Strategies and Tactics in Organic Synthesis, Volume 7 (Strategies and Tactics in Organic Synthesis)

Dedication This volume is dedicated to Professor Paul A. Wender on the occasion of his 60th birthday and in recognitio...

Author: M. Harmata

167 downloads 1514 Views 30MB 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

Dedication

This volume is dedicated to Professor Paul A. Wender on the occasion of his 60th birthday and in recognition of the beautiful chemistry he has created.

CONTRIBUTORS

JEFFREY AUBI~, Department of Medicinal Chemistry, University of Kansas, 1251 WescoeHall Drive, Lawrence, KS 66045- 7852 VERONIQUE BELLOSTA, Laboratoire de Chimie Organique, ESPCI, CNRS 10 rue Vauquelin, 75231 Paris Cedex 05, France KAYM. BRUMMOND, Department of Chemistry, University of Pittsburgh, Pittsburgh, PA 15260 CAMERON M. BURNETT, Department of Chemistry, Colorado State University, Fort Collins, CO 80523 JANINE COSSY, Laboratoire de Chimie Organique, ESPCI, CNRS 10 rue Vauquelin, 75231 Paris Cedex 05, France SIMON R. CRABTREE, Research School of Chemistry, Australian National University, Canberra, Australia 0200 XING DAI, Department of Chemistry, University at Buffalo, The State University of New York, Buffalo, NY 14260 HUW M. L. DAVIES, Department of Chemistry, University at Buffalo, The State University of New York, Buffalo, NY 14260 SCOTT E. DENMARK, Department of Chemistry, University of lUinois, at Urbana-Champaign, Urbana, IL 61801 KELLY A. FAIRWEATHER, Research School of Chemistry, Australian National University, Canberra, Australia 0200 KEVIN J. FRANKOWSKI, Department of Medicinal Chemistry, University of Kansas, 1251 WescoeHall Drive, Lawrence, KS 66045-7852 SHINJI FUJIMORI, Department of Chemistry, University of Illinois, at Urbana-Champaign, Urbana, IL 61801

XV

xvi

CONTRIBUTORS

BAUDOUIN GERARD, Department of Chemistry, Boston University and

Center for Chemical Methodology and Library Development, Boston, MA 02215 FRANK-GERRIT ~ E R , Institut fiir Organische Chemie, Fachbereich Chemie, Universitiit Duisburg-Essen, Essen, Germany MIREIA CAMPANA KUCHENBRANDT, Institut fiir Organische Chemie, Fachbereich Chemie, Universitiit Duisburg-Essen, Essen, Germany LEONARD R. MACGILL1VRAY, Department of Chemistry, University of Iowa, Iowa City, IA 52242 LEWIS N. MANDER, Research School of Chemistry, Australian National University, Canberra, Australia 0200 BRANKO MITASEV, Department of Chemistry, University of Pittsburgh, Pittsburgh, PA 15260 THOMAS PETTUS, Department of Chemistry and Biochemistry, University of California at Santa Barbara, Santa Barbara, CA 93106-9510 JOHN A. PORCO, Jr., Department of Chemistry, Boston University and Center for Chemical Methodology and Library Development, Boston, MA 02215 JON D. RAINIER, Department of Chemistry, University of Utah, Salt Lake City, UT 84112 CATHERINE TAILLIER, Laboratoire de Chimie Organique, ESPCI, CNRS 10 rue Vauquelin, 75231 Paris Cedex 05, France TODD WENDERSKI, Department of Chemistry and Biochemistry, University of California at Santa Barbara, Santa Barbara, CA 93106-9510 DAVID R. WILLIAMS, Department of Chemistry, Indiana University, Bloomington, IN 4 7405 ROBERT M. WILLIAMS, Department of Chemistry, Colorado State University, Fort Collins, CO 80523 AARON WROBLESKI, Department of Medicinal Chemistry, University of Kansas, 1251 WescoeHall Drive, Lawrence, KS 66045- 7852

Preface

I believe I first saw Paul Wender speak at the University of Illinois-Champaign/Urbana on December 8, 1981. Having an intense love affair with organic chemistry at the time, I recall being utterly thrilled by the chemistry I saw. Photochemistry to do a [5+2]-cycloaddition, taking simple starting materials to complex structures in one step!! It remains very powerful chemistry. Though not an easy thing to do, with the help of Scott Denmark and the NIH, I eventually became a postdoc in Paul's lab, and was, I believe, the first person he introduced to the neocarzinostatin chromophore. There was great science in those labs and many very talented people. It was fun. Paul was a very interesting mentor. It became clear very early that he wanted his co-workers to think and think deeply about their research. More important, he often seemed willing to "wait it out" as someone struggled with a problem he believed they could handle. I always felt as though he knew an answer, but expected us to know, learn or discover an answer too. This is a rather daring way to run a research group, but it produces real thinkers. Some things that really shocked me about Paul can be related in two stories. A few weeks into my postdoc I decided to head up to Muir Woods one weekend to see what I could see. While sitting and enjoying some ice cream, I looked up to see Paul standing right in front of me. Ouch! I should have been in lab, or so I thought. Paul happily introduced me to his companions and we went our separate ways. I expected some comments on Monday regarding my absence from lab. They never came.

xvii

xviii

PREFACE

Much later, I informed Paul that my girlfriend would be visiting me for a week and I would not be doing my 80 reactions per week as he had come to expect from me (editor's privilege !). Without hesitation, he suggested that I take the week off and do some touting of California. I accepted his offer. Paul has always impressed me as someone who has an incredibly deep interest in chemistry and science in general. So deep, in fact, that he is willing to spend a great deal of time with people talking about it. I am one of those people. His ability to see in ways that are often unique, from my point of view at least, have no doubt opened new roads, not only for me, but for others as well. Especially important from my perspective is my observation that at meetings, unless a very tight travel schedule calls him away, Paul will visit with people at posters, empty tables, you name it, and pass along insights and give encouragement, whether they are doing a complex natural product synthesis or relatively simple chemistry. He does not have to do that; he chooses to and the chemistry community at large benefits from it. I have heard physical chemists refer to synthetic organic chemists as reptiles who eat their own children. Paul is not that way. He takes great pleasure in seeing his "children" reach their highest potential and is willing to take steps to help them achieve that goal. Happy Birthday, Paul!

As the editor of this book, I have many people to thank. The authors did a spectacular job in presenting some very nice chemistry. One of the perks of being an editor is getting to enjoy all of that science. My wife, Judy Snyder, (the "girlfriend" above) helped with proofing and for that I am extremely grateful. Thanks to Joan Anuels and all at Elsevier for their encouragement. It is time to think about another volume in this series, but I also have papers and especially proposals to write to keep my own research alive. I hope to see another volume in 2009 or so. Until then, take care of your science and yourselves.

Michael Harmata

Foreword

This 7 th volume of Strategies and Tactics in Organic Synthesis presents an extraordinary range of superb chemistry in the 13 chapters that follow. It also makes clear the great human effort involved in the construction of complex targets, and teaches us how a combination of perseverance and imagination can conquer the problems that always arise as the price of creativity. For some 33 years, the combination of great art with great teaching has been the hallmark of Professor Paul Wender's chemistry. It is most fitting that this volume is dedicated to him. Gilbert Stork

Columbia University, NY May 2007

xix

STRATEGIES AND TACTICS IN ORGANIC SYNTHESIS, VOL. 7 9 2008 Elsevier Ltd. All rights reserved.

Chapter 1 TOTAL SYNTHESIS OF RK-397 Scott E. Denmark and Shinji Fujimori Department of Chemistry University of Illinois at Urbana-Champaign Urbana, IL 61801, USA

I. II. III.

Introduction and Background Retrosynthetic Analysis Synthesis of RK-397 A. Preparation of Fragment 41 B. Preparation of the Fragment 42 C. Aldol Addition of Trichlorosilyl Enolate Derived from 41 D. Synthesis of C(11)-C(33) Fragment E. Preparation of Polyene Fragment 40 IV. Completion of the Synthesis V. Summary Acknowledgements References and Footnotes

1 8 10 10 13 16 20 26 28 30 32 32

I. Introduction and Background

RK-397, a member of a large family of polyene macrolides that include amphotericins, nystatin, mycoticins, and roxaticins, was isolated and structurally characterized by Osada et al. in 1993 (Figure 1). ~ Most of these macrolides are produced by soil actinomycetes, mainly belonging to the genus Streptomyces, and are isolated by extraction. 2 The structural features include a large lactone ring (20-44-membered), three to eight conjugated double bonds and, in most cases, a sugar moiety. Because of the presence of the polyene, these compounds show strong UV absorption, and characteristic absorption maxima are observed depending on the length of polyene, allowing classification of over 200 macrolides into subcategories. This class of compounds is particularly known for their antifungal properties. Nystatin was the first polyene macrolide antibiotic to be isolated and continues to be used in antifungal therapy. 2 Amphotericin B has

SCOTT E. DENMARK AND SHINJI FUJIMORI

O~ ~

Me

HO'"L'~Meo

~

~

?H ?H

Me

~

~

O: OH _ OH NH2le

Me M2~~v-~~~v~.~,,,OH

?H ?H ? ' ~ '''CO2H Me) ....~ ~ r

-OH Amphotericin B

OH

OH

~"

_~ _

~H o ~

Me"' '-,,,~""-,~ " ~ __'-,,,~" , ~ V : OH OH OH OH OH Roxaticin

.,/k--..0.~--.-.-Me ~ O H " OH NH2

O

-

Me

~

~

HO"'L~']~Meo

~

~

~

~

?H ?H ?H

M2 ~ i ~,~,,,O

?H ?,,,~,,,CO2H

Me

-OH

OH

OH

R

Nystatin

M

E

O

W

,

,

Me...-J..... 0 M

e

,

M

~

~

~

~

,''OH i Med,~,,,OH

e ' " ~ OH OH OH OH OH OH Mycoticin A (R = H) g (R= Me) OH

L.,. ,,,OH " ' ~ OH OH OH OH OH OH RK-397

FIGURE 1. Representative polyene macrolides.

been used extensively in treatment of serious infections for the past 40 years and remains drug of choice for systemic fungal infections. 3 These antifungal agents have gained growing interest in the last decade since the occurrence of fungal infections has increased due to AIDS and immunosuppression associated with organ transplants. 4 However, prolonged parenteral administration of amphotericin B often leads to various adverse effects, especially in the kidney. 5 RK-397 possesses relatively weak antifungal activity. However, unlike amphotericin B and nystatin, it exhibits potent antibacterial activity along with some cytotoxicity against human leukemia cells. ~The mechanism of action for amphotericin B and nystatin has been studied extensively, and these antibiotics are known to associate with sterols and to form channels on cell membranes, inducing leakage of cellular contents. 6 However, due to the great difference in biological activity, RK-397 and several other polyene macrolides are believed to have different mechanisms of action. 7 The structural elucidation and configurational assignment of these macrolides initially relied on degradation studies combined with spectroscopic analysis. Amphotericin B was the first polyene macrolide to have

1

TOTALSYNTHESISOF RK-397

its three-dimensional structure established by X-ray crystallography. 8 More recently, the advances in NMR spectroscopy greatly improved its ability to analyze structures, and Rychnovsky's 9,1~ acetonide analysis is particularly useful in determining the configuration of polyols. These techniques allowed determination of molecular structure and absolute configuration of several other polyene macrolides including RK-397. Because of their potent biological activity and structural complexity, these natural products have attracted interest as targets for total synthesis by many synthetic chemists. ~~ So far, total syntheses of amphotericin B, 12 mycoticin A, 13 roxaticin, ~4 filipin III, ~5 roflamycoin, ~6 dermostatin ~7 and RK-397 ~a have been reported. The obvious challenges in the synthesis of polyene macrolides lie in the stereocontrolled assembly of the polyol chain and construction of the polyene backbone. Various approaches are available for construction of the 1,3-polyol: iterative Sharpless asymmetric epoxidation/hydride reduction, 12two-directional chain synthesis, 13aalkylation of cyanohydrin acetonides 1~ and aldol/anti (or syn) reduction sequence (Figure 2). 14d Nicolaou: SAE/reduction (+)-DET Ti(OiPr)4 O R ~ O H t-BuOOH R ~ O H CH2Cl2 1

1. Swern 2. Ph3PCHCO2Et

2

R i~/..O ""~ " ' ~ ~COOEt

R~

OH

1. DibaI-H 2. PivCl 3.TBDPSCI 4. DibaI-H 1. Red-AI 2. TBSCI

(+)-DET Ti(Oi-Pr)4 t-BuOOH

R'.I~OH OTBDPS

CH2CI2

4 D,,

FI~

OTBS

TBDPSO

OTBDPS 5

OH 6

Rychnovsky: Alkylation of cyanohydrin acetonides Me

Me

Me

O-~-O R

CN 7

Me

O,~-O + B r ~

Me R'

THF

Me Me

Me

oXo oXo

LiNEt2 "

R

~

8

FIGURE 2. Syntheticstrategies for 1,3-polyol chains.

R CN

' 9

SCOTT E. DENMARK AND SHINJI FUJIMORI Schreiber: Two-directional chain synthesis

M

e

O

1. RuCI2[(R)-BINAP ] 2. (EtO)2CHMe, H+ OMe 3. Li, NH3; then 0 3

~

~.~

oo

~ 1. RuCl2[(R)- BINAP] 2. (MeO)2CMe2, H+ 3. DibaI-H; VinylMgBr 4. (MeO)2CMe2, H+ =

10 O

M

e O ~ M e II II I II 0 0 O\ I0 0

T

Me 0 0 MUMe

II 0

11

0 0 MUMe

0 0 MUMe

12

Mori: Epoxide opening with lithiodithane

Ph Ph

Ph

Ph'J~-0

S/~

o

n-BuLi

Ls.J + 13

THF

v

v

v

-a

15

14

1. Hg(CIO4)2 Ph 2. NaBH4 Ph,,~ 3. (MeO)2CMe2 / O =- O ~

F" o, A . s x s j . "

Me

Me

O'/X~'O R

16 Evans: Aldol/carbonyl reduction

O R'~H

Bu2B0

+

O_Pg

~ R '

Et20

17 Me4NHB(OAc)3 MeCN, AcOH

OH 0 I II R~

OPg R'

18

OH OH OPg N

N' 19

FIGURE 2. (continued)

Stereoselective polyene synthesis has historically relied primarily on olefination reactions (Figure 3). Nicolaou and Moil employed conjugated phosphorus-based olefination reagents 20 and 21 to extend a polyolefin by three double bonds at a time. ~2,18 Rychnovsky employed Wollenberg's ~9 reagent 23 followed by elimination to install the dienal unit. Schreiber and

l

TOTAL SYNTHESIS OF RK-397

Nicolaou/Mori: O

R.,JQH

1.20 or 21 2. DibaI-H 3. MnO2 ,. R ~ C H O 22 20= (EtO)20P~-~/~.../~/COO Et 21 Ph3P~COOEt

Rychnovsky" O

1.23 2. MsCI, Et3N

H 23= B r M g ~ o E

t

24

Schreiber/Evans: [ ~

H .OAc Hg(OAc)2 ~ , O LiAIH4;then 02 = = O H C ~ c H O "

25

Ac

26

27

FIGURE 3. Synthetic strategies for polyene.

Evans took advantage of oxidative electrocyclic closure of cyclooctateraerie followed by reduction to provide triene dialdehyde 27. ~3 Another challenge arises from the macrocyclization in the presence of both polyene and polyol units. The most popular method to construct a macrocycle is the Yamaguchi method, which employs an activated anhydride derived from 2,4,6-trichlorobenzoyl chloride (Figure 4 ) . 20 The Yamaguchi macrolactonization works for a variety of large rings; however, the reaction sometimes requires vigorous heating to effect cyclization. An alternative method to close rings having more than 20 members is the intramolecular Horner-Wadsworth-Emmons olefination, originally developed by Stork and Nakamura. 21 The application of Roush-Masamune conditions to the macrolactonization makes this approach applicable to substrates with base-sensitive functional groups. 22 The first total synthesis of RK-397 was completed by Burova and McDonald in 2004. ~ Their approach was based on iterative coupling of


Macrolactonization:

Me Me Me

1.2,4,6-trichlorobenzoylchloride 2. DMAP O

Me

--- "~

- "~

28 Me

Me./l....[/O M

O~_.J e

'

"

~

Intramolecular Olefination:

Me Me

Me

_

O LiCI, DBU

Me O

O

O O MUMe PO(OEt)2

O O MUMe 30

O O MUMe

OR

MeCN

H

Me _

A

M~

\

\

\

\

Me...J.... 6

\

~

,%~,OR

o. J

oFr

o~o

o

Me Me

Me Me

81

FIGURE 4. Macrolactone formation strategies.

six-carbon units by epoxide opening with an acetylide anion (Scheme 1). Chiral epoxide 33 was prepared by Jacobsen's hydrolytic kinetic resolution, and the attachment of acetylide anion derived from 32 furnished homopropargyl alcohol 34. The installation of 1,3-diol functionality was accomplished by hydration of the alkyne and subsequent stereoselective reduction. For the construction of polyene fragment, a Stille coupling of stannyltriene (C(9)-C(10)) was employed, and the closure of the macrocycle was accomplished by the intramolecular olefination (C(2)-C(3)) (Figure 5).

1


,/

intramolecular HWE ~ Stille coupling \

Me

"]

\

"

" 11

I'

Me.,P,,, 0

., OH

OH OH OH ~ O H epoxide opening

FIGURE 5. Key disconnections in McDonald's synthesis of RK-397.

H

O

O

32

Me'Me

~,~,.,,,.TMS

n-BuLi BF3-Et20

+

,.TMS

THF

0 0 Me.Me

o, -J

OPMB 33

34 1. HN(SiMe2H)2 2. Pt(DVDS) 3. H202, KF, KHCO3 4. Et2BOMe, NaBH4 ..TMS

0 0 Me'Me

OH OH OPMB 35

SCHEME 1

Another synthetic approach toward RK-397 has been reported by Schneider et al., ~ who employed a thermal oxy-Cope rearrangement of a silyloxy diene to construct a precursor for the polyol unit (Scheme 2). The diene 36 was readily prepared by auxiliary-controlled aldol addition in a stereoselective manner. The rearrangement was accomplished by heating the diene at 170 ~ and the subsequent acidic hydrolysis of silyl enol ether provided the aldehyde 37. After a few manipulations of 37, the silyl group was oxidized under conditions developed by Fleming to the corresponding diol 39. Using this method, Schneider constructed the


polyol chain fragment of RK-397, however, the completed total synthesis has not yet been reported. 53 SO PhMe2Si~N 36

O O

~ t-Bu

OBn OH *~~[V[v~~OMe

1. 170~ 2. p-TsOH 70 - 75%

O

39

1. BF3(AcOH)2 ?" H202' KF ~ 72%

O SiMe2Ph O A ~~J'~~'~ N O 37

iI t-u--

OBn SiMe2Ph O O M e 38

SCHEME 2. Schneider's approach for the polyol fragment of RK-397.

The challenges of stereoselective construction of the 1,3-polyols as well as the all-E-polyene fragment in RK-397 could be efficiently addressed by ongoing methodological programs in these laboratories on the enantioselective, Lewis-base-catalyzed aldol addition 23 and siliconbased cross-coupling reactions. 24This chapter describes an efficient enantioselective total synthesis of RK-397 utilizing the methods mentioned above as key strategic steps. II. Retrosynthetic Analysis

To maximize synthetic convergence, the target was divided into four modules (Scheme 3). The disconnections at the lactone linkage and the C(10)-C(ll) bond provide a known polyene phosphonate fragment 40.13'14c'25 Obviously, the polyol fragment lends itself to myriad aldol/ reduction disconnections. 26 However, by carefully examining the pattern of stereogenic centers on the polyol chain, the most convergent synthesis might be achieved by the disconnections between the C(18)-C(19) and C(26)-C(27) bonds. This provides both the C(11)-C(18) and C(19)-C(26) fragments as an identical building block. In the forward sense, these disconnections require an aldol addition with 1,5-anti stereoinduction from the methyl ketone 41. The aldehyde functionality at C(11) and C(19) was masked as an alkenylsilane for better functional group compatibility. The remaining C(27)-C(31) fragment would be synthesized by use of Evans' chiral acyl oxazolidinone technology. 27

1


MeO 1 ~

~

~

10 ,~.~v.~,,,OH

~

OH OH OH OH OH OH

PO(OEt)2 0.~-~.~~~~~ EtO 40

10

BnMe2Si

..0

..Ph ,,',"10

41 Me Me.,,,J3,j/OPMB 26 41 Me'"~ H Me~SiMe2Bn II -42 0 0 O~v~O_

O 19

_

#h SCHEME 3

The polyene phosphonate 40 can be further subdivided into three units by sequential palladium-catalyzed cross-coupling reactions of bis-silyl diene 45 (Scheme 4). 28 The phosphonate 40 would be derived from hydroxy-ester 43 by functional group conversions at C(10), and the construction of the tetraene moiety in 43 would be accomplished through the sequential coupling of 45 with protected 3-iodo-allyl alcohol 46 and with iodoacrylate 44. The order of these two couplings can be changed depending on the reactivity of these iodides. However, ester functional groups are usually not compatible with TMSOK-promoted coupling conditions. Therefore, the coupling of 46 should be carried out with TMSOK activation, and the coupling of 44 should be performed with fluoride activation. The key building block 41 was envisioned to arise from vinylogous aldol addition of dienol ether 48. 29The aldol addition of 48 is catalyzed by a chiral dimeric phosphoramide, and the reaction should provide the 3~-addition product selectively. The conversion of 47 to 41 involves transformation of the C(25) ester to a methyl ketone and installation of the syn-diol. The disconnection at C(21)-C(22) in enoate 47 reveals dienol ether 48 and aldehyde 49 as substrates for the Lewis base-catalyzed vinylogous aldol

10


addition. The fragment 42 would be derived from known aldehyde 5030 by olefination with methyl triphenylphosphanylideneacetate 51. The stereogenic centers at C(30) and C(31) would be established by Evans aldol addition with isobutyraldehyde.

