)

R

R

H

C

O

R

C(2)

C

O

O

H

H

O

O

C

O R

S

R

D

R R2 (8) 2 Scheme 1

A hydrogen-bond pattern containing one unique type of hydrogen bond (distinguished from other types of hydrogen bonds by chemical nature and/or the symmetry relation of the donor and acceptor atoms used in the hydrogen bond) is referred to as a motif. Although each example in Scheme 1 contains only one motif, a graph set may define a single motif or it may define a pattern containing two or more motifs. A listing of all the hydrogen-bond motifs present in a given structure is known as the first level network. Hydrogen-bond patterns containing two different motifs are called second level networks. 4.2

Analysis of packing architectures of NSAIDs crystal lattices

The parameters of NSAIDs crystal lattices received from diffraction experiments are summarized in Table 2. Two molecules in both crystal lattices of (+)- and of (±)-Ibuprofen form a cyclic dimer through hydrogen bonds of their carboxylic groups (Fig. 6a,b). However, in the unit cell of (+)-IBP, both molecules are in the S configuration, and they are in different conformational

C12A

C11A

C5A

C10A C7A C8A

C13A

C9B

O2B

O1A

C6A C4A

C2A

C2B

C1A

C4B

C7B

C10B C11B

C5B C3B

C3A

C9A

C12B

C8B C1B

O2A

C6B

C13B

O1B

(a)

(b) b c a

C6 C12

C11

C13

C10

C7

O1

C5 C4

C8

C9

C2

C1 O2

C3 PowderCell 2.0

Figure 6 A perspective view of the cyclic dimer of (+)-Ibuprofen (a) and (±)-Ibuprofen (b); fragmentation of the molecule for calculations.

R 22 (8)

R 22 (8)

Graph set assign.

R 22 (8)

0.046

25 (2)

Cu Kα

1.400

4

854.2 (4)

90.00

95.68 (1)

90.00

11.395 (2)

6.591 (1)

11.430 (1)

P21/c

monoclinic

R 22 (8)

0.0378

25 (2)

Mo Kα

1.293

2

624.7 (6)

106.94 (3)

107.28 (4)

82.99 (4)

5.786 (5)

12.721(4)

9.299 (3)

P1

triclinic

Perlovich et al.(e)

FBP

R 22 (8); S (5)

0.045

25 (2)

Mo Kα

1.324

8

2519.4 (4)

90.00

110.57 (2)

90.00

20.737 (4)

3.743 (1)

34.666 (6)

C2/c

monoclinic

Kim et al.(f)

DIF

R 22 (8)

0.0661

10–33

Mo Kα

1.28

2

657.639

88.78 (4)

94.56 (4)

89.61 (3)

6.136 (2)

7.741 (3)

13.893 (8)

P1

triclinic

Briard et al.(g)

KETO

C (4)

0.042

10–33

Mo Kα

1.25

2

611.7 (3)

90

93.91 (3)

90

7.914 (3)

5.793 (2)

13.375 (5)

P21

monoclinic

Kim et al.(h)

NAP

In brackets, standard deviations; (b)(CSD-IBPRAC02) Ref. [59]; (c)Ref. [31]; (d) Ref. [60]; (e) Ref. [50]; (f)Ref. [61]; (g)(CSD-KEMRUP) Ref. [62]; (CSD-COYRUD11) Ref. [63].

(a) (h)

0.0385

0.077

R[F2 >2σ (F 2)]

25 (2)

–173

T, °C

Mo Kα

1.098

4

1248.2 (5)

90.00

112.86 (2)

90.00

13.533 (3)

8.0362 (11)

12.456 (4)

P21

monoclinic

pulsed neutron

1.175

ASA

Perlovich et al.(c) Kim et al.(d)

(+)-IBP

Radiation

Dcalc,

4

Z

g⋅cm–3

90.00

1165.6 (11)

99.70 (3)

β, °

Volume, Å3

90.00

α, °

γ, °

7.818 (4)

10.506 (6)

b, Å

14.397 (8)

a, Å

c, Å

P21/c

monoclinic

Crystal system

Space group

Shankland et al.(b)

(±)-IBP

Crystal lattice parameters of some NSAIDs(a).

Crystal data

Table 2

304 German L. Perlovich and Annette Bauer-Brandl


305

states, in contrast to (±)-IBP, where the dimer is formed by hydrogen bonds across the center of inversion (space group P21/c), one molecule being R and the other S. Therefore, the topology of the hydrogen bonds networks of both ibuprofens can be described by R22 (8) graph set assignment. Some special characteristics of the conformational states and geometry of hydrogen bonds of both (+)- and (±)-IBP are presented in Table 3. As the two molecules of (+)-IBP are situated in the asymmetric unit, the geometries of the two hydrogen bonds are not equivalent: one of them is shorter than the other. Comparison of the data leads to the following conclusions: a) the angle of the hydrogen bonding of the racemate is more planar in comparison to the enantiomer; b) the D…A distance of one of the hydrogen bonds of the (+)-IBP is approximately equal (within experimental errors) to the analogous value of the enantiomer, whereas the second hydrogen bond is longer; c) the H…A distance of the (+)-IBP is the average of the analogous values of the enantiomer. The conformational states of the enantiomer molecules in the asymmetric unit cell are different. As it follows from Table 3, one of the two molecules in the (+)-IBP asymmetric cell has approximately the same conformational state as in the racemic IBP crystal. Table 3 Some parameters, which characterize the conformational states and hydrogen bond geometry of (+)- and of (±)-ibuprofen molecules in the crystal lattice. (+)-IBP

∠C5-C4-C2-C3 [°] ∠C7-C10-C11-C12 [°] ∠C4-C2-C1-O1 [°] O2-C1 [Å] O1-C1 [Å] C1-C2 [Å]

(±)-IBP

A

B

144.4 (4) –67.9 (5) 81.7 (4) 1.219 (3) 1.302 (4) 1.496 (5)

–29.1 (4) 68.0 (5) –83.5 (3) 1.226 (3) 1.302 (4) 1.518 (4)

140.9 (4) –67.3 (4) 88.7 (3) 1.222 (3) 1.305 (3) 1.509 (3)

Hydrogen bond geometry D⎯H…A (+)-IBP (A) (+)-IBP (B) (±)-IBP (a) Symmetry

O1A-H1AO…O2B O1B-H1BO…O2A O1-H1O…O2(a)

D⎯H [Å]

H…A [Å]

D…A [Å]

D⎯H…A [°]

0.94 (5) 1.07 (5) 0.963 (13)

1.73 (6) 1.58 (5) 1.664 (10)

2.651 (4) 2.634 (4) 2.627 (7)

169 (5) 168 (4) 179.5 (7)

code: 1-x,1-y,1-z

The packing architectures of flurbiprofen, ketoprofen and acetylsalicylic acid crystal structures are shown in Fig. 7a,b and c, respectively. Both FBP and KETO structures have the same symmetry of the crystal lattices – triclinic, whereas ASA has the monoclinic symmetry. However, the molecules of outlined compounds form dimers packing architecture. Therefore, the hydrogen bond networks can be described by R22 (8) graph set assignment. It should be noted that for FBP and KETO the dimers in the crystal lattices have conformational states like a “chair”. Projection of the naproxen crystal lattice along the OX axis is presented in Fig. 8a. In contrast to the previous structures, the hydrogen bonds create infinite chains (helicoids)

306


(a)

(b)

(c) Figure 7 The packing architecture of flurbiprofen (a), ketoprofen (b) and acetylsalicylic acid (c) crystal structures.

(a) Figure 8

(b) The packing architecture of naproxen (a) and diflunisal (b).

with graph set assignment C (4) along the two-fold screw axis. The packing architecture of diflunisal is shown in Fig. 8b and may be characterized by motifs R22 (8); S (6).


307

with graph set assignment C (4) along the two-fold screw axis. The packing architecture of diflunisal is shown in Fig. 8b and may be characterized by motifs R22 (8); S (6). Based on the carried out analysis, it can be concluded that the topology of hydrogen bond networks in the crystal lattices of the outlined class of compounds differs. For example, ASA, (+)- and (±)-IBP, KETO and FBP molecules are packed with only dimer aggregates with the graph set assignment R22 (8). The DIF molecules in the crystal lattice create additional multicenter hydrogen bonds (combination of inter- and intramolecular hydrogen bonds). Due to this fact, the topology of hydrogen bond networks has a higher level, including several motifs: R22 (8); S (6). NAP molecules form only a C (4) motif. In order to find out correlations between the crystal lattice parameters and the thermochemical and thermodynamic characteristics, let us consider some experimental methods and approaches which give an opportunity to analyze the outlined functions correctly.

5

Thermochemical and thermodynamic properties of NSAIDs

5.1

Thermodynamic characteristics of sublimation of NSAIDs

The saturated vapor pressure data and thermodynamic functions of sublimation, fusion and vaporization processes calculated from the experimental data are summarized in Tables 4 and 5, respectively. Table 4 Coefficients and correlation parameters of the regression equation ln (P [Pa]) = A + B/T for the studied NSAIDs. ASA

(+)-IBP

(±)-IBP

FBP

DIF

NAP

[t1–t2](a) 40.0–89.0 32.0–45.0 40.0–67.0 68.5–104.5 76.0–137.0 68.0 –124.0 A 38.2±0.2 38.1±0.2 40.4±0.2 33.8±0.2 36.4±0.2 39.7±0.2 13190±65 12920±60 13927±73 13040±60 14400±800 15431±65 –B 0.999 0.999 0.999 0.9998 0.9997 0.99987 r (b) σ (c) 3.74·10–2 9.1·10–3 2.54·10–2 1.62·10–2 3.72·10–2 3.22·10–2 F (d) 41808 48662 36648 52994 36261 58613 23 14 25 10 16 15 n(e)

KETO 68.0–91.5 33.0±0.2 13250±60 0.999 1.35·10–2 47818 16

(a)experimental temperature interval, °C; (b)pair correlation coefficient; (c)standard deviation; (d)calculated Fisher distribution value; (e)number of experimental points.

The Gibbs energy of the sublimation process at room temperature can be separated into relative fractions of the enthalpic and entropic terms by the following parameters: 298 298 298 ς H = (∆H sub /(∆H sub + T ∆Ssub )) ⋅ 100%,

(15)

298 298 298 ς TS = (T ∆Ssub /(∆H sub + T ∆Ssub )) ⋅ 100%.

(16)

Results of these calculations are also shown in Table 5. It is not difficult to see that in all studied cases the sublimation process is an enthalpy controlled process because the enthalpy exceeds the entropy by a factor of approximately 1.8. However, the relative fraction

298

298

21.7 72.9 88.5

13.6

45.8

76.9

298

30.2±0.2

(e)

141.0±0.5(e)

18.1±0.2

122.3±0.5

298

66.1

298

93.5

48

14.2

15.4±0.4

(h)

50.3±0.4(h)

222±2 62.0 38.0

66.6

223±2 62.3 37.7

56.1

188 ±2 61.7 38.3

41.6 107.7±0.5

43.6

(+)-IBP(g)

110.2±0.5

90.5± 0.3

34.4

ASA(d)

298

96.2

67

19.8

23.1±0.4

[49]; (k)Ref. [50]; [29]; (m)Ref. [51]; (n)Ref. [28]; (o)Ref.[52]

(j)Ref.

T

T

(q)∆ H 298 fus

= ∆ H fus – ∆ S fus (Tm – 298.15)

298

298

(p)∆ H 298 vap

=∆ H sub −∆ H fus

[32]; (d)Ref. [26]; (e)Ref. [47]; (f)Ref. [48]; (g)Ref. [31]; (h)Ref. [42]; (i)Ref. [27];

= ∆Hfus/T(f);

(c)Ref.

(l)Ref.

(f)

74.0±0.4(f)

241±2 61.8 38.2

71.8

116.0±0.6

44.2

(±)-IBP(g)

= (∆ H sub /(∆ H sub + T∆ S sub ))·100%; ςTS = (T∆ S sub /(∆ H sub +T∆ S sub ))·100%;

fus

H

(b)∆S

(a)ς

298 ∆ H fus [kJ·mol–1] 298 ∆ H fus [kJ·mol–1](q) 298 ∆ S fus [J·mol–1 ·K–1](b) 298 ∆ H vap [kJ·mol–1]

T (f) [°C]

298 ∆ H sub [kJ·mol–1] 298 T⋅∆ S sub [kJ·mol–1] 298 ∆ S sub [J·mol–1 ·K–1] ςH [%](a) ςTS [%](a)

BA(c)

89.9

68.3

20.4

26.4±0.5

(k)

113.5±0.2(k)

191±1 65.9 34.1

56.9

110.2±0.5

53.3

FBP(i)

Thermodynamic characteristics of processes of sublimation, fusion and vaporization of NSAIDs.

∆ G sub [kJ·mol–1]

298

Table 5

98.1

73.9

22.0

35.9±0.2

(j)

212.8±0.2(j)

210±2 65.8 34.2

62.5

120.1±0.6

57.6

DIF(i)

(m)

108.1

74

22.0

31.5±2.1

154.4(m)

240±2 64.5 35.5

71.6

130.1±0.5

58.5

NAP(l)

94.8

57

17.1

21.0±0.8(o)

93.9±1.3(o)

184±1 67.1 32.9

54.9

111.9±0.5

57.0

KETO(n)

308 German L. Perlovich and Annette Bauer-Brandl


309

of the enthalpic term varies between 61.7% for benzoic acid and 67.1% for ketoprofen. For a better comparison, benzene [39], biphenyl [39], benzophenone [40] and naphthalene [41] as the non-substituted analogues of the studied NSAIDs (which in their crystal lattices exclusively interact non-specifically by van-der-Waals forces) were investigated. The value of solid state benzene ςH (Ben) = 53% differs significantly from the benzene derivatives of the NSAIDs (ςH (BA) = ςH (ASA) = ςH ((+)-IBP) = ςH ((±)-IBP) ≈ 62%). The analogous value of biphenyl ςH (BiPh) = 60% differs also from the biphenyl substituted compounds: ςH (FBP) = ςH (DIF) ≈ 66%. The same situation is observed for naphthalene ςH (Naph) = 59% and naproxen ςH (NAP) ≈ 65% and also for benzophenone ςH (BenzPhen) = 62% and ketoprofen ςH (KETO) ≈ 67%. The causes of the non-systematic share between enthalpy and entropy with respect to the skeletal structures are supposed to be a different distribution of the total crystal lattice energy between van-der-Waals interactions and hydrogen bonds depending on the functional groups in the molecules. 5.1.1 Differences of racemate and enantiomer ibuprofen crystal lattices It is interesting to note that the difference between the Gibbs energies of (±) and (+)-IBP at 25°C is 2.6 kJ·mol –1 (Table 5). This value practically coincides with the value of heat fluctuation, RT = 2.5 kJ·mol –1. This fact once more confirms that the problem of separation of the enantiomers is very delicate. Using the thermodynamic cycle (whilst neglecting the differences of the heat capacities between the racemate and chiral substances which is supposed to be very small), the thermodynamic functions of evaporation of both (±)- and (+)-IBP may be estimated as follows: 298 298 298 ∆H vap = ∆H sub − ∆H fus ,

(17)

298 298 298 ∆S vap = ∆Ssub − ∆Sfus .

(18)

298 for (±)-IBP is slight higher It is not difficult to see (Table 5) that the value of ∆H vap than that for (+)-IBP. The standard value of the entropy of sublimation of (±)-IBP at 298 K exceeds that of (+)-IBP by 19 J·mol –1 ·K –1. The difference between the entropies of fusion at 298 K between the racemate and the enantiomer is also 19 J·mol –1 ·K –1 (Table 5). Therefore, the entropy of evaporation for both compounds considered coincides at 174 J·mol –1 ·K –1. If we take into account the thermodynamic cycle presented in Scheme 2 then the difference between the entropies of (±) and (+)-IBP crystal lattices may be calculated as:

∆∆S = ∆Ssub (± ) − ∆Ssub (+ ) − R ln 2 = 13 J ⋅ mol-1 ⋅ K -1 .

(19)

This value quantifies the difference between the entropies of the crystal lattices of the racemate and the enantiomer and is only caused by particularities of the respective crystal lattice structures. For a deeper understanding of the nature of interaction of IBP molecules both in the racemate and the enantiomer in the crystal lattice, the packing energies are calculated. For this purpose X-ray data for (+)-IBP obtained by us [31] and the neutron diffraction data for (±)-IBP from Shankland et al. [59] (refcode CSD – IBPRAC02) were used. Molecular crystals consist of discrete molecules, which interact with each other by intermolecular oooooooo

310


chiral gas

racemic gas Rln2

∆Ssub((+)-IBP)

1/2 (+)-IBP 1/2 (-)-IBP chiral crystal

∆Ssub(IBP)

∆∆S racemic IBP crystal

Scheme 2

non-bonded interactions. Therefore the crystal lattice energy, Elatt, may conditionally be divided into three main terms: van der Waals, E vdw; electrostatic (Coulombic), E coul; and hydrogen bonds energy, E HB: E latt = E vdw + E coul + E HB .

(20)

The results of calculations of the energetic terms of the (±)- and the (+)-IBP crystal lattices for both the Mayo et al. [65] (M) and Gavezzotti [66] (G) force field are presented in Table 6. As it follows from Table 6, the van der Waals terms of both (+)- and (±)-IBP are approximately equal if calculated by the same force field. It should be noted that this value is slightly higher (absolute value) for the G force field compared to the analogous value for the M force field. In contrast, the terms related to energies of hydrogen bonds show an opposite trend: for the G force fields the term on approximately 3 kJ·mol –1 less than for the M force field. The opposite trends sum up to approximately the same total values of the Table 6 The calculation results of the various energetic terms of (+)- and (±)-ibuprofen crystal lattices obtained by the two types of the force fields (Mayo et al. [65] and Gavezzotti et al. [66]). Terms(a)

(±)-IBP

(+)-IBP

∆((±)-(+))

–78.0 (71.0) –2.8 (2.6) –29.0 (26.4) –109.8

–78.8 (72.8) –2.5 (2.3) –26.9 (24.9) –108.2

0.8 –0.3 –2.1 –1.6

Evdw Ecoul EHB Elatt

–71.7 (65.2) –2.8 (2.5) –35.5 (32.3) –110.0

–73.1 (68.2) –2.5 (2.1) –31.9 (29.7) –107.5

H sub

116.0 ±0.6

107.7±0.5

Gavezzotti et al. Evdw Ecoul EHB Elatt Mayo et al.