PO(OEt)2 10

0~'~~~~~ EtO

OR

.

40

EtO

0.~-~ I EtO 44 1

43 OR SiMe20HI-~-./J 10

BnMe2Si~ 45

46

Ph

0 0"~0 19 Me'JL~[~~~~SiMe2Bn I 26 41

0 OH -21 19 EtO~'~SiMe2Bn 22

~

47

TBSO EtO~

0 22 H21~~SiMe2Bn

48

Me Me...J;~,.OPMB Ue,,,~~.~H i 0 42

PMBO

0

~ Me.~Me~Me 50

49

0 Ph3P~-~~OMe 51

SCHEME 4 III.

Synthesis of R K - 3 9 7

A. P R E P A R A T I O N OF F R A G M E N T 41

This synthesis plan was initiated with the preparation of key building block 41 (Scheme 5). Stereoselective Red-A1| reduction of 3-benzyldimethylsilylpropargyl alcohol 5231 provided allyl alcohol (E)-53, 32 which was oxidized to the 3-silyl-2-propenal 49 in good yield (61%, two steps).

1

11


BnMe2Si~'~-.~/OH

Red-AI Et20 " BnMe2Si~--"~~OH

52

74%

DMSO, (COCl)2 O Et3N 0H2012 ~" BnMe2SiA~.~H

53

83%

49

SCHEME 5

The vinylogous aldol addition was initially attempted with dioxanonederived dienol ether 54 (Scheme 6). The aldol addition of 54 was studied previously under catalysis with bisphosphoramide (R,R)-56, and the reaction proceeded smoothly with a broad range of aldehydes, including aliphatic aldehydes, and afforded good-to-excellent enantioselectivities. 29 The aldol addition of 54 to [3-trimethylsilyl acrolein 55 provided a good yield of the aldol product, however, the enantioselectivity was only modest. The additions to 3-alkoxypropanals, 58, were also considered as alternative approach to 41. However, additions to these aldehydes did not provide a synthetically useful enantioselectivity.

Me Me O,/~O

O + TMS ~ ' / " ' ~ ~ H 55

~OTBS 54 Me

(R,R)-56 (5 mol %) SiCl4

Me

O./~.O

OR O

Me Me OH O/~/~'O

0H2012

T

(R,R)-56(5 mol %) SiCI4, TBAI (20 mol %)

M

S ~ O 57: 63%, er 82/18 Me Me OH O/~/~'O

+

0H2012 54

R = Bn 58b: R = TBS 58c: R = TIPS

59a: 79%, er 64/36 (R = Bn) 59b: 58%, er 73/23 (R = TBS) 59e: 86%, er 85/15 (R = TIPS)

58a:

(R,R)-56 = ~k]~ ~ _ ~

... CH2 N' i 'MeMe / 2 SCHEME 6

Therefore, the crotonate-derived silyl ketene acetal was chosen as an alternative nucleophile (Scheme 7). Although the additions of 48 to 58

12


or to 61 were low-yielding and unselective, the vinylogous aldol reaction with 49 (using chiral bisphosphoramide (R,R)-56) efficiently provided 47 in good yield with excellent ?-selectivity and enantioselectivity (75%, er 98/2). The beneficial effect of a silyl group at the [3-position was illustrated by the successful aldol additions of 48 to 55 and 63, whereas reactions with acrolein resulted in polymerization of the aldehyde. OTBS ~ o e t 48

OR O +

[''~

H

Bn 58b: R = TBS 58a: R =

OTBS ~ o e t 48

(R,R)-56 (5 mol %) SiCl4

O +

~"~"/~ H 61

OTBS ~ o e t 48

(R,R)-56 (5 mol %) OH O SiCI4,TBAI(20moI%) RO/~ -- v " ~ v / ~ -,4,,~ O- E t -" 0H2012 60a: 32%, er 85/15 (R = Bn) 60b: 44%, er 81/19 (R = TBS)

0H2012

O + BnMe2Si~ - ~ ' H 49

-"

(R,R)-56 (1.5 mol %) SiCI4 CH2C!2

"~'~~oet 48

O +

R3Si~ H 55: R = TMS 63: R = SiMe2Ph

SiCl4 0H2012

O

~~~~'~----------------------~~oet 62: not observed

OH

O

= BnMe2Si~'~'~~~----------------------~~oet 47, 75%, er 98/2

(R,R)-56 (5 mol %) OTBS

OH

OH

0

R 3 S i ~ O E t 64: 86%, er 97/3 (R = TMS) 65: 62%, er 98/2 (R = SiMe2Ph)

SCHEME 7

Although the additions to other 13-silyl enals also showed excellent enantioselectivities, we chose to proceed with compound 49 because the benzyldimethylsilyl group can be oxidized under very mild conditions. The vinylogous aldol addition could be performed on a 50-mmol (9.5 g of 49) scale without loss in yield or selectivity. The protected syn-diol functionality was installed using the tandem alkoxide addition/conjugate addition protocol developed by Evans and Gauchet-Prunet 33 (Scheme 8). This transformation proceeded with excellent diastereoselectivity to provide the benzylidene-protected diol 66 in good yield. The complete consumption of the starting material

1

13

TOTALSYNTHESIS OF RK-397

with minimal formation of the side products was achieved by maintaining the reaction temperature between 0 and 5 ~ Carrying out the reaction at lower temperature (-20 ~ yielded a significant amount of the aldol product between the resulting ester enolate and excess benzaldehyde. Ph OH

PhCHO KHMDS

O

BnMe2Si~

OEt

THF 74%

47

THF 87%

O

BnMe2Si~OEt 66: dr >19/1

Ph

HN(OMe)Me

/-PrMgCI

O~O

,.

MeMgBr

O~O O gnMe2Si~N..OMe 67

Me

Et20 90%

Ph 0~0

BnMe2Si~

0

Me 41 SCHEME 8

Finally, construction of the key building block 41 was completed by conversion of the ester functional group into a methyl ketone via Weinreb's amide in excellent yield. 34 First, the formation of the Weinreb amide from 66 was accomplished by using isopropylmagnesium chloride as a base. The controlled addition of methylmagnesium bromide at room temperature provided the desired methyl ketone without the formation of the over-addition product. B. PREPARATION OF THE F R A G M E N T 42

Several different routes employing Lewis base catalysis to construct the stereogenic centers in 42 were attempted (Scheme 9). The first approach involved the catalytic, asymmetric addition of a Z-crotylsilane to establish the two stereogenic centers in one transformation (path b, Scheme 9). The addition of (Z)-2-butenyltrichlorosilane catalyzed by

14


chiral dimeric phosphoramides provides excellent diastereoselectivity and good enantioselectivity for aromatic aldehydes and alkenyl aldehydes. 35 Thus, it was envisioned that the addition of (Z)-71 to methacrolein would give the syn-crotylation product. Then, the selective reduction of the C(32)-C(33) double bond would provide 70, followed by the oxidative cleavage of the terminal methylene group to provide the desired aldehyde 50. However, the addition of (Z)-71 to methacrolein catalyzed by 74 only produced a trace amount of the crotylation product (Scheme 10). The remainder of the reaction mixture showed a small amount of the unreacted aldehyde and unidentifiable materials.

Me Me/J-..~.,,OPMB 42

a

Me Me@OH

O

68

PMBO O M e ~ H Me Me 50

Me

OTBS

Me~ O

~Ot-Bu Me 69

O

I

~ Me

Me Me 70

H Me/~~--SiCI3 33 (Z)-71

OH 0 Me Me 72

H

R (Z)-73

SCHEME 9

O Me ? H

74 (10 mol %) DIPEA

Me'~...

Me

SiCl3 CH2CI2,-78 ~

Me

Z-71 74=

H-~N //O

~

Ue/ SCHEME 10

OH - ~

75:trace

1

15


The second approach employed the aldol addition of a trichlorosilyl enolate to set the two stereogenic centers (path c, Scheme 9). The reaction of a Z-trichlorosilyl enolate under phosphoramide catalysis should provide the desired syn-aldol product. Three trichlorosilyl enolates 73 derived from propiophenone, 2,2-dimethyl-3-pentanone and propanal were prepared, and the aldol additions of these enolates to methacrolein were investigated (Scheme 11). The additions of the ketone-derived enolates were catalyzed by 76, and the addition of 73c was catalyzed by dimeric phosphoramide 56. Aldol products were obtained in modest yields, but the diastereoselectivities were very low for 77 and 79. The addition of 73b was syn (relative) selective, and the enantiomeric ratio of the corresponding benzoate was only 1.5/1.

Me~SiCI3~ Ph

(R,R)-76 (10 mol %)

O +

Me

H

CH2CI2

73a

Ph

-78 ~

OSiCI3 Me~t_B 73b

OH O

u

O +

-~ Me

77: 51%, syn/anti = 1/1.3

(R,R)-76 (10 mol %) H

CH2CI2 -78 ~

,-

OH O M e ~

t-Bu

78: 34%, syn/anti = 10/1 Me Ph,~1~ +O (R,R)-76 = ip. a..~....., Ph'

N,Me~

1. (S,S)-56 (10 mol %)

Me

OSiCI3 O 2. MeOH H Me? H CH2CI2 73c -78 ~

Me

OH OMe OMe

" Me 79: 54%, syn/anti = 1/2

SCHEME 11

The third option utilized the vinylogous aldol of dienol ether 69 to construct the whole carbon framework of 42 (path a, Scheme 9). The preliminary studies on aldol addition of 69 to benzaldehyde showed excellent site-, diastereo- and enantioselectivity. 29 The expected aldol product 68

16

SCOTT E. DENMARKAND SHINJI FUJIMORI

would have anti-configuration at the newly created stereogenic centers, therefore, the inversion at C(31) hydroxyl would be necessary. The aldol addition of 69 catalyzed by (R,R)-56, however, provided no desired aldol product (Scheme 12). It is most likely that methacrolein underwent the undesired side processes under these conditions. TM O Me~..H

+

M e ~ O T B

S

Ot-Bu

(S,S)-56 (5 mol % ) CH2CI2,-78~

69

OH O Me..~l.~.~Ot_Bu_ Me 80: not observed

SCHEME 12

After many unsuccessful attempts, it was decided that the C(27)- C(31) fragment would be prepared from the known aldehyde 50 (Scheme 13).3o The stereogenic centers in 50 were established by the aldol addition of chiral acyl oxazolidinone to isobutyraldehyde. From this aldehyde, Wittig olefination with methyl triphenylphosphanylideneacetate provided the conjugated ester 81. The reduction of the ester to the allylic alcohol with Dibal-H was followed by allylic oxidation with manganese(IV) oxide to provide the aldehyde 42 in good yield (61%, three steps, Scheme 13). PMBO

O

Me

H Me

Me 5O

Ph3PCHCO2Et " MeCN, reflux 90%

PMBO_ Me

I Me

_-Me 82

PMBO DibaI-H Me \ . / ~ . , , ~ / C O O E t ,] --_- Et20 Me Me 97% 81 MnO2 PMBO_ = M e ~ C H O OH 0H2CI2,reflux Me Me 70% 42 SCHEME 13

C. ALDOL ADDITION OF TRICHLOROSILYL ENOLATE DERIVED FROM 41

With the methyl ketone and the aldehyde in hand, the key aldol coupling was attempted. On the basis of previous studies, 36 the aldol addition

1

17


of the trichlorosilyl enolate derived from 41 was tested (Scheme 14). TMS enol ether 83 was prepared from methyl ketone 41 with diisopropylethylamine and TMSOTf ~3b(94%). The transsilylation of 83 to the corresponding trichlorosilyl enolate 84 was accomplished using SiC14 and mercury(II) acetate. 37 After 30 min, the volatile components were removed to provide the crude trichlorosilyl enolate 84, and this crude product was used directly for the Lewis-base-catalyzed aldol addition. With (R,R)-76 as the catalyst, aldol product 85 was obtained in good yield (81%), but the diastereoselectivity was only 2/1 favoring the desired diastereomer (27R)-85.

Ph

O~O

O

BnMe2Si~Me

CH2CI2

41

O/~O

OTMS

" BnMe2Si~ 83

94%

SiCI 4 I Hg(OAc)2 (2 mol % ) CH2Cl2

Ph

TMSOTf DIPEA

Ph 0~0

OSiCI3

/ B n M e 2 S i ~ 84

42, (R,R)-76 (10 mol %) CH2CI2 -78 ~ 81%, dr 2/1

Ph PMBO

M

e ~ s Me Me

OH 0

i

M

OH O

0~0

e

2

B

n

(27R)-85

,. ~

t

(27S)-85 SCHEME 14

Interestingly, by use of the enantiomeric catalyst (S,S)-76, diastereomer (27S)-85 was obtained in good yield (72%) with a 4/1 diastereomeric ratio. These results indicate that (27S)-85 forms in the matched case arising from 1,5-syn stereoinduction in the aldol addition. To improve the diastereoselectivity of the aldol addition, various chiral phosphoramides including dimeric species, chiral N-oxides, achiral Lewis bases such as phosphines, formamides, phosphine oxides and sulfides, and ureas were surveyed as the promoter, however, the diastereoselectivities were uniformly low (Chart 1 and Table 1).

18


Ph0

Ph Ph%.IN, 0

IN'phN~ Ph

Me K'"-~N ,,0

N::N 0 H"'L [P--N/" ' ~N L}

Ph ~

86

87

MeMe / 2

~']N ,0

88

89

Ph" N~lel~e ] 2

(R,R)-56

90

~

Me

/2

74

Me

Me t-Bu n-BuO

91

~~/ t-Bu On-Bu

N'Me0 MeMe 92

0 L~/

0 93

CHART 1

The lack of selectivity in this aldol addition may be due to the coordination of benzylidene acetal oxygen to the trichlorosilyl moiety (Figure 6). The chelation by this additional oxygen will allow only one phosphoramide to bind, and the stereochemical information cannot be effectively transmitted from the chiral catalyst to the newly created stereogenic center. A dramatic difference in enantioselectivity between the one- and twophosphoramide pathways has been documented in the aldol addition of cyclohexanone-derived trichlorosilyl enolates where the one-phosphoramide pathway not only altered the diastereoselectivity but also led to reduction in enantioselectivity. 38 Another possibility is that the chelation may facilitate an uncatalyzed process by activating the trichlorosilyl moiety of the enolate. The product arising from the uncatalyzed pathway can compromise the observed diastereoselectivity of the aldol addition. This type of chelation was also proposed in analogous titanium(IV) enolates, and these enolates were found to give extremely low levels of stereoinduction. 39

TABLE 1 Aldol addition of trichlorosilyl enolate catalyzed by various Lewis bases Ph

1. SiCI4, Hg(OAc)2 OTMS 2.42, cat.

"""u ~ J BnMe2Si~

83

CH2012

---

Ph _

PMBO_

_OH 0

OH 0

0_~_0

Me

SiMe2Bn + " Me

Me

"

(27R)-85

(27S)-85

Entry

Lewis base a

Loading (tool%)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17b

(R,R)-76 (S,S)-76 90 86 92 (R,R)-56 91 89 74 88 (R,R)-87 (S,S)-87 93 Ph3PO Ph3PS DMF Uncatalyzed

20 20 10 20 10 10 8 10 11 10 10 10 100 100 100 100

Yield (%)

No No

No No No

27R/27S

2/1 1/4 2/1 1/1

81 72 61 57 reaction 73 reaction 46 20 42 57 72 reaction reaction 33 reaction 41

2/1 1/1 1/1 2/1 2/1 1/3

1/1 1/1

aSee Chart 1. bReaction run at 0 ~

CI CI Ph si_Cl

oAo " b

Ph LB CI c'l

(R,R)-76

-

RoHo

BnMe2Si ~

] ci

e

,,,st< o'

0~0

0

\\

BnMe2Si 84 !

i background reaction !

85

' "one phosphoramide" , pathway |

85

FIGURE 6. Possible scenarios for the low diastereoselectivity.

20

SCOTT E. DENMARKAND SHINJI FUJIMORI

The aldol addition of TMS enol ether 83 catalyzed by bisphosphoramide 56 and SiC14 was also examined (Scheme 15). However, the reaction did not yield the desired product and only led to the elimination of the alkoxy group from aldehyde 42. Replacement of the PMBO group with other ether functions such as benzyloxy or naphthylmethoxy groups, or to noncoordinating functionality such as tert-butyldimethylsilyloxy or triisopropylsilyloxy groups did not prevent the elimination. In the aldol additions catalyzed by bisphosphoramide 56, simple olefinic aldehydes such as crotonaldehyde and acrolein were problematic due to the competitive conjugate addition of the ionized chloride. 23c In this case, however, the ionized chloride may act as a base to eliminate the alkoxide from activated 42 by deprotonation at 3,-position resulting in formation of 95. PMBO

Me_.x~x~CHO+Me 42

Me

~

0

Ph : ~ 0 OTMS

BnMe2Si

(R,R)-56

SiCI4,DIPEA . M e ~ C H O CH2CI2

83

Me

Me

95

SCHEME 15

Thus, both types of Lewis-base-catalyzed aldol additions were not able to provide the desired aldol product with high.diastereoselectivity. 23b The aldol addition of trichlorosilyl enolate 84 provided the aldol product in good yield, however, the diastereoselectivity is very low. Therefore, an alternative plan to connect these modules was considered. D. SYNTHESIS OF C(11)-C(33) FRAGMENT

On the basis of the studies by Evans et al. 39 and Paterson et al., 4~ the aldol additions of boron enolates derived from methyl ketones bearing a 13-oxygen stereogenic center provide high 1,5-anti stereoinduction. Indeed, better 1,5-anti stereoinduction could be achieved by substratecontrolled aldol addition using the dibutylboron enolate derived from 41 (Scheme 16). The boron-mediated aldol addition of 41 to 42 afforded 85 with excellent diastereoselectivity (85%, dr > 19/1). Although the origin of the 1,5-anti stereoinduction is not clear, the empirical analysis of these types of aldol additions indicated the nature of enolate metal, solvent and protecting group greatly affects the diastereoselectivity. Evans et al. 39

1

21


proposed the intermediacy of multiple sets of competing transition states based on the nonlinear relationship between the reaction temperature and diastereoselectivity. The transition state I leads to the observed 1,5-anti diastereomer. The conformation of the boron enolate can be rationalized by the avoidance of lone-pair repulsion between the [3-oxygen (benzylidene acetal) and the enolate oxygen. This repulsion would force the C(23)-O bond to be antiperiplanar to the C(24)-C(25) bond, and the two enolate faces would be sterically differentiated. The aldehyde will approach from the less sterically hindered side. A similar analysis has been provided for aldol addition of a boron enolate beating a [3-benzyloxy group. 4~ Also, Lee et al. 42 proposed an analogous model to explain the 1,5-anti stereoinduction in the addition of a lithium enolate bearing [3-alkoxy stereogenic center. Ph O~O

19

BnMe2Si ~

M

O

PMBO

e

+M e ~ C - H O Me Me 42 Ph

41 PMBO Me ~ Me

OH O

S

i

M

e

O1"1"O

2

B

n

Bu2BOTf DIPEA Et20,-78 ~ 85%, dr >19/1

H , 24

,H", 'u2

Me

85 SCHEME 16

The configuration at C(27) of 85 was confirmed to be R by Mosher ester analysis of the aldol product (Figure 7). 43 The Mosher ester was prepared by the reaction of 85 with 1-methoxy-l-trifluoromethylphenylacetyl (MTPA) chloride in pyridine/CDC13 and was analyzed by ~H NMR spectroscopy of the crude reaction mixture. Both diastereomers of the R- and S-MTPA esters (96 and 97, respectively) were analyzed, and their chemical shift differences at C(26)-methylene and C(28)-vinyl protons were measured. In the most relevant conformer of these esters, the CF 3 group, the ester carbonyl group and the C(27) hydrogen are all in the same plane (I and II, Figure 7). 44 The phenyl

22


Ph

(R)-MTPA PMBo

H ~o

,~ o

o

o

PMBo

Me. ~ .,.~28~.,26,~ ~ .~/~ T ___v ~-.~ 27v v v ~ Me

Me H 96: 5H2C(26 ) = 2.70 ppm 5HC(28) = 5.48 ppm

(R)-MTPA:

Ph

(S)-MTPA \ .

,,,J,,., o

o

o

Me

o F~c~ ~-~__~,,,~ o OMe I

/~

-SiMe2Bn

Me H 97:~H2C(26 ) = 2.77 ppm ~HC(28) = 5.35 ppm

(S)-MTPA:

shielding

o

Me. A ..~28,~26~.,,~ ~ ~ -SiMe2Bn ~r~ ___v ~r~ 27. . . .

"

F~c~ ,/28

MeO'"~?

"~28

/,~, II shielding

FIGURE 7. Mosher ester analysis of the aldol product 85.

substituents of the MTPA ester can influence the chemical shifts of either the C(26) or the C(28) protons. In the R-MTPA ester, the C(26) protons of 96 are shielded (I) and therefore their chemical shift should be lower than that of the S-MTPA ester. On the other hand, the C(28) proton in 97 is shielded (II), and the chemical shift in S-MTPA ester should be lower than that of the R-MTPA ester. These shifts were observed and the assignment of configuration was made accordingly. The C(25) carbonyl group in 85 was reduced in the presence of diethylmethoxyborane and sodium borohydride, and the resulting syn-diol was protected as a benzylidene acetal to give 98 (86%, two steps, Scheme 17).45 The origin of the stereocontrol in the syn reduction arises from the formation of a boron chelate 100 with [~-hydroxy ketone and diethylmethoxyborane (Figure 8). The axial attack of hydride provides borate 101 leading to the observed syn-diol. The oxidative unmasking of the alkenylsilane to reveal the aldehyde functionality at C(19) was accomplished by the protocol developed by Tamao et al. 46 (73%). The benzyl group of the silane was cleaved by the action of fluoride, and the resulting silanol was oxidized by hydrogen peroxide under basic conditions. It is essential to completely remove the excess peroxide during the workup because the concentration of the crude 99 with trace amount of peroxide leads to decomposition. In addition, aldehyde 99 slowly decomposed even when it was stored at - 15 ~ and it was therefore necessary to use this aldehyde immediately after purification. 47

1

23

TOTALSYNTHESIS OF RK-397 o.~

Et.B/

,,

J

0

' '

H 100

ii

..m

~

Et

H

101

LJ hydride attack

FIGURE 8. syn Reduction of the aldol product 85.

!

PMBO Me. ~

/~..~

Me Me Me

Ph

OH 0 O/X'O ~ ~ ...~//~ 27 25 19 SiMe2Bn 85

Me~j .... ~.OPMB

M

e

98

,

"

OvO

Ph

15h

]

86%, dr >19/1

TBAF; Me2Bn H202'KHCO3D. Si THF/MeOH

~

O~/0

1. Et2BOMe, NaBH4 2. PhCH(OMe)2 CSA

73%

Me Me) . . .' (. OPMB ,,L / / ~ 2 7

Me'

25/~

---~ "/-

"1(

v

O\/0

OvO

Ph

Ph

]

99

/~19H

~

"-j

-11-" 0

SCHEME 17

With anticipation of 1,5-anti stereoinduction to establish the C(19) stereogenic center, the second iteration of the aldol addition was carried out using 41. The aldol addition product 102 was obtained in high yield and excellent selectivity (88%, dr > 19/1, Scheme 18). The aldol addition showed excellent 1,5-anti stereoinduction, and the configuration at C(19) was confirmed by Mosher ester analysis (Figure 9). 43 The analysis of the configuration was carried out in a similar manner. The chemical shifts of C(18)-methylene and C(21)-methyne protons in.R- and S-MTPA esters were measured. The results indicated that the configuration

24

SCOTT E. DENMARKAND SHINJIFUJIMORI

Me Me...,j. . . . OPMB ',~_.../ ~ 2 7 ~

Me" ~

y

y

O\/0 I Ph

99

Me t Me.,P. . . . OPMB '~ 27 M e ' " O\ O 102 II" Ph e Me....J....r/OPMB

~

~

~19

=Y -7-- Y OvO

0

15h

41, Bu2BOTf DIPEA Et20,-78 ~ 88%, dr > 19/1

BnMe2Siv~'~,,,O-,,,1,,,Ph 1. Me4NHB(OAc)3 11 I " 2. Me2C(OMe)2 " k.,. ,,O CSA 19 "1' ~ OvO OH O 87%,dr>19/1 Ph BnMe2Siv ~ , , . O . ~ l , , , Ph ~ I TBAF: "" L,~,,,O H202, KHCO3

Me" ~ 103

O\/O "1 Ph

OvO 15h

O .O Me'Me

THF/MeOH 86%

Me O.~--.~,.,O...],,, Ph / Me..P. . . . . OPMB H k.....,O DDQ ~..,./~27~ ~ ~ 1 9 ~ .,~ Me"' CH2CI2/H20 rt O~/O O.,.~fO 0 .0 90% 104 ][ _ Me'~Me Ph 15h iMe O ~ , , , O . . . ] , , , Ph Me..P.... r/OH H L..~,,.O I,,,. ~,,,,,,.27~ ~ ~ ~19~ .,,,,,I Me'" O\/O OvO O .O 105 | ae'~ Ph Ph Me SCHEME 18

at C(19) was S. The carbonyl group at C(17) was reduced using tetramethylammonium triacetoxyborohydride, and the resulting anti-diol was protected as an acetonide to give 103 (87%, two steps). 48 The stereochemical course of the anti reduction can be rationalized by the transition state illustrated in Figure 10. The formation of alkoxy borohydride III (Figure 10)

l

25


Me

BnMe2Si" v ' ~ ' " O " ' ] '''Ph

/

MeJ .... /OPMB [~ M

21

e

'

"

O\ / 0

t-~ ,,,O ']

~

OvO

][

-

Ph

Ph

106: (R)-MTPA 5H2C(18) = 2.86 ppm a HC(21) = 3.80 ppm

O, 0 Ror S-MTPA 107: (S)-MTPA 8H2C(18) = 2.90 ppm ,5HC(21)= 3.62 ppm i.

(S)-MTPA:

shielding

(R)-MTPA:

19

O

F ~_ 0 3C ~~/\

O

ncjL

0, a ~~'" l 8 ....

,.~

""

.H--~l8

j73. IX. Epilogue

Organic chemists have made enormous progress over the past four decades in developing new methods and strategies for the construction of increasingly complicated molecules. However, the efficiency, reliability, and predictability of the methodology still leaves much to be desired as confirmed by the twists and turns in the present synthesis. The assembly of polycyclic molecules is especially challenging, since extended, linear routes tend to be necessary, resulting in an inexorable decline in the

2

TOTALSYNTHESIS OF MARINEDITERPENOID,DIISOCYANOADOCIANE

57

quantities of intermediates. It is therefore important that we continue to attempt difficult syntheses and in this way define the scope and limitations of our chosen procedures, as well as develop new strategies and methodology. Apart from some very notable exceptions, progress has been and will continue to be incremental, and relies on the collective efforts of a large number of synthesis groups beavering away and sharing information. This synthesis and the knowledge gained in its execution is offered as one such minor contribution to the common enterprise. Acknowledgments

We are indebted to Professor Mary Garson (University of Queensland) for the provi' sion of authentic samples of diisocyanoadociane (10) and diamine 96; to Dr Jamie Simpson (Monash University) for helpful advice; to Bruce Twitchin and Tony Herlt (ANU) for technical assistance; to Tony Willis and Alison Edwards (ANU) for X-ray studies; and to Chris Blake (ANU) for assistance with high field NMR spectra. References and Footnotes