298

(a)[kJ⋅mol–1];

in brackets E term/Elatt in % is presented.

1.4 –0.3 –3.6 –2.5 8.3


311

- (+)-IBP - (±)-IBP

30

EX-X / E

vdw

[%]

40

20

10

0

H-H

C-H

H-O

C-C

C-O

O-O

Types of interactions

Figure 9 The energetic terms of different types of nonbonded van der Waals interactions of the (+)and (±)-ibuprofen crystal lattices (Mayo et al. force field).

crystal lattice energies obtained by the two force fields considered. It should be mentioned that for (±)-IBP the ratio between hydrogen bonding energy and common crystal lattice energy is sensitive to the choice of the force field: 26.4% for the G force field and 32.3% for the M. The same tendency is observed for (+)-IBP: 24.9% for G and 29.7% for M. It should also be noted that the van der Waals term of (+)-IBP is higher (absolute value) compared to the (±)-IBP irrespective of the force field used. Therefore, it may be assumed that for (±)-IBP the loss of van der Waals energy (in comparison with (+)-IBP) is compensated by hydrogen bonding energy. Probably, these two energies are competing when enantiomer and/or racemate crystals are growing. Comparative analysis of the energetic terms of different types of nonbonded van der Waals interactions for the crystal lattices considered (Mayo et al. force field) was carried out. The results thereof are presented in Fig. 9. As it follows from Fig. 9, both for (+)- and (±)-IBP, the dominating contributions are interactions between: C–C > C–H > C–O. Moreover, a transition from (+)- to (±)-IBP makes the relative contributions of the following terms decrease: C–H, C–C and H–O, while a slight increase of the H–H, C–O and O–O interaction is noted. Because the contribution of the C–H interactions for both compounds contribute by more than 25% to the common energy of the nonbonded van der Waals interactions, it should be expected that if the positions of the hydrogen atoms in the unit cells could be resolved more accurately by neutron diffraction experiments, the accuracy of the final result should be essentially increased. Lacking these data, we tried to estimate the influence of the C–H distance on the van der Waals term of the crystal lattice energy. For this purpose, in the calculation procedure, only the C–H distance is changed (from 0.95 to 1.20 Å), and the same coordinates of the other atoms in the unit cell retained. The results of the calculation are presented in Fig. 10. In Fig. 10 the filled symbols mark the E vdw values corresponding to the C–H bonds obtained from diffraction experiments. From Fig. 10, it follows that within the noted variation interval of the C–H bond lengths, the E vdw values for (+)-IBP are changed within 4.5 kJ·mol –1, whereas for (±)-IBP – within 17.6 kJ·mol –1. This fact confirms once more that the van der Waals energy of the enantiomer is approximately two times less sensitive to C–H bond variations (6.2%) in comparison to the racemate (24.4%). This

312

German L. Perlovich and Annette Bauer-Brandl (±)-IBP

-54 -56

Evdw [kJ⋅mol-1]

-58 -60 -62 -64 -66 (+)-IBP

-68 -70 -72 -74 0.95

1.00

1.05

1.10

1.15

1.20

RC-H [Å]

Figure 10 The dependence of the van der Waals term of the crystal lattice energy on the length of the C–H bond (the filled symbols mark the Evdw values corresponding to the C–H bonds obtained from the diffraction experiments).

fact also stresses that the estimation of crystal lattice energy for (+)-IBP does not too much depend on the accuracy of hydrogen atoms coordinates (by X-ray or neutron diffraction). 5.2

Thermochemical characteristics of NSAIDs

The thermodynamic functions of fusion and vaporization processes for the NSAIDs studied are presented in Table 5. In the next step of the investigations we tried to analyze the influence of the topology of hydrogen bond networks on the thermodynamic functions of the fusion process. The dependence of the fusion entropies on the fusion enthalpies of the compounds discussed is shown in Fig. 11. It should be noted that the drugs with the complex structure of hydrogen bond network deviate from the common trend line, with a tendency of decreased fusion entropy. Probably, the hydrogen bond networks are kept maximally in the liquid state after the fusion process. Substances situated on the trend line, probably, realize the degrees of freedom completely (if one takes into account the hydrogen bond networks in the melt). The compounds, which deviate from the trend line to the side of fusion entropy increase, are conformationally more flexible in comparison with the others. Probably, in the fusion process, as the hydrogen bonds are broken off, additional degrees of freedom appear due to conformational (structural) disordering. In order to find out a regularity between the characteristics of the fusion process and crystal lattice parameters, we calculated the free molecular volumes, V free, of the considered drugs in the crystal lattices. For this purpose, the following algorithm was used. The unit cell volumes, Vcell, were estimated based on X-ray diffraction experiments. After them, the van der Waals molecular volumes, V vdw, into the crystal lattice by GEPOL package [67] and Kitaigorodsky radii [68] were calculated and the free volumes were derived by the equation: V free = (Vcell – Z·V vdw )/Z,

(21)


313

Figure 11 Dependence of fusion entropy on fusion enthalpy of drugs (the graph set assignment is shown under the drug name). DIF: Diflunisal; NAP: Naproxen; IBP: Ibuprofen; ASA: Acetylsalicylic acid; FBP: Flurbiporfen; Keto: Ketoprofen; SA: Salicylamide; i-OH-BA: i-hydroxybenzoic acids; Me-, Et-, Pr-, BuPB: Parabens.

Figure 12 Dependences of melting points on molecular free volumes in crystal lattices (symbols correspond to Fig 11).

where Z is the number of molecules (structural units) in unit cell. Dependences of the melting points on molecular free volumes in the crystal lattices of drugs under investigation are presented in Fig. 12. It is not difficult to see a regularity: an

314


increase of V free values leads to a decrease of the melting temperature, with an essential exception for naproxen due to structural disordering of the molecules in respective crystal lattices.

6

The difference between partitioning and distribution of NSAIDs from the thermodynamic point of view

The thermodynamic cycle of the relationships between the thermodynamic parameters of a drug molecule HD and its dissociate D – + H + is shown in Scheme 3. The thermodynamic parameters of solution and solvation are presented in Tables 7 and 8. HD(gas)

∆Hsolv ∆Hsolv

∆Hsub HD(buffer pH 2.0) ∆Hsol

HD(solid)

∆Hdep ∆Hsol

(D- + H+) (buffer pH 7.4

Scheme 3 Table 7 Coefficients and correlation parameters of the regression equation ln (X2) = A + B/T for the studied NSAIDs in buffers with pH 2.0 and pH 7.4. (+)-IBP

(±)-IBP

[t1°C – t2°C](a) 20.0 − 42.0 20.0 − 42.0

FBP

DIF

NAP

KETO

20.0 − 42.0

20.0 − 42.0

20.0 − 42.0 20.0 − 42.0

pH 2.0 –A –B r(b)

σ(c)

N(d)

2.3±0.2 2874±76 0.9990 1.45·10–2 5

1.0±0.3 3518±101 0.9988 1.24·10–2 5

–2.9±0.2 5195±65 0.9998 1.24·10–2 5

6.0±0.4 2582 ±118 0.9969 2.27·10–2 5

2.33±0.02 3367±6 0.9999 1.09·10–3 5

–1.8±0.2 4123±48 0.9998 9.26·10–3 5

6.0±0.2 1670±55 0.9984 1.06·10–2 5

4.1±0.2 2345±48 0.9994 9.22·10–3 5

8.8±0.2 811±54 0.9934 1.04·10–2 5

8.6±0.2 952±62 0.9938 1.19·10–2 5

9.08±0.04 628±13 0.9994 2.41·10–2 5

4.4±0.1 1929±32 0.9996 6.22·10–3 5

pH 7.4 –A –B r(b)

σ(c) n(e)

(a) experimental temperature interval; (b) (d) number of the experimental points

pair correlation coefficient; (c) standard deviation;

(a)accuracy,

0

53.9 46.1

53.8 46.2

–10.3 –34.5±1.5 14.4 96.3±1.0 81.9 274.7 ±3.3 54.0 46.0

–14.9

–50.0±1.0

12.8

93.5±1.0

80.7

270.7 ±3.3

53.7 46.3 0

19.5±0.4

13.9 ± 0.5

29.8

248.2±4.0

240.1±3.7

28.8

74.0

71.6

0

57.3 42.7

278.4±3.7

83.0

111.4±1.1

28.4

–71.4±1.7

–21.3

7.9±0.5

29.2

56.1 43.9

256.6±5.4

76.5

97.8±1.6

21.3

12.5 86.5±1.2

11.9

83.5±1.1

–49.6±3.4

–8.0±2.7

–19.5±2.1

–14.8

21.5±1.0

36.3

–2.4

29.3±0.8 (9.6)(e)

(34.85)(e)

DIF

56.9 43.1 0

258.6±3.4

77.1

101.7±1.0

24.6

–73.8±1.7

–22.0

6.7±0.5

28.7

57.6 42.4

160.7±3.4

47.9

65.2±1.0

17.3

24.1±1.7

7.2

43.2±0.5 (12.5)(e)

36.0 (35.7)(e)

FBP

0

59.5 40.5 0

215.0±2.7

64.1

94.1±0.8

30.0

–36.9±1.0

–11.0

16.0± 0.3

27.0

60.9 39.1

163.0±3.0

48.6

75.8±0.9

27.2

15.1±1.3

4.5

34.3±0.4 (26.4)(e)

29.8 (28.5)(e)

KETO

57.1 42.9

309.6 ±2.0

92.3

123.1±0.6

30.8

–75.5±0.4

–22.5

5.2± 0.1

27.7

57.0 43.0

253.6±2.0

75.6

100.3±0.6

24.7

–19.5±0.4

–5.8

28.0±0.1 (21.3)(e)

33.8 (33.72)(e)

NAP

2 %; (b)Ref. [77]; (c)ςH = (⏐ ∆H solv ⏐/(⏐ ∆H solv ⏐+⏐ T∆S solv ⏐)) ·100%; (d)ςTS = (⏐ T∆S solv ⏐/(⏐ ∆H solv ⏐+⏐ T∆S solv ⏐)) ·100%; (e)Ref. [53].

0 ∆H solv [kJ·mol–1] 0 T∆S solv [kJ·mol–1] 0 ∆S solv [J· K–1 ·mol–1] 0 – ∆G solv [kJ·mol–1](b) 0 – ∆H solv [kJ·mol–1] 0 – T∆S solv [kJ·mol–1] 0 – ∆S solv [J⋅K–1 ·mol–1] [%](c) ςHsolv ςTSsolv [%](d)

∆G solv [kJ· mol–1](a)

0

pH 7.4

0 T∆S solv [kJ·mol–1] 0 ∆S solv [J·K–1 ·mol–1] 0 – ∆G solv [kJ · mol–1](b) 0 – ∆H solv [kJ · mol–1] 0 – T∆S solv [kJ · mol–1] 0 – ∆S solv [J·K–1 ·mol–1] ςHsolv [%](c) ςTSsolv [%](d)

31.7

(±)-IBP

–5.8

23.9±0.6

∆H solv [kJ· mol–1]

0

29.7

(+)-IBP

Thermodynamic characteristics of solubility and solvation processes of some NSAIDs in aqueous buffers at pH 2.0 and 7.4 at 25 °C.

∆G solv [kJ · mol–1](a)

0

pH 2.0

Table 8


315

316

6.1


Solvation characteristics of dissociated and non-dissociated (+)- and (±)-IBP

The respective solution enthalpies, ∆H 0sol , were calculated using the van’t Hoff relationship, which – in contrast to other works [69] – was found to be satisfactorily linear. Further it was found that the dissolution of ibuprofen in the buffers, both the racemate and the pure enantiomer, is endothermic (Table 8). Moreover, the values of the entropy of the dissolution process, ∆S 0sol , (calculated from solubility and enthalpy) are negative. Probably, while a molecule is transferred from the solid state into the solution, some structure is built in the solvation shell and in the surrounding solvent, overcompensating the increase in entropy caused by the dissolution due to a “hydrophobic effect” [70–73] of solvation. Analysing the solvation process in more detail (Table 8), in both buffers the solvation is found to be exergonic, and Gibbs energy of solvation, ∆G 0solv , to comprise negative values for both its enthalpic and entropic terms, ∆H 0solv , and ∆S 0solv . In the current case, the main driving force of solvation is enthalpy, which is regarded as a “classical” hydrophobic interaction as the mechanism of solvation [70–73]. The significance of entropy for the solvation process, which is decreasing and working in the opposite direction by probably creating solvent cages around the solute molecules, is not much smaller than enthalpy at room temperature. When further comparing the racemate with the pure enantiomer, all the (absolute) values of the thermodynamic solvation functions for (+)-IBP are slightly smaller (taking into account the experimental errors) compared to (±)-IBP, for both the dissociated and non-dissociated form. This means that the solvation of racemic IBP molecules is slightly stronger compared to the pure enantiomer. This behaviour is probably connected with a difference in the molecular association states for the racemate and the pure enantiomer in solutions. The molecules may be exposed to interaction of neighboring molecules /solvation shells present in the buffers, like dimers (or multimers). These would probably be of different symmetry in the case of the racemate compared to the pure enantiomer (similarly to the symmetry of dimers in crystals). Different symmetry determines small variations of the structure of the solvation shells and their thermodynamic characteristics: the more symmetrical (±)-IBP dimer has a stronger ability of solvation. However, the solubility is considerably higher (by approx. a factor of 2) for the pure enantiomer than for the racemate, which means that the distance to neighboring molecules is smaller. It is difficult to decide whether this effect accounts for the difference in solvation energy, considering the generally very low solubility. Comparison of the enthalpy values in the respective solutions of different pH, as presented in Scheme 3 and Table 8, enables one to calculate the enthalpy of deprotonation of the molecules. Deprotonation is exothermic, the absolute value of enthalpy of protonation /deprotonation is 10 kJ·mol –1, coinciding within experimental error both for (+)- and (±)-IBP. It should be noted that the ionic state of molecules in general is more important for the solvation thermodynamics compared to non-ionic molecules (Table 8). This difference is higher than the differences between the racemate and the pure enantiomer at the same pH. The total solvation abilities of (+)- and (±)-IBP for the different states of protonation over0 0 0 0 ((±)- IBP-)|. Particlap: | ∆Ysolv ((+)-IBP)| < | ∆Ysolv ((±)-IBP)| < | ∆Ysolv ((+)-IBP-)| < | ∆Ysolv ularly for the entropy, the protonated / deprotonated state of the molecule is of significance. Therefore, it may be speculated that also the interaction of the solvated IBP-molecules with membranes and receptors is widely dependent on the properties of the surrounding medium in terms of pH, possibly also on ion strength and composition.


317

- pH 2.0 - pH 7.4 60

50

ζΗ [ %]

40

30

20

10

0

(+)-IBP

IBP

DIF

FBP

KETO

NAP

Figure 13 Comparative analysis of the relative enthalpic parameter, ςH, of the studied drugs for buffers with pH 2.0 and 7.4.

6.2

Solvation characteristics of dissociated and non-dissociated forms of the other NSAIDs

As it follows from Table 7 and as is shown in Table 8, dissolution in aqueous buffers both at pH 2.0 and pH 7.4 is endothermic for all drugs studied. This is evidence for solvation enthalpies not overweighing the respective crystal lattice energies (as was discussed above for IBP). The entropies of dissolution are, as a rule (with the exception of FBP and KETO at pH 2), negative. Therefore, the degree of order in the solvation shells and in the solvent structure increases (i.e. there is a hydrophobic effect). In all cases, the enthalpic term of the Gibbs energy of solvation exceeds the entropic term. The relative enthalpic parameter (ςHsolv) is between 53.8% (for (+)-IBP at pH 2.0) and 60.9% (for KETO at the same pH). The comparative analysis of ςHsolv values of the studied drugs for the two buffers considered is shown in Fig. 13. It is not difficult to see that the noted parameter is sensitive to the variation of buffers only for KETO, DIF and FBP. It should be mentioned that for every drug under investigation and for the two buffers considered the solvation process is an enthalpy-driven process. It is interesting to note that the enthalpy of transition from pH 7.4 to 2.0, ∆Htr (pH 7.4 → pH 2.0), which characterises the protonation process, is endothermic in all cases. The values vary from a minimum for (±)-IBP (9.8 kJ·mol –1) to a maximum for FBP (36.5 kJ·mol –1) by a factor of more than three. All these values exceed by far the enthalpy of proton ionisation in dilute aqueous solutions, which have been reported for some aromatic acids [74]. Probably, in the present case, the solvation effects play an essential role in the transfer of molecules from one buffer to the other. It can be assumed that fluorine atoms (as an electron acceptor) in molecules of diflunisal and flurbiprofen induce an essential redistribution of the electron density from the COO – group to the phenyl ring (by conjugation effects). As a consequence thereof, solvation effects are increased by both specific and non-specific interaction (electrostatic interactions and hydrogen bond energy terms). Probably, the outlined effect of F atoms is the cause of an extraordinarily large increase in sol-

318


5.5 (+)-IBP

IBP

5.0 KETO

FBP

4.5

pKa

NAP

4.0 3.5

DIF

3.0 0

1

2

3

4

5

6

7

8

-1

∆Gtr(pH 7.4 → pH 2.0) [kJ⋅mol ]

Figure 14 Dependence of pKa value on the transfer of Gibbs energy of solvation from the buffer at pH2.0 – ∆G0 pH7.4. pH 7.4 to the buffer at pH 2.0; ∆Gtr(pH 7.4 → pH 2.0) = ∆G0solv sol

ubility with pH that is found for DIF and FBP (Table 7) compared to the other compounds studied. It is not difficult to see a regularity between transfer energy ∆Gtr (pH 7.4 → pH 2.0) values and pKa (Fig. 14): the weaker the acid, the lower the value of the driving force for the transfer process (and the easier it is to protonate the respective base). It should also be noted that a compensation effect is observed between the thermodynamic functions of transfer, which can be described by the following equation: T∆Str = (–8.2 ± 0.2) + (0.8 ± 0.1) ·∆Htr ,

(22)

σ = 2.29; r = 0.970; F = 64.4; F 2.5% tab = 9.365; n = 6. In other words, the entropic term of the Gibbs energy is 0.8 times less compared to the enthalpic term. It should also be kept in mind that determination of pKa values of poorly soluble drugs is a delicate experiment, and the values may differ considerably according to the method used. In the case of IBP, for example, using apparent pKa values in different solvent/water mixtures and extrapolating to 0% solvent content, the pKa value varies between 5.2 and 4.3 [75]. The pKa values reported for other NSAIDs vary within the similar ranges [76]. In the present study, this was not taken into consideration, neither for correlation analysis nor in Fig. 14, because it would not affect the key messages anyway. 6.3

Solvation characteristics of transfer process of dissociated and non-dissociated molecules from buffer to n-octanol

Taking into account the thermodynamic data of solvation in octanol measured earlier for the NSAIDs discussed [77], it is possible to analyse the transfer of dissociated and non-dissociated molecules from a respective buffer solution into the octanol phase. The thermody-


319

namic relations between the parameters are illustrated in Scheme 4. Knowledge about hydration and solvation characteristics of drug molecules exclusively enables the use of an absolute energetic scale. The discussed thermodynamic parameters together with related literature data are presented in Table 9. Buffer pH 2.0 Buffer pH 7.4 T⋅∆Str ∆ Gt

T⋅∆Str

∆ Gt

∆Ht n- Octanol

∆Ht

Scheme 4

From Table 9 and Scheme 4 it follows that the relationship between the outlined functions can be described as follows: Buffer pH 2.0 → octanol: ∆Htr < 0; ∆Str > 0; |∆Htr | 0; ∆Str > 0; |∆Htr | 1.0 > 1.5 > 1.75 > 2.0 > 3.0 > 5.0 wt. %. The drug release profiles of PHB were registered in water and phosphate buffer (pH = 7.4) by the UV technique using a Beckman DU65 UV spectrophotometer at 25°C.