1. Chang, W. C., Progress in the Chemistry of Natural Products, Vol. 80, Springer, Wien: New York, 2000. 2. The numbering system for "isocycloamphilectane" (4) derived from its presumed biosynthetic relationship to structures 2 and 3, and will be used for intermediates throughout this chapter. 3. Baker, J. T., Wells, R. J., Oberh~insli, W. E., Hawes, G. B., J. Am. Chem. Soc. 1976, 98, 4010. 4. Kazlauskas, R., Murphy, E T., Wells, R. J., Tetrahedron Lett. 1980, 21, 315. 5. K6nig, G. M., Wright, A. D., Angerhofer, C. K., J. Org. Chem. 1996, 61, 3259. 6. Wright, A. D., Wang, H., Gurrath, M., K6nig, G. M., Kocak, G., Newmann, G., Loria, E, Florey, M., Tilley, L., J. Med. Chem. 2001, 44, 873. 7. Corey, E. J., Magriotis, E A., J. Am. Chem. Soc. 1987, 109, 287. 8. Fairweather, K. A., Mander, L. N., Org. Lett. 2006, 8, 3395. 9. Cossey, A. L., Gunter, M. J., Mander, L. N., Tetrahedron Lett. 1980, 21, 3309. 10. Piers, E., Llinas-Brunet, M., J. Org. Chem. 1989, 54, 1483. 11. Schultz, A. G., Macielag, M., Podhorez, D. E., Suhadolnik, J. C., Kullnig, R. K., J. Org. Chem. 1988, 53, 2456. 12. Stork, G., Darling, S. D., J. Am. Chem. Soc. 1960, 82, 1512. 13. Evans, D. A., Stereoselective Alkylation Reactions of Chiral Metal Enolates. In: Asymmetric Synthesis, Vol. 3, Morrison, J. D. (Ed.), Academic Press: New York, 1984, p. 2. 14. Dalcanale, E., Montanari, F., J. Org. Chem. 1986, 51,567. 15. Hook, J. M., Mander, L. N., Woolias, M., Tetrahedron Lett. 1982, 23, 1095. 16. Crabtree, S. R., Chu, W. L. A., Mander, L. N., Synlett 1990, 169. 17. Sharpless, K. B., Lauer, R. E, Teranishi, A. Y., J. Am. Chem. Soc. 1973, 95, 6137. 18. Tietze, L. E, Chem. Rev. 1996, 96, 115.

58

KELLYA. FAIRWEATHER,SIMONR. CRABTREE,AND LEWIS N. MANDER

19. Dess, D. B., Martin, J. C., J. Org. Chem. 1983, 48, 4155. 20. The X-ray structure was solved by Dr Alison Edwards, whose expertise we gratefully acknowledge. 21. Ireland, R. E., Mander, L. N., J. Org. Chem. 1969, 34, 142. Ziegler, E E., Kloek, J. A., Tetrahedron 1977, 33, 373. 22. Charette, A. B., Beauchemin, A., Org. React. 2001, 58, 1. 23. Suda, M., Synthesis 1981, 714. 24. Wenkert, E., Acc. Chem. Res. 1980, 13, 27. 25. Cotsaris, E., Paddon-Row, M. N., J. Chem. Soc., Perkin Trans. 2 1984, 1487. 26. Greeves, N. In: Comprehensive Organic Synthesis, Vol. 8. Trost, B. M. (Ed.), Pergamon Press: Oxford, 1991, pp. 1-24. 27. Mander, L. N., Partial Reduction of Aromatic Rings by Dissolving Metals and by other Methods. In: Comprehensive Organic Synthesis, Vol. 8. Trost, B. M. (Ed.), Pergamon Press: Oxford, 1991, pp. 489-521. 28. Fairweather, K. A., Ph.D. Dissertation, Australian National University, 2006. 29. Ito, Y., Hirao, T., Saegusa, T., J. Org. Chem. 1978, 43, 1011. 30. Luche, J. L., J. Am. Chem. Soc. 1978, 100, 2226. 31. Corey, E. J., Ohno, M., Mitra, R. B., Vatakencherry, P. A., J. Am. Chem. Soc. 1964, 86, 478. 32. The X-ray structure was solved by Tony Willis, whose expertise we gratefully acknowledge. 33. Mander, L. N., Sethi, P. A., Tetrahedron Lett. 1983, 24, 5425. 34. Yang, Z., Lorenz, J. C., Shi, Y., Tetrahedron Lett. 1998, 39, 8621. 35. Furrow, M. E., Myers, A. G., J. Am. Chem. Soc. 2004, 126, 5436. 36. Hutchins, R. O., Maryanoff, B. E., Milewski, C. A., J. Am. Chem. Soc. 1971, 93, 1793. 37. Barton, D. H. R., McCombie, S. W., J. Chem. Soc., Perkin Trans. 1 1975, 1574. 38. Bartlett, P. A., Johnson, W. S., Tetrahedron Lett. 1970, 4459. 39. The authentic sample was provided by Professor Mary Garson. 40. Simpson, J. S., Garson, M. J., Org. Biomol. Chem. 2004, 2, 939. 41. Mander, L. N., Synlett, 1991, 134. 42. Stork, G., Rosen, P., Goldman, N. L., J. Am. Chem. Soc. 1961, 83, 2965. 43. A more direct strategy for elaborating a primary amino group at a quaternary center by adding tosyl nitrene to a methylene group has been reported by Wood (White, R. D., Keany, G. E, Slown, C. D., Wood, J. L., Org. Lett. 2004, 6, 1123). As applied to our substrate, however, we would have expected this method to afford the undesired C-20 epimer.


Chapter 3 TOTAL SYNTHESES OF ZOAPATANOL Janine Cossy, Vdronique Bellosta, and Catherine Taillier Laboratoire de Chimie Organique, ESPCI, CNRS 10 rue Vauquelin, 75231 Paris Cedex 05, France

I. II. III. IV. V. VI. VII. VIII.

Introduction Nicolaou's Synthesis Chen's Synthesis Cookson's Synthesis Kocienski's Synthesis Kane's Synthesis Trost's Synthesis Our Approaches for the Total Synthesis of (+)-Zoapatanol A. Ring-Closing Metathesis Approach B. Homer-Wadsworth-Emmons Approach IX. Conclusion References and Footnotes

59 60 64 67 69 73 78 83 83 90 95 95

I. Introduction

(+)-Zoapatanol 1, montanol 2, tomentanol 3 and tomentol 4 are diterpenoid oxepanes isolated from the leaves of the Mexican zoapatle plant Montanoa tomentosa, which Mexican women have been using for centuries to prepare "tea" to induce menses, labor and to terminate early pregnancy. 1 Recent studies support the belief that zoapatanol and its metabolites might be responsible for the observed antifertility activity. 2 In 1979, the isolation and the structure of zoapatanol were described. 3 Due to its biological profile and its challenging structure, several synthetic approaches have been described 4 and seven total syntheses of zoapatanol have been reported 5-11 but only two of them were enantioselective, l~ Key issues for a successful synthesis of zoapatanol 1 are the stereocontrolled construction of the oxepane ring containing the two stereogenic centers, the introduction of the (E)-exocyclic double bond and the installation of the nonenyl side chain. Since (+)-zoapatanol was isolated as

60

JANINECOSSY,VI~RONIQUEBELLOSTA,AND CATHERINETAILLIER

4p

5p

OR'

HO,,.

O 1 2 3 4

R R R R

= = = =

(CH3)2C=CHCH 2(CH3)2CHC(CH3)=CHH2C=C(CH3)CH(CH3)CH 2(CH3)2C(OH)CH=CH-

R'= R'= R'= R'=

H H H Ac

Zoapatanol Montanol Tomentanol Tomentol

FIGURE 1. Oxepanederivatives isolated from M.

tomentosa.

a 1/1 mixture of epimers at C6, control of this stereocenter is not required (Figure 1).12 In 1980, the first two syntheses of (_+)-zoapatanol were disclosed, one by Nicolaou 6 and the other by Chen and Rowand, 5 and six other syntheses have been disclosed since then. 4,7-11 For all of them, the construction of the oxepane ring was achieved through the formation of the O1'-C7' or O1 '-C2' bonds except for one ~ in which the C4'-C5' bond was formed. II. Nicolaou's Synthesis

A convergent synthesis of (+)-zoapatanol was achieved. 6 In the retrosynthetic analysis, the oxepane ring would be obtained by nucleophilic attack of the tertiary alcohol on the epoxide present in compound I via a 7-exo-tet process that would allow the formation of the O 1 '-C7' bond. The syn-1,2-diol present in compound I would come from a chelation-controlled addition of a methyl Grignard reagent to the ~-benzyloxymethoxy ketone II according to the Cram-chelate model A (Figure 2). ~3 Ketone II would be synthesized from glycidol 5 as the starting material and bromide 10 as the precursor of the long side chain of zoapatanol (Scheme 1). Bromide 10, which would allow the introduction of the side chain of zoapatanol, was synthesized from 5-hydroxy-2-pentanone. By reaction with an excess of (methoxymethyl)triphenylphosphorane, ketone 6 was transformed in 75% yield to methoxy enol ether 7, which was directly and quantitatively converted to dithiane 8 [HS(CHz)3SH, HC1 gas, CHC13, 0-25 ~ Alkylation of the dianion derived from 8 (2.2 equiv, of n-BuLi, THF, - 7 8 ~ ---> - 1 5 ~ with 1-bromo-3-methyl-2-butene (1.1 equiv., - 7 8 ~ ---> - 1 5 ~ led to 9 (85% yield), which was converted to 10 in five steps, with a 60% overall yield, by using classical transformations: acetylation (Ac20, Pyr, DMAP, CHzC12, 0 ~ followed by removal of the

3

61

TOTALSYNTHESESOF ZOAPATANOL

Nu~% ~.O,,,,

Ph O.si...t_gu

O---M

Ph

CH2Ph A

FIGURE 2. Cram-chelate model.

CH2OH

O/-Ph k_O,,,

0 (+_)-Zoapatanol 1

OH O

I

0 ~~ ~ ~ 0

Br

/--Ph

OSi(Ph)2t-Bu

10

k~oH II

5

SCHEME 1. The Nicolaou strategy.

dithiane by using HgC12 (CaCO 3, MeCN/H20, reflux), formation of the dioxolane (HOCHzCHzOH , TsOH, benzene, reflux) then reduction of the acetate using LiA1H4 (ether, 0 ~ Finally, transformation of the resulting alcohol by using CBr 4 in the presence of PPh 3 ( - 4 0 ~ ~ 0 ~ produced the desired bromide 10 (Scheme 2). The construction of the oxepanic fragment began with the preparation of the protected aldehyde 11 from glycidol 5 employing classical chemistry (Scheme 3). Incorporation of the side chain of zoapatanol was accomplished by the coupling of the Grignard reagent derived from bromide 10 (Mg, THF, 25 ~ with aldehyde 11 at - 7 8 ~ providing a secondary alcohol (80% yield), which was oxidized to ketone 12 (Collins' reagent, 0 ~ CH2C12). This ketone was then treated with MeMgC1 (THE -105 ~ to afford the Cram-chelate-derived syn-1,2-diol 13, as the only detectable product (Scheme 3 and Figure 2).

62

JANINECOSSY,Vt~RONIQUEBELLOSTA,ANDCATHERINETAILLIER MeO

~

PPh3 M e O ~ O H

75%

7

6

quant.

I

.s/

sH

HCI (g)

r%

1. n-BuLi

S

~

O

H

85%

1. Ac20, Pyr, 4-DMAP 2. HgCI2, CaCO3, MeCN/H20 3. HOCH2CH2OH,TsOH

@

V

~

4. LiAIH4, Et20 5. CBr4, PPh3 60%

B

r

10

SCHEME 2. Synthesisof the side-chain precursor.

After deprotection of the silyl ether (n-Bu4NF), a chemoselective epoxidation of the allylic alcohol in the presence of the unactivated olefin was performed with tert-butyl hydroperoxide in the presence of vanadyl bisacetoacetate [VO(acac) 2] to produce the epoxy alcohol 14 (80%). ~4 This selectivity is due to the complexation of the vanadium derivative [VO(acac)2] by the free hydroxyl group of the allylic alcohol and by the oxidative reagent t-BuOOH, according to a six-membered ring pseudochair transition state of type B (Scheme 3). The construction of the oxepane ring by an internal epoxide ring-opening was realized with a high degree of regioselectivity by subjecting epoxide 14 to KCH2SOCH 3 in MezSO, leading to the desired oxepane 15 (75%). This latter compound was then transformed to the key ketone 16 by oxidative cleavage of the 1,2-diol using NaIO 4 via intermediate C (95%) (Scheme 3). The oxepanone 16 then underwent a condensation with the lithium salt of

3

63

TOTAL SYNTHESES OF ZOAPATANOL

1. DHP, H+

~ H

.OBn r/ l O"]""~'~

~

OH L__/

~~

3. TBDPSCl 4. PhCH2OCH2Cl,iPr2NEt 5. AcOH, THF/H20 6. Pyr.SO3 56%

5

|

r . . . o v Ph

~-~o., n-~u,i

OTBDPS

I 1.10, Ug, THF 2. CrO3, Pyr OBn

MeMgBr OTBDPS EHF,-100~ "

r" l~

i.

~

95%

~176 0

13

~

" 12

L I o~Vi ' ',L R~.~O-O-t-Bu H

1. n-Bu4NF 2. t-BuOOH VO(acac)2

80%

BnO

BnO k---O,,,

OH KCH2SOCH3 DMSO 75%

~J

14

OH

15

OH

l'-/4-~- 11ol" "~

95%

NalO4 EtOH/H20

C

| ~

BnO __ k--O'"~k/-~ --'J

~ O

L__/

CO2Me "1 (MeO,2P(O)CH2CO2Me| n-BuLi, THE - "

17

BnO ~ ~ - - O ' " ~ ~ :~O ~ ~ 0 ''J k_J

16

E / Z = 2.5/1 /

1. DIBAL-H 1 70% 2. Separation BnO ~--O . . . .

CH2OH 1. Li/NH3(liq) 2. AcOH/THF/H20

k~/

18

=

(+)-Zoapatanol1

80% SCHEME 3. Completion of the Nicolaou synthesis.

64

JANINE COSSY, VI~RONIQUE BELLOSTA, AND CATHERINE TAILLIER

trimethylphosphonoacetate (LDA, - 2 0 ~ to afford selectively the unsaturated methyl ester 17 as a 2.5/1 mixture of geometric isomers. After reduction of this mixture of isomers by DIBAL-H, the corresponding allylic alcohols were obtained as a mixture of E/Z-isomers (2.5/1) that were separated by chromatography to provide the desired stereoisomer 18 of E-configuration. Finally, deprotection of the secondary alcohol of the obtained oxepane 18 under Birch conditions [Li~H3(liq), - 7 8 ~ according to the mechanism presented in Scheme 4, followed by acidic treatment (AcOH, THF/H20) to release the carbonyl function, furnished ( +)-zoapatanol.

RCH2O_ C H 2 ~

f--"~e- ,. RCH2O~-CH2~--~,

1 RCH2OH =H20 RCH2O@ + H2C==~" eH3C

H20 SCHEME 4. Birch reduction of a benzyl group.

The Nicolaou total synthesis of racemic zoapatanol required 16 steps, and was accomplished in an overall yield of 12%. III.

Chen's Synthesis

Instead of using basic conditions to build the oxepane ring from an co-hydroxy epoxide, Chen and Rowand 5 chose to utilize acidic conditions to transform an e,~-epoxy alcohol to an oxepane by creating the O1'-C2' bond of (_)-zoapatanol. The synthesis of (_)-zoapatanol was envisioned starting from the ~,~-epoxy alcohol III. This compound would be prepared by selective epoxidation of 2-methyl-6-methylene-(E)-2,7-octadien-l-ol 19, which could be synthesized from myrcene (Scheme 5).

3

65

TOTALSYNTHESESOF ZOAPATANOL OH

~

(_+)-Zoapatanol1

RO;"~CH3LOH III

H

O

~

Myrcene

'

19

SCHEME5. The Chen strategy.

The regioselective oxidation of myrcene by SeO 2 (Scheme 6) produced the allylic alcohol 19, ~5 which after treatment with m-CPBA was transformed to the epoxy alcohol 20. Reaction of 20 with bromine gave an allylic dibromide (Scheme 7, via intermediate D), which was subjected to potassium acetate to give a diacetate intermediate. This latter was then tosylated to afford, after treatment under basic conditions (KzCO 3, CH3OH/H20 ), the pure diol 21. Reaction of 21 with the Grignard reagent 22 (3.3 equiv.) in the presence of a catalytic amount of LizCuC14 ( - 2 0 ~ ----> 0 ~ gave a complex mixture that was directly treated with acetic anhydride. After purification by chromatography followed by a basic hydrolysis, the key intermediate 23 was isolated in 32% yield. Next, treatment of 23 with trifluoroacetic acid (0.1 equiv.) in methylene chloride afforded

)••,•e=o

ene reaction= ~~e -OH

H~._.,/O

I [2,3]-sigmatropic

rearrangement

,, H20 OH

~/R O-Se-OH

SCHEME6. Allylicoxidationby seleniumdioxide.

1. i) SeO2

ii) NaBH4 He

2. m-CPBA 16%

Myrcene

//i Br2

/~~Br-

20

Br 14%

1. Br2, CH2012,0~ 2. AcOK, Acetone reflux 3. TsCI, Et3N 4. K2003, MeOH/H20

OH

OH TsO ~ 0::i

~k....

:

. THPO ~

23

(''"

O

H

O 22

MgBr

21

2.,,~:~O014

OTHP

1. CF3CO2H CH2CI2, 0 ~

3. K2CO3, MeOH/H20 32%

path b

iOAc

,~ THPO

17% 25

2. Ac20, Pyr

/iOAc

30%

THPO_ ~/

~

"-,,s-.---O 24

24%

1. Jones reagent (CrO3, H2SO4) 2. ~ L i

(+)-Zoapatanol 1 SCHEME 7. Chen's synthesis of (+__)-zoapatanol.

3

67

TOTALSYNTHESES OF ZOAPATANOL

the oxepane 24 as the major product (30% yield), resulting from nucleophilic attack of one of the primary hydroxy groups on the more substituted carbon of the epoxide (Scheme 7, path a). Under these conditions, tetrahydropyrane 25 was also formed in 17% yield, resulting from the nucleophilic attack of the hydroxy group on the less substituted carbon of the epoxide, according to a 6-exo-trig process (Scheme 7, path b). The final transformation of 24 to (_+)-zoapatanol was accomplished in two steps. After a deprotection-oxidation process of the primary alcohol using Jones reagent, the corresponding carboxylic acid was formed and its treatment with an excess of 3-methyl-2-butenyllithium led to (+_)-zoapatanol (24% yield, two steps). This synthesis of racemic zoapatanol spanned 13 steps from myrcene and gave an overall yield of 0.15%. A similar approach to Chen's strategy was used by Cookson et al. and Kocienski et al. to build up the oxepane ring of ( +)-zoapatanol as the O 1'-C2' bond of the oxepane was formed by treatment of an e,~-epoxy alcohol under acidic conditions. IV. Cookson's Synthesis

Cookson et al. have envisaged the synthesis of the key intermediate e,~epoxy alcohol IV from the homoallylic alcohol V, which would come from a methylalumination of the acetylenic compound VII, followed by functionalization of intermediate VI. The acetylenic compound VII would be obtained by alkylation of the enolate resulting from deprotonation of tert-butyl propionate (Scheme 8). //---x~

(+_)-Zoapatanol 1

", R O ~ c H 3 ~ j O H IV

VI

OH3

)

t-BuO2C VII SCHEME 8. The Cookson strategy.

oH

68


The synthesis of (_+)-zoapatanol started with the alkylation of the enolate of tert-butyl propionate with 5-iodopent-1-yne. The straightforward reduction of the obtained acetylenic ester with LiA1H4 and subsequent benzylation of the resulting alcohol gave the terminal acetylenic compound 27 (68%). The transformation of the acetylenic compound 27 to the homoallylic alcohol 28 was achieved by methylalumination with trimethylaluminium in the presence of bis(cyclopentadienyl)zirconium dichloride (CpzZrCI2) (Scheme 9). 16The conversion of the alane intermediate of type VI into the more reactive aluminate with n-BuLi, followed by quenching with ethylene oxide allowed the formation of the desired homoallylic alcohol 28 with an overall yield of 62%. ~7 One proposed mechanism for the zirconium-catalyzed carboalumination involves the methylenation of CpzZrC12 with Me3A1 producing MeZrCpzCI and MezA1C1, followed by methylzirconation of the alkyne. After transmetalation of the resulting alkenylzirconium derivative with MezA1C1, the alkenyldimethylalane is formed and CpzZrC12 is regenerated. 18

Me3AI + Cp2ZrCI 2

_

-

Me2A I -

,,CI. ~ ZrCP2Me CI

R

R'

Me/~-~Z,rCP2 Cl ,Cl ~,[Me 2 R

1

IR R~-~R' = '+8 -5 Me--ZrCP2--CI- -AIMe2CI

m

R'

i

R'

+ CP2ZrCI 2 AIMe 2 SCHEME 9. Proposed mechanism for the Zr-catalyzed carboalumination.

Compound 28 was then transformed to the e,~-epoxy alcohol 31 in seven steps using classical chemistry. After mesylation of 28 (MsC1, Et3N), nucleophilic displacement of the mesylate using NaI in acetone, the corresponding iodide derivative was produced. Alkylation of the lithium anion of diethyl 2-triphenylphosphoranylidene butadienoate 29 by the previously synthesized iodide under basic conditions (LDA) afforded, after benzoic acid-catalyzed elimination of triphenylphosphine, the E-diester 30 in 63% yield. Reduction of 30 with DIBAL-H resulted in the formation of the corresponding unsaturated diol (97% yield). After acetylation of this diol

3


69

(Ac20, Et3N), epoxidation of the more electron-rich olefin with m-CPBA 19 and saponification with K2CO3 in MeOH, the desired e,~-epoxy alcohol 31 was isolated in 88% yield over four steps. This latter compound cyclized when treated with SnC14 in THF leading to the oxepane derivative 32 in 79% yield and to the corresponding tetrahydropyran derivative 33 in 7% yield. The epoxide ring-opening had presumably occurred stereospecifically with inversion of configuration. We have to point out that the use of SnC14 is crucial for the success of this cyclization and appears to be much better than CF3COzH, previously used by Chen et al., in terms of regioselectivity, as the oxepane derivative is formed preferentially to the pyrane derivative. The end of the synthesis of (_+)-zoapatanol from 32 was straightforward. After protection of the hydroxy groups as tert-butyldimethylsilyl ethers (TBSOTf, 2,6-1utidine), selective cleavage of the benzyl ether using Birch conditions [Li/NH3(liq), -78 ~ followed by oxidation of the resulting primary alcohol with PCC on alumina, and subsequent treatment of the corresponding aldehyde with silver(I) oxide, carboxylic acid 34 was isolated with an overall yield of 61%. Addition of prenyllithium to acid 34 then allowed the introduction of the desired [~,y-unsaturated ketone moiety. Finally, treatment with 25% HF in acetonitrile led to the cleavage of the silyl ethers, producing (+_)-zoapatanol (Scheme 10). Cookson's approach to (_+)-zoapatanol took 19 steps from tert-butyl propionate, and was achieved with an overall yield of 4.9%. V. Kocienski's Synthesis

Kocienski et al. envisaged the synthesis of zoapatanol by forming the O1'-C2' bond according to the same procedure developed by Cookson et al., i.e., the formation of the oxepane ring by SnC14-induced intramolecular cyclization of an ~,~-epoxy alcohol. The synthesis of the ~,~-epoxy ether VIII precursor of (_+)-zoapatanol was envisioned from diene IX, which would be obtained by the ring-opening of dihydrofuran 35. The stereochemistry of the exocyclic double bond would be controlled during the formation of the oxepane ring as the hydroxyl group at C 1' would be the only one to be able to open the oxirane ring (Scheme 11). Construction of the C3-C4 double bond of the key homoallylic ether IX was performed via a Ni(0)-catalyzed coupling between MeMgBr and 2-alkyl dihydrofuran 352o (Scheme 12). At first, the 5-1ithio-2,3dihydrofuran was alkylated with the appropriate alkyl iodide affording 35, which reacted with MeMgBr in the presence of (PPh3)zNiC12 to produce

1. LDA, THF,-78~ then ~ " " ' - ' " ~ I, HMPA

CO2t-Bu

.)

Bn

-78~ to rt 2. LiAIH4, Et20 3. Nail, DMF then PhCH2Br 68%

26

27

1. Me3AI, CP2ZrCI2 CICH2CH2CI, rt 2. n-BuLi then O

62%

/___&

I/OBn

/

CO2Et

1. MsCI, Et3N 2. Nal, Acetone

.OBn

3.C.O2Et 30

CO2Et

28

, ~ P P h 3 , LDA

29 CO2Et

88%

then PhCO2H Phil, reflux 63%

1. DIBAL-H, PhMe -78~ to -30~ 2. AcCI, NEt3 3. m-CPBA 4. K2003, MeOH

.OH

0 ~ ,

,--OH

OH

SnCI4

o:;'

33 (7%)

THF ,. BnO - - /

H O , , , ~

BnO

31

OH

oJ 32 (79%)

60%

1. ~ L i ~)-Zoapatanol 1

1. TBSOTf, 2,6-1utidine 2. Li, NH3(liq) THF, -78~ 3. PCC, AI203, rt 4. AgNO3, NaOH, MeOH OTBS

2. 25% HF, MeCN 44% 34 SCHEME 10. Cookson's synthesis of (_+)-zoapatanol.

3

71


(+)-Zoapatanol 1

>

1/--R OR "~2 1

, O',' VIII

2

35

IX SCHEME 11. The Kocienski strategy.

alcohol 36. The C2-C3 double bond of the key intermediate of type VIII was then introduced by using carbomagnesiation of butyn-l,4dio121 with Grignard reagent 37, which was obtained by standard methods from alcohol 36 (mesylation, transformation to the corresponding bromide by displacement of the mesylate by LiBr, and treatment with magnesium). The high trans-stereoselectivity (>95%) is probably due to an internal coordination of the alkenylmagnesium bromide 38. Protection of the hydroxy groups as acetates (Ac20, Et3N), and treatment of 39 with 1 equiv, of m-CPBA afforded epoxide 40. It is worth mentioning that the reaction was not totally chemoselective, as competitive epoxidation of the disubstituted terminal alkene also took place. After methanolysis of the acetate, opening of epoxy alcohol 41 occurred when this compound was subjected to SnC14 in THF at - 2 0 ~ affording the required oxepane 42. Conversion of the oxepane 42 to (_+)-zoapatanol started with the protection of the hydroxy groups as t-butyldimethylsilyl ethers (TBSOTf, 2,6-1utidine), followed by selective hydroboration of the terminal double bond, and Swern oxidation. Reaction of the aldehyde thus formed with dimethylsulfonium methylide 22 gave epoxide 43 via intermediate E. Nucleophilic cleavage of the resulting epoxide 43 was achieved with the homocuprate derived from 2,2-dimethylvinyllithium and CuI to afford the alcohol 44. Finally, Swern oxidation of the secondary alcohol and removal of the protecting groups with HF produced (+)-zoapatanol. The synthesis of (_+)-zoapatanol developed by Kocienski's group proceeded over 16 steps from dihydrofuran in 6.6% overall yield (Scheme 12).

1. n-BuLi

89%

i

rl sc"Et3

35 MeMgBr 92% (PPh3)2NiCI2 Et20/PhH

MgB ~

/

\

OMgBr Et20

36

OMgBr OMgBr

.--~~ ~ . . ~ / O A c

/

Ac20

OMgBr J

Et3N 50%

~ ~ / # M g B E 38

OH

2. LiBr 3. Mg, Et20 92%

~

OH

~OAc

39

I m-CPBA @ @ . . O R

SnCI4

64%

"ON

42

K2CO3 ~ 40 R = Ac MeOH 41 R=H quant.

L_..

1. TBSOTf, 2,6-1utidine 78% 2.9-BBN-H, H202, NaOH 3. Swern 4. Me2S=CH2

[R'~fH| OTBS

|174 .] ~ J'~S(Me)2

(~S(Me)2

/~2

R

vk_A

CuLi , Et20 -10oC 64%

43

..OTBS

~)-Zoapatanol 1

TBSO,,,/----~~

1. Swern 2. HF/MeCN ~ 73%

\

OH

J //O 44

SCHEME12. Kocienski'ssynthesisof (_+)-zoapatanol.

3

73


VI. K a n e ' s

Synthesis

In their retrosynthetic analysis, Kane and coworkers envisaged access to (+_)-zoapatanol from the seven-membered ring lactone X, which would come from a regioselective oxidation of the cx,cx-disubstituted cyclohexanone XI under Baeyer-Villiger conditions. The synthesis of XI was envisaged from the Wieland-Miescher ketone 45 (Scheme 13).

(+)-Zoapatanol 1

) o x

OR'

45

XI SCHEME 13. The Kane strategy.

The synthesis of (_+)-zoapatanol was achieved from the Wieland-Miescher ketone (45), which was transformed to ketal 46 in five steps using classical transformations (Scheme 14). After acidic hydrolysis, the obtained cx,[3-unsaturated ketone 47 was treated with H202/NaOH to produce a mixture of diastereomeric epoxides 48 that were transformed to the monocyclic co-acetylenic ketone 50 via tosylhydrazone 49, by using an Eschenmoser fragmentation under acidic conditions via intermediate F (Scheme 14). 23 Ten steps were then necessary to transform ketone 50 to the key seven-membered ring lactone of type X (Scheme 13). After protection of the carbonyl group of 50 [HO-(CH2)2-OH, TsOH] followed by oxidative hydroboration of the acetylenic moiety using an excess of 9-BBN-H, alcohol 51 was produced as the major product. 24 This latter compound was then oxidized with Collins' reagent (CrO 3, Pyr) and the resulting aldehyde was treated with CH3Li to give a secondary alcohol. After a second oxidation with Collins' reagent, the resulting methyl ketone led to olefin 52 upon treatment with methyltriphenylphosphonium iodide/Nail in DMSO.

O

_OCH2Ph

1. NaBH4

2. Ac20, Pyr 45

I

3. HOCH2CH2OH, Ph, H+ 4. LiAIH4 5. BnBr, PhH/DMSO < 60%

46 I AcOH MeOH

; ?c.2Phl

_OCH2Ph H+

j

H

H

90% 1 _OCH2Ph

_OCH2Ph

H202, NaOH CH3OH 74%

48 I

H p-TsNHNH 2 CH3COOH CH2CI2

+s _OCH2Ph OCH2Ph

83%

Ts

5O

49

V--] 1. HO OH, H+ 80%

HO

2.9-BBN-H NaOH, H202 _OCH2Ph ~ :

1. CrO3, Pyr 2. CH3Li 3. CrO3, Pyr

O_CH2Ph

4. Ph3PCH31 68% 51

SCHEME 14. Preparation of the Kane intermediate.

52

3

75

TOTALSYNTHESESOF ZOAPATANOL

The alkene 52 was converted to a primary alcohol upon oxidative hydroboration (9-BBN-H/NaOH, H202), hydrolysis of the ketal function (H2SO4), and protection of the alcohol as its THP derivative to afford the ketone 53 in 81% yield (Scheme 15). This ~,~-disubstituted ketone was oxidized under Baeyer-Villiger conditions with m-CPBA, to produce the

1.9-BBN-H~ NaOH/H202

ttt

BnO

2. H2SO4, 0.002N 3. DHP, TsOH/EteO "O

O

L._/

52

OTHP

81%

O 53

m-CPBA NaOAc m

BnO ,,,/ T

H

P

O

~

o

\ ~

O

70%

OTHP

Ar ,(~.O O)

54

85%

o

LDA, (EtO)2POCI TMEDA, HMPA,THF ~BnO ,,,/~~ ~

THPO

Na/NH3(liq) HO,, O

r

II

O

O'~-OEt OEt

..•

~174

~"R

%" :

Et

H

55

57%1 H+

.j

THPO

57

~

1. NaH, BnBr 2. BH3then

O -

3. CrO3, Pyr 66%

HO,,,

OTHP 56

SCHEME 15. Kane's approach to the key oxepanone intermediate.

--~

76


desired lactone 54. This reaction is highly regioselective as the more electron-rich bond preferentially migrates via a concerted mechanism (intermediate G, Scheme 15).25 In order to transform lactone 54 to the oxepanone 57, an 1,2-transposition of a carbonyl group was needed. This transposition was achieved in five steps. After formation of the lithiated anion of 54 using LDA and trapping of the enolate intermediate by diethyl chlorophosphate, the obtained enol phosphate 55 was reduced, via intermediate H, to an oxepene using an electron transfer process induced by Na in NH3(liq). Under these conditions, a Birch reduction took place and at the same time the benzyl ether was cleaved. The hydroxy group of 56 was then reprotected (Nail, BnBr) and the enol ether was regioselectively hydroborated to produce a secondary alcohol that was then oxidized using CrO 3 in pyridine. The desired oxepanone 57 was thus obtained in 66% yield (Scheme 15). The next stage of the synthesis focused on the elaboration of the side chain. Ketalization of 57 and concomitant deprotection of the primary hydroxy group followed by oxidation with Collins' reagent, produced an aldehyde intermediate. Next, addition of the Grignard reagent generated from 4-bromo-2-methyl-l-butene provided an alcohol that was acetylated to give acetate 59 (Scheme 16). Isomerization of the double bond of 59 under acidic conditions (TsOH, Phil), basic hydrolysis of the acetate (K2CO3,MeOH/H20), THP ether formation and reductive cleavage of the benzyl ether under Birch conditions [Na/NH3(liq), t-BuOH/THF] afforded alcohol 60. Acetylation of the secondary alcohol and nonselective hydrolysis of the ketal moiety (0.002N H2SO4) followed by reprotection of the secondary alcohol afforded the tetrahydropyran ether 61. Completion of the synthesis of ( +)-zoapatanol required the transformation of the ketone at C6' to the exocyclic (E)-2-hydroxyethylidene group and a functional group transformation on the side chain at C5. Thus, compound 61 was condensed onto triethylphosphonoacetate under basic conditions to provide the unsaturated ester 62 as an inseparable mixture of (E)- and (Z)-isomers (Scheme 16). Reduction of 62 with LiA1H4 gave diastereomeric diols (E/Z = 2/3 ratio) that were separated by chromatography to give the desired diol 63 (28% yield). Diol 63 was diacetylated, the tetrahydropyranyl protecting group removed (CH3CO2H/H20/THF) and the alcohol function oxidized (Collins' reagent) to give the corresponding ketone. Finally, treatment with an excess of tetrabutylammonium hydroxide was used to remove the acetate groups thus leading to (+_)-zoapatanol.

",

O

HO HO, TsOH P h H

L

87%

OH

OTHP

B~ nO

"~"--7~

57

58 1. CrO3, Pyr 77%

2 / ~ , - , MgBr

3. Ac20, Pyr

H O , , , / / - ~ O ~] THPO % ~ ~ ~ . -

O'~O. ZsOH,phH

6~0 / .

-2. Kf;flou31MeOH

1. Ac20, Pyr 63% 2. H2SO4(0.002N) 3. DHP,TsOH .

k k

.

BnO,,,~----~O ~

AcO~~~z ~

59

3. DHP/TsOH 4. Na/NH3(liq) t-BuOH 66%

"

CO2Et

',,

AcO,,, ~ ~ ~ - - - - O

61

o>O

(EtO)2F~C;H2CO2Et Nail = 98%

62 28% 1. LiAIH4 2. Separation

(+)-Zoapatanol1

1. Ac20, Pyr 2. CH3CO2H 3. CrO3, Pyr 4. n-Bu4NOH

HO,,, THPO

71%

SCHEME 16. Completion of the Kane synthesis.

,1OH "-~\~./

/-

/ 63

78


Kane's racemic synthesis of zoapatanol required 40 steps from Wieland-Miescher ketone and proceeded in 0.1% overall yield. VII. Trost's Synthesis

The first enantioselective synthesis of zoapatanol was reported by Trost et al. in 1994. l~ The key step was an intramolecular Williamson reaction

to build up the oxepane ring. The control of the contiguous stereogenic centers at C2' and C3' was achieved via a Sharpless asymmetric epoxidation of an allylic alcohol using t-BuOOH, chiral diethyl tartrate (DET) and Ti(OiPr)4 .26 The configuration of the epoxide formed is dependent upon the enantiomer of DET used and can be predicted according to Scheme 17. The configuration of the (E)-exocyclic double bond was controlled by using a palladium-catalyzed 1,4-addition of triphenylsilanol to a vinyl epoxide. 27 (-)-DET ,,,-OH R 0 OH

.0. ,-OH R

(+I-BET SCHEME 17. Sharpless asymmetric epoxidation of allylic alcohols.

The synthesis of (+)-zoapatanol was envisioned from alcohol XII by using an intramolecular Williamson cyclization. Compound XII would be obtained from a nucleophilic attack of an alcohol on a ~-allylic palladium complex generated from the vinyl epoxide XIII. This compound would be synthesized from epoxide XIV, which would come from a Sharpless epoxidation applied to the allylic alcohol XV, which in turn would be obtained from methallyl alcohol 64 (Scheme 18). Methallyl alcohol 64 was transformed to allylic alcohol 66 in two steps (Scheme 19). After treatment of 64 with 2.2 equiv, of n-BuLi in the presence of tetramethylethylenediamine (TMEDA), the resulting dianion was quenched at first with 1-bromo-3-methylbut-2-ene and then with methoxymethyl chloride, allowing the isolation of dienic compound 65 (92%). The regioselective oxidation of one of the methyl groups in 65,

3

79


(+)-Zoapatanol 1

) "/

~

~ -~ "OH H3C

LG

XII

11 R'O/~/~OMOM "/ XlV

HO~

~

~ -~ "OH H3C XIII

"•OH

"

64

XV SCHEME 18. The Trost strategy.

achieved by using 4 mol% SeO 2 in the presence of t-BuOOH and 10% salicylic acid led to alcohol 66 in 53% yield. Use of L-(+)-diethyl tartrate [(+)-DET] as the chiral ligand gave epoxide 67 in 90% yield and with an ee greater than 95%. 2s The transformation of this latter epoxide 67 to the terminal epoxide 68 was accomplished sequentially, first by a titanium-promoted regioselective opening of the epoxide 29 with acetic acid as the nucleophile. A tosylation of the primary alcohol of the resulting 1,2-diol followed by treatment under basic conditions were used to build up epoxide 68. This sequence of reactions set the two contiguous stereogenic centers present on the oxepane ring of zoapatanol. In order to introduce the side chain present at C2', epoxide 68 was opened by using a cuprate derived from 3-butenylmagnesium bromide and the resulting product was transformed to diol 69 after acidic hydrolysis. The straightforward conversion of allylic alcohol 69 to the vinyl epoxide 70 was achieved in two steps. The first one was an oxidation with Dess-Martin periodinane 3~ (DMP) (Scheme 20), and the second one was Corey's epoxidation using the dimethylsulfonium methylide. 22 The key step, which is the vinyl epoxide ring-opening to stereoselectively produce the monoprotected diol 71, was achieved by treatment of

1. n-BuLi (2 eq) TMEDA 2. ~~-Br "~OH

~~A'OMOM

then MOMCI 92%

65

64

I SeO , t-BuOOH t-BuOOH L-(+)-DET Ti(O/-Pr4)

~OMOM

53% salicylic acid HO I

v ~

~OMOM

90% 67 (ee > 95%)

66

1. AcOH, Ti(O/-Pr4) 66% 2). TsCI 3. OH resin

A C O ~ o M O M ~'0 "ell

1. ~ M g B r Cul, THF

-/

k-)3I:- OH

/~OH

2. HCI 6N, MeOH 68

69 71% / 1. DMP 2. (CH3)2S=CH2

,.

OH

=

"/

~,-}3I: "OH 70

72

[Pd2(dba)3].CHCI3 67%

P(O/-Pr)3 Ph3SiOH, THF

1. Ac20 H 73

2. KF 79%

71

SCHEME19. PreparationoftheTrostintermediate.

3

81


AcC~ OAc

OH

0

0

RI~"-R 2

RI-~R 2

.,. OAc .OAc

AcO~)I

AcO,

0 ~ H

OyR2 O

R1

0

/OAc +

O J] R1 / \ R 2

+

AcOH

0 SCHEME 20. Dess-Martin periodinane oxidation.

70 with [Pd2(dba)3].CHC13 in the presence of triphenylsilanol in THF at room temperature. 27 Under these conditions only the product resulting from cleavage of the distal bond of the epoxide was obtained, and exclusively with the required E-configuration. It is noteworthy that the formation of oxepane 72, which could arise from an intramolecular attack of the tertiary alcohol of 70 on the vinyl epoxide, was not observed. Acetylation of the primary allylic alcohol of 70, followed by the cleavage of the silyl ether using KF, led to the diol 73. The primary alcohol of 73 was then transformed to the corresponding primary triflate [(CF3SO2)20), 2,6-1utidine], and this latter spontaneously cyclized in situ under the reaction conditions, affording oxepane 74 (Scheme 21). At this stage the elaboration of the side chain was required to complete the synthesis. After oxidation of the terminal double bond under Wacker's conditions 31 (LiC1, PdC12, CuC1, DMF, H20, 0 2, rt) generating a methyl ketone (Scheme 22), the acetyl protecting groups were exchanged with tert-butyldimethylsilyl (TBS) groups in order to minimize problems of chemoselectivity at the end of the synthesis. The resulting ketone 75 was transformed to vinylstannane 76, via a vinyl triflate

82

JANINE COSSY, V]~RONIQUE BELLOSTA, AND CATHERINE TAILLIER

~ -/

OAc

(0F3SO2)20 ~ A~cO'"~/--~ -~~OAc 2'6"17~i~ ~ / / ~ v ~ - " O

\-)31~ un [ 73

74

1. LiCI,PdCI2, DMF,H20, 02, rt 71% 2. K2CO3,MeOH 3. TBSCI, imidazole

TBSO

1. KHMDS, -= CI~ | ' ) ; ~',.O T B S N(SO2CF3)2 TBSO

T ~ O ,,,.: ~ Me3Sn" ~3 "~-0 76

76%

2. [(CHs)sSn]2, Pd(PPh3)4, LiCI 67%

O 77

~'0 75

~CI 34 atm CO Pd2(dba)3.CHCI3 PPh3, Phil, 60~

~

TBSO,,, ~~~~Jl~~_

0~~

S O

/OTBS 11[Ph3PCuH]6,H20, Phil .~ (+)-Zoapatanol1 2 HF, H20, CH3CN,rt 85%

SCHEME 21. Completion of the Trost synthesis.

intermediate, in order to produce the 0~,[3-unsaturated ketone 77 by a carbonylative alkylation under 34 atm of CO in the presence of Pdz(dba) 3. CHC13 and prenyl chloride. The obtained 0~,[3-unsaturated ketone 77 was then reduced chemoselectively by conjugate addition of hydride by using the Strycker reagent [Ph3PCuH]632 and, after cleavage of the protecting groups using HF, (+)-zoapatanol was obtained. The Trost sequence led to the first enantioselective total synthesis of (+)-zoapatanol in 20 steps from methallyl alcohol in 1.6% overall yield.

3

83


R~

cat PdCI2, H20 CuCI 2, 02

~176

"f~

PdCl2 ~ (

Pd(0)

Hc, .'-m /

CI"

H-Pd-CI

Regeneration of Pd(ll) Pd(0) + 2CuCI 2 ~

~

OH I[,- _PdCI

PdCI2 + 2CuCI

Regeneration of Cu(ll) ..~.-R

H..O..H

fl-hydride ~ elimination f ~ OH

a'~ olefin "~complexation

c,,p

reductive ~ elimination \

O

O

_

-- R,,~,"

2CuCI + 1/202 + 2HCI-~,'- 2CuCI 2 + H20

nucleophilic attack

S C H E M E 22. Wacker oxidation.

VIII. Our Approaches for the Total Synthesis of (+)-Zoapatanol

Two synthetic strategies have been examined to construct the oxepane ring of (+)-zoapatanol, one utilizing a ring-closing metathesis (RCM) and the other one an intramolecular Horner-Wadsworth-Emmons (HWE) reaction. A. R I N G - C L O S I N G

METATHESIS APPROACH

The first strategy envisioned to synthesize (+)-zoapatanol relies on a RCM reaction to produce the seven-membered ring. When a diene is treated with a transition metal alkylidene complex, a metallacyclobutane intermediate is formed and a succession of [2+2]-cycloadditions and cycloreversions takes place (Scheme 23). Each step is reversible and the driving force is the elimination of a molecule of ethylene and the formation of a cyclic adduct. The retrosynthetic analysis revealed that an oxidation of oxepene XVII should lead to ketone XVI (Scheme 24), which is a precursor of zoapatanol according to Kane's synthesis. 70xepene XVII could be obtained by using a RCM applied to the unsaturated enol ether XVIII in which the two stereogenic centers could be controlled through application of a Sharpless asymmetric dihydroxylation 33 of the trisubstituted (Z)-olefin XX. The result of the enantioselective cis-dihydroxylation of olefins using an osmium catalyst [K2OsOz(OH)4 ] in the presence of K2CO 3, K3Fe(CN)6

84


[M]--CH2 "~[M]

~[[M]

SCHEME 23

.f/--OH

HO,,,..//---',~ 0

9

{ R'01,.~O

1

R

'

O

XVI

,

RO~ O

,

,

.

~

~

"

XVIII

. Na + > K + > Cs +) is reversed in this case for K + and Cs +. Based on the single-crystal structures (Figure 7), the distance between the naphthalene sidewalls in the dimethylene-bridged clip 5b has to be compressed from 11.4 ,~ in empty 5b to 8.3 A in the complex of the Kosower salt KS@5b. 3~ The increase in steric strain resulting from this compression is certainly one reason why the complexes of 5 are usually less stable than those of the tetra- and trimethylene-bridged host molecules 2 and 4. A larger and even more impressive compression of this distance from 14.5 to 6.5 A is observed for the formation of the TCNB complex of anthracene clip 7b. 32According to force-field calculations, the expansion and compression of the sidewalls by angle distortion and outof-plane deformation of the aromatic sidewalls in 5a and 7a are lowenergy processes. For example, the compression from 10 (the global minimum) to 8 A in the parent naphthalene clip 5a and from 12.4 to 6.5

126

FRANK-GERRIT KL~i.RNER AND MIREIA CAMPAIq/~ KUCHENBRANDT

Kal + TCN B _.. -

Ka2 [TCNB@DNN]

21 ~

+ DN N _ -

[ T C N B @ 2 DNN]

CHCI3:

Kal = 360 M -1, Ka2 = 24 M -1 Kg = [ T C N B @ 2 D N N ] / [ T C N B ] = 8600 M -2

[DNN] 2

A(~max = 6.0 ppm T C N B @ 2 DNN

FIGURE 8. Formation and single-crystal structure of the (2:1) complex TCNB@2 DNN. 3

in the parent anthracene clip 7a, is calculated to require an energy of approximately 1.5 and 3.5 kcal/mol, respectively, which is, apparently, more than compensated by the noncovalent attractive host-guest interactions in the complexes of g and 7. DNN 9 ("the monomethylene-bridged clip") forms a 2:1 complex with TCNB in the crystal and in solution as well (Figure 8). In the crystalline state, the complex shows an optimal arrangement of the TCNB molecule between two DNN molecules without any distortion of the receptor geometry. There are attractive CH-rc and slipped face-to-face rt-rt interactions between TCNB and the naphthalene rings of DNN. Evidently, the gain in enthalpy resulting from the arene-arene interactions overcomes the unfavorable entropy term for the formation of a termolecular associate. Thus, the 2:1 complex is also stable in solution. 3~ Besides the single-crystal structures, the complexation-induced chemical 1H NMR shifts, A6max, of the guest protons provide important information on the complex structures, as has been recently shown for the complexes of p-DCNB, TCNB, and TCNQ as guest molecules with the parent naphthalene tweezer 2a as host. In the solid state NMR spectra of these complexes, the signals of the guest as well as host protons could be assigned by the use of ~H DQ MAS and 1H-~3C correlation spectra (Figure 9, bottom). 52,53 In the spectrum of complex p-DCNB@2a, two signals at & = 5.6 and 2.0 are observed for the nonequivalent p-DCNB protons H a and H b pointing either out of the tweezer's cavity or toward the benzene tings of the sidewalls (Figure 9, top, left), whereas in each spectrum of the two other complexes,

4

127

SYNTHESIS OF M O L E C U L A R TWEEZERS AND CLIPS

,

,

,

d~

#

...... ,

e ...

............

......

:+=.~ ..........~

~

. , , . ...... ,..

ji

2

4

i!!::~i:= i

I0

,. 150

140

....

...:=

.... ~ .

~ ......

~ ~'.~,o~

ppnl ............................................... 0

'

130

120 ppm