2

Results and discussion

Typical kinetic profiles of Fr release from PHB films are shown in Fig.1. As is clear from the graph, for PHB release systems loaded by the drug at concentrations exceeding 1% there are no constant limiting values of equilibrium concentration that would be typical of a diffusion picture according to Fick’s law. These kinetic curves are characterized by the initial nonlinear range and final range where the drug release profile is linear relative to time (zero-order kinetics). Analysis of the data in Fig. 1 shows that the superposition of the proper diffusion and a linear kinetic process defines the complicated character of release. Most clearly the linear ranges are manifested after compilation of drug diffusion and are observed for 8–10 days. 0.45 5

0.40 0.35 0.30

3

Dt

0.25 0.20 0.15 1

0.10 0.05 0.00 0

5

10

15 20 time, days

25

30

Figure 1 Kinetic profiles of aseptic release. The figures show the initial concentrations of the drug (wt. %); Dt is the optical density of the drug in a liquid medium at 373 nm.

344

Controlled Release of Aseptic Drug from Poly(3-hydroxybutyrate) Films

1

1.0

2

0.8 Gt/G00

3 0.6

5

0.4 0.2 0.0 0

5

10

15

20

25

(Time)1/2,

(h)1/2

30

35

40

Figure 2 Impact of diffusion in the total drug release kinetics for PHB films (thickness, 180 µm). Initial concentrations of the aseptic are (1) 1.75 wt. %, (2) 2.5 wt. %, (3) 3 wt. %; (5) 5 wt. %.

Based on aforesaid, the release profile is described by the following equation ∂Ct/∂t = D[∂2Ct/∂x2] + k,

(1)

where D is the drug diffusion coefficient, cm2/sec; k is the kinetic constant of hydrolytic destruction, sec –1; Ct is the concentration of the drug in the polymer, wt. %; x (cm) and t (sec) are the coordinate and time of diffusion, respectively. After subtracting the linear kinetic term (k · t) from the integral profile of release Ct – kt ≡ Gt,

(2)

the diffusion equation has the classical view ∂Gt/∂t = D[∂2Gt/∂x2].

(1a)

The solution of Eq. 1a provided that Gt/Goo > 0.5 has the classical Fick’s form Gt/Goo = 1 – (8/π2⋅exp (–π2Dt/L2),

(3)

where L is film thickness, cm; Gt is the concentration of the drug available for diffusion, and other parameters were determined in Eq. 1. Finding the logarithms of the left- and right-hand side parts of Eq. 3 and solving this equation graphically in coordinates log (1 – Gt/Goo) – t: log (1 – Gt/Goo) = log (8/π2 ) – π2Dt/L2 ,

(4)

we can calculate the diffusion coefficients of Fr in the PHB films. Numerical subtraction of the linear contribution to the release kinetics from the total concentration of the drug

Mobile fraction of furacillin, wt. %

R.Y. Kosenko, Y.N. Pankova, A.L. Iordanskii, A.P. Bonartsev, and G.E. Zaikov

345

3.0 2.5 2.0 1.5 1.0 0.5 0.0 0

Figure 3

1

2 3 4 Total furacillin concentration, %

5

Relation between total and mobile concentrations of the drug in PHB films.

(shown in Fig. 1) was performed to separate the diffusion and kinetic components. After computer calculation, the proper diffusion contribution is shown in Fig. 2. Here the results are presented in classical diffusion coordinates: Gt/Goo – t1/2. We emphasize two features of drug release profiles: 1) the linear character of the data in the above diffusion coordinates holds within the limits 0 < Gt/Goo < 0.65 that attests a predominance of diffusion at the initial stage of release; 2) the slope of the lines in the same coordinates depends on the initial concentration of the drug that demonstrates its concentration diffusivity dependence. Limiting (equilibrium) values of the drug (Goo) dissolved in PHB films are needed to plot the curves in Fig. 2. Additionally, these values show the portion of the drug taking part in molecular diffusion. The isotherm of Fr sorption as the dependence of free diffusing concentration of the drug on the total loaded drug concentration is shown in Fig. 3. From this figure it follows that up to 3 wt. % both concentrations are related by a linear function. The deviation from linearity is observed at the maximal concentration (5 wt. %) of the loaded Fr. At this point the drug forms the proper phase in the polymer as yellow crystals, while the effect of phase formation does not distort the manner of the kinetic curve (See Fig.2, the curve for 5 wt. %). Summarizing the results presented in Figs. 1 and 2, we can estimate effective diffusion coefficients for all initial concentrations of the drug. Figure 4 shows the concentration dependence of diffusivities which has a maximum defined clearly in the drug concentration area 1.0–1.5 wt. %. The rising branch of the curve D(C) results likely from disordering of the PHB structure after drug loading. In contrast, the dropping branch is related to drug crystal formation in the PHB matrix that occurs during the decrease of low-molecular-weight component mobility. The formation of Fr crystals in PHB has been observed recently in our work by Krivandin with WAXS technique [21]. Above we have mentioned that, simultaneously with the diffusion kinetics, the linear kinetic process of Fr release is observed. In this case, the greater the initial concentration of the loaded drug is, the higher the constant rate of drug release is. More informatively this effect is shown in Fig. 5, where the exponential dependence of the degradation constant (k) on the drug concentration is observed. Simultaneously with the measurement of the


Mobile fraction of furacillin, wt. %

346

3.0 2.5 2.0 1.5 1.0 0.5 0.0 0

1

2 3 4 Total furacillin concentration, %

5

Figure 4 Dependence of release destruction constant (k) on loaded drug concentration for PHB films.

Drug diffusion coefficient × 1010, cm2/s

1.6

1.2

0.8

0.4

0.0 0.0

0.5

1.0 1.5 2.0 2.5 Aseptic concentration, wt. %

3.0

Figure 5 Diffusivity dependence on aseptic concentration in the drug release process.

concentration of the drug desorbing PHB from films, we have determined the loss of weight for polymer samples. The gravimetric measurements have shown that the polymer sample loses its weight also in accordance with the linear law. Initial polymer without the drug has the stable weight throughout the entire time of release. Preliminary experiments show that, in contrast to enzymatic biodegradation of PHB going on the polymer surface [19], hydrolytic destruction involves all accessible volume of PHB that is supported by the increase in brittleness and the decrease in the strength of PHB films.

R.Y. Kosenko, Y.N. Pankova, A.L. Iordanskii, A.P. Bonartsev, and G.E. Zaikov

3

347

Conclusion

We proposed a new polymer system for aseptic controlled release that includes films of PHB and furacillin. The release goes on simultaneously in accordance with the kinetic (polymer degradation) and diffusion mechanism. The rates of the kinetic mode for release obey zero degree curves relative to time. The diffusion mode, which determines the profile of release in the initial range of time (about the first week), was analyzed in greater detail. The dependences both of diffusion coefficients and kinetic constant (k) of release on the drug concentrations were demonstrated. These results are requisite for the further development of drug release systems with multi-component action when several drugs simultaneously have a local action on biological tissues and cells.

4

Acknowledgement

The authors thank the Biomer Company (Krailing FRG) for the donation of PHB. Funding for this research was provided partly by the Russian Foundation for Basic Research with grants Nos 06-04-49339 and 08-03-00929-a.

References 1. R.W. Lenz and R.H. Marchessault, Biomacromolecules, 6 (1), 1–8 (2005). 2. D. Williams, Medical Device Technology, 16 (1), 9–10 (2005). 3. Y. Doi and C. Abe, Macromolecules, 23 (15), 3705–3707 (1990). 4. Y. Liu and M. Wang, Current Appl. Phys., 7 (5), 547–554 (2007). 5. R. Sodian, S.P. Hoerstrup, J.S. Sperling, D.P. Martin, S. Daebritz, J.E. Mayer Jr. and J.P. Vacanti, ASAIO J., 46 (1), 107–110 (2000). 6. Biomaterials Science: An Introduction to Materials in Medicine, ed. by B.D. Ratner, A.S. Hoffman, F.J. Schoen and J.E. Lemons, San Diego: Academic Press (1996). 7. G.X. Cheng, Z.J. Cai and L. Wang, J. Mater. Sci., 14, 1073–1078 (2003). 8. V.I. Sevastianov, N.V. Perova, E.I. Shishatskaya, G.S. Kalacheva and T.G. Volova, J. Biomater. Sci. Polymer Ed., 14, 1029–042 (2003). 9. D.P. Martin and S.F. Williams, Biochem Eng. J., 16, 97–105 (2003). 10. R. Sodian, M. Loebe, A. Hein, D.P. Martin, S.P. Hoerstrup, E.V. Potapov et al., ASAIO J., 48, 12–16 (2002). 11. R. Sodian, S.P. Hoerstrup, J.S. Sperling et al., Ann. Thorac. Surg., 70, 140–144 (2000). 12. M. Unverdorben, A. Spielberger, A. Schywalsky et al., Cardiovasc. Intervent. Radiol., 25, 127–132 (2002). 13. G.T. Köse, S. Ber, F. Korkusuz et al., Biomaterials, 24, 4998–5007 (2003). 14. L.J. Chen and M. Wang, Biomaterials, 23, 2631–2639 (2002). 15. G.-Q. Chen and Q.Wu, Biomaterials, 26 (33), 6565–6578 (2005). 16. D.P. Martin, F. Skraly and S.F. Williams, US Patent 403242 (2003). 17. C.W. Pouton and S. Akhtar, Adv. Drug Deliver. Rev., 18, 133–162 (1996). 18. D. Sendil, I. Gürsel, D.L. Wise and V. Hasirci, J. Controlled Release, 59, 207–217 (1999). 19. A.L. Iordanskii and P.P. Kamaev, Polym. Science, 41B (1–2), 39–43 (1999). 20. R.H. Marchessault, K. Okamura and C.J. Su, Macromolecules, 3 (6), 735–740 (1970). 21. O.V. Shatalova, A.V. Krivandin, and A.L. Iordanskii, 6th Eur. Symposium on Polymer Blends. Program and Abstracts. May 16–19, Max-Planck-Institut fur Polymerforschung (1999).

348


Авторам: Нет упоминания рис. 2 Обратить внимание на выделенное красным, на нумерацию уравнений

11

Transport of Water as a Structurally Sensitive Process Characterizing the Morphology of Biodegradable Polymer Systems A.L. Iordanskii1, Yu.N. Pankova1, R.Yu. Kosenko1, A.A. Ol’khov1 and G.E. Zaikov2 1N.N.

Semenov Institute of Chemical Physics, Russian Academy of Sciences, 4 Kosygin Street, Moscow 117977, Russia email: [email protected] 2N.M. Emanuel Institute of Biochemical Physics, Russian Academy of Sciences, 4 Kosygin Street, Moscow 119991, Russia email: [email protected]

Investigations of water transport in polymer mixtures (where, as a consequence of a partial compatibility of components, the influence of structural factors on the mechanical, barrier and other operational properties is especially pronounced) take on special significance. In this work, we try to show how the morphological characteristics of polymer mixtures influence the water diffusion parameters and how the latter correlate with the operational characteristics of mixtures by the example of mixed compositions of a natural polymer product of bacterial vital functions, poly(3-oxybutyrate) [PHB], and synthetic polymers of various polarities (hydrophilicities). Such investigations seem necessary for a detailed description of ecologically compatible (bio)degradable systems functioning in water media.

350

A.L. Iordanskii, Yu.N. Pankova, R.Yu. Kosenko, A.A. Ol’khov and G.E. Zaikov

Keywords: diffusion, structure of polymers, morphology, biodegradation, bacteria, natural and synthetic polymers

Introduction Professor Nikolai M. Emanuel very attentively and, one could say, tremulously treated young scientists. Direct contacts with Professor Emanuel were infrequent but became turning and, sometimes, key points of the scientific life of an inexperienced researcher. In our case, his conceptions about diffusion processes playing not an accessory but a dominating role in multi-phase biosystems served the basis of a scientific direction “Transport Phenomena in Structurally Organized Polymer Systems”. Why is it, in particular, diffusion? Why is it the diffusion namely of water? The human habitat including its biological structures contains a considerable amount of water. Variation of its concentration and its functionalization determine all processes of vital functions [1]. Concerning polymer systems, especially if they are used as biomedical systems (vascular prostheses, matrixes for the controlled release of medicines, osteoprosthesis, newgeneration vascular stents, etc.) the necessity arises to obtain information about the state water is in and the rate of its motion in intercrystal regions of polymer products [2]. From the fundamental point of view the diffusion of bipolar water molecules may be considered as the motion of a specific probe extremely sensitive to polar functional groups of macromolecules forming the polymer matrix. Moreover, water has a noticeable affinity to various hydrophilic impurities occurring in the polymer at some concentration (remnants of catalysts, plasticizers, monomers and other accompanying components of synthesized materials or the residual content of albumens, lipids and slats in polymers of natural origin [3]). Due to the small size of water molecules, such a probe overpasses nonpolar polymer regions quickly enough. But its diffusion rate is significantly decreased in polar regions (for example, near albumens’ and polyamides’ amine groups [4]) due to participation in the formation and redistribution of hydrogen bonds in such hydrophilic polymers as cellulose, PVA, etc. [5]. Investigations of water transport in polymer mixtures (where, as a consequence of partial components’ compatibility, the influence of structural factors on the mechanical, barrier and other operational properties is especially pronounced) take on a special significance. In this work we try to show how the morphological characteristics of polymer mixtures influence the water diffusion parameters and how the latter correlate with the operational characteristics of mixtures on the example of mixed compositions of: • a natural polymer product of bacterial vital functions – poly(3-oxybutyrate) [P3HB], • synthetic polymers of various polarities (hydrophilicities). Such investigations seem necessary for a detailed description of ecologically compatible (bio)destructible systems functioning in water media.

1

Experimental

Materials. A superfine powder of poly-3-oxybutyrate (PHB) was kindly provided by Biomer (Krailing, Germany). For preparation of mixed films from the solvent the powder was preliminary dissolved in chloroform and the polymer solution was filtered through a glass Schott 160 filter. The solubility of initial polymer powder in chloroform turned to be

Transport of Water as a Structurally Sensitive Process

351

close to 0.01 k/ml. The thickness of films was 100±5 µm. X-ray crystallinity of polymer was 70% and according to the data of the DSC method it was 75%. After filtration, the molecular mass of polymer films was determined by viscosimetry using the following equation [6]: 0.82 [η ] = 7.7·10 –5 M ω ,

(1)

where [η ] is the intrinsic viscosity of the polymer solution in chloroform. The viscosityaverage molecular mass of polymer, MW, turned to be equal to 150·103 g/mol. Granular low density polyethylene (LDPE) (State Standard 583-020) was characterized by a molecular mass of 250·103 and density 0.92 g/cm3. In this work, we also used polyvinyl alcohol (PVA) of Russian brand 8/27 with the residual content of non-saponated acetate groups 8.2% and the concentration of sodium acetate 0.04 wt. %. The molecular mass of PVA is equal to 64,000 g/mol and the melting temperature is 146°C. Mixing in extruder. Mixture of LDPE with PHB: firstly, the preliminary mixing in a ball mill was carried out. The ratios of the components (PHB/LDPE) were: 2/98, 4/96, 8/92, 16/84, and 32/68 wt. %. Blending was carried out in a single-auger extruder (ARP-20) with the frequency of auger rotation 100 min –1 provided that the temperature of the melt was not higher than 182°C. The plane-parallel slot was regulated in such a way so as to obtain films 50–60 µm thick. A similar procedure was carried out when obtaining mixtures of PHB/polyvinyl alcohol films (PVA, Russian brand 8/27) of structures 10/90, 20/80, 30/70, and 50/50 wt. %. A similar procedure of extrusion was carried out for initial individual polymers: PVA, LDPE and PHB. Methods of analysis. The kinetic curves of water sorption / desorption and equilibrium sorption were measured by vacuum gravimetry using a quartz McBen’s microbalance with the quartz spiral sensitivity of 0.67 mg/mm. The permeability of water vapors was registered in a traditional two-chamber cell described in [7]. The results of X-ray diffraction at large angles were obtained with the help of a diffractometer with a linear-coordinate detector constructed at the Joint Institute of Nuclear Research (Dubna, Russia). Since extrusion films are characterized by the orientation of both macromolecules and structural elements in particular crystallites, such orientation was estimated by measuring their polarization spectra. In investigations, the plane of the catalyst was oriented either along or transversely to the direction of extrusion. Spectral characteristics and round dichroism were measured on an IFS-48 Bricker IR spectrometer.

2

Hydrophobization of poly-(3-hydroxybutyrate)

Differences in water diffusion behavior when comparing synthetic (gas and oil derivatives) and natural (products of animals’ and plants’ vital functions) polymers are more clearly revealed on the molecular and crystal level. For hydrophobic polymers (low density PE, polypropylene, polysiloxane, synthetic rubbers, etc.), analysis of the chemical structure of macromolecules shows the damage of the primary chemical structure and accumulation of oxygen-containing functional groups in the polymer chain as a result of imperfect conditions of polymer production. The appearance of groups “introduced” into the hydrophobic matrix leads to an increase of equilibrium sorption of water and to a decrease of its effective diffusion coefficient. The diffusion mobility is reduced as a result of water molecules’


1/Pw 109, cm2 h mmHg/g cm

352

a

8 6 4 2 0

0

5

10 PHB, %

2.6

15

30

35

b

1/R (1228 cm_1)

2.4 2.2 2.0 1.8 1.6 1.4 0

5

10

15 20 PHB, %

25

30

35

Figure 1 Comparison of inverse permeability (a) and dichroism (b) values as a function of the structure of the polymer mixture for films based on LDPE–PHB.

interaction with the damaged polar groups of the polymer. For hydrophobic polymers of natural origin, the structure of which is formed on the cell and enzyme levels, the damage of the primary structure is less typical and may occur only at later technological stages, e.g., in processing or mixing in extruders. Under normal conditions, the solubility of water in saturated hydrocarbons is limitedly low (several mils of percent) [8]. In hydrocarbons C7 –C16 the mole fraction of water (Xw) is written as the following equation: ln(Xw –1) = Vw (∆δ)2/RT,

(2)

where Vw is the water molar volume, ∆δ is the difference between the solubilities of water and hydrocarbon, R and T are the gas constant and temperature, respectively. The authors of [8] by extrapolation of eq. (1) showed that at 25°C the values of Xw did not exceed 10–6 mol/l (~2·10–5 wt. %), so this value might be considered as an ideal solubility of water in the hydrocarbon matrix.


353

In both synthetic (polyolefines) and natural hydrocarbon polymers the limit of the solubility of water is significantly higher and is significantly changed depending on the polymer production method. Systematic investigations of water diffusion in films of low and high density PE showed that the emergence of damaged fragments on the polymer chain influenced water molecules’ sorption and diffusion [9]. Immobilization of water molecules on such polar groups statistically distributed in the hydrocarbon nonpolar matrix leads to a noticeable decrease of the effective diffusion coefficient of water. Dw –1 = Dwf–1(1 + [Spf KL/KH]),

(3)

where Dwf is the diffusion coefficient of a single water molecule (free diffusion), Spf is the concentration of introduced (damaged) groups, KH and KL are the Henry and Langmuir constants of sorption, respectively. With the aim of obtaining biodegradable polymer materials able to destruct in soil but to keep their transport and mechanical properties during long exploitation period we investigated new polymer compositional mixtures of poly-3-oxybutyrate and LDPE [10]. For such mixtures, water permeability depends on PHB/PE, the ratio of components’ concentrations, see Fig. 1a where the value of inverse permeability is presented, i.e. water diffusion resistance as a function of PHB concentration. Introduction of carbonyl groups of POB into the nonpolar matrix of polyethylene (LDPE) enables regulation of transport parameters. The dependence “water resistance vs. POB concentration” described by the extreme and non-additive function is in good agreement with the segment orientation acquired during the extrusion process during the melting of PE/POB mixtures. The degree of orientation is determined by the method of Fourier spectroscopy under construction of the dichroism function [11] and is presented in Fig. 1b. The reason for the extremums to appear on the “dichroism–mixture structure” curves may be the covering of two processes oppositely influencing the orientation of macromolecules in crystal regions (and consequently the orientation of crystals). On the one hand, with the increase of the content of rigid molecules of PHB in the melt the tendency of ordering all molecules while passing through the slot of the extruder is increased. On the other hand, the viscosity of the mixture in this case is increased and consequently the time necessary for structural elements’ (including crystals’) orientation is also increased. At low concentrations of PHB (>8 wt. %) the orientation process prevails. And at higher PHB concentrations the tendency of disordering of the structural organization begins to be exhibited more clearly. The decrease of the degree of crystallinity of both components observed independently by the DSC method confirms this fact. The results of X-ray analysis for large and small angles showed the presence of crystal axial a-texture realized in investigated films for both PE and PHB [12]. Crystallites belonging to both polymers are oriented preliminary in such a way that axes a of elementary crystal cells of both mixture components coincide with the direction of extrusion, and axes c coinciding with macromolecules’ orientation in crystallites’ elementary cells are oriented transversely to this direction. Such a position of crystallites is characteristic of all compositions studied. That is why a combination of earlier obtained X-ray conceptions about the crystal structure of the POB/PE mixture and the results of polarization IR Fourier spectroscopy suggest that the cause of inversion of the dichroism function is probably the difference in the angles of transition moment for methyl groups belonging to PE (=0°C) and POB (=90°C), respectively.

354

3


Hydrophilization of poly-(3-hydroxybutyrate)

The results obtained by the DSC method for mixtures and initial samples of PHB and PVA are presented in Table 1. Table 1 Glass-transition temperature of mixture compositions with various contents of PHB. PHB concentration, %

0

10

20

30

50

100

Tg, ºC

53.9

32.7

28.5

25.3

25.9

24.1

As we mentioned above, all samples were obtained by the extrusion method as described in the experimental part. Analysis of the thermograms showed that there was a stable low-temperature transition in the interval 24–54°C reflecting, most probably, the transition from the glassy to rubberlike state at the glass-transition temperature of ~37°C. We should note that for initial polymer components the corresponding glass-transition temperatures lie in the region of 24.1°C and 53.9°C for PHB and PVA, respectively. The existence in the mixture compositions of only one phase transition of the second order should testify to the interaction between two polymers [13] forming the mixture and to their possible partial compatibility in inter-crystallite regions of mixture films. A combination in polymer crystallite regions (their part is large and is equal to more than 70%) most likely does not occur since on DSC thermograms two independent peaks belonging correspondingly to PHB and PVA are observed. The occurrence of independent crystallite phases of each of mixture-forming polymers was observed by the method of wide angle X-ray scattering (WAXS). In the interval of PHB concentration from 0 up to 30% each of components forms its own, slightly oriented crystal phase that is confirmed by X-ray diffractograms of the mixtures; they represent a superposition of diffractograms of individual polymers: PHB (degree of crystallinity ~70%) and PVA (degree of crystallinity ~30%). At all polymer ratios forming the given mixtures, the parameters of the elementary orthorhombic crystal cell of PHB remain constant and have the following values, respectively: a = 0.576 nm, b = 1.32 nm, and c = 0.596 nm [14]. X-ray reflexes for PVA show that crystalline regions of polyvinyl alcohol are in gamma form well described earlier in domestic literature by L.L. Razumovskaya et al. [15]. In particular, the reflex S = 2.21 nm –1 belongs to this modification of PVA structure. The form of diffractograms obtained at various angles of X-ray irradiation incidence in relation to sample plane shows that for all samples containing PVA the axial cylindrical texture is observed where the axis of texture coincides with direction of extrusion through the slot. In samples containing 10 and 20% of PHB, the axial texture is observed also for poly-3-hydroxybutyrate. However, in contrast to PVA where macromolecules are oriented along the direction of extrusion, in crystal regions of poly-3-hydroxybutyrate the axis a of elementary crystal cell coincides with the direction of extrusion and, consequently, biopolymer macromolecules are preliminarily arranged transversely to the axis of extrusion [16]. Diffractograms of mixture compositions containing 30 and especially 50% of PHB show the absence of texture of the latter and so, these sample are preliminarily isotropic. The transition from textured to isotropic crystal structures observing in the region of 30% of PHB is probably related to the phase inversion of samples occurring in the mentioned region of mixture structure. A sharp change of structural and physical characteristics of polymer mixtures occurs namely in the region of phase inversion, see, e.g., [17].


355

4 PHB(020)

I, rel. units

3 2

PHB(110) 1 PVA(110) 2 PHB(111)

1 0

1.0

1.5

2.0

2.5

3.0

3.5 S, 1/nm

3.0

3.5 S, 1/nm

6 PHB(020)

I, rel. units

PHB(110) 4 2

3

2

PVA(110) 2

1 1, 3 0

1.0

1.5

2.0

2.5

Figure 2 WAXS diffraction of extrusion films prepared by mixing PHB and PVA (upper, 80:20% and lower, 70:30%). The vector of X-ray irradiation is directed along the direction of extrusion (1), transversely (2) and at an angle of 20º (3) to the direction of extrusion. Results and Discussion were prepared by O.V. Shatalova and A.V. Krivandin (IBCS RAS).

Taking into account the results presented above, we studied the physicomechanical characteristics of polymer mixtures with various contents of PHB. It is important to note that on the curves “rupture stress–structure of mixture” and “elongation at rupture– structure of mixture” the sharp changes of values of mechanical parameters are observed, and in the same concentration interval (~ 30% PHB), where the above mentioned transition occurs. Besides, the minimum of the extreme dependence of elasticity modulus on PHB concentration in the mixture is situated in the same concentration interval. A decrease of elasticity modulus in the region of low concentrations of PHB is caused by the reduction of segment mobility of interaction macromolecules of PHB and PVA. A further increase of elasticity modulus values reflects the increase of structural units’ content responsible for material strength and namely of crystallites belonging to poly-3-hydroxybutyrate. The totality of the results of mechanical tests for mixture films presented in Figure 3 confirms the physical concept of structural inversion, which concerns both crystal and inter-crystal (conditionally amorphous) regions of mixture compositions. Transformation of amorphous polymer regions as a result of phase inversion should be reflected on the level of diffusion processes and especially under diffusion and

356


100

200

75

150 50

100

Elongation at rupture, %

Rupture stress, MPa

250

50

25

0 0

20

40

60

PHB content, wt. %

Water solubility, g/cm3 . molecular mass of mercury

Figure 3 Dependence of the tensile strain and rupture stress on the structure of a PHB–PVA polymer mixture. 40 35 30

T = 20°C

25 20 15 10 5 0 0

Figure 4

20

40 60 80 PHB concentration, wt. %

100

Change of water solubility as a function of the structure of a mixture.

permeability of structurally-sensitive water molecules. It is well known that even so small-size molecules as those of water are not capable of penetrating crystal regions of any polymer and, consequently, information about diffusion flows, diffusion mobility and water sorption reflects first of all the state of amorphous regions of the polymer matrix. In mixture compositions, the content of water is determined by the ratio of the components if polymers of various hydrophilicities are mixed. In our case, the mixture is formed by moderately hydrophilic complex polyester (PHB) and typical hydrophilic polyalcohol (PVA). The increase of the concentration of the latter in the mixture corresponds to a monotonous growth of hydroxyl groups in the system. The equilibrium concentration of water is increased analogously, since in accordance with the theory of group contributions [18] there is a definite amount of water molecules interacting with the OH group per each

Water permeability . 108, g cm/cm2 hour mm of mercury


300

357

Pw 108

Hypothetically additive dependence

250 200 150 100 50 0 0

20

40

60

80

100

PHB concentration, wt. % Figure 5

Extreme dependence of water permeability through films based on PHB–PVA mixtures.

of these OH groups of polymer. The contribution of the ester group of PHB is significantly lower and in this particular case we can neglect it. Actually, the equilibrium sorption of water in investigated mixtures similarly increases monotonously with the rise of PVA concentration or with the decrease of the PHB content (Figure 4). Characteristic inflections and extreme points on the water sorption–mixture structure dependence are not observed and consequently the equilibrium sorption is sensitive mainly to the primary chemical structure of polymer, to be more exact, to the chemical structure of the polymer mixture. Thus, a monotonous increase of water equilibrium sorption with the change of the mixture structure is determined by the substitution of ester functional groups belonging to PHB and possessing a low affinity to water molecules by more hydrophilic groups (hydroxyl) belonging to PVA. In contrast to equilibrium sorption data the values of normalized water flows (permeability) have a characteristic degraded maximum in the concentration region corresponding approximately to 40% of PHB as shown in Figure 5. Obviously, in this case as a result of rebuilding the structure the maximum accumulation of structural defects occurs due to which the increase of water flow is observed. In the interval of concentrations preceding this point (10–30% PHB) there is no additivity of the permeability parameter and even a small addition of polyester (10%) significantly reduces the values of water permeability first of all at the expense of segment interaction between PHB and PVA macromolecules forming the mixture. This conclusion is also confirmed by the shift of mixture glass-transition temperature. We want to note once again that such interactions are realized in intercrystalline or better to say amorphous regions and do not affect the crystalline regions of polymer. The descending line of the water permeability–mixture structure dependence is approximately described by a linear function and consequently obeys a property–structure additive dependence. So, in this region of structure each of components independently leaks water flow and, thus, segment interaction revealed in the concentration region below the point of structural transition is practically absent. When one analyzes the dependence of water permeability on structure, one should remember that from the physical point of view permeability represents the product of two other physical parameters: sorption capacity and diffusion mobility. We considered water


Diffusion coefficients 108, cm2/sec

30

100

25 75

20 = Dw

=σ

15

50

10 5

Rupture stress, MPa

358

25

0 0

10

20 30 40 PHB concentration, wt. %

50

Figure 6 Similarity of the behavior of the diffusion coefficients and rupture stress for PHB–PVA mixture compositions.

sorption above, and as the parameter characterizing the mobility of water molecules in mixture films we chose the coefficient of water diffusion. In Figure 6 we show these coefficients in the form of a dependence on mixture structure. An analogous dependence for rupture stress is presented in the same figure for clarity. Both the diffusion and mechanical characteristics of mixture compositions are characterized by clear inflection points situated in the region of structural transition at 30% of PHB that coincides with the results of X-ray analysis. As in the case of mechanical tests in the region of low content of PHB the diffusion behavior of the system resembles the diffusion in pure PVA (the region 0–20%) whereas at concentrations exceeding 30% of PHB the diffusion coefficients of mixture compositions sharply exceed the diffusion coefficients of water in PHB (Dw = 1–10-8 cm2/sec [19]). In accordance with the results of water permeability measurements we assume that high values of the diffusion coefficients are caused by structure defects more clearly revealed as a result and after phase transition. Earlier, by WAXS, EPR, Fourier IR spectroscopy and vacuum gravimetry methods we showed [20, 21] that crystalline transition from anisotropic into isotropic state of PHB led to the change of water diffusion mobility. In the case of the change of mixture structure we also observe the transition from the textured state into a more disordered one, i.e. isotropic. Besides, an increase of defects of mixture compositions at a relatively high content of PHB does not make it possible to exclude the possibility of partial water leakage through the defect zones in the matrix analogously to the case observed for PHB–LDPE mixtures [22]. Intensive studying of particularities of water transfer in mixture compositions on the basis of poly-3-hydroxybutyrate represents a necessary stage in the investigation of such fundamental processes as physical ageing at high dampness, corrosion stability of polymer membranes; these results should also promote the development of novel ecologically and biologically compatible materials for medicine and agriculture. A complex of methods combining structural (DSC, IR Fourier polarization spectroscopy), mechanical and diffusion experiments was used in this work. By varying the nature


359

of the second mixing component (its hydrophilicity) we may control the morphology of the mixture and, consequently, the rate of diffusion processes. For example, mixing of PHB with a representative of the class of hydrophobic polymers (LDPE) does not give compatible compositions. Matrices of PHB–LDPE mixtures are partially crystalline, oriented in relation to extrusion direction. The direct correlation between the diffusion resistance of film to water flow and the degree of orientation of macromolecules is shown. On the contrary, hydrophilization of PHB by mixing with PVA gives the possibility of preparing novel mixture films partially compatible in intercrystalline regions. The region of structural transition (at 30% of PHB) where mechanical and transport characteristics are sharply changed was found.

References 1. E.A. Vogler, Biological Properties of Water, in: Water in Biomaterials. Surface Science, Wiley: Chichester–New York–Weinheim–Brisbane–Singapore–Toronto, Ch. 1, 3 (2001). 2. Water in Polymers, ACS Symposium Series 127, ed. by Rowland (1980). 3. J.H. Lee, T. Li and K. Park, Solvation Interactions for Protein Adsorption to Biomaterial Surfaces, in: Water in Biomaterials. Surface Science, ed. by M. Morra, Wiley: Chichester–New York–Weinheim–Brisbane–Singapore–Toronto, Ch. 5, 127 (2001). 4. G.E. Zaikov, A.L. Iordanskii and V.S. Markin, Diffusion of Electrolytes in Polymers, Ser. New Concepts in Polymer Science, VSP Science Press: Utrecht–Tokyo (1988). 5. A.L. Iordanskii, P.P. Kamaev, A.A. Ol’khov, and A.M. Wasserman, Desalination, 126, 139, (1999). 6. R.H. Marchessault, K. Okamura and C.J. Su, Macromolecules, 3 (6), 735 (1970). 7. J. Crank and G.S. Park, Diffusion in Polymers, Academic Press: London–New York (1968). 8. B. Schatzberg, J. Phys. Chem., 3 (67), 775 (1963). 9. D.W. McCall, D.C. Douglass and L. Blyler, Macromolecules, 17, 1644 (1984). 10. A.A. Ol’khov, A.L. Iordanskii, S.V. Vlasov et al., Russian Polymer News, 5 (1), 34 (2000). 11. R. Zbinden, Infrared Spectroscopy of High Polymers, Academic Press: New York (1966). 12. A.A. Ol’khov, S.V. Vlasov, L.S. Shibryaeva and A.L. Iordanskii, Polymer Sci. Ser. A., 42 (4), 676 (2000). 13. D.R. Paul and C.B. Bucknal, Polymer Blends, Wiley: New York, 1 (2000). 14. A.L. Iordanskii, A.A. Ol’khov, O.V. Shatalova, G.E. Zaikov, and U.J. Hanggi, Water Diffusion, Crystalline Structure, and Mechanical Properties of Novel PHB–PVA Blends, ed by A.L. Iordanskii, O.V. Startsev and G.E. Zaikov, Nova Science: New York, Ch. 9, 213 (2004). 15. L.L. Razumova, O.V. Shatalova, A.L. Iordanskii and G.E. Zaikov, Vysokomol. Soed., 38A, 1271 (1976) (in Russian). 16. A.A. Ol’khov, S.V. Vlasov, A.L. Iordanskii, G.E. Zaikov and V.M. Lobo, J. Appl. Polymer, 90, 1471 (2003). 17. J.P. Runt, Crystalline Polymer Blends, in: Polymer Blends, ed. by D.R. Paul and C.B. Bucknall, Wiley: New York, 1, 167 (2000). 18. A.E. Chalykh, Diffusion in Polymers, Khimiya: Moscow (1984) (in Russian). 19. P.P. Kamaev, I.I. Aliev, A.L. Iordanskii and A.M. Wasserman, Polymer (UK), 42, 515 (2001). 20. P.P. Kamaev, A.L. Iordanskii, I.I. Aliev, A.M. Wasserman and U.J. Hänggi, Desalination, 126, 153 (1999). 21. A.L. Iordanskii and P.P. Kamaev, Polymer. Sci. Ser. B, 41 (1–2), 39 (1999). 22. A.L. Iordanskii, From Traditional to Novel Environmentally Friendly Polymers, in: Water Transport in Synthetic Polymers, ed. by A.L. Iordanskii, O.V. Startsev, and G.E. Zaikov, Nova Science: New York, Ch. 1, 1 (2004).