ppm

ppm i

-2 q

-2 ~I

01 2J

0 2

4~

4

6q

6

8q 1o -t

8 10 1150 140

130

1120 ppm

..... - - ~ ' ~ r * .... ~v-~'~---~'r9. . . . . t50 140 130 120 ppm

F I G U R E 9. Top and middle: single-crystal structure analyses of the host-guest complexes p - D C N B @2a (left), TCNB@2a (middle), T C N Q @ 2 a (right). Bottom: 1H rotor-synchronized 1H-13C R E P T - H S Q C

solid-state NMR spectra. (See color insert.)

only one signal is observed for the guest protons (in the case of TCNB @2a at 6 = 1.8 and of TCNQ@2a at 6 = 3.4). In addition, some of the signals assigned to the host protons, especially those that are close to the arene units of a neighbor tweezer molecule in the crystal lattice, show unusually large upfield shifts in the solid-state 1H NMR spectra. The 1H NMR chemical shifts of the guest and host protons were computed by quantum chemical ab initio methods for the monomeric complex as well as for larger segments of the crystal lattice (containing up to five complexes). These calculations were in good agreement with the solid-state ~H NMR data. The ~H NMR chemical shifts of the guest protons in the solution-state spectra of the complexes TCNB @2a and TCNQ@2a were very similar to those observed in the solid-state spectra (c5 = 2.0 and 3.9, respectively). Evidently, the complex structures in solution closely resembled those in the crystal. In the solution-state, however, the 1H NMR spectrum of

128

FRANK-GERRIT KLARNER AND MIREIA CAMPAlq/~ KUCHENBRANDT

p-DCNB @2a showed only one signal at 6 - 3.5 for the guest protons H a and H b e v e n at low temperature ( - 7 0 ~ This indicates that in solution the exchange of the nonequivalent protons H a and H b, resulting from mutual complex dissociation-association and/or rotation of the guest molecule p-DCNB inside the tweezer cavity, is fast with respect to the NMR timescale over a broad range of temperature (from + 21 to - 7 0 ~ This assumption is supported by a solid-state NMR study at different temperatures. In the solid-state 1H NMR spectrum of p-DCNB@2a, a broadening and finally a coalescence of the separated signals of H a and H b w e r e observed upon heating to 137 ~ From this observation, an activation barrier of AG ~ = 17.2 kcal/mol could be estimated for the exchange of H a and H b inside the tweezer cavity in the crystalline state. Two dynamic processes can be envisaged, which are consistent with the exchange of H a and H b, namely either a 180 ~ "rotation" around the long axis of the guest molecule or a 60 ~ "flip" between the two equivalent sites in the complex (Figure 10). According to quantum chemical and force-field calculations for the isolated complex in the gas-phase, the 60 ~ "flip" has a very low activation barrier (

,::

,$

,,

> .........

.......

FIGURE 17. The shuttling process of tweezer 2a along the axis of the N,N-dimethyl-bipyridinium dication calculated by force field MMFF94 (AEc,,I # c -- 8 kcal/mol). 49

of a cation radical (B 2+ --->B +) and a neutral (B + ~ B) species. The cyclic voltammetric (CV) pattern for reduction of the dendritic cores is affected by the addition of tweezer 2b. In particular, both the cathodic and anodic peaks corresponding to the first one-electron reduction process of the bipyridinium core progressively move to more negative values upon addition of tweezer 2b (e.g., half-wave potential for the reduction of D1B 2+ and D1BZ+@2b: B 2+ ---->B+: - 0 . 2 9 or - 0 . 3 6 V), whereas the peaks corresponding to the second reduction process were practically unaffected (B + ~ B: - 0 . 7 7 or - 0 . 7 7 V). This finding indicates that the bipyridinium core is stabilized by complexation and that the complex of the radical cation resulting from the first reduction dissociates before the radical cation is further reduced. The behavior of the tweezer-bipyridinium complexes could also be determined in the gas phase by using mass spectroscopy (Figure 18). 55 The monoisotopic complex ions were isolated in a FT-ICR cell and subjected to collision with argon as collision gas. In the tweezer complex of (D0)2 B2+, the benzylic C - N bond is exclusively cleaved under these conditions leading to the tweezer-monocation complex, which subsequently dissociates the tweezer molecule, whereas the complex of the tweezer with the dendrimer of second generation (G2) loses the tweezer in the first step and then in a second step, the benzylic C - N bond in the bipyridinium core is cleaved. In the complex with dendrimer of first generation (G1), a competition between these two reactions - C - N bond cleavage and tweezer loss - is observed in the first step. Evidently, dendritic substituents of increasing generation at the bipyridinium core provide increasing stability to the naked dication in the gas phase so that the host-guest complex dissociation is favored over C - N bond cleavage with increasing dendrimer generation. This finding can be explained by force-field calculations (Figure 19). The G2

136

FRANK-GERRIT KL,g,RNER AND MIREIA CAMPAIqA KUCHENBRANDT

major pathway for

~ ~

(,N-2 R

///-- RCH?

- Tweezer"~

/~~-Tweezer

-RCH?/4R.,,.N..~ (~)

both pathways observed f o r- ~- ~ Ar LO Ar^O''~''N'~

G r~N~OvAr %

(D0)2B2+

| r~-N---R ~ R.,,.N/~~'~ Q

major pathway for Ar~ tAr O,1~O

=

~O

OAr N~-O'X3"O"Ar

r..O

Ar

Ar"~

"Ar

FIGURE 18. Pattern of fragmentation determined for tweezer-bipyridinium complexes by the use of electrospray ionization Fourier transform ion cyclotron resonance (ESI-FT-ICR) mass spectroscopy in the gas-phase: competition between tweezer loss and benzylic C-N bond cleavage in dependence of the size of the dendritic substituents. 55

FIGURE 19. Monte-Carlo simulation of bipyridinium structures (MMFF94). Left: (D0)2B2+, middle: (D1)(Me)B 2+, and right: (D2)(Me)B 2+. (See color insert.)

dendron is calculated to completely engulf the bipyridinium dication by back folding of the dendritic 'arms' leading to an 'intramolecular solvation' and hence to a stabilization of the bipyridinium core. The bipyridinium core is, however, calculated to be less efficiently contained by

4

137

SYNTHESIS OF MOLECULAR TWEEZERS AND CLIPS

folding of the G1 and GO dendron, respectively, resulting in the instability observed for naked dications in the gas phase. The decrease in the binding constants observed 49 for the complex formation between tweezer 2b and the dendrimers of increasing generation in solution (Table 3) may be the result of the stabilization of the bipyridinium core by back folding of the dendritic groups, which is evidently also efficient in CH2C12solution. The molecular tweezers and clips are able to influence chemical reactions as well. One example, which has been already discussed, is the reduction of bipyridinium salts in the presence of tweezer 2b. Another example is the methylation of pyridine derivatives such as 4-cyanopyridine. This reaction is highly accelerated in the presence of molecular tweezer 2b, but in chloroform equimolar amounts of the tweezer are required to run the reaction. 56 Thus, the tweezer does not function as a catalyst because the p r o d u c t - N-methyl-4-cyanopyridinium iodide - forms a highly stable host-guest complex with 2b that causes product inhibition. The product inhibition can be avoided by using a stirred two-phase solvent system consisting of water and chloroform. The starting material and the host-guest complex of the pyridinium salt with tweezer 2b resulting as primary product of the reaction are soluble in chloroform, whereas the decomplexed s a l t - N-methyl-4-cyanopyridiniumiodide - is only soluble in water. After the reaction, the free salt is transferred to the water phase, the empty tweezer remains in the chloroform phase and is ready to bind the starting material and accelerate its reaction with methyl iodide once again. Under the two-phase conditions, the tweezer catalyzes the reaction substantially (Figure 20). A complete kinetic and thermodynamic analysis shows that this reaction is accelerated by the tweezer by a factor of kcat/kunc~ t = 3010 and CN

CN 25 ~

100 h, H20/CHCI 3

+ CH3--1 @ I

CH3 in the presence of 2a (5 mol%) in the absence of 2a

I

yield: 95% 9:1

CHCH 2

12

-

0

5" 1

CHCH 2

12

57

-40

>95:5

CH2CCH

13

78

0

>95:5

PhMgC1

Ph

14

78

-60

1"1

Ph2CuLi

Ph

14

84

0

>95:5

15

78

16

85

Nucleophile

Entry

MgBr OMe

{ M: MgBr

M~/J"'OMe

M: CuMgBr OEt

Nu

Temperature

[~:a

BrMgCu " ~ ~ O E t MgBr ~'MgBr

~

/MgC1

10

MgBr

CH2CHCH2CH2

-60 to rt >95:5

0

1"1

generation of a mixture of cyclic enol ether 19 and acyclic enol ether 18 in 80% yield. 15,~7 Although we initially separated 18 and 19, we eventually found it best to carry a mixture of the two substrates into the subsequent ring-closing reaction. Enol ether-olefin RCM employing the Schrock catalyst 2020 converted the remaining acyclic enol ether 18 into cyclic enol ether 19. Subsequently, Shawn and Jason Cox found that the second-generation Grubbs catalyst could also be used to convert 18 into 19. 2~

160

JON D. RAINIER

OBn

OBn CH2Br2, Zn TiCI4, PbCIz

BnO~ ....~O..) .... Ac20 NEt3 DMAP (95%)

OBn BnO- ~ -O ..Me

.

BnO~..~OR

_

Bn:n~..O~,~ O "

Me

BnO~,

(80%)

g:R=H

19

18:19 = 65:15

17: R=Ac

i-Pr~'~kpr

RO No_____/C(CH3)2Ph RO"

20

R = (CF3)2CH3C SCHEME 3

TABLE 2

Optimization studies on the cyclization of 17

OBn

OBn OBn BnO- -i- -O Me BnO- ~ -0. /Me conditions ~- BnO~,,,,~?8~,,, ~]~ + ~ B n O.,,~O~,,. ~, 9 , 19

Bno~BnO~'r~OAc ....95:5). In addition to the preparation of the anhydride and the choice of Grignard counterion, Utpal found that the choice of the C(25) substituent

5

APPLICATION OF C-GLYCOSIDE IN THE SYNTHESIS OF (-)-GAMBIEROL

185

TABLE 11 The addition of 2-methylpropenyl nucleophiles to the anhydride from 144

OTBDPS OTBDPS OTBDPS ~O"si(t-Bu)20~O_ cond,tlons H'H O ~ -~O'si( H .. 9 | H o"~~~O'si(t-Bu)2 t-Bu)2 Me/~"O-~14 O ~MgX "~~2"~ O H- ~./O additive 147 148 Entry

X

Conditions a

Additive

147:148

Yield (%)

1 2 3 4 5

C1 C1 C1 C1 Br

A A A B B

None ZnC12 CuI None None

95:5

a | H R'H-O"~'~O"si(t-Bu)2,

+ ~~MeO ~ ' ~ O a

Ketoside

Yield (%)

152 153 154

60 80 91

was key to the success of the ketoside forming reaction (Table 12). The use of C(25)-deoxy substrate 149 gave a 60% yield of 152 as a 2:1 ]3:a mixture and the C(25) TBDMS ether 150 gave 153 as a 4:1 ]3:~ mixture in 80% yield. The difference in selectivity between the TBDPS and TBS groups appears to be steric in nature as C(3) TIPS-protected glycal 151 also underwent a stereoselective coupling reaction to give 154 in 91% yield. When a mixture was formed, it was most interesting to us that the

186

JON D. RAINIER

mixture did not lie at the C(24) hydroxyl group but instead at the newly formed C(23) C-C bond. Thus, the C(25) substituent was not only influencing the oxidation reaction but, to our surprise, was also playing a role in the subsequent formation of the C(23) C-C bond. We currently believe that the reaction requires a group at C(25) that is of sufficient size (i.e. >OTBDMS) to serve as a protecting group for the adjacent anhydride enabling it to avoid decomposition via oxocarbenium chemistry prior to C-C bond formation. Having demonstrated that ~-methylglycals having the appropriate C(3) substituent can be used to stereoselectively synthesize C-ketosides, we became interested in determining whether glycals of higher synthetic utility might also undergo these transformations and settled upon ~-benzyloxyglycal 155. Scott Roberts examined this and was extremely pleased to find that the anhydride from the oxidation of 155 with DMDO (i.e. 156) was efficiently converted into [3-C-ketosides 157 and 158 when subjected to propenyl and propynylmagnesium chloride, respectively (Table 13). In contrast to related substrates, anhydride 156 was surprisingly robust, as exemplified by its recovery following aqueous workup and its stability to silica gel chromatography. Next, Scott examined the efficiency with which non-allylic and propargylic Grignard nucleophiles coupled with the anhydrides from 144 and TABLE 13 The coupling of A-benzyloxy anhydride 156 with carbon nucleophiles

TBDPSO [],~O H . Si(t_Bu)2 B n O v ~ o -~""/O H 155

TBDPSO ~"" H' s i ( .... ~ ~ / O O-O=_ u'~J"'c~-/~d t-Bu)2 BnO--/ "" 1:4

156 TBDPSO Nucleophile _-- HO'"~"O"si(tgu)2

R+o BnO /

Entry

Nucleophile

~ H

Ketoside

R

Yield (%)

13:~

~MgCI

157

CH2CHCH 2

91

>95:5

~ ""~ MgBr

158

CH2CCH

99

>95:5

5


187

TABLE 14

159-161 PO HO, H O. " _ ~ _ H O~si(tBu)2 O

The generation of C-ketosides

R Me

PO H r~O"si(tBu)2 ae'/J'LO-'~/OH

~,'"' O-O; ~ Nucleophile

144: P = TBDPS 151 P = TIPS

Entry

Glycal

P

1

144

TBDPS

2 3

144 151

TBDPS TIPS

~MgBr PhMgBr PhMgBr

Me'/~O~ O H 144

[3 PO HO,,,[..~O..si(tBu)2 Me...~---o.~O

Nucleophile

TBDPSO ~ H O..Si(tBu)2

H

C-Ketoside

R

Yield (%)

[3:0~

159

CHCH 2

45

1"13.5

160 161

Ph Ph

75 72

95:5) SCHEME 26

151 (Table 14). Interestingly, vinyl magnesium bromide and phenylmagnesium chloride both coupled with the anhydrides to give the 0~-anomers 159-161, respectively. Scott also employed the anhydride from 144 in a Mukaiyama-type addition reaction with ketene acetal 162 (Scheme 26). The addition of 162 to the anhydride from 144 in the presence of TBSOTf gave ~-C-ketoside 163 in 77% yield as a single diastereomer. To the best of our knowledge, the generation of 163 is novel in that it represents the first example of the addition of any ketene acetal to a glycosyl anhydride. The disparate stereochemical outcomes described above can be rationalized (Figures 6-8). For propargyl and allyl nucleophiles, coordination of

188

JON D. RAINIER

BoPso ., 6 ] L

MgXR j

FIGURE 6. Rationale for the addition of allyl and propargyl nucleophiles to 145.

I

imXMgO--" o so.--.r"--(~/~''%" \ x

[--IM OTBDPS eO~~(~)

,-Bu)2

FIGURE 7. Rationale for the addition of non-allyl and propargyl Grignards to 145.

Nu:

OTBDPS

Me TBDMSO

t-Bu)2 H 167

u

FIGURE 8. Rationale for the addition of ketene acetal 162 to 145.

the Mg counterion to the axial lone pair of the pyranyl oxygen and ligand transfer via six-membered transition structure 164 would lead to the observed 13-addition products. As mentioned above, the importance of C(3) substitution on these reactions is presumably tied to the ability of the substituent to sterically protect the epoxide and inhibit the competitive formation of oxocarbenium ion intermediates. Alternatively, this group could simply serve to lock the glycosyl anhydride into the conformation shown having the pyran ring in a bow tie with the TBDPS or TIPS ethers in pseudo-equatorial positions. Nucleophiles (phenyl, vinyl, and silyl ketene acetals) that are incapable of forming a six-membered transition structure presumably react through

5


189

oxocarbenium ion intermediates. Curiously, while ketene acetal 162 adds to the si-face of the oxocarbenium ion intermediate, Grignard nucleophiles add to the re-face. The Grignard additions can be rationalized either by invoking the conformation having all groups equatorial and a chair transition state as depicted for 165 or through a directed addition and a boat conformer as illustrated for 1 6 6 . 42 In an effort to explain si-face addition for 162, Scott calculated the low energy conformer for the oxocarbenium resulting from the interaction of anhydride 145 with TBSOTf and found it to exist in a boat as depicted by 167 (Figure 8). 43 Assuming 167 to also be the reactive conformer, approach of the nucleophile to the oxocarbenium from the face opposite the adjacent pseudo-axial TBDMS ether would give the observed product. Presumably, this conformation minimizes relatively severe gauche interactions between the C(2) TBDMS and C(3) TBDPS ethers. Interestingly, as evidenced by its C(3) and C(4) 1H NMR coupling constants (Ju(2),H(3) = 1.0 Hz, JH(3~,H(4) = 4.2 Hz), the pyranyl ring in 163 exists in a boat conformation. Having explored the scope of our C-ketosides efforts, we returned to gambierol. Utpal found both the esterification and acyclic enol ether forming chemistry of 147 to be sluggish. Esterification of 147 with 168 required a large excess of acid and prolonged reaction times to deliver 169 in 75% yield (Scheme 27). The conversion of 169 into the corresponding acyclic enol ether 170 was also relatively slow and gave a 35% yield of 170 (47% recovered starting material). Although clearly concerned about the yield of 170, Utpal was equally anxious about the generation of the F-ring so he decided to examine its formation prior to attempting to optimize the formation of 170. In light of

OTBDPS H H|H O~O"si(t-Bu)2 e

I~

O HO~j.....~......~/O p M B 168

= RO~"~'~

DCC, DMAP (75%)

OTBDPS 0 H"~O"si(t-Bu)2 e

TiCI 4, Zn, PbCI2, CH2Br2 ("- 169: X = O TMEDA, THF, 65oC k,~ 170: X = CH2 (35%) SCHEME 27

190

JON D. RAINIER TABLE 15 Tetrasubstituted enol ether 171 from RCM

OTBDPS

OTBDPS

PMBO"~'~]/0,, O"si(t-Bu)2 catalyst ,. P M B O I r~r 2 ~ . . . 0 conditions .~'~'Me "170

MsN.

MeO..~O..si(t_Bu) 2 H _ ~0 171

NMs

CI "]/ F3C~ I CI/ au_-~. I Ph Me/~" 0 ....,Mo~C(Me)2Ph PCY3 0d 90 Me__~CF3 20 OF3

Entry 1 2 3 4 5

Catalyst (mol%)

Conditions

Yield (%)

20 (20) 90 (20) 90 (20) 90 (45) "~ 90 (45) a

Hexanes, 65 ~ Phil, rt Phil, rt Phil, 65 ~ Phil, 80 ~

0 0 0 5 82

a90 was added in three portions (15% + 15% + 15%).

the fact that the F-ring required the generation of a tetrasubstituted enol ether it was not surprising that these reactions were also challenging. Subjecting 170 to either the Schrock Mo alkylidene catalyst 20 at 65 ~ or the second-generation Grubbs catalyst 90 at room temperature resulted in the complete recovery of starting material (Table 15, entries 1-3). The stability of the Grubbs catalyst at elevated temperatures turned out to be critical. 44 When 170 was subjected to 90 (45 mol%, added in three portions) at 65 ~ a small amount (ca. 5%) of tetrasubstituted enol ether 171 was isolated (entry 4, Table 9). When the temperature of the reaction of 170 was increased to 80 ~ Utpal isolated 171 in 82% yield (entry 5). In contrast to the enol ether RCM reactions of the substrates that have been described previously in this chapter, we believe that the reactions of the more sterically encumbered substrate 170 proceed through a less reactive Fischer carbene intermediate (i.e. 172); thus, the need for elevated temperatures (Scheme 28). 45 To the best of our knowledge, this reaction represented the first use of enol ether-olefin RCM to build a tetrasubstituted olefin.

5

191

APPI,ICATION OF C-GLYCOSIDE IN THE SYNTHESIS OF (-)-GAMBIEROL

OTBDPS H H TO'~O"si(t-Bu)2

90

~eeO~ e H"~0

[OTBDPS R' H H I' - ~ ~ 1 7 6

7

LI Ru, ~JM~:'e- -l l0H~ 0

]/

170

I

172

OTBDPS R'

O

M

2 /.~.O..si(t_Bu) | _

~...0 H

171

R'= P M B O ~ r r r SCHEME 28

Although clearly pleased with the generation of 171, the described sequence and especially the Takai-Utimoto reaction would need to be much more efficient for it to be of use in our gambierol work. In light of the sensitivity of the Takai-Utimoto protocol to the steric environment of the substrate and that the C(25) alcohol needed to be removed for the generation of gambierol, Utpal decided to examine whether the C(25) deoxy substrate might undergo the Takai procedure more efficiently than had 169 (Scheme 29). Unfortunately, the most direct route to the deoxy substrate (i.e. removal of the TBDPS group from ester 169 and deoxygenation) proved unworkable as it required forcing conditions that resulted in competitive removal of the silylene and/or the decomposition of 169. Removing the TBDPS group prior to ester formation circumvented this problem.

OTBDPS HO'-.~O'- Si(t-Bu)2

OTMS ~.H f .H 1. Nail, HMPA(92%) ~ H"'~/~-.[..'O-si(t_Bu)2 2. TMSOTf,i-Pr2NEt(90%) J'-- i :.~....O ~M2"H Me 172 OH H

-~'O"si(t-Bu)2

1. 168, DCC,DMAP(90%) 2. HOAc,H20 (98%)

o/~..o.:80:1 diastereomeric ratio

SCHEME 33

Having completed the F-ring, Utpal moved to the seven-membered H-ring. Required was not only the incorporation of the 3 ~ alcohol and the olefin but also a handle to attach the side chain. He initially opted to employ an acid-mediated cyclization and elimination approach to this problem with the notion that the side chain would be incorporated in a subsequent C-glycoside-forming sequence. The olefin and 3 ~ alcohol would come from the corresponding ketone and a Saegusa oxidation followed by an axial attack of a methyl nucleophile on the ketone. To examine this, the cyclic silylene from 182 was removed using HF.pyridine and the resulting triol was transformed into the corresponding 1o triflate and 2 ~ TBS ether (183, Scheme 34). To incorporate the acetal required for the hydroxy acetal cyclization reaction, Utpal examined the coupling of the triflate with acetal magnesium cuprate 187. Unfortunately, this reaction was completely unsuccessful, resulting in either recovered starting material or decomposition of the triflate. This result forced him to examine a more circuitous route to the same compound. In contrast to 187, allyl