360


статья одобрена автором больше не смотреть Всё сделано - только номера страниц

12

A Novel Technique for Measurement of Electrospun Nanofiber M. Ziabari, V. Mottaghitalab and A.K. Haghi* University of Guilan, P.O. Box 3756, Rasht, Iran *[email protected]

In electrospinning, fiber diameter is an important structural characteristic because it directly affects the properties of the produced webs. In this chapter, we have developed an image analysis based method called direct tracking for measuring electrospun fiber diameter. In order to evaluate its accuracy, samples with known characteristics have been generated using a simulation scheme known as µ-randomness. To verify the applicability of the method, some real webs obtained from electrospinning of PVA have been used. Due to the need of binary input images, micrographs of the real webs obtained from scanning electron microscopy were segmented using local thresholding. The method was compared with the distance transform method. Results obtained by direct tracking were significantly better than distance transform, indicating that the method could be used successfully for measuring electrospun fiber diameter.

Keywords: electrospinning, fiber diameter, image analysis, direct tracking, distance transform

Introduction Conventional fiber spinning (like melt, dry and wet spinning) produce fibers with diameters in the range of a micrometer. In recent years, electrospinning has gained much attention as a useful method to prepare fibers in nanometer diameter range. These ultrafine fibers are classified as nanofibers. The unique combination of high specific surface area, extremely small pore size, flexibility and superior directional strength makes nanofibers a preferred material form for many applications. Proposed uses of nanofibers include wound dressing, drug delivery, tissue scaffolds, protective clothing, filtration, reinforcement and microelectronics.

362

M. Ziabari, V. Mottaghitalab and A.K. Haghi

Figure 1 Electrospinning setup.

d

Figure 2

α O

Procedure for µ-randomness.

In the electrospinning process, a polymer solution held by its surface tension at the end of a capillary tube is subjected to an electric field. Charge is induced on the liquid surface by an electric field. Mutual charge repulsion causes a force directly opposite to the surface tension. As the intensity of the electric field is increased, the hemispherical surface of the solution at the tip of the capillary tube elongates to form a conical shape known as the Taylor cone. When the electric field reaches a critical value at which the repulsive electric force overcomes the surface tension force, a charged jet of the solution is ejected from the tip of the cone. Since this jet is charged, its trajectory can be controlled by an electric field. As the jet travels in the air, the solvent evaporates, leaving behind a charged polymer fiber which lays itself randomly on a collecting metal screen. Thus, continuous fibers are laid to form a nonwoven fabric. Figure 1 illustrates the electrospinning setup [1–6]. The properties of electrospun nonwoven webs depend on the nature of the component fiber as well as its structural characteristics such as fiber orientation [7–12], fiber diameter [13], pore size [15], uniformity and other structural features [16]. Analyzing the electrospun webs yield results and information, which will help researchers in improving the quality and predicting the overall performance of electrospun webs. Some of the reasons for characterization may be process control, process development and product or quality control. Fiber diameter is the most important structural characteristics in electrospun nonwoven webs. Depending on the process and material variables, the diameter of fibers produced by electrospinning varies. Almost all researches who have done characterization, have reported the effects of processing variables on electrospun fiber diameter. There is no standard technique to measure the fiber diameter and analyze its distribution. This explains the need to study the fiber diameter of electrospun webs. Recently, image analysis has been used to identify

A Novel Technique for Measurement of Electrospun Nanofiber

363

fibers and measure structural characteristics in nonwovens. However, the accuracy and limitations of these techniques have not been verified. The objective of our research is to use image analysis for measuring electrospun fiber diameter.

1

Methodology

1.1

Simulation of electrospun web

In order to reliably evaluate the accuracy of the developed methods, samples with known characteristics are required. Since this end cannot be met with experiment, a simulation algorithm has been employed for generating nonwovens with known characteristics. The use of simulation is not a new idea. It was used by Abdel-Ghani and Davis [17] and Pourdeyhimi et al. [7] for simulation of nonwovens with both continuous and discontinuous fibers by the use of idealized straight lines. The most important component of simulation is the way in which lines or curves are generated. Abdel-Ghani and Davis [17] presented three methods for generating a random network of lines: 1) Surface randomness known as S-randomness 2) Mean free path known as µ-randomness 3) Internal randomness known as I-randomness. It is assumed that the lines are infinitely long (continuous filament) so that, at least in the image plane, all lines intersect the boundaries. The aim is to obtain unbiased arrays spatially homogeneous for infinitely long lines. Lately it was discovered by Pourdeyhimi et al. [7] that the best way to simulate nonwovens of continuous fibers is through the second method. Under this scheme, a line with a specified thickness is defined by the perpendicular distance d from a fixed reference point O located in the center of the image and the angular position of the perpendicular α. Distance d is limited to the diagonal of the image [7, 17]. Figure 2 demonstrates this procedure. Several variables are allowed to be controlled during the simulation: 1) Web density: can be controlled using the line density which is the number of lines to be generated in the image. 2) Angular density: useful for generating fibrous structures with specific orientation distribution. The orientation may be sampled from either a normal or a uniform random distribution. 3) Distance from the reference point: varies between 0 and the diagonal of the image, restricted by the boundary of the image and is sampled from a uniform random distribution. 4) Line thickness (fiber diameter): is sampled from a normal distribution. The mean diameter and its standard deviation are needed. 5) Image size: can also be chosen as required. 1.2

Fiber diameter measurement

The electrospinning process produces very fine fibers, and this is one of the few methods from which fibers of submicron size can be produced. So it becomes immensely important

364


Figure 3

Manual method.

to understand the behavior of fiber diameter and fiber diameter distribution in the electrospun web as impacted by the processing parameters. The first step in determining fiber diameter is to produce a high quality image of the web at a suitable magnification using electron microscopy techniques, called micrograph. The methods for measuring electrospun fiber diameter are described in the following sections. 1.2.1

Manual method

The conventional method of measuring the fiber diameter of electrospun webs is to analyze the micrograph manually. The manual analysis usually consists of the following steps, determining the length of a pixel of the image (setting the scale), identifying the edges of the fibers in the image and counting the number of pixels between two edges of the fiber (the measurements are made perpendicular to the direction of fiber axis), converting the number of pixels to nm using the scale and recording the result. Typically 100 measurements are carried out. Figure 3 illustrates this process. However, this process is tedious and time-consuming especially for a large number of samples. Furthermore, it cannot be used as an on-line method for quality control since an operator is needed to perform the measurements. Thus, developing automated techniques which eliminate the need for an operator and can be employed as on-line quality control is of great importance. 1.2.2

Distance transform

The distance transform of a binary image is the distance from every pixel to the nearest nonzero-valued pixel. The center of an object in the distance transform image will have the highest value and lie exactly over the object’s skeleton. The skeleton of the object can be obtained by the process of skeletonization or thinning. The algorithm removes pixels on the boundaries of objects but does not allow objects to break apart. This reduces a thick object to its corresponding object with one pixel width. Skeletonization or thinning often produces short spurs which can be cleaned up automatically with a pruning procedure [18]. The algorithm for determining fiber diameter uses a binary input image and creates its skeleton and distance transformed image. The skeleton acts as a guide for tracking the


a)

365

b)

c) Figure 4 a) A simple simulated image, b) its skeleton overlaid on its distance transform, c) a histogram of fiber diameter distribution obtained by distance transform.

distance transformed image by recording the intensities to compute the diameter at all points along the skeleton. This method was proposed by Pourdeyhimi et al. [13]. Figure 4 shows a simple simulated image, its skeleton overlaid on its distance transform and the histogram of fiber diameter obtained by this method. 1.2.3

Direct tracking

In order to measure the electrospun fiber diameter, we developed an image analysis-based method called direct tracking. This method in which a binary image is used as an input determines the fiber diameter on the basis of two scans; first a horizontal scan and then a vertical scan. In the first scan, the algorithm searches for the first white pixel adjacent to a black. Pixels are counted until reaching the first black. The second scan is then started from the mid point of a horizontal scan and pixels are counted until the first black is encountered. If the black pixel is not found, the direction changes. Having the number of horizontal and vertical scans, the number of pixels in perpendicular direction which is the fiber diameter could be measured from a geometrical relationship. The process is illustrated in Fig. 5. In electrospun webs, fibers cross each other at intersection points and this brings about some untrue measurements in these areas. To circumvent this problem, black regions are labeled and it is identified which couple of regions consists of fiber. Then, in order to enhance the processing speed, the image is cropped to the size of selected regions. Afterwards, fiber diameter is measured according to the explained algorithm. Finally, the data in pixels are converted to nm and the histogram of fiber diameter distribution is plotted. Figure 6 shows a labeled simple simulated image and the histogram of fiber diameter obtained by this method.

366


x y a)

d

b)

Figure 5 a) Direct tracking, b) fiber diameter from the number of vertical and horizontal pixels.

a)

b)

Figure 6 a) A simple simulated image which is labeled, b) histogram of fiber diameter distribution obtained by direct tracking.

1.3

Real webs treatment

Both the distance transform and direct tracking algorithm for measuring fiber diameter require binary image as input. Hence, micrographs first have to be converted to black and white. This is done by thresholding (known also as segmentation), which produces a binary image from a grayscale (intensity) image. This is a critical step because segmentation affects the result. In the simplest thresholding technique, called global thresholding, the image is partitioned using a single constant threshold. One simple way to choose a threshold is by trial and error. Then each pixel is labeled as an object or background depending on whether the gray level of that pixel is greater or less than the value of the threshold, respectively. The main problem here is that global thresholding can fail in the presence of nonuniform illumination or local gray level unevenness. An alternative to circumvent this problem is to use local thresholding instead. In this approach, the original image is divided to subimages, and different thresholds are used for segmentation. Another variation of this approach which has been used in this study consists of estimating the background illumination using morphological opening operation, subtracting the obtained background from the original image and applying a global thresholding to produce the binary version of the image. In order to automatically compute the appropriate threshold, Ostu’s method could be employed. This method chooses the threshold to minimize interaclass variance of black and white pixels. Prior to segmentation, an intensity adjustment operation and a two dimensional median filter were applied in order to enhance the contrast of the image and remove noise


a)

367

b)

c) Figure 7

a) A real web, b) global thresholding, c) local thresholding.

[18–21]. As it is shown in Fig. 7, global thresholding resulted in some broken fiber segments. This problem was solved using local thresholding.

2

Experimental

Electrospun nonwoven webs used as real webs in image analysis obtained from electrospinning of PVA with an average molecular weight of 72,000 g/mol, purchased from Merck, at different processing parameters. The micrographs of the webs were obtained using a Philips XL-30 environmental scanning electron microscope under magnification of 10,000× after being gold coated. Figure 8 shows micrographs of electrospun webs used as real webs.

3


Two sets each composed of five simulated images generated by µ-randomness procedure were used as samples with known characteristics to demonstrate the validity of the techniques. The first set had random orientation with increasing constant diameters; the second was also randomly oriented but with a varying diameter sampled from normal distributions with a mean of 15 pixels and standard deviations ranging from 2 to 10 pixels. Tables 1 and 2 show the structural features of these simulated images, which are shown in Figs. 9 and 10. Moreover, the applicability of the techniques was tested using five real webs obtained from electrospinning of PVA. Mean and standard deviation of the simulated images in the first and second set are shown in Tables 3 and 4, respectively. Table 5 shows the results for real webs in terms of pixel and nm. Figures 11 and 12 show histograms of fiber diameter distribution for simulated images in the first and second set, respectively. Histograms for real webs are given in Fig. 13. In the first set, for simulated images with the line thickness of 5 and 10 pixels, the distance transform presents mean and standard deviation of fiber diameter closer to the

368


R1

R2

R3

R4

R5

Figure 8 Micrographs of the electrospun webs.

simulation. For the line thickness of 15, the standard deviation of the diameter obtained from the direct tracking method is closer to that of the simulation. However, in this case the distance transform measured the average diameter more accurately. For the simulated webs with a line thickness of more than 15 in the first set, the direct tracking method resulted in a better estimation of the mean and standard deviation of fiber diameter. This is due to the fact that as the lines get thicker, there is a higher possibility of branching during the skeletonization (or thinning), and these branches remain even after pruning. Although these branches are small, their orientation is typically normal to the fiber axis, thus causing a widening of the distribution. Table 1 Structural characteristics of first set images. Image No

Angular range

Line density

Line thickness

C1 C2 C3 C4 C5

0–360 0–360 0–360 0–360 0–360

30 30 30 30 30

5 10 15 20 25

369


C1

C2

C3

C4

C5

Figure 9 Simulated images with constant diameter.

Table 2 Structural characteristics of second set images. Image No

V1 V2 V3 V4 V5

Angular range

0–360 0–360 0–360 0–360 0–360

Line density

30 30 30 30 30

Line thickness Mean

Standard

15 15 15 15 15

2 4 6 8 10

Furthermore, the distance transform method fails in measuring the diameter of fiber in intersections. Intersections cause to overestimate fiber diameter. Since in the direct tracking method an image is divided into parts where single fibers exist, the effect of intersections which causes an inaccurate measurement of fiber diameter is eliminated. Therefore, there will be a better estimate of fiber diameter.

370


V1

V2

V3

V4

V5

Figure 10

Simulated images with varying diameter.

In the second set, irrespective of the line thickness in the simulation, for all simulated webs direct tracking resulted in a better measurement of mean and standard deviation of fiber diameter. For real webs, mean and standard deviation of fiber diameter for direct tracking were closer to those of the manual method, which concurs with the trends observed for the simulated images. Table 3 Mean and standard deviation for series 1. C1

C2

C3

C4

C5

Simulation

mean standard

5 0

10 0

15 0

20 0

25 0

Distance transform

mean standard

5.486 1.089

10.450 2.300

16.573 5.137

23.016 6.913

30.063 10.205

Direct tracking

mean standard

5.625 1.113

11.313 2.370

17.589 4.492

22.864 5.655

29.469 7.241

371


C1

C2

C3

C4

C5

Figure 11 Histograms for simulated images with constant diameter.

Table 4 Mean and standard deviation for series 2. V1

V2

V3

V4

V5

Simulation

mean standard

15.247 1.998

15.350 4.466

15.243 5.766

15.367 8.129

16.628 9.799

Distance transform

mean standard

16.517 5.350

16.593 6.165

17.135 7.597

17.865 9.553

19.394 11.961

Direct tracking

mean standard

16.075 2.606

15.803 5.007

16.252 6.129

16.770 9.319

18.756 10.251

372


V1

V2

V3

V4

V3

V4

V5 Figure 12 Histograms for simulated images with varying diameter.

Table 5 Mean and standard deviation for real webs.

Manual

mean standard

Distance transform

mean standard

Direct tracking

mean standard

R1

R2

R3

R4

R5

pixel nm pixel nm

24.358 318.67 3.193 41.77

24.633 322.27 3.179 41.59

18.583 243.11 2.163 28.30

18.827 246.31 1.984 25.96

17.437 228.12 2.230 29.18

pixel nm pixel nm

27.250 356.49 8.125 106.30

27.870 364.61 7.462 97.62

20.028 262.01 4.906 64.18

23.079 301.94 7.005 91.64

20.345 266.17 6.207 81.21

pixel nm pixel nm

27.195 355.78 4.123 53.94

27.606 361.15 5.409 70.77

20.638 269.99 4.148 54.27

21.913 286.68 4.214 55.14

20.145 263.55 3.800 49.72


R1

R2

R3

R4

373

R5 Figure 13 Histograms for real webs.

4

Conclusion

Fiber diameter is the most important structural characteristics in electrospun nonwoven webs. The typical way of measuring electrospun fiber diameter is through the manual method which is a tedious, time-consuming and an operator-based method and cannot be used as an automated technique for quality control. We investigated the use of image analysis for determining the fiber diameter and developed an automated method called direct tracking. Since this is a new technique, its accuracy needs to be evaluated using samples with known characteristics. To that end, the µ-randomness procedure was used in order to simulate electrospun nonwoven webs. Based on this scheme, two sets of simulated images, each containing 5 webs, were generated. The first set had random orientation with increasing constant diameter. For the second set, the diameter values were sampled from normal distributions with a mean of 15 and standard deviation ranging from 2 to 10 pixels. We

374


compared our method with the distance transform method. For all the simulated webs with varying diameter and for those with constant diameter more than 15, the direct tracking method resulted in the mean and standard deviation closer to the simulation. However, for the simulated webs with smaller constant diameter, the distance transform measured the mean and standard deviation of fiber diameter more accurately. The results suggest that the direct tracking method is an accurate, direct measurement technique, because it extracts the fiber diameter for the samples by tracking fixed segment of the fiber and eliminates the effect of intersections. We have demonstrated the general applicability of the method using real webs. Five real electrospun nonwoven webs obtained by electrospinning of PVA were used. Since the methods needed binary images as input, the images first had to be segmented. A local thresholding method together with Ostu’s method was employed in order to automatically compute the appropriate threshold. The results obtained for real webs confirm the trends suggested by simulated images. The results show that the use of image analysis in order to determine the fiber diameter in electrospun nonwoven webs has been successful.

References 1. A. K. Haghi and M. Akbari, Phys. Stat. Solidi (a), 204, 1830–1834 (2007). 2. J. Doshi and D.H. Reneker, J. Electrostatics, 35, 151–160 (1995). 3. D.H. Reneker and I. Chun, Nanotechnology, 7, 216–223 (1996). 4. H. Fong and D.H. Reneker, Electrospinning and the Formation of Nanofibers, in: D.R. Salem, Structure Formation in Polymeric Fibers, Hanser, Cincinnati, pp. 225–246 (2001). 5. Th. Subbiah, G.S. Bhat, R.W. Tock, S. Parameswaran and S.S. Ramkumar, J. Appl. Polymer Sci., 96, 557–569 (2005). 6. W. Zhang, Z. Huang, E. Yan, Ch. Wang, Y. Xin, Q. Zhao and Y. Tong, Mater. Sci. Eng. A, 443, 292–295 (2007). 7. B. Pourdeyhimi, R. Ramanathan and R. Dent, Textile Res. J., 66, 713–722 (1996). 8. B. Pourdeyhimi, R. Ramanathan and R. Dent, Textile Res. J., 66, 747–753 (1996). 9. B. Pourdeyhimi, R. Dent and H. Davis, Textile Res. J., 67, 143–151 (1997). 10. B. Pourdeyhimi and R. Dent, Textile Res. J., 67, 181–190 (1997). 11. B. Pourdeyhimi, R. Dent, A. Jerbi, S. Tanaka and A. Deshpande, Textile Res. J., 69, 185–192 (1999). 12. B. Pourdeyhimi and H.S. Kim, Textile Res. J., 72, 803–809 (2002). 13. B. Pourdeyhimi and R. Dent, Textile Res. J., 69, 233–236 (1999). 14. B. Xu and Y.L. Ting, Textile Res. J., 65, 41–48 (1995). 15. A.H. Aydilek, S.H. Oguz and T.B. Edil, J. Comput. Civil Eng., 280–290 (2002). 16. R. Chhabra, Int. Nonwoven J., 43–50 (2003). 17. M.S. Abdel-Ghani and G.A. Davis, Chem. Eng. Sci., 40, 117–129 (1985). 18. R.C. Gonzalez and R.E. Woods, Digital Image Processing, 2nd edn, Prentice Hall (2001). 19. B. Jähne, Digital Image Processing, 5th revised and extended edn, Springer (2002). 20. M. Petrou and P. Bosdogianni, Image Processing: the Fundamentals, Wiley (1999). 21. J. Serra, Image Analysis and Mathematical Morphology, Academic Press, London (1982).

статья одобрена автором больше не смотреть Всё сделано - только номера страниц

13

Image Analysis of Pore Size Distribution in Electrospun Nanofiber Webs: New Trends and Developments M. Ziabari, V. Mottaghitalab and A.K. Haghi* University of Guilan, P. O. Box 3756, Rasht, Iran *[email protected]

Nanofibers produced by the electrospinning method are widely used for drug delivery, as tissue scaffolding materials and filtration purposes where specific pore characteristics are required. For continued growth in these areas, it is critical that nanofibers be properly designed for these applications to prevent failure. Most of the current methods only provide an indirect way of determining the pore structure parameters and contain inherent disadvantages. In this study, we developed a novel image analysis method for measuring the pore characteristics of electrospun nanofiber webs. Five electrospun webs with different pore characteristics were analyzed by this method. The method is direct and rapid and presents valuable and comprehensive information regarding the pore structure parameters of webs. Two sets of simulated images were generated to study the effects of web density, fiber diameter and its variations on pore characteristics. The results indicated that web density and fiber diameter significantly influence the pore characteristics whereas the effect of fiber diameter variations was insignificant.

Keywords: image analysis, pore size, nanofibers, electrospinning

Introduction Fibers with a diameter of around 100 nm are generally classified as nanofibers. What makes nanofibers of great interest is their extremely small size. Nanofibers compared to conventional fibers, with higher surface area to volume ratios and smaller pore size, offer an opportunity for use in a wide variety of applications. To date, the most successful method of

376


Figure 1 Electrospinning setup.

producing nanofibers is through the process of electrospinning. The electrospinning process uses high voltage to create an electric field between a droplet of polymer solution at the tip of a needle and a collector plate. When the electrostatic force overcomes the surface tension of the drop, a charged, continuous jet of polymer solution is ejected. As the solution moves away from the needle and toward the collector, the solvent evaporates and the jet rapidly thins and dries. On the surface of the collector, a nonwoven web of randomly oriented solid nanofibers is deposited [1 – 5]. Figure 1 illustrates the electrospinning setup. Material properties such as melting temperature and glass transition temperature as well as structural characteristics of nanofiber webs such as fiber diameter distribution, pore size distribution and fiber orientation distribution determine the physical and mechanical properties of the webs. The surface of electrospun fibers is important when considering end-use applications. For example, the ability to introduce porous surface features of a known size is required if nanoparticles need to be deposited on the surface of the fiber, if drug molecules are to be incorporated for controlled release, as tissue scaffolding materials and for acting as a cradle for enzymes [6]. Besides, filtration performance of nanofibers is strongly related to their pore structure parameters, i.e., percent open area (POA) and pore-opening size distribution (PSD). Hence, the control of the pore of electrospun webs is of prime importance for nanofibers produced for these purposes. There is no literature available about the pore size and its distribution of electrospun fibers and in this work, the pore size and its distribution was measured using an image analysis technique. Current methods for determining PSD are mostly indirect and contain inherent disadvantages. Recent technological advancements in image analysis offer great potential for a more accurate and direct way of determining the PSD of electrospun webs. Overall, the image analysis method provides a unique and accurate method that can measure pore opening sizes in electrospun nanofiber webs.

1

Methodology

The porosity, εV, is defined as the percentage of the volume of the voids, Vv, to the total volume (voids plus constituent material), Vt, and is given by

εV =

Vv ×100 . Vt

Image Analysis of Pore Size Distribution in Electrospun Nanofiber Webs

Figure 2

377

Equivalent opening size, Oi, based on (a) equivalent area, (b) equivalent size.

Similarly, the Percent Open Area (POA), εA, that is defined as the percentage of the open area, Ao, to the total area At, is given by

εA =

Ao × 100 . At

Usually porosity is determined for materials with a three-dimensional structure, e.g., relatively thick nonwoven fabrics. Nevertheless, for two-dimensional textiles such as woven fabrics and relatively thin nonwovens it is often assumed that porosity and POA are equal [7]. The size of an individual opening can be defined as the surface area of the opening, although it is mostly indicated with a diameter called Equivalent Opening Size (EOS). EOS is not a single value and may differ for each opening. The commonly used term in this case is the diameter, Oi, corresponding to the equivalent circular area, Ai, of the opening. Oi = (4 Ai / π )1/ 2 . This diameter is greater than the side dimension of a square opening. A spherical particle with that diameter will never pass the opening (Fig. 2a) and may therefore not be considered as an equivalent dimension or equivalent diameter. This will only be possible if the diameter corresponds to the side of the square area (Fig. 2b). However, not all openings are squares, yet the equivalent square area of openings is used to determine their equivalent dimension because this simplified assumption results in one single opening size from the open area. It is the diameter of a spherical particle that can pass the equivalent square opening, hence the equivalent opening or pore size, Oi, results from Oi = ( Ai )1/ 2 . From the EOSs, Pore Size Distribution (PSD) and an equivalent diameter for which a certain percentage of the opening have a smaller diameter (Ox, pore opening size that x percent of pores are smaller than that size) may be measured. The PSD curves can be used to determine the uniformity coefficient, Cu, of the investigated materials. The uniformity coefficient is a measure for the uniformity of the openings and is given by Cu = O60 O10 .

378


The ratio equals unity for uniform openings and increases with decreasing uniformity of the openings [7]. Pore characteristic is one of the main tools for evaluating the performance of any nonwoven fabric and for electrospun webs as well. Understanding the link between processing parameters and pore structure parameters will allow for better control over the properties of electrospun fibers. Therefore there is a need for the design of nanofibers to meet specific application needs. Various techniques may be used to evaluate pore characteristics of porous materials including sieving techniques (dry, wet and hydrodynamic sieving), mercury porosimetry and flow porosimetry (bubble point method) [8, 9]. As one goes about selecting a suitable technique for characterization, the associated virtues and pitfalls of each technique should be examined. The most attractive option is a single technique which is nondestructive, yet capable of providing a comprehensive set of data [10]. 1.1

Sieving methods

In dry sieving, glass bead fractions (from finer to coarser) are sieved through the porous material. In theory, most of the glass beads from the first glass bead fraction should pass. As larger and larger glass bead fractions are sieved, more and more glass beads should become trapped within and on top of the material. The number of pores of a certain size should be reflected by the percentage of glass beads passing through the porous material during each glass bead fraction sieved; however, electrostatic effects between glass beads and between glass beads and the material can affect the results. Glass beads may stick to fibers making the pores effectively smaller and they may also agglomerate to form one large glass bead that is too large to pass through the any of the pores. Glass beads may also break from hitting each other and the sides of the container, resulting in smaller particles that can pass through smaller openings. In hydrodynamic sieving, a glass bead mixture is sieved through a porous material under alternating water flow conditions. The use of glass bead mixtures leads to results that reflect the original glass bead mixture used. Therefore, this method is only useful for evaluating the large pore openings such as O95. Another problem occurs when particles of many sizes interact, which likely results in particle blocking and bridge formation. This is especially a problem in hydrodynamic sieving because the larger glass bead particles will settle first when water is drained during the test. When this occurs, fine glass beads which are smaller than the pores are prevented from passing through by the coarser particles. In wet sieving, a glass bead mixture is sieved through a porous material aided by a water spray. The same basic mechanisms that occur when using the hydrodynamic sieving method also take place when using the wet sieving method. Bridge formation is not as pronounced in the wet sieving method as in the hydrodynamic sieving method; however, particle blocking and glass bead agglomeration are more pronounced [8, 9]. The sieving tests are very time-consuming. Generally two hours are required to perform a test. The sieving tests are far from providing a complete PSD curve because the accuracy of the tests for pore sizes smaller than 90 µm is questionable [12]. 1.2

Mercury porosimetry

Mercury porosimetry is a well known method which is often used to study porous materials. This technique is based on the fact that mercury as a non-wetting liquid does not intrude into pore spaces except under applying sufficient pressure. Therefore, a relationship can be


379

found between the size of pores and the pressure applied. In this method, a porous material is completely surrounded by mercury and pressure is applied to force the mercury into pores. As mercury pressure increases the large pores are filled with mercury first. Pore sizes are calculated as the mercury pressure increases. At higher pressures, mercury intrudes into the fine pores and when the pressure reaches a maximum, total open pore volume and porosity are calculated. The mercury porosimetry thus gives a PSD based on total pore volume and gives no information regarding the number of pores of a porous material. Pore sizes ranging from 0.0018 to 400 µm can be studied using mercury porosimetry. Pore sizes smaller than 0.0018 µm are not intruded with mercury and this is a source of error for porosity and PSD calculations. Furthermore, mercury porosimetry does not account for closed pores as mercury does not intrude into them. Due to applying high pressures, sample collapse and compression is possible, hence it is not suitable for fragile compressible materials such as nanofiber sheets. Other concerns would include the fact that it is assumed that the pores are cylindrical, which is not the case in reality. After the mercury intrusion test, sample decontamination at specialized facilities is required as highly toxic mercury is trapped within the pores. Therefore this dangerous and destructive test can only be performed in well-equipped laboratories [6, 8, 9]. 1.3

Flow porosimetry (bubble point method)

The flow porosimetry is based on the principle that a porous material will only allow a fluid to pass when the pressure applied exceeds the capillary attraction of the fluid in largest pore. In this test, the specimen is saturated with a liquid and continuous air flow is used to remove liquid from the pores. At a critical pressure, the first bubble will come through the largest pore in the wetted specimen. As the pressure increases, the pores are emptied of liquid in order from largest to smallest and the flow rate is measured. PSD, number of pores and porosity can be derived once the flow rate and the applied pressure are known. Flow porosimetry is capable of measuring pore sizes within the range of 0.013–500 µm. As the air only passes through the through pores, the characteristics of these pores are measured while those of closed and blind pores are omitted. Many times, 100% total flow is not achieved. This is due to porewick evaporation from the pores when the flow rate is too high. Extreme care is required to ensure the air flow does not disrupt the pore structure of the specimen. The flow porosimetry method is also based on the assumption that the pores are cylindrical, which is not the case in reality. Finding a liquid with low surface tension, which could cover all the pores, has no interaction with material and does not cause swelling in material is not easy all the times and sometimes is impossible [6, 8, 9]. 1.4

Image analysis

Because of its convenience to detect individual pores in a nonwoven image, it seemed to be advantageous to use image analysis techniques for pore measurement. Image analysis was used to measure pore characteristics of woven [11] and nonwoven geotextiles [12]. In the former, successive erosion operations with increasing size of structuring element was used to count the pore openings larger than a given structuring element. The main purpose of the erosion was to simulate the conditions in the sieving methods. In this method, the voids connected to border of the image which are not complete pores are considered in measurement. Performing opening and then closing operations proceeding pore measurement cause the

380


pore sizes and shapes deviate from the real ones. The method is suitable for measuring pore sizes of woven geotextiles with fairly uniform pore sizes and shapes and is not appropriate for electrospun nanofiber webs of different pore sizes. In the later case, cross sectional image of nonwoven geotextile was used to calculate the pore structure parameters. A slicing algorithm based on a series of morphological operations for determining the mean fiber thickness and the optimal position of the uniform slicing grid was developed. After recognition of the fibers and pores in the slice, the pore opening size distribution of the cross sectional image may be determined. The method is useful for measuring pore characteristics of relatively thick nonwovens and cannot be applied to electrospun nanofiber webs due to extremely small size. Therefore, there is a need for developing an algorithm suitable for measuring the pore structure parameters in electrospun webs. In response to this need, we have developed a new image analysis based method and presented in the following. In this method, a binary image of the web is used as an input. First of all, voids connected to the image border are identified and cleared using morphological reconstruction [13], where mask image is the input image and marker image is zero everywhere except along the border. Total area which is the number of pixels in the image is measured. Then the pores are labeled and each considered as an object. Here the number of pores may be obtained. In the next step, the number of pixels of each object as the area of that object is measured. Having the area of pores, the porosity and EOS regarding to each pore may be calculated. The data in pixels may then be converted to nm. Finally PSD curve is plotted and O50, O95 and Cu are determined. 1.4.1

Real webs

In order to measure pore characteristics of electrospun nanofibers using image analysis, images of the webs are required. These images called micrographs usually are obtained by a scanning electron microscope (SEM), transmission electron microscope (TEM) or atomic force microscope (AFM). The images must be of high-quality and taken under appropriate magnifications. The image analysis method for measuring pore characteristics requires the initial segmentation of the micrographs in order to produce binary images. This is a critical step because the segmentation affects the results dramatically. The typical way of producing a binary image from a grayscale image is by global thresholding [13], where a single constant threshold is applied to segment the image. All pixels up to and equal to the threshold belong to object and the remaining belong to the background. One simple way to choose the threshold is picking different thresholds until one is found that produces a good result as judged by the observer. Global thresholding is very sensitive to any inhomogeneities in the gray-level distributions of object and background pixels. In order to eliminate the effect of inhomogeneities, the local thresholding scheme [13] could be used. In this approach, the image is divided into subimages where the inhomogeneities are negligible. Then optimal thresholds are found for each subimage. A common practice in this case, which is used in this study, is to preprocess the image to compensate for the illumination problems and then apply a global thresholding to the preprocessed image. It can be shown that this process is equivalent to segment the image with locally varying thresholds. In order to automatically select the appropriate thresholds, Otsu’s method [14] is employed. This method chooses the threshold to minimize interaclass variance of the black and white pixels. As it is shown in Fig. 3 global thresholding resulted in some broken fiber segments. This problem was solved


a)

381

b)

c)

Figure 3

a) A real web, b) Global thresholding, c) Local thresholding.

using local thresholding. Note that, since the process is extremely sensitive to noise contained in the image, before the segmentation, a procedure to clean the noise and enhance the contrast of the image is necessary. 1.4.2

Simulated webs

In is known that the pore characteristics of nonwoven webs are influenced by web properties and so are those of electrospun webs. There are no reliable models available for predicting these characteristics as a function of web properties [15]. In order to explore the effects of some parameters on pore characteristics of electrospun nanofibers, simulated webs are generated. These webs are images simulated by straight lines. There are three widely used methods for generating random network of lines. These are called S-randomness, µ-randomness (suitable for generating a web of continuous filaments) and I-randomness (suitable for generating a web of staple fibers). These methods have been described in detail by Abdel-Ghani et al. [16] ] and Pourdeyhimi et al. [17]. In this study, we used µ-randomness procedure for generating simulated images. Under this scheme, a line with a specified thickness is defined by the perpendicular distance d from a fixed reference point O located in the center of the image and the angular position of the perpendicular α. Distance d is limited to the diagonal of the image. Figure 4 demonstrates this procedure.

d

Figure 4

α O

Procedure for µ-randomness.

382


No.1

No.2

No.3

No.4

No.5

Figure 5 Micrographs of the electrospun webs.

One of the most important features of simulation is that it allows several structural characteristics to be taken into consideration with the simulation parameters. These parameters are: web density (controlled as line density), angular density (sampled from a normal or random distribution), distance from the reference point (sampled from a random distribution), line thickness (sampled from a normal distribution) and image size.

2

Experimental

Nanofiber webs were obtained from electrospinning of PVA with average molecular weight of 72000 g/mol (Merck) at different processing parameters for attaining different pore characteristics. Table 1 summarizes the electrospinning parameters used for preparing the webs. The micrographs of the webs were obtained using a Philips XL-30 environmental scanning electron microscope under magnification of 10000× after being gold coated. Figure 5 shows micrographs of the electrospun webs. Table 1 Electrospinning parameters used for preparing nanofiber webs. No

Concentration (%)

Spinning distance (Cm)

Voltage (kV)

Flow rate (ml/h)

1 2 3 4 5

8 12 8 8 10

15 20 15 10 10

20 15 20 15 15

0.4 0.2 0.2 0.3 0.2

383


Figure 6

3

PSD curves of electrospun webs.