5


H H H R2~O1 ,~.~O..si(t_Bu)2 1. HF.pyridine (100%) HO,.MU _ _ , ~ o ~ O Me H

195

=_ R, ,H, ~ O H . ~ ~ OHT B S

2. Tf20, 2,6-1utidine(62%) 3. TBSOTf, 2,6-1utidine(92%)

HO,,.~../~. O..~-.,..jOTf Me Me H

182

183

R= PMBO~#( ~ C u M2g C l

H H H T BS R,,,~O,...~~O '

~_

(88%)

H O ~ e O ~ 184

H H H R~O..~~OTBS

1. MeOH, HCI .

o q. "H

185

60 ~ pyridine,

"]

CHO

135 ~

1. TMSCI,i-Pr2NEt (100%)

2.BH3~ H202,KOH (62%) 3. (COCI)2, DMSO, NEt3 (85%) H H H,, R]~O~--~u~k /

T SO- e eO186

(37%,2 steps)

OMe BrMgCu'~..~eMe)2 187 SCHEME 34

cuprate underwent an efficient coupling with triflate 183 to give adduct 184. 47 Hydroboration and oxidation of the olefin gave aldehyde 185 after Swern oxidation. Treatment of the aldehyde with methanol and HC1 resulted in the formation of the methyl acetal along with the removal of both the TMS and TBS ethers. Treatment of this compound with PPTS, pyridine, and heat gave the H-ring enol ether in 37% yield for the last two transformations. Although we were pleased to have generated the H-ring using the acetal cyclization chemistry shown in Scheme 34, we were not satisfied with the low yield in the 185 to 186 conversion; nor was it clear that 186 was ideal for the incorporation of the H-ring side chain. After much deliberation, we came to the conclusion that the incorporation of the H-ring side chain could best be accomplished using an enol ether olefin RCM reaction as this would enable us to incorporate a precursor to the side chain into the H-ring precursor (Scheme 35). The H-ring alkene would come from an elimination reaction of the appropriately substituted C(30) ketone. Subsequent to coupling, the addition of MeMgBr to the ketone

196

JON D. RAINIER

X

..."5

,o,

POlO

_...,~,,,h/OR, ". . . . . . . . . . . . . . . . . .

IVle IVle- H L 187

PO"MUMeO'H'~t" 188 OR'

14

II X

FI H

H

X

H.-. H )

H

H

H.-.H/~

ba, e

o

PO

189

OR'

MeMgBr ..............

X H H H~H) R,, ~ , , , . O ~ U ~ ' . OH 191

SCHEME 35

would give the H-ring. Coupling with the A-C ring would then be followed by the attachment of the side chain. Because he was concerned with our ability to incorporate the requisite H-ring functionality, Utpal decided to examine the route in a model substrate, namely vinyl C-glycoside, 192 (Scheme 36). 17 He oxidatively cleaved the alkene using ozone to give aldehyde 193 after reductive workup with DMS. The coupling of 193 with allyl magnesium bromide gave 2 ~ alcohol 194 as a 3.8:1 mixture of diastereomers in a 75% unoptimized yield. Since he felt that both isomers would be amenable to elimination and enone formation, he carried both diastereomers forward into the RCM chemistry after conversion of the alcohol into the corresponding methyl ether. Removal of the TBS ether and esterification gave the metathesis precursor 195. By sequentially subjecting 195 to the Takai protocol and then the second-generation Grubbs catalyst 90, he was able to form the oxepene 196 in an unoptimized 50% yield for the two steps. With 196 in hand, Utpal examined its conversion into the corresponding enone. Although not optimized, he found that he could convert the intermediate anhydride into ketone 197 in 40% yield when 196 was treated with DMDO followed by MgC12 in THE Unfortunately, he was unable to find conditions to eliminate the methoxy group in 197 to give 198 without decomposing the remainder of the substrate.

5

197

APPLICATIONOF C-GLYCOSIDE IN THE SYNTHESIS OF (-)-GAMBIEROL BnO_ BRO~OTBS

BnO

03~DMS

BRO~.~OTBS

BnO~ ....L'O'~'"Jl 192 '~

BnO~ ....L...O.,P,.,CHO 193 BnO BnO%..,,L..,y~OTBS

1. ~ M g B r 2. Nail, Mel (75%, 3 steps)

1. TBAF

BnO~ .... I 1 ~ . . JO ....~

,,

0

194 OMe

2. HO OTBS,...~/ (80%, 2 steps)

/ -OTBS BnO '~:O BnO@..~O

BnO H

1. TiCI4, Zn, CH2Br2, PbCI2, TMEDA

DCC

TBSO ~

.

BnO~ ....t.,,O.J ....~ 195

2.90 (20 mol%)

OMe

(50%, 2 steps)

BnO~ ....~176HJ~oMe

TBSO BnO H ) BnO - : 0

DMDO; MgCI2 =BnO~, (40% yield)

TBSO BnO H ) BnO : : 0 0

197

//

OMe

"

0

BnO~, 198

H

SCHEME 36

While we probably could have overcome our inability to generate the enone in model substrate 198, we decided to use Yamamoto's Saegusa oxidation approach to the H-ring enone. 2sb In addition to its precedent, this approach would allow us to avoid carrying a mixture of diastereomers through much of the H-ring sequence as we had done in our model work to 197. To carry out this plan Utpal returned to the real gambierol F-H system and 182 (Scheme 37). Removal of the TBS group, esterification, and C(21) TMS ether formation gave olefinic-ester 201. As had our model substrate 195, 201 underwent a successful RCM reaction when exposed to the Takai-Utimoto conditions and Schrock's molybdenum catalyst 20 (Scheme 38). In contrast to our previous oxepene syntheses, ~7 the second-generation Grubbs catalyst 90 was less successful than the Schrock catalyst 20, giving a 39% yield of 202 over the two steps. With the H-ring skeleton in place, we were now prepared to examine the formation of the C(30) and C(31) stereocenters. To our delight the use of the DMDO oxidation, DIBAL-H reduction sequence worked nicely

198

JON D. RAINIER H H H R'"~2g~O"si(t-Bu)2

2/LL ..L ..L .o

1. HF,pyridine (100%) 2. Tf20, 2,6-1utidine -40 ~ TBSOTf, 2,6-1utidine 0 ~ (79%)

HO'~e~eO'H~ 182

R : PMBO~/ H

H

H ~

H

H H

.

,,

-40~

Me Me 199

H

(88%)

200

1. TBAF (85%)

3. TMSCI, i-Pr2NEt (100%)

O'~/'~OTBS H ~ O

R H H ,,, O

2. TBSOCH2CO2H, DCC, DMAP (95%)

TMSO 201

SCHEME 37

O~OTB

9,F -t- -1~ 1

1

1

~

TBSO

s 1. TiCl4, Zn, PbCI2, CH2Br2 T EO ,THF, 6 ~ (71% (10% recovered s.m.))

H

H H DDH~ . . y / OO~ '~

/ .

TMSO~eOH~~

2.20, hexanes, 65~ (88o/o)

201

TMSO -

or 90, Phil, rt (50-55%)

H 202

R = PMBO~/ _ H _ H

i-Bu2AI" (92%)

H ,.., H / ~ O T B S

T M S O " M U MeO" H ~ 203

H

SCHEME 38

and resulted in the generation of 203 in 92% yield as a single diastereomer (Scheme 38). Interestingly, substitution at the ~-position on the oxepene is critical to achieving high diastereoselectivity. When unsubstituted oxepene 204 was subjected to DMDO, we isolated a mixture of diastereomers in spite of the presence of an angular methyl group at the junction between the six- and seven-membered rings (Scheme 39). Scott Roberts and Anita Orendt calculated transition states for the oxidation of both substituted and unsubstituted oxepenes and found that

5

199


H

Me Me

H

O.A..:_zO \ - - ~ : _ z O--~

t-Bu , ~ "Si

,

T

I

Me Me

o.,--,~_o ~ o - - - ~ o

O-O

t-Bu, , ~ 'Si ,

I

I

V'

205 2:1 mixture of diastereomers by ~H NMN

204

SCHEME 39

substitution helps to make the bond formation more asynchronous. 33 The result of this is that interactions between the axial hydrogen on the allylic carbon and DMDO are accentuated in the disfavored transition state. In the oxidation of the unsubstituted substrate, the reaction is not as asynchronous and the interaction between DMDO and the axial hydrogen is not as important, leading to lower levels of diastereoselectivity. From 203, Utpal found Yamamoto's TPAP and Saegusa oxidation conditions to result in the incorporation of the requisite enone and to give 206 (Scheme 40). Addition of methylmagnesium bromide gave 3 ~ ether 207 following silyl ether formation. The stereoselectivity in this transformation is interesting; we believe that axial attack of methylmagnesium bromide is dictated by developing eclipsing interactions between the C-O bond and the adjacent C(30) silyloxymethyl substituent during the transition state that would lead to the undesired axial alcohol. 48 The completion

R H A H

TPAP, NMO (90%) H ,,..,H / ~ O T B S 1. 2. LiHMDS, NEt3,

TMSO"MU 1~12" H ~ 203 R = PMBO~rF

H H H ~ H/"~OTBS R~ / O ~ ' - ' - ~

TMSCI ,. ~ 3. Pd(OAc)2, CH3CN TMSO

H

2

(90%, 2 steps)

~

O

206

r R H A H

H ,.,, H/~OTBS

1. MeMgBr,-70 ~ (94%) 2. TBSOTf, CH2CI2, rt (96%)

~ ~ -/~'OTBS T M S O ' M U MeO" H ~ " Me 207

1. DDQ, CH2CI2, H20 (98%) 2. TPAP, NMO, CH2CI2 3. NaCIO2, 2-methyl-2-butene NaH2PO4, H20, t-BuOH (90%, 2 steps)

HO2C,.

H O....~~O.~

~ . . ~ ~ TM~ MeO" H~ 208

SCHEME 40

OTBS F~. OTBS Me

200

JON D. RAINIER

of the synthesis of the F-H coupling precursor 208 involved oxidative hydrolysis of the PMB group, TPAP oxidation of the resulting 1~ alcohol, and sodium chlorite oxidation to the corresponding carboxylic acid. C. SUBUNIT COUPLING AND COMPLETION

Having completed the A-C and F-H precursors, we were finally prepared to examine their coupling chemistry as was outlined in Scheme 6. Henry Johnson took on this task. His initial efforts were wildly successful. He began with the generation of 209 from the esterification of 208 with 142 (Scheme 41). The two-step enol ether-olefin RCM reaction gave dihydropyran 210. Oxidation of the cyclic enol ether using DMDO and directed reduction using DIBAL-H provided the corresponding 2 ~ alcohol as a 3:1 mixture of diastereomers. Of note is that the use of hydroboration/ oxidation on 210 resulted in the competitive reduction of the H-ring olefin. That the reaction gave a mixture of C(17) diastereomers favoring the undesired ~-isomer was not a problem; we took advantage of the thermodynamic stability of the desired C(17) [3-stereochemistry by oxidizing the C(16) alcohol and equilibrating the C(17) stereocenter to give 211. 49 Equilibration resulted in a 4:1 mixture of isomers that could be separated and recycled. With 211 in hand, it remained to form the E-ring. As was mentioned above, prior to carrying out this work, we were much more confident about our ability to generate the E-ring using acidic conditions than we had been about generating the D-ring using metathesis. As often happens in situations like this, our analysis was off the mark. All attempts to effect the cyclization of 211 were unsuccessful. Included were attempts to generate the corresponding mixed thioketal through the use of EtSH and various acids and the generation of the cyclic ether directly through the use of BiBr 3 and Et3SiH or TMSOTf and PhzMeSiH. 5~ Based upon the lack of olefinic protons in the ~H NMR spectra of recovered samples, we believe that the H-ring olefin was undergoing competitive decomposition under the reaction conditions. In an attempt to avoid the olefin decomposition problem, Henry examined the corresponding C(28)-C(29) saturated substrate (Scheme 42). Although somewhat less than ideal in that the use of this substrate would require that the olefin be introduced post-coupling, at the very least these experiments would enable us to determine the overall feasibility of the approach. Subunit coupling, metathesis, and oxidation/reduction were carried out as described previously for the generation of 211 to give 215. Unfortunately, all of the conditions that were attempted to convert (211) into the corresponding

5


201

PMBO

H O ~ HO ~H ~

O

I

I

u ~OTBS H On/

~(,

I

142

~IOTBS

DCC, DMAP, CH2CI2, rt (90%)

""

208 PMBO ~Me Me

H

H _ H

u ~OTBS H On /

1. TiCI 4, TMEDA, THF CH2CI2, Zn, PbCI2, CH2Br2

%OTIS 2.20 (20 mol %) (75%, 2 steps) 121 " i-21 PMBO M

t11 " 121 I TMSO" I~e

I~e" I11 ~ 209

e

M

e

H

H

B n O ~ o ~ O ~ T M S O

H

"'

H fOTBS

Hon./

%OTIS

-~"v~: O~"Me e H

(80%, 3:1 0~:13mix) 2. TPAP, NMO (80%)

210 PMBO

wMe

Me

H

H

B n O ~ o ~ o ~ . . . t-21

H

H

1. O-O (acetone free); DIBAI-H, -65 ~

,, f O T B S

H On /

%OTIS

conditions~/Z

HO- - * ~ " ' ~ : O ~ " M e e H

conditions DBU

f

211 C(17) [3-isomer

EtSH, acid

Phil, 80 ~ ~ 212: C(17) or-isomer (4:1 mixture)

BiBr3, MeCN, Et3SiH, rt Ph2MeSiH, TMSOTf, CH2CI2, 0 ~

PMBO

- O O o. OTOS 213: R=SEt

Me "0 ~--~2 8 H ~9~ ~

214: R = H

Me/~.

11

O ...., M o ~ C ( M e ) 2 P h

ol M e + C F 3 20 CF3

SCHEME 41

O,S-ketal failed. The main products were either the acyclic dithiane 217 or decomposition when attempts were made to push the reaction. Also unsuccessful were our attempts to convert hydroxy ketone 215 directly into the octacycle using BiBr 3, Et3SiH or TMSOTf, Et3SiH. Clearly, the use of the C(21) 3 ~ alcohol as a nucleophile to generate gambierol's E-ring was problematic in our hands.

202

JON D. RAINIER PMBO ~Me

Me

H

H

H

u ~OTBS H O~. /

H

00o. o

EtSH, H+

OTiS

215 HO ~Me

O

Me

H

H

O

O

u ~OTBS H O n./

H

BnO

TBS !:1

e PMBO ~Me

co

ns=_

Me

H

H 216 H

B n O \ ~ . / ~ - ~ _ ,, _..-~../~_ ,~_.~,,./1L._ _/

Et21;-

conditions BF3*Et20, PhCl

~

H

H

Me- " O ~ o T B

S

H

AgNO 3, NCS, 2,6-1utidine, 3A MS, SiO2 AgCIO4, NaHCO3, 3 A MS, SiO2, CH3NO2

SCHEME 42

From the efforts described above, it was clear that an alternate coupling protocol was needed. Because of our continued belief that it could become a highly efficient means of generating polycyclic ethers, we opted to continue to pursue an enol ether-olefin RCM strategy. However, instead of employing metathesis to generate the "easier" D-ring, we would use it to generate the seven-membered E-ring (Scheme 43). Subsequently, a ketal cyclization and reduction sequence similar to that employed by Sasaki in his gambierol synthesis would be employed to generate the D-ring. TM To examine the new approach, our syntheses of both the A-C and the F-H subunits required modification. The generation of the A-C substrate 222 was carried out from 140 according to the sequence of reactions illustrated in Scheme 44. To simplify the removal of the alcohol protecting groups subsequent to the coupling of the A-C and F-H precursors, we opted to incorporate silyl ethers and exchanged the C(1) benzyl ether for a TBDPS ether. Acid-catalyzed hydrolysis of the 1~ TIPS ether in the presence of the 2 ~ TIPS ether and 1o TBDPS ether gave 221 after bis-TES ether formation. Selective hydrolysis of the 1o TES ether and oxidation provided coupling precursor 222. Henry constructed the new gambierol F-H precursor 225 according to the sequence illustrated in Scheme 45. Oxidative hydrolysis of the PMB group

5


HO -

Me Me

H

203

H

/:-q,. o...I..,--..~ o.,F.'--- ~ a HO

0 H

r, '-'..

'-'.

r,

~;~-',_/,. __7-~ a~

_--~=/--

O.-7~ / H" -'--- : OH Me

PO Me Me OH -

eo

O

:-

H

oH

H-

H0 H ROe "

PO

Me Me

H

p

H

Me H

H

H R

99:1 er

Ph Ph~o

H

NH2

(-)-5

SCHEME 5. SharplessAD reaction

on trans-stilbene.

272

ROBERT M. WILLIAMS AND CAMERON M. BURNETT

O ph"'~H 14

HCN

OH

,.._

O ~

oxynitrilase~Ph~'.CN "-ph.,,~CN

OTHPH O ..~Nv.,LLoMe Ph Ph 18

15

NH 1. PhMgBr ... THPO,_~)._~,_,

2. MeOH "-16, (> 99 % er)

Pn

r'h

1.

H2NCH2COOMe

2. NaBH4

17

1. p-TsOH OH Cbz O Ph ,.__ p h . ~ ~ v . . U . . . O M e p-TsOH ~ Ph~oII21 MeOH "cyclohexane C b z N ~ H I O 2. CbzCI Ph 19 (-)-1

SCHEME 6. Alternative oxynitrilase-based synthesis of the oxazinone by Brussee et al.

An alternative and potentially preparative synthesis of the N-Cbzprotected Williams oxazinone has been reported by Brussee and co-workers (Scheme 6). 13 The key step involves the oxynitrilase-mediated asymmetric cyanohydrin-forming reaction on benzaldehyde. Phenyl Grignard addition to the protected nitrile (16) provided an imine (17) that underwent transamination with glycine methyl ester followed by a stereospecific sodium borohydride reduction, providing the protected syn-(erythro) amino alcohol (18) that was acylated and cyclized to the oxazinone ((-)-1). The procedure reported in this paper furnished a little more than 9 g of the final oxazinone; the overall yield from benzaldehyde was 48%. The oxynitrilase utilized in this study (E.C. 4.1.2.10) is present in almond meal. llI. Glycine Electrophile

The diphenyloxazinone template was originally conceived as an electrophilic glycine equivalent. The methodology developed for this application also allowed conversion of the lactones to glycine phosphonate equivalents. A. D I R E C T S U B S T I T U T I O N

Reaction of oxazinone ( - ) - 1 with N-bromosuccinimide in refluxing carbon tetrachloride gave the anti-bromide 20 in quantitative yield. The reaction of 20 with a variety of nucleophiles in the presence of mild Lewis acids gave ~-substituted oxazinones (21) in good yield and with generally excellent diastereoselectivities (Scheme 7, Table 1). 14'15 The key C-C bond-forming reaction is believed to proceed through the agency of a highly reactive N-acyl iminium ion species that is attacked from the leasthindered face of the oxazinone, anti to the two phenyl substituents.

8

DIPHENYLOXAZINONES

27 3

Ph Ph Ph P~,~O NBS,COl4D.P OCbzN h ~O ~L Nu ~ P h ~ o deprotection + OCbZN.v~0 quant. : conditionsCbZN@o_ ,. HaN-@OR (-)-1

Br

A

20

21

22

SCHEME 7. Bromoglycinate homologations.

TABLE 1 Electrophilic reaction of N-Cbz-oxazinone-bromide 20 Entry

Nucleophile

a

OTBS --~

b

OEt //-.../SiMe 3

c

//~./SiM%

d

MeZnC1

Conditions

% Yield (21)

Amino acid (22)

% Yield 22 (ee)

ZnC12/THF, 25 ~

74

Ethyl aspartate

85 a (>98:2)

ZnC1JTHF, 25 ~

66

Norvaline

93 a (>99:1)

ZnC1JTHF, 25 ~

66

Allylglycine

908 (>95.5:4.5)

THE 78 ~

46

Alanine

100a (>98:2)

THF/Et20, 78 ~

48

Norleucine

52 a (>99.5:0.5)

72

Homophenylalanine

91a (>98:2)

-

e

Bu2Cu(CN)Li

-

f

E~,

g

/OSiM% ZnC1JMeCN, 25 ~ ,SiMe,

ZnClJTHF, 25 ~

82

Cyclopentylglycine

91 a (>98:2)

.SiM%

ZnC1JTHE 25 ~

82

Cyclopentenylglycine

94 b (>98"2)

ZnC1JTHF, 25 ~

64

(2-Tetrahydrofuryl)glycine

89b (>98"2)

ZnC12/THF, 25 ~

66

Dihydrofuranomycin

89 a (n.d.)

(~ h (~ i

j

1~_/CH 3

apdCl2 (0.3 equiv.), H 2 (20 psi), ETCH. bLi/NH3.

While the C-C couplings generally produced the anti configuration, the TBS enol ether of ethyl acetate gave the syn adduct. Further experimentation revealed that electron-poor nucleophiles react with the putative iminium ion from the less hindered face, while electron-rich nucleophiles react via direct SN2 displacement of bromide. The more

274


basic organometallic reagents showed reduced yield due to a competing one-electron reduction of the bromolactone. The auxiliary was cleaved in most cases via catalytic hydrogenation over palladium chloride in good to excellent yields to give amino acids 22 (Table 1). Dissolving-metal conditions were also successfully employed to produce the amino acid derivatives and, in the case of unsaturated side chains, obviated saturation of olefinic residues (cf Table 1, entries b/c and g/h). Dissolving-metal reduction of the corresponding homologated N-Bocoxazinones 24 provided the first direct asymmetric synthesis of the corresponding N-Boc-protected amino acids 25 (Scheme 8, Table 2). This methodology was applied to a straightforward synthesis of [3carboxyaspartic acid (Asa), a natural, post-translationally modified amino acid first reported by Koch and co-workers. ~6This amino acid is unstable, being sensitive to decarboxylation and elimination of ammonia. The silyl enol ether of dibenzyl malonate was condensed with the bromide 20 to give the syn-coupling product 26 in modest yield and diastereoselectivity (Scheme 9). Catalytic hydrogenation of the major diastereomer under mildly acidic conditions gave a mixture of Asa as well as the decarboxylation product D-aspartate; purification with ion-exchange chromatography gave Asa in 30% yield and high optical purity. Importantly, no racemization was seen during the catalytic hydrogenation or workup.

ph_o y .~L . Ph

NBS,CCI4 P h ~

BocN.v~o

quant.

Ph O

B~

THF 13r 23

(-)-2

RM, ZnCI2 P h ~

Ph O

Li/NH 3

BooHN...,~OH ,...../-. 0 B~ _ ~ - " O EtOH / THF I~ 15, 24 25

SCHEME 8. Direct asymmetric synthesis of Boc-protected amino acids.

TABLE 2 Electrophilic reaction of Boc-oxazinone-bromide 23 Entry

Nucleophile

24 % Yield

Amino acid

25 % Yield (er)

~'-....~SiMe3 'SiMe3

63

Boc-allylglycine

70 (>98:2)

59

Boc-cyclopentenyl-glycine

70 (>99)

8

275

DIPHENYLOXAZINONES

Ph

OTMS

Ph~o

BnOJ~CO2Bn~

CbzN_~L-" O ZnOI2,THF [3r 53% (5.6:1 dr)

P/~

H2 (40 psi)

PhcbzN.7. ~O

O-

PdCI2(0.3 equiv.) ~ H3~I~cICO3

T 'O EtOH/THFo(0.04M) HO2C BnO2C"''CO2Bn 30

20

2H

6-carboxyaspartic acid (Asa)

26

SCHEME 9. Asymmetric synthesis of ~-carboxyaspartic acid (Asa).

Ph Phi..? CbzN~ O

Ph

2 ZnCI2 P h i " - O + Bu3SnC-C-R C014 = CbzN_~_..O