Due to previously mentioned reasons, sieving methods and mercury porosimetry are not applicable for measuring pore structure parameters in nano-scale. The only method which seems to be practical is flow porosimetry. However, since in this study, the nanofibers were made of PVA, finding an appropriate liquid for the test to be performed is almost impossible because of solubility of PVA in both organic and inorganic liquids. As an alternative, image analysis was employed to measure pore structure parameters in electrospun nanofiber webs. PSD curves of the webs, determined using the image analysis method, are shown in Fig. 6. Pore characteristics of the webs (O50, O95, Cu, number of pores, porosity) measured by this method are presented in Table 2. It is seen that decreasing Table 2 Pore characteristics of electrospun webs. No

1 2 3 4 5

O95

O50 pixel

nm

pixel

nm

39.28 27.87 26.94 22.09 19.26

513.9 364.7 352.5 289.0 252.0

94.56 87.66 64.01 60.75 44.03

1237.1 1146.8 837.4 794.8 576.1

Cu

Pore No

Porosity

8.43 5.92 3.73 3.68 2.73

31 38 64 73 69

48.64 34.57 26.71 24.45 15.74

the porosity, O50 and O95 decrease. Cu also decreases with respect to porosity, that is to say increasing the uniformity of the pores. The number of pores has an increasing trend with decreasing porosity. The image analysis method presents valuable and comprehensive information regarding to pore structure parameters in nanofiber webs. This information may be exploited in

384


preparing the webs with needed pore characteristics to use in filtration, biomedical applications, nanoparticle deposition and other purposes. The advantages of the method are listed below: 1) The method is capable of measuring pore structure parameters in any nanofiber webs with any pore features and it is applicable even when other methods may not be employed. 2) It is so fast. It takes less than a second for an image to be analyzed (with a 3 GHz processor). 3) The method is direct and so simple. Pore characteristics are measured from the area of the pores which is defined as the number of pixels of the pores. 4) There is no systematic error in measurement (such as assuming pores to be cylindrical in mercury and flow porosimetry and the errors associated with the sieving methods which were mentioned). Once the segmentation is successful, the pore sizes will be measured accurately. The quality of images affects the segmentation procedure. High-quality images reduce the possibility of poor segmentation and enhance the accuracy of the results. 5) It gives a complete PSD curve. 6) There is no cost involved in the method and minimal technical equipments are needed (SEM for obtaining the micrographs of the samples and a computer for analysis). 7) It has the capability of being used as an on-line quality control technique for large scale production. 8) The results obtained by image analysis are reproducible. 9) It is not a destructive method. A very small amount of sample is required for measurement. In an attempt to establish the effects of some structural properties on pore characteristics of electrospun nanofibers, two sets of simulated images with varying properties were generated. The simulated images reveal the degree to which fiber diameter and density affect the pore structure parameters. The first set contained images with the same density varying in fiber diameter and images with the same fiber diameter varying in density. Each image had a constant diameter. The second set contained images with the same density and mean fiber diameter while the standard deviation of fiber diameter varied. The details are given in Tables 3 and 4. Typical images are shown in Figs. 7 and 8. Table 3 Structural characteristics of first set images. No

Angular range

Line density

Line thickness

1 2 3 4 5 6 7 8 9

0–360 0–360 0–360 0–360 0–360 0–360 0–360 0–360 0–360

20 30 40 20 30 40 20 30 40

5 5 5 10 10 10 20 20 20

385


Table 4 Structural characteristics of second set images. No

Angular range

1 2 3 4

0–360 0–360 0–360 0–360

Line density

30 30 30 30

No.1

No.2

No.3

No.4

No.5

No.6

No.7

No.8

No.9

Figure 7 Simulated images of the first set.

Line thickness Mean

Standard

15 15 15 15

0 4 8 10

386


No.1

No.2

No.3

No.4

Figure 8

Simulated images of the second set.

Pore structure parameters of simulated webs were measured using the image analysis method. Table 5 summarizes the pore characteristics of the simulated images in the first set. For webs with the same density, increasing the fiber diameter resulted in a decrease in O95, number of pores and porosity. No particular trends were observed for O50 and Cu. Figures 9 and 10 show the PSD curves of the simulated images in the first set. As the web density increases, the effects of fiber diameter are less pronounced since the PSD curves of the webs become closer to each other.For webs with the same fiber diameter, the density increase resulted in a decrease in O50, O95, Cu and porosity whereas the number of pores increased with the density. Table 5 Pore characteristics of the first set of simulated images. No

O50

O95

Cu

Pore No

Porosity

1 2 3 4 5 6 7 8 9

27.18 15.52 13.78 36.65 17.89 12.41 24.49 16.31 13.11

100.13 67.31 52.32 94.31 61.64 51.60 86.90 56.07 45.38

38.38 22.20 18.71 43.71 22.67 16.70 33.11 21.66 17.75

84 182 308 67 144 245 58 108 126

79.91 71.78 69.89 66.10 53.67 47.87 41.05 32.53 22.01

Table 6 Pore characteristics of the second set of simulated images. No

O50

O95

Cu

Pore No

Porosity

1 2 3 4

14.18 13.38 18.14 15.59

53.56 61.66 59.35 62.71

18.79 20.15 22.07 20.20

133 136 121 112

35.73 41.89 41.03 37.77


387

a)

b)

c) Figure 9 PSD curves of the first set of simulated images; effect of density, images with the diameter of a) 5, b) 10, c) 20 pixels.

Table 6 summarizes the pore characteristics of the simulated images in the second set set. No significant effects for variation of fiber diameter on pore characteristics were observed. Suggesting that average fiber diameter is determining factor not variation of

388


a)

b)

c)

Figure 10 PSD curves of the first set of simulated images; effect of fiber diameter, images with the density of a) 20, b) 30, c) 40 lines.

diameter. Figure 11 shows the PSD curves of the simulated images in the second set.

4

Conclusion

The evaluation of electrospun nanofiber pore structure parameters is necessary as it facilitates the improvement of the design process and its eventual applications. Various


389

Figure 11 PSD curves of the second set of simulated images, the effect of fiber diameter variation.

techniques have been developed to assess pore characteristics in porous materials. However, most of these methods are indirect, have inherent problems and are not applicable for measuring pore structure parameters of electrospun webs. In this investigation, we have successfully developed an image analysis-based method as a response to this need. The method is simple, comprehensive and fast and directly measures the pore structure parameters. The effects of web density, fiber diameter and its variation on pore characteristics of the webs were also explored using some simulated images. As the fiber diameter increased, O95, the number of pores and porosity decreased. No particular trends were observed for O50 and Cu. Increasing the density resulted in a decrease in O50, O95, Cu and porosity whereas the number of pores increased with density. The effects of variation of the fiber diameter on pore characteristics were insignificant.

References 1. A.K. Haghi and M. Akbari, Phys. Stat. Solidi (a), 204, 1830–1834 (2007). 2. D.H. Reneker and I. Chun, Nanotechnology, 7, 216–223 (1996). 3. D.R. Salem, Structure Formation in Polymeric Fibers, Hanser, Cincinnati, Chapter 6, H. Fong and D.H. Reneker, Electrospinning and the Formation of Nanofibers (2001). 4. Th. Subbiah, G.S. Bhat, R.W. Tock, S. Parameswaran and S.S. Ramkumar, J. Appl. Polymer Sci., 96, 557–569 (2005). 5. A. Frenot and I.S. Chronakis, Curr. Op. Coll. Interface Sci., 8, 64–75 (2003). 6. Ch.L. Casper, J.S. Stephens, N.G. Tassi, D.B. Chase and J.F. Rabolt, Macromolecules, 37, 573–578 (2004). 7. W. Dierickx, Geotext. Geomembr., 17 (4), 231–245 (1999). 8. S.K. Bhatia and J.L. Smith, Geosynthetics Int., 3 (2), 155–180 (1996). 9. S.K. Bhatia, J.L. Smith and B.R. Christopher, Geosynthetics Int., 3 (3), 301–328 (1996). 10. S.T. Ho and D.W. Hutmacher, Biomaterials, 27, 1362–1376 (2006). 11. A.H. Aydilek and T.B. Edil, Geotechn. Testing J., 27 (1), 1–12 (2004). 12. A.H. Aydilek, S.H. Oguz and T.B. Edil, J. Comput. Civil Eng., 280–290 (2002). 13. R.C. Gonzalez and R.E. Woods, Digital Image Processing, Prentice Hall, New Jersey, second

390


edn (2001). 14. B. Jähne, Digital Image Processing, Springer, England, 5th revised and extended edn (2002). 15. H.S. Kim and B. Pourdeyhimi, Int. Nonwoven J., 15–19 (Winter 2000). 16. M.S. Abdel-Ghani and G.A. Davis, Chem. Eng. Sci., 117 (1985). 17. B. Pourdeyhimi, R. Ramanathan and R. Dent, Textile Res. J., 66 (11), 713–722 (1996).

AUTHORS: - Figure 13 still not mentioned in the text - a list of keywords is missing - two Figures 22 (the second Fig. 22 -- > Fig. 23?) - check the items highlighted in red; approve the complete paper

14

Electrospun Biodegradable and Biocompatible Natural Nanofibers: A Detailed Review A.K. Haghi and R.K. Haghi University of Guilan, P.O. Box 3756, Rasht, Iran email: [email protected]

This chapter presents a detailed review on processing of biodegradable and biocompatible natural nanofibers.

1

Introduction

An emerging technology of manufacturing of thin natural fibers is based on the principle of the electrospinning process. In conventional fiber spinning, the mechanical force is applied to the end of a jet, whereas in the electrospinning process the electric body forces act on an element of charged fluid. Electrospinning has emerged as a specialized processing technique for the formation of submicron fibers (typically between 100 nm and 1 µm in diameter), with high specific surface areas. Due to their high specific surface area, high porosity and small pore size, the unique fibers have been suggested for a wide range of applications. Electrospinning of natural fibers offers unique capabilities for producing novel natural nanofibers and fabrics with controllable pore structure. Current research effort has focused on understanding the electrospinning of natural fibers in which the influence of different governing parameters are discussed. Electrospinning [1–10] is an economical and simple method used in the preparation of polymer fibers. The fibers prepared by this method typically have diameters much smaller than is possible to attain using standard mechanical fiber-spinning technologies [11]. Electrospinning of polymer solutions has gained much attention in the last few years as a

392

A.K. Haghi and R.K. Haghi

Figure 1 Schematic of the electrospinning setup.

cheap and straightforward method to produce nanofibers [12 – 16]. Electrospinning differs from the traditional wet/dry fiber spinning in a number of ways, of which the most striking differences are the origin of the pulling force and the final fiber diameters. The mechanical pulling forces in the traditional industrial fiber spinning processes lead to fibers in the micrometer range and are contrasted in electrospinning by electrical pulling forces that enable the production of nanofibers [14]. Depending on the solution properties, the throughput of single-jet electrospinning systems ranges around 10 ml/min. This low fluid throughput may limit the industrial use of electrospinning. A stable cone-jet mode followed by the onset of the characteristic bending instability, which eventually leads to a great reduction in the jet diameter, necessitate the low flow rate [6, 17]. When the diameters of polymer fiber materials are shrunk from micrometers (e.g. 10 – 100 µm) to submicrons or nanometers, there appear several amazing characteristics such as a very large surface area to volume ratio (this ratio for a nanofiber can be as large as 103 times of that of a microfiber), flexibility in surface functionalities, and superior mechanical performance (e.g. stiffness and tensile strength) compared with any other known form of material. These outstanding properties make the polymer nanofibers optimal candidates for many important applications [18]. These include filter media [19–21], composite materials [18, 22], biomedical applications (tissue engineering scaffolds [23–27], bandages [28], drug release systems [29–30]), protective clothing for the military [31–34], optoelectronic devices and semiconductor materials [35–38], biosensors/chemosensors [39–42]. 1.1

Electrospinning setup

A schematic diagram to interpret electrospinning of polymer nanofibers is shown in Fig. 1. There are basically three components to fulfill the process: a high voltage supplier, a capillary tube with a pipette or needle of small diameter, and a metal collecting screen. In the electrospinning process, a high voltage is used to create an electrically charged jet of polymer solution or melt out of the pipette. Before reaching the collecting screen, the solution jet evaporates or solidifies, and is collected as an interconnected web of small fibers [43–45]. One electrode is placed into the spinning solution/melt or needle and the other is attached to the collector. In most cases, the collector is simply grounded. The electric field is applied to the end of the capillary tube that contains the solution fluid held by its surface tension. This induces a charge on the surface of the liquid. Mutual charge repulsion and the

Electrospun Biodegradable and Biocompatible Natural Nanofibers

393

contraction of the surface charges to the counter electrode cause a force directly opposite to the surface tension [46]. As the intensity of the electric field is increased, the hemispherical surface of the fluid at the tip of the capillary tube elongates to form a conical shape known as the Taylor cone [47]. Further increasing the electric field, a critical value is attained with which the repulsive electrostatic force overcomes the surface tension and the charged jet of the fluid is ejected from the tip of the Taylor cone [18, 48]. The jet exhibits bending instabilities due to repulsive forces between the charges carried with the jet. The jet extends through spiraling loops, as the loops increase in diameter the jet grows longer and thinner until it solidifies or collects on the target [49].

2

Effect of systematic parameters on electrospun nanofibers

It has been found that the morphology such as fiber diameter and uniformity of electrospun polymer fibers are dependent on many processing parameters [50]. These parameters can be divided into 3 groups: a) solution properties, b) processing conditions, c) ambient conditions. Each of the parameters was found to affect the morphology of electrospun fibers. 2.1

Solution properties

Solution parameters such as viscosity of solution, polymer concentration, molecular weight of polymer, electrical conductivity, elasticity and surface tension, which attribute to polymer and its solution characteristics, have an important effect on the morphology [18]. 2.1.1

Viscosity

The viscosity range of different polymer solutions which are spinnable is different. One of the most significant parameters influencing the fiber diameter is the solution viscosity. A higher viscosity results in a large fiber diameter. Figure 2 shows the representative images of beads and beaded fibers for solutions with a range of viscosity. Beads and beaded fibers are less likely to be formed for more viscous solutions. The diameter of beads becomes bigger and the average distance between beads on fibers longer as viscosity increases. Meanwhile, the shape of beads gradually changes from spherical to spindle-like [51]. 2.1.2

Solution concentration

The change of the diameter of electrospun fibers as a function of electrical conductivity of the solutions is shown in Fig. 3. The diameter dropped significantly when the conductivity increased. Beads were also observed due to the low conductivity of the solution, which resulted in an insufficient elongation of a jet by electrical force to produce uniform fibers. As the solution concentration increase, a mixture of beads and fibers is obtained [52]. Higher solution concentration may result in fewer beads. The shape of the beads changed from spherical to spindle-like when the solution concentration varied from low to high levels [18, 53]. The fiber distribution became gradually broader with increasing concentration, which is consistent with the results obtained by Ryu et al. and Kidoaki et al. [54]. The fiber diameter increased with increasing solution concentration because the higher viscosity resisted the extension of the jet [50, 55]. However, it is impossible to electrospin if the

394


Figure 2

The morphology of beaded fibers versus solution viscosity [51].

20

Figure 3 Solution conductivity effects on the diameter of the electrospun P(LLA-CL) (70/30 wt. %) fibers [50].

solution concentration or the corresponding viscosity is too high due to the difficulty in liquid jet formation [56–58]. Figure 4 shows the morphology of fibers obtained from solutions of 6, 8, 10 and 12 wt. % concentration at applied voltage 10 kV with a constant tip-to-collector distance of 10 cm. Spindle-like beads formed for fibers obtained from solution of 6 wt. %.


a

b

c

395

d

Figure 4 The morphology of nanofibers at concentrations from 6% to 12% with a constant tip-tocollector distance of 10 cm at applied voltage 10 kV. (The values below the images and the brackets show the average fiber diameter (nm) and the standard deviation of fiber diameter. The scale bars for fibers from solution from 6 to 10 wt.% and 12 wt.% are 2 mm and 5 mm, respectively.). above: Distribution of fiber diameters obtained at 10 kV with a tip-to-collector distance of 10 cm. The concentrations for (a), (b), (c) and (d) are 6, 8, 10 and 12 wt. %, respectively [55].

2.1.3

Molecular weight

Molecular weight also has a significant effect on the rheological and electrical properties such as viscosity, surface tension, conductivity and dielectric strength. It has been reported that too low molecular weight polymers tend to form beads rather than fibers and high molecular weight polymers yield fibers with a larger average diameter [59]. Note that in Figure 5, as the molecular weight increases, several changes can be observed in the electrospun structure of PVA. At a molecular weight of 9000–13,000 g/mol,

Figure 5 Photographs showing the typical structure in the electrospun polymer for various molecular weights. The samples were obtained at the center of the deposition area. (a) 9000–10,000 g/mol; (b) 13,000–23,000 g/mol; and (c) 31,000–50,000 g/mol (solution concentration: 25 wt. %) [58].

396


Figure 6 [60].

Solution viscosity and surface tension as a function of polymer solution concentration

the fibrous structure is not completely stabilized and a bead-on-string structure is obtained, indicating the resistance of the jet to extensional flow.

2.1.4

Surface tension

Surface tension seems more likely to be a function of solvent compositions, but is negligibly dependent on the solution concentration. Different solvents may contribute different surface tensions. However, not necessarily a lower surface tension of a solvent will always be more suitable for electrospinning. Generally, surface tension determines the upper and lower boundaries of electrospinning window if all other variables are held constant. Formation of droplets, beads and fibers can be driven by the surface tension of the solution and lower surface tension of the spinning solution helps electrospinning to occur at a lower electric field [57].

2.1.5

Number of entanglements

The above variables are not independent of each other. Solution viscosity is a function of both concentration and polymer molecular weight. These two parameters also affect the number of entanglements [52]. Figure 6 shows the viscosity and surface tension of solution dissolved in formic acid; they were 400–5550 cps and 49–68 mN/m, respectively. These values increased with increasing of concentration. Specifically, the viscosity increased from 1400 to 5100 cps for solution concentration from 25 to 30 wt. % [60]. 2.1.6

Solution conductivity

Jun et al. [56] have shown that increasing the solution conductivity by addition of a salt can significantly aid fiber formation (Fig. 7). Addition of a salt has a positive effect on the electrospinning number (Vq/γR2). Specifically, the electrical energy (Vq) increases. On the other hand, the change in the surface free energy due to a change in surface tension is not


397

Figure 7 Effect of NaCl amount in the PVA solution on fiber morphology (DH = 98%, voltage = 5 kV, tip–target distance = 10 cm; flow rate = 0.2 ml/h). NaCl amount based on H2O: (a) 0.05%, (b) 0.10%, (c) 0.15%, (d) 0.2%. Original magnification 10,000 [55].

Figure 8 Electric conductivity as a function of polymer solution concentration [60].

expected to be significant. As the charges carried by the jet increased, higher elongation forces were imposed to the jet under the electrical field, resulting in smaller beads and thinner fiber diameters (Fig. 8). A comparison of the diameter of the electrospun fibers with the electrical conductivity of the solutions is shown in Fig. 9. There was a significant drop in the diameter of the electrospun polymer fibers when the electrical conductivity of the solution increased. Beads were also observed due to low conductivity of the solution, which results in insufficient elongation of a jet by electrical force to produce uniform fiber. Electrospun polymer nanofibers with the smallest fiber diameter were obtained with the highest electrical conductivity. This suggests that the drop in the size of the fibers was due to the increased electrical conductivity [50]. 2.1.7

Effect of salt addition

Figure 10 shows the SEM images of PAA nanofibers fabricated by electrospinning into solutions with different ionic strengths. Magnified PAA nanofibers are also shown. At a concentration of 1 M NaCl, no PAA nanofibers were synthesized. Bead structures were

398


Figure 9 Solution conductivity effects on the diameter of the electrospun P(LLA-CL) (70/30 wt. %) fibers [50].

Figure 10

SEM images of PAA nanofibers with different concentrations of NaCl [61].

observed from all prepared PAA nanofibers. With increasing the concentration of NaCl, the relative viscosity slightly decreased. Variations in viscosity may be caused by the chain conformation change of PAA in solution upon adding NaCl. With increasing the ionic strength of the solution, the chain conformation of PAA changes from the extended linear conformation to the coil conformation because PAA is a polyelectrolyte. Although relative viscosity decreases slightly with increasing the amount of NaCl [61]. The effect of NaCl addition on the morphology of electrospun PVA fibers was shown in Fig. 11. Even a little sodium chloride added in the solution made its conductivity to increase sharply. Net charge density carried by the jet in the electrospinning process can be


399

Figure 11 Effect of NaCl amount in the PVA solution on fiber morphology (DH = 98%, voltage = 5 kV, tip–target distance = 10 cm; flow rate = 0.2 ml/h). NaCl amount based on H2O: (a) 0.05%, (b) 0.10%, (c) 0.15%, (d) 0.2%. Original magnification 10,000 [56].

affected by the conductivity of the solution. PVA fiber diameters were gradually decreased from 214±19 nm to 159±21 nm with increasing content of NaCl from 0.05% to 0.2%, for the higher net charge density increased the electrical force exerted on the jet and led to decreased fiber diameter. When 0.2% NaCl was added in PVA solution, many small particles, perhaps sodium chloride crystals, were observed on the SEM micrograph (Fig. 11d). 2.1.8

Solvent

The effects of solvents and their properties on electrospinnability of as-prepared polystyrene (PS) solutions and the morphological appearance of as-spun PS fibers were investigated qualitatively by means of a scanning electron microscope (SEM) in [62]. The eighteen solvents used were benzene, t-butyl acetate, carbon tetrachloride, chlorobenzene, chloroform, cyclohexane, decahydronaphthalene (decalin), 1,2-dichloroethane, dimethylformamide (DMF), 1,4-dioxane, ethylacetate, ethylbenzene, hexane, methylethylketone (MEK), nitrobenzene, tetrahydrofuran (THF), 1,2,3,4-tetrahydronaphthalene (tetralin) and toluene. DMF was the best solvent to dissolve PS pellets within six hours, while others were found to dissolve PS pellets within one to three days, with the exception of hexane which was not able to dissolve PS pellets after seven days. Only PS solutions in 1,2-dichloroethane, DMF, ethylacetate, MEK, and THF could produce fibers with high enough productivity, while PS solutions in benzene, cyclohexane, decalin, ethylbenzene, nitrobenzene, and tetralin were not spinnable [62]. 2.2 2.2.1

Processing condition Applied voltage

In the case of electrospinning, the electric current due to the ionic conduction of charge in the polymer solution is usually assumed small enough to be negligible. The only mechanism of charge transport is the flow of polymer from the tip to the target. Thus, an increase in the electrospinning current generally reflects an increase in the mass flow rate from the

400


Figure 12 Effect of voltage on morphology and fiber diameter distribution from a 7.4% PVA/water solution (DH = 98%, tip–target distance = 15 cm, flow rate = 0.2 ml/h). Voltage: (a) 5 kV, (b) 8 kV, (c) 10 kV, (d) 13 kV. Original magnification 10,000 [56].

capillary tip to the grounded target when all other variables (conductivity, dielectric constant and flow rate of solution to the capillary tip) are held constant [16]. With the increase of the electrical potential the resulting nanofibers became rougher. It was already reported that a diameter of electrospun fibers was not significantly affected by an applied voltage. This voltage effect was particularly diminished when the polymer concentration was low. According to earlier works, higher voltage was reported to induce not only a larger diameter but also a smaller diameter. Applied voltage may affect some factors such as the mass of polymer fed out from a tip of needle, elongation level of a jet by an electrical force, morphology of a jet (a single jet or multiple jets), etc. A balance among these factors may determine a final diameter of electrospun fibers. It is also noted that bead-


Figure 13

401

Main effect plots of applied voltage on average fiber diameter [55].

Figure 14 a) SEM micrographs of P(LLA-CL) fibers electrospun from a polymer concentration of 5 wt. % at different applied voltage: (a) 9 kV, (b) 12 kV and (c) 15 kV. b) Relation between fiber diameter and applied voltage in the electrospinning with 5 wt. % P(LLA-CL) solution [64].

ed fibers have been found to be electrospun with too high a level of applied voltage. Although voltage effects show different tendencies, the voltage did not show a significant role in controlling the fiber morphology [59]. A series of experiments were carried out when the applied voltage was varied from 5 to 13 kV and the tip to target distance was held at 15 cm. The results are shown in Fig. 12. There was a slightly increase in average fiber diameter with increasing applied electric field. A considerable amount of thin fibers with diameters below 150 nm were found when the applied voltage was above 10 kV. A narrow distribution of fiber diameters was observed at a lower voltage of 5 kV, while a broad distribution in the fiber diameter was obtained at higher applied voltages of 10 – 13 kV. Increasing the applied voltage, i.e., increasing the electric field strength will increase the electrostatic repulsive force on the fluid jet which favors the thinner fiber formation. On the other hand, the solution will be removed from the capillary tip more quickly as jet is ejected from the Taylor cone. This results in the increase

402


of the fiber diameter. Corona discharge was observed at voltages above 13 kV, making electrospinning impossible [56]. Also, an increase in voltage from 9.2 kV to 25 kV did not induce a significant change in the mean diameter of the fibers [54]. The diameter of electrospun PAN fibers did not change significantly over the range of applied voltage for the various solution concentrations in the experimental region [56]. SEM micrographs of nanofibers electrospun at different electrospinning voltage from a constant polymer concentration of 5 wt. % are shown in Fig. 14a. A fiber diameter tended to decrease with increasing electrospinning voltage, although the influence was not as great as that of polymer concentration. Figure 14b shows fiber diameter as a function of electrospinning voltage. The increase of the electrospinning voltage causes an increase of the electrostatic stress on the jet, which may be analogous to an increase of the draw rate in conventional fiber spinning [64]. 2.2.2

Feed rate

The morphological structure can be slightly changed by changing the solution flow rate as shown in Fig. 15. At the flow rate of 0.3 ml/h, a few big beads were observed on the fibers. The flow rate could affect the electrospinning process. When the flow rate exceeded a critical value, the delivery rate of the solution jet to the capillary tip exceeded the rate at which the solution was removed from the tip by the electric forces. This shift in the mass balance resulted in a sustained but unstable jet and fibers with big beads were formed [56]. Figure 16 shows that the diameter of the electrospun HM-PLLA fibers was not significantly changed with the varied volume feed rate. The influence due to the volume feed rate also diminished when the polymer concentration was low. The solution’s electrical conductivity was found to be the dominant parameter to

Figure 15 Effect of flow rate of 7% PVA water solution on fiber morphology (DH = 98%, voltage = 8 kV, tip–target distance = 15 cm). Flow rate: (a) 0.1 ml/h; (b) 0.2 ml/h; (c) 0.3 ml/h. Original magnification 10,000 [56].


403

Figure 16 Volume feed rate effects on the diameter of the PLLA (Mw: 300 K) fibers electrospun from solutions with different polymer concentration [50].

control the morphology of electrospun polymer fibers [9]. In the case of low-molecular-weight liquid, when a high electrical force is applied, formation of droplets can occur. A theory proposed by Rayleigh explains this phenomenon. As evaporation of a droplet takes places, the droplet decreases in size. Therefore, the charge density of its surface is increased. This increase in charge density due to Coulomb repulsion overcomes the surface tension of the droplet and causes the droplet to split into smaller droplets. However, in the case of a polymer solution (high molecular weight liquid) the emerging jet does not break up into droplets, but is stabilized and forms a string of beads connected by a fiber. As the concentration is increased, a string of connected beads is seen, and with a further increase there is reduced bead formation until only smooth fibers are formed. And sometimes spindle-like beads can form due to the extension caused by the electrostatic stress. The changing of fiber morphology can probably be attributed to a competition between surface tension and viscosity. As concentration was increased, the viscosity of the polymer solution increased. The surface tension attempted to reduce the surface area per unit mass, thereby caused the formation of beads/spheres. Viscoelastic forces resisted the formation of beads and allowed for the formation of smooth fibers. Therefore formation of beads at lower polymer solution concentration (low viscosity) occurred where surface tension had a greater influence than the viscoelastic force. However, bead formation was reduced and finally eliminated at higher polymer solution concentration, where viscoelastic forces had a greater influence in comparison with surface tension. But when the concentration was too high, high viscosity and rapid evaporation of solvent made the extension of jet more difficult, thicker and nonuniform fibers were formed [55]. 2.2.3

Distance of needle tip to collector

Tip–target distance had no significant effect on the electrospun fiber morphology of fully hydrolyzed PVA, as shown in Fig. 17. Micrographs are indistinguishable for electrospin-

404


Figure 17 Effect of tip–target distance on fiber morphology from a 7.4% PVA / water solution (DH = 98%, voltage = 5 kV, flow rate = 0.2 ml/h). Tip–target distance: (a) 8 cm, (b) 10 cm, (c) 12 cm, (d) 15 cm. Original magnification 10,000 [56].

Figure 18 Processing map obtained based on the systematic parameter study: (a) jet elongation / an electrical force (affected by electrical conductivity of solvents, applied voltage), (b) mass of polymer (affected by polymer concentration, applied voltage, volume feed rate) [50].

ning at 8 –15 cm of the tip–target distance. It was assumed that solution jets were elongated and solidified quickly after they flowed out of the needle tip because of the high conductivity of fully hydrolyzed PVA used [56]. Based on the processing parameter studies, all the parameters’ effects on the morphology of the electrospun nanofibers were summarized in a processing map (Fig. 18). A suitable level of processing parameters must be optimized to electrospin polymers into


405

nanofibers with a desired morphology and the parameters’ levels are dependent on the properties of polymers and solvents used in each electrospinning process. Understanding of how each of processing parameter affects the morphology of the electrospun nanofibers is essential. All the parameters have been divided into two groups; i.e. one with parameters which affect the mass of polymer fed out from a tip of needle, and the other with parameters which affect an electrical force during electrospinning. Polymer concentration, applied voltage and volume feed rate were considered to affect the mass of polymer. Polymer concentration and feed rate directly reflect the mass of polymer. Increased polymer concentration and feed rate tend to bring more mass of polymer into the polymer jet. It is noteworthy that the minimum polymer concentration to electrospin uniform fibers was determined by the molecular weight of polymer. High molecular weight of polymer provides a sufficient level of solution viscosity to produce a uniform jet during electrospinning even when polymer concentration is relatively low. Applied voltage reflects the force to pull a solution out from the needle hence higher applied voltage causes more solution coming out. On the other hand, it was considered that solution electrical conductivity and applied voltage affect a charge density thus an electrical force, which acts to elongate a jet during electrospinning. Hence, higher solution electrical conductivity and applied voltage increase the jet elongation. Therefore, it is summarized that electrospun fibers with a smaller diameter can be produced with lower polymer concentration, feed rate and applied voltage when the effect of the mass of polymer dominates to determine the final diameter of electrospun fibers, while a smaller diameter of fibers can be electrospun with higher solution electrical conductivity and applied voltage when the effect of the jet elongation is dominant. For both cases, non-uniform / beaded fibers were found if the parameters were either too high or too low. In fact, applied voltage affects both the polymer mass and jet elongation; however, the effect is not as dominant as the other parameters for controlling the morphology of electrospun fibers. It must be noted that polymer concentration, molecular weight and solution electrical conductivity play a prime role in determining the morphology of electrospun fibers. Polymer fibers with a smaller diameter can be electrospun with higher electrical conductivity of solution and lower polymer concentration which can be further decreased by higher molecular weight of polymer [50].

3

Theory and modeling

Though easily realizable in the laboratory, electrospinning is a complex phenomenon to analyze because of the coupling between the electric field and the deformation of the fluid, the latter in turn determined by the rheology of the material. Typically, electrospinning has two stages. In the first stage, the polymer jet issues from a nozzle and thins steadily and smoothly downstream. In the second stage, the thin thread becomes unstable to a nonaxisymmetric instability and spirals violently in large loops. The enormously increased contour length produces a very large stretch ratio and a nanoscale diameter [9]. For the steady stretching in the first stage, Spivak and Dzenis [65] published a simple model that assumes the electric field to be uniform and constant, unaffected by the charges carried by the jet. Reneker et al. modeled the viscoelasticity of the jet by a linear Maxwell equation [9]. Hohman et al. [67, 68] developed a slender-body theory for electrospinning that couples jet stretching, charge transport and electric field. The model encounters difficulties, however, with the boundary condition at the nozzle. For the second stage, the bending instability has been carefully documented by two groups (Reneker et al. [69, 70] and Shin et al. [71]); each has proposed a theory for the

406


Figure 19 Momentum balance on a short section of the jet [9].

instability. Reneker et al. modeled the polymer jet by a linear Maxwell equation. Like-charge repulsion generates a bending force that destabilizes the jet. Hohman et al. [67] built an electrohydrodynamic instability theory, and predicted that under favorable conditions, a nonaxisymmetric instability prevails over the familiar Rayleigh instability and a varicose instability due to electric charges. In theoretical work to date, the rheology of the polymer jet has been represented by a Newtonian viscosity [67, 68], a power-law viscosity [66] and the linear Maxwell equation [9, 69, 71]. The jet is governed by four steady-state equations representing the conservation of mass and electric charges, the linear momentum balance and Coulomb’s law for the field E [9]. Mass conservation requires that π R2υ = Q,

(1)

where Q is a constant volume flow rate. Charge conservation may be expressed by πR2KE + 2 π Rυσ = I,

(2)

where E is the z component of the electric field, K is the conductivity of the liquid, and I is the constant total current in the jet. The momentum equation is formulated by considering the forces on a short segment of the jet (Fig. 19):

d d γ (π R 2 ρυ 2 ) = π R 2 ρ g + [π R (− P + τ zz )] + ⋅ 2π RR ′ + 2π R (tte − tne R ′) , dz dz R

(3)

where τzz is the axial viscous normal stress, p is the pressure, γ is the surface tension, and t et and t en are the tangential and normal tractions on the surface of the jet due to electricity. The prime indicates the derivative with respect to z, and R´ is the slope of the jet surface. The ambient pressure has been set to zero. The electrostatic tractions are determined by the surface charge density and the electric field:

tne =

ε 2

( En2 − Et2 ) ≈

σ 2 ε′−ε 2 − E , 2ε 2

tte = σ Et ≈ σ E ,

(4)

(5)

where ε and ε are the dielectric constants of the jet and the ambient air, respectively, En


Figure 20

407

A vast classification chart for fibers [72].

and Et are the normal and tangential components of the electric field at the surface, and ||*|| indicates the jump of a quantity across the surface of the jet. We have used the jump conditions for En and Et: ||ε En|| = ε E – εEn = σ, ||Et|| = E t – Et = 0, and assumed that εEn

Chemical Processes with Participation of Biological and Related Compounds

Compounds of Uranium and Fluorine (Chemical Compounds)

Asymmetric Synthesis with Chemical and Biological Methods

Asymmetric Synthesis with Chemical and Biological Methods

Chemical and Biological Approaches,

Physical Properties of Chemical Compounds

Optimization of chemical processes

Optimization of Chemical Processes

Fusion of Biological Membranes and Related Problems

Fusion of Biological Membranes and Related Problems

Stable Processes and Related Topics

Biological Processes Associated with Impact Events

Markov processes and related problems of analysis

Biological Chemical and Radiological Terrorism

Markov processes and related problems of analysis

Alkaloids: Chemical and Biological Perspectives

Alkaloids: Chemical and Biological Perspectives

Handbook of Chemical and Biological Warfare Agents

Handbook of chemical and biological warfare agents

Handbook of chemical and biological sensors

Biological treatment processes

Anaerobic Biological Treatment Processes

Handbook of Chemical and Biological Sensors

Handbook of Chemical and Biological Warfare Agents

Nuclear Biological Chemical Handbook

Conceptual Design of Chemical Processes

Kinetics of Soil Chemical Processes

Elementary Principles of Chemical Processes

Conceptual Design of Chemical Processes

Elementary Principles of Chemical Processes

Thermal Safety of Chemical Processes

Chemical Processes with Participation of Biological and Related Compounds

Compounds of Uranium and Fluorine (Chemical Compounds)

Asymmetric Synthesis with Chemical and Biological Methods

Asymmetric Synthesis with Chemical and Biological Methods

Chemical and Biological Approaches,

Physical Properties of Chemical Compounds

Optimization of chemical processes

Optimization of Chemical Processes

Fusion of Biological Membranes and Related Problems

Fusion of Biological Membranes and Related Problems

Stable Processes and Related Topics

Biological Processes Associated with Impact Events

Markov processes and related problems of analysis

Biological Chemical and Radiological Terrorism

Markov processes and related problems of analysis

Alkaloids: Chemical and Biological Perspectives

Alkaloids: Chemical and Biological Perspectives

Handbook of Chemical and Biological Warfare Agents

Handbook of chemical and biological warfare agents

Handbook of chemical and biological sensors

Biological treatment processes

Anaerobic Biological Treatment Processes

Handbook of Chemical and Biological Sensors

Handbook of Chemical and Biological Warfare Agents

Nuclear Biological Chemical Handbook

Conceptual Design of Chemical Processes

Kinetics of Soil Chemical Processes

Elementary Principles of Chemical Processes

Conceptual Design of Chemical Processes

Elementary Principles of Chemical Processes

Thermal Safety of Chemical Processes

Recommend Documents