H2 (30 psi) + O PdCl2 (30-50 mol%) ~- H3N_.y~ O_ THF/EtOH (0.02- 0.06 M)

13r

R 273, R = Ph: 55% 27a, R = 06H13: 53%

20

""7 R

283, R = Ph: 57%, 97:3 er 28b, R = 06H13: 68%, 99:1 er

SCHEME 10. Alkynylation of bromo-oxazinone 20.

Ph

Ph Ph

~-

? + Bu3SnC-C-Me Br

Ph~o 2 ZnCI2 B o c N . ~ O ~ :

cc,. 61%

M~

_~

OH BooHN. ~ - . . O

iiI Me

23 29

M = Li: 79%, 82:18 er M = Na: 18%, >99:1 er

_

Me 30

SCHEME 11. Asymmetric synthesis of E-vinylglycine derivative 29.

Bromide 20 could also be reacted with organotin acetylides in the presence of zinc chloride to give the alkynyllactones 27 in moderate yield (Scheme 10); catalytic hydrogenation yielded the fully saturated amino acids 28.17 The analogous N-Boc-alkynyl oxazinone 29 was subjected to Birch reduction to yield the E-vinylglycine derivative 30 (Scheme 11). The Birch reduction with lithium proceeded in good yield but caused partial epimerization at the ~-stereogenic center (82:18 er). Switching to sodium eliminated the epimerization (>99:1 er) but also drastically reduced the yield. 18

276

ROBERTM. WILLIAMSAND CAMERONM. BURNETT

Ph Ph,,,1,/~O

Ph ArM o r ArH Ph,,,1,/~,O

B~

B~

ZnCI 2

Br

Ar

31

32

Ph

1. TMSI

Ph'"~OH

NalO4

+

O -

H20 / THF H3N,~ CIHoHNyCO2H =O 2.10% HCI pH 3 Ar THF, A Ar 33

34

SCHEME 12. Oxidative cleavage of the chiral auxiliary.

The arylated oxazinones 32 could be prepared via reaction of arylmetal or electron-rich aryl species with 31 (Scheme 12), but liberation of the amino acids required a new method for removal of the auxiliary, as neither catalytic hydrogenation nor Birch reduction were expected to selectively cleave the biphenyl auxiliary over the benzylic C-N bond of the arylglycine unit. The oxidative protocol reported by Weinges worked well: removal of the Boc group with TMS iodide allowed acidic ringopening of the lactone to produce acid 33, and periodate cleavage liberated two equivalents of benzaldehyde to give the free amino acids 34 in moderate yield. ~9 Unfortunately, some epimerization of the diastereomerically pure adducts was observed during biphenyl removal (Table 3). The furyl adducts 36a,b could be d!rectly hydrogenated to the free amino acids without reduction of the furan ring (Scheme 13), though

TABLE 3 Synthesis of arylglycines Entry

ArM/ArH

a

~

34

Conditions

33 % Yield a (er)

34 % Yield (er)

Et20/THF, -78 ~

56 (1:0)

52 (>91:9)

Et20/THF'-78 ~

55(1:0)

29(>97"3)

ZnClz/THF,25 ~

838(1:0)

62 (>95.5:4.5)

ZnC12/THF, 25 ~

50 (1:0)

26 (>95:5)

ZnClJMeCN, 25 ~

39 (1:0)

73 (>96.5"3.5)

CuLi

CuLi

b

(

~

c

MeO\~..OMe

d

I~/

e

o

OMe

ayield over three steps, byield after removal of Boc group.

8

277

DIPHENYLOXAZINONES Ph Ph Ph,,,,/~ O / CbzN

yA Br

~O O

R

ZnCI2 THF

"

Ph',,1,~-ON~ O H2 (1 atm.) 5% Pd/C (4 mol%) Cbz O

+ H3

O-~i ~o O

"

THF (0.07 M)

R

R 36a, R = H: 64% 36b, R = Me: 66%

35

37a, R = H: 57% 37b, R = Me: 82%

SCHEME 13. Removalof the chiral auxiliary without furan decomposition.

Ph

Ph O

1. D2 (40 psi) H,,Ph O Peel2 (0.30 equiv.) Ph~ O'D O ~H 5:1 THF/D20 (0.046 M) ,-" --

CbzN"~O13r

2. Dowex 54%

20

|

o|

H3N~ O H D

" L~

39, 84-90% atom-D 88.5:11.5-91:9 er

38

SCHEME 14. Asymmetricsynthesis of s-deuterium-labeled glycine. Ph 1.T2 (1 atm.) PdCI2 (096 equiv') | O,~ ~ Ph ~ O CbzN j,,O 6:1 THF/T20 (0.032 M) ~ H3N O -

13r 20

2. ion-exchange column 31%

H

T

40, 0.78 Ci/mmol 88:12 er

SCHEME 15. Asymmetricsynthesis of s-tritium-labeled glycine.

direct hydrogenation of the Cbz-adduct resulted in decomposition in other cases. Selective removal of the Cbz group from 36 followed by the ringopening/periodate protocol also gave the free amino acid without furan decomposition. The bromolactone 20 also served as a convenient template for the synthesis of chiral, isotopically labeled glycine. Reduction of putative iminium 38 under D 2 gas (Scheme 14) gave ~-deuterioglycine 39 with high isotopic incorporation of deuterium and good enantioselectivity. 2~Vederas showed that the procedure could be adapted to produce chiral, monotritium-labeled glycine 40 (Scheme 15). 21 This amino acid is difficult to produce in high isotopic purity by pyridoxal-dependent enzymatic exchange due to the large preponderance of the doubly-labeled byproduct.

278


B. D I P H E N Y L O X A Z I N O N E S

AS GLYCINE PHOSPHONATES

Williams and co-workers used the lactone as a glycine phosphonate equivalent to synthesize the cyclopropyl amino acids. 22 Bromide 23 was treated with trimethylphosphite in THF at reflux to afford crystalline phosphonate ester (41) in 86% yield with the anti stereochemistry (Scheme 16). Upon treatment with base and an aldehyde, 41 provided the E-~,13-dehydrolactone adducts (42) in generally high yields (Table 4). The unexpected olefin stereochemistry probably results from minimization of steric interaction between the aldehyde R-group and the N-Boc residue in the betaine transition states. Cyclopropanation with (diethylamino)phenyloxosulfonium methylide gave cyclopropyloxazinones (43) in excellent chemical yields and high diastereomeric excess with addition

Ph Ph O BooNO \/~

Ph Ph Phi.. Ph~..-[... 1. NBS, CCI4, reflux O DMSO BocN~ O conditions BocN~ -1"-.-O ,, | 2. (MeO)3P,THF -HO ~" O 86% (2 steps) (MeO)2P"~O R'"/~ R R''''LL"R Ph-S-CH 2

(-)-2

41

Et/(~. Et

42

Ph Ph~

O

Li, NH3

B~ ~-_-__~ RO R' 43

-~

BocHN@,,~R'

HO2(f

""R 44

S C H E M E 16.

(~)

HCI, MeOH;

O~..-Me EtOH

HaN~H

""R 020E) 45

Asymmetric synthesis of aminocyclopropane carboxylic acids.

TABLE 4

Preparation of ~,[3-dehydrolactones (42), cyclopropanations (43), and ACC derivatives (44/45) Entry

R

R'

Conditions

42 % Yield

a b

H 2H

H 2H

NaH/THF LDA/THF

Quantitative 97

c

Me

H

LDA/THF

93

82 (1:0)

63

100 (1:0)

d e f g h

Et Pr i-Pr Ph p-NO2Ph

H H H H H

LDA/THF LDA/THF LDA/THF Nail/Phil Nail/Phil

92 82 19 96 84

79 (1:0) 88 (1:0) 96 (1:0) -

64 61 -

100 (1:0) 99 (1:0) -

43 %

Yield (dr)

96 (11:1)

44 % Yield 65

45 %

Yield (dr)

100 (92:8)

8

DIPHENYLOXAZINONES

279

occurring unexpectedly on the same face of the oxazinone as the sterically demanding phenyl groups, presumably due to rt-stacking interactions with the ylide phenyl group. Birch reduction of the adducts gave N-Boc-cyclopropane amino acids (44). Acidic removal of the N-Boc group and treatment of the resulting hydrochlorides with excess propylene oxide in EtOH generated the free cyclopropane amino acids (45) in essentially quantitative yield. The phenylcyclopropyl oxazinone 46 decomposed under both attempted dissolving-metal and periodate conditions previously employed for removal of the biphenyl auxiliary. Interestingly, a lead tetraacetate oxidative cleavage protocol gave the free amine, which was not isolated but immediately protected with a Boc group to give 48 (Scheme 17). Ester hydrolysis and Boc removal yielded the desired (E)-cyclopropylphenyl-alanine 49. 23 Ph

1. TFA, CH2CI2 2. LiOH (aq) EtOHI reflux

1. Pb(OAc)4 MeOH,CH2CI2 Ph O OH 2.1 M HCI (aq.), THF BocN.~ph O ,, HN /___~~O2Me ,,3. CH2N2,Et20/MeOH 3. Boc20, Et3N, THF H 85% (3 steps) H h 41% (3 steps) ~.

46

Ph

Ph~

47

BocHNU,%,, ph ..-

1.2.LiOHHcl,(aq)'MeoHEtOHCIH~

-.

MeO2C

_

H

87% (2 steps)

%

HO2C

48

H 49

SCHEME 17. Oxidative cleavage of the chiral auxiliary for synthesis of a phenylcyclopropyl amino acid.

Williams and Fegley used a diastereoselective [1,3]-dipolar cycloaddition of an azomethine ylide on the N-Cbz-~,13-dehydrolactone 50 (obtained from bromide 20) in their synthesis of S-(-)-cucurbitine. 24 A single diastereomer of cycloadduct 51 (Scheme 18) was obtained. Subsequent hydrogenation and ion-exchange chromatography yielded the free amino acid S-(-)-cucurbitine.

Ph

Ph O

1 (MeO)3P,THF Ph "2. Nail, THF;" Ph CbzN~ O O

CbzN_ " ~ O (CH20)n 13r 67% (2 steps) 20

v

50

(~ [/Ph //N~Q

hP].. P h .,y. O O CbzN

CH2CI2 0 ~ -> r.t.94% Ph--/ 51

(~) N(~,,4 H3 O

H2(60psi 5% Pd/C)

HCI (3.0 equiv.) HN--' ;

90%

SCHEME 18. Asymmetric synthesis of S-(-)-cucurbitine.

S-(-)-cucurbitine

280

ROBERT M. WILLIAMSAND CAMERON M. BURNETT

IV. Glycine Enolate

While originally conceived as an electrophilic template, the oxazinone was quickly adapted for use as a glycine enolate equivalent. A variety of alkali metal enolate conditions have been exploited for alkylation reactions and the corresponding boron, silicon, titanium, and aluminum enolates have been deployed in diastereoselective aldol condensation reactions. A. MONOALKYLATION

Treatment of the oxazinones (2) with either sodium or lithium hexamethyldisilazane gave a stable enolate. 25 Reaction with halides yielded the anti-alkylation products 52 (Scheme 19, Table 5), and the alkylated oxazinones could be deprotected as before to yield the free amino acids. 26 The N-Boc-oxazinone (2) generally gave higher yields and cleaner reactions than the N-Cbz-oxazinone (1). 27 The synthesis of N-Boc-alanine (1) (Table 5, entry a) was greatly improved via the enolate alkylation methodology, obviating the bromolactone reduction seen with the electrophilic coupling. N-Boc-allylglycine 52 (entry b) 28 served as a template for the synthesis of protected 2,7-diaminosuberic acid (see Schemes 55-56 a l s o ) . 29

Ph Ph

O

MHMDS RX

BooN.,.~ O

THF

Ph ~-

Ph

Li/NH 3

O B~

~

(-)-2

OH ~_ B o c H N . ~ o O EtOH / THF I~

52

53

SCHEME 19. Asymmetric synthesis of monoalkylated Boc-amino acids. TABLE 5 Production of amino acids via enolate alkylation of Boc-oxazinone 2 Entry

RX

M

52 % Yield

53 % Yield (er)

MeI I/-...~ Me

Na Li Na

91 86 84

54 (98.5:1.5) 50-70 (99:1) 52 (>99:1)

Na

70

76 (99:1)

Br/~'-Me

Br/~

8

281

DIPHENYLOXAZINONES

Ph Ph,,,

0

~

THF Br~CO2Et

CbzN"-~O (+)-1

CbzN

61%

H2 (20-50 psi) ~"

0

0

EtOH 2Et 71%, 98:2 er

2Et

54

Ph

Ph

,

55

OSiMe2t-Bu OEt

CbzN'...~ O quant. (+)-1

O-

CbZN@o Br 35

Ph",~'%0

H2, Pd=

ZnCl2,THF " C b z N ~ - " O 56 ~CO2Et

O+ H3N_~--. O .

~CO2 Et 57

SCHEME 20. Complementary syntheses of 13-ethyl aspartic acid by enolate and electrophilic templates.

An illustration of the complementarity of this approach is illustrated with the synthesis of R- or S-ethyl aspartate 55 or 57, respectively. Thus, via enolate alkylation of (+)-1 with ethyl bromoacetate and catalytic hydrogenation (Scheme 20) one obtains R-j3-ethyl aspartate. The electrophilic methodology yields S-[3-ethyl aspartate via the SN2 displacement of the bromide from 35 (Table 1, entry a). Thus, the same antipode of the oxazinone allowed synthesis of either enantiomer simply by the use of either electrophilic or nucleophilic chemistry. Williams and Aoyagi used the 13C/~SN-labeled N-Cbz-oxazinone 58 for the synthesis of ~3C/15N-labeled alanine. 3~Alkylation with methyl iodide proceeded in excellent yield to give 59 (Scheme 21), which was catalytically hydrogenated to give the doubly-labeled amino acid 60. Dong extended the reported alkylation of (-)-1 with ~,o~-alkyl diiodides to the five- and six-methylene units to give adducts 61, whose iodide was displaced to azides 62 (Scheme 22). Catalytic hydrogenation provided the desired ~,~o-diaminoalkanoic acids 63 in good yield. 31

Ph

NaHMDS,Mel Ph THF ,, Ph~15N~.c~O

Ph~o 15N. _ .~... Cbz~ -t~,--'O 58

-78 ~ 91%

0 = 13C

H2 (45 psi) PdCl2 (25 mol%)

" H315N+O'x~ -O

Cbz" "~NH Me 2 /| II ~NH O 79 F3CCO2|

H H o2 N_ ...J.....J>..OBn l. NagH4, EtOH; "]// " ~ ~"~ Pd(OH)2,H2 ,J N_ /.N r ~ 2. conc. HCI, reflux (63% + C-8-diast. 2 steps) Y "~---OEt OBn : II L..._N 78

SO3.pyr,DMF, 3,&, (33%, C-8-diast. 33%)

r_

|

7 , ~ N . . . , ~ N H HN...,~NH Me . ~| "]]I" --NH

O

7-deoxFcvlindrospermopsin

SCHEME 25. Asymmetric synthesis of 7-deoxycylindrospermopsin.

alkylations gave a single diastereomer. 35 The axial benzyloxy group of iodide 80 would presumably block enolate attack from the back face of the iodide, while the all-equatorial orientation of the benzyloxy groups in 83 could allow some attack from the front face.

284


BnO~_102n 1. Li/NH 3 EtOH / THF

Ph BnO-3~~80. BnO /

Ph Ph

O

B~

Ph~

i ,.

~O (-)-2

LiHMDS THF/HMPA -78 ~ -> r.t. BnO 87%, >99:1 dr

A co [ ~ O 2 0

o

2. CH2N2 ,, A c O ~ , , , , ~ BocN\_~ O AcO OBn 3. Ac20, py. 77% (3 steps)

NHBoc OMe O

82

B

OBn

Ph

Ph Ph@o

!~

BocN...v~ O (-)-2

LiHMDS THF / HMPA -78 ~ -> r.t. 71%, 93:7 dr

OO Bn

~

i 84

SCHEME 26. Asymmetric synthesis of amino acid sugar derivatives.

Ph Ph

1) TiCI4, Et3N

ph ~

Ph

h"" O

H2 (95 psi) 20% Pd(OH)2/C

O " CbzNL CbzN\...~O~ 2) (RO)3CH, CH2CI2 __ O -78 -> 0 ~ RO/~OR (-)-1

85a, R = Me (94%) 85b, R = Et (85%)

99%

O-

RO/~OR 86a, R = Me (99%) 86b, R = Et (99%)

SCHEME 27. Asymmetric synthesis of protected serinal derivatives.

DeMong and Williams reported the first use of the titanium enolate of (-)-1, alkylating with trimethyl orthoformate to give the corresponding dimethyl acetal (Scheme 27). 36 Auxiliary removal completed the first asymmetric synthesis of (2R)-~-formylglycine dimethyl acetal 86a. The total synthesis of capreomycin IB (see Scheme 62) necessitated use of the diethyl acetal 86b, synthesized by following the procedure with (+)-1 and triethyl orthoformate. 37 Lee arylated (-)-2 with the phenylmanganese complex 87 (Scheme 28); oxidative removal of the metal with NBS yielded aryloxazinone 89 in good overall yield. 38 Alkylation with KHMDS and an alkyl halide gave the disubstituted oxazinone 90. Removal of the N-Boc group with TFA allowed selective catalytic hydrogenation of the auxiliary to give ~-alkyl~-phenylglycines 92 in good yields. 39

8

285

DIPHENYLOXAZINONES

Ph Ph

O

NaHMDS Ph HMPA/ THF BocN -78 ~ /~ PF6 ~

Ph O LO

BocN...~O

Ph O BocN.,~..O

NBS

-Mn( L)3

59% (2 steps) 90% d.e. Mn(CO)3

D,

KHMDS -78 ~ -> r.t.

Et20

(-)-2

RX, THF;

89

88

Ph Ph~o

Ph TEA ~ P h ~ o

B~ P..~'-R O CH2CI2

H2 (1 atm.) Pd(OAc)2(50 mol%). H3N+ Oh..~ R -O

HNh P .~-"-RO

90

EtOH

p

91

92

SCHEME 28. Asymmetric synthesis of ~-alkyl-a-phenyl amino acids.

Ph

Ph

Ph O

c0z... o

O H2 (60 psi) NaHMDS,THF CbZN~o_ 10% Pd/C(50 mol%) -78 ~ : =I ~ 3:1 MeOH/THF (0.01 M) 93

OMe ~ Me

58% (-)-1

0-

84% Me~ --NB~ Me

Me.~-NBoc Me

94

95

SCHEME 29. Asymmetric synthesis of protected 5-hydroxylysine.

Brussee alkylated ( - ) - 1 with the chiral iodide 93 to give acetonide 94, whose catalytic hydrogenation yielded the protected (2S,5S)-5-hydroxylysine 95 (Scheme 29). 40 As either enantiomer of the iodide 93 could be obtained, this methodology provides access to all four possible diastereomers of 5-hydroxylysine. Allevi and Anastasia pursued the same target starting with ( - ) - 2 , which was alkylated with homoallyl iodide (Scheme 30). Epoxidation with mCPBA and opening with sodium azide gave azido-oxazinone 97 as a 1:1 mixture of diastereomers (epimeric at the 2 ~ alcohol stereogenic center). Removal of the N-Boc group and hydrogenation yielded the free 5-hydroxylysine 98. 41 Singh alkylated ( - ) - 1 with cyclopentyl iodide to give adduct 99 as a single diastereomer after recrystallization (Scheme 31). Catalytic

286


Ph Ph~ - . . 0

Ph Ph~o B~ - ' ~ O

LiHMDS I ~

oc.. o _T

Ph Ph~...O ~__ BooN_...~O

c

19:1TFA/H20;H2 + _ O10%Pd/C(23mol%) H3N,..~ O__

8:3:1 MeOH/ H20 ,,..'~".../ HCI (0.0026M) H N3 88%

60%

U

(-)-2

96 S C H E M E 30.

HO" ~

NH2

v

97

98

Asymmetric synthesis of ~-hydroxy-S-lysine.

NaHMDS

Ph ~ ~ I Ph O ,,, CbzN..O/`~' '.. THF / HMPA -

P|h Ph

O

~00O o rt

+ H3N

O,~:O

1:2 MeOH / THF

,,4O,o

(-)-1 S C H E M E 31.

H2 (60 psi) PdCi2

@

99

100

Asymmetric synthesis of cyclopentylglycine via enolate alkylation.

Ph

Ph. ~

j[,.h P .O

NaHMDS,THF -78 ~

CbzN.v~O (-)-1

(Br ~ 101 MeO"Av~ L "O"L.~-O 40%

S C H E M E 32.

Ph~ ~ O CbzN O.....i--.,.> j.. :

O-/ "O" v 102

H2 (50 psi) ...~O+ PtCI2 (50 mol%) H3N -,,rO _ _THF 80%

"

O

~ OMe

"OMe 103

Asymmetric synthesis of a coumarin-containing amino acid.

hydrogenation gave S-cyclopentylglycine 100, though in lower yield than achieved with the electrophilic lactone methodology (Table 1, entry g).42 Leblanc alkylated ( - ) - 1 with 4-bromomethyl-7-methoxycoumarin (101) to give the ~,]3-unsaturated ester 102 as a single diastereomer (Scheme 32). 43 Catalytic hydrogenation in a mixed MeOH/THF system not only removed the auxiliary but also saturated the olefin. Fortunately, reaction in pure THF yielded the desired unsaturated product 103, presumably due to precipitation of the amino acid before the alkene could react. Combination of the phosphonate and enolate alkylation methods allowed a stereocontrolled synthesis of 2,3-methano-2,6-diaminopimelic

S

DIPHENYLOXAZINONES

287

Ph

Ph

,

2.03; DMS 79% (2 steps)

i"~

(-)-2

: (MeO)2P"~O

Et/o" Et

H~" O 105 BOCNv~,,, Ph (~ Ph O

O

104 OO

27%

41

Ph Ph~ " " O -

DMSO 83%

OH Li, NH3 63%

106

_ l z

: "Ph Ph

:

O HCI, THF

OH O ~ NHBoc 97% 107

-

O NH3 108 (2S,3S,6S)

SCHEME 33. Asymmetric synthesis of DAP stereoisomers.

acids. 44 Alkylation of ( - ) - 2 with 4-iodo-l-butene and ozonolysis of the alkene gave aldehyde 104 (Scheme 33). Horner-Wadsworth-Emmons condensation of 104 with phosphonate ester 41 produced bis-lactone 105 in poor yield, apparently due to retro-Michael decomposition of aldeyhde 104 under the basic conditions employed. The alkene of 105 was cyclopropanated as before, and Birch reduction and N-Boc removal yielded the cyclopropyldiaminopimelic (DAP) acid 108. Two other diastereomers were prepared by starting from (+)-2 in the synthesis of either aldehyde 104 or phosphonate 40. B. DIALKYLATION

Monoalkylated lactones can be deprotonated with potassium hexamethyldisilazane to effect a second alkylation with activated alkyl halides (Scheme 34, Table 6). The second alkylation proceeded anti to the phenyl rings as well, shifting the first substituent to the same side as the phenyl rings. Baldwin sequentially dialkylated the N-Boc-oxazinone (+)-2 with benzyl bromide and allyl bromide, using NaHMDS for both enolate formations, to give the disubstituted lactone 116 (Scheme 35). Birch reduction yielded amino acid 117, which was carried on to the bicyclic y-lactam dipeptide analog 118. 45

288


KHMDS Ph OH R2X ~. P h~~ , . O Li/NH3 BocHN~, THF BOCNy~. EtOH/THF R( "R20 alrl-,,R20

Ph Ph@o

BocN_ @ O 109

Ph

i

Ph~o CbzN _ @

110 111 KHMDS Ph R2X Ph~.~ O Li/NH3 +~OCbzN.~~ ~" H3N. . . ~ THF RI~,,R:O EtOH/THF R1r "R20

O

i

al 112

113

114

SCHEME 34. Asymmetric synthesis of a,~-disubstituted amino acids.

TABLE 6 Production of a,0~-disubstituted amino acids from monosubstituted lactones 109 and 112 Template

R~

R2X

Alkylation % yield

Amino acid % yield (er)

a b

Boc Boc

Me n-Pr

i,/'-,,,,~ i,,,,---,,,~

87 90

70 (>99.5:0.5) 60 (>99.5"0.5)

c

Boc

Me

Me

80

65 (>99.5:0.5)

d

Cbz

Me

84

93 (>99.5:0.5)

e

Cbz

Me

80

95 (>99.5:0.5)

Entry

Br/"..,,~ Me Br/"'.,.]~ Br~

q J

Ph Ph NaHMDS Ph,,,~O ph,,,l~ O BnBr,THE B ~ O B~ 93% Bn 115

(+)-2

Li/NH3

,. t-BuOH/ THF 65%

BocHN

O,,,~

..,_ O Bn L 117

Ph NaHMDS Ph,,,1,~O BE~ BocN,,~ O 96%THE Bn" L 116 Bn.~//~,,,S,

" "~.N-...~

~.~ BocHN

O 118

/~NHi-Bu O

SCHEME 35. Asymmetric synthesis of a 7-1actam dipeptide analog.

8

DIPHENYLOXAZINONES

289

Ph

Ph,,, B~

O v"~'O

, SEMCI, THF

BocN

Ph _

O

76%

L..../TM S

(+)-2

Boc

15-crown-5, THF 2. Nal,refluxaCetone

119

O

/O120~

70% (2 steps)

I

TMS"

Ph _

Ph,,,./i~. O

O

O

T

B~

o

J'~L"ONa= 15-crown-5 75%

Ph .

.

.

.

Ph-

.

.

.

,TF

Ph

O--~__ TMs"P h

o

2. H 2, Pd/C 83% (2 steps)

O

-

H3N

121

~oa

I

NH3

122

SCHEME 36. Asymmetric synthesis of 6-hydroxymethyl-2,6-diaminopimelic acid.

Baldwin also reported sequential alkylation of Boc-lactone (+)-2 with SEMC1 and 3-chloro-l-iodopropane to give the disubstituted lactone chloride (Scheme 36), which was converted to the iodide 120 via a Finkelstein reaction. The iodide was used to alkylate the sodium enolate of a second oxazinone molecule, giving the bis-oxazinone 121. Removal of the protecting groups and hydrogenation gave 6-hydroxymethyl-2,6-diaminopimelic acid (122), a constituent of a naturally occurring antibiotic isolated from

Micromonospora chalcea. Aoyagi and Williams explored the synthesis of (S)-2-methylasparagine from either antipode of 1 2 3 . 46 Alkylation of the methylated oxazinone 123 with tert-butyl bromoacetate gave the disubstituted lactone 124 in good yield (Scheme 37). Cleavage of the tert-butyl ester to the acid and amidation gave amide 125, whose hydrogenolysis yielded (S)-2-methylasparagine (126). Sequential alkylation of the antipode (+)-1 in reverse order gave the disubstituted lactone 128, with the same stereochemistry at the ~-carbon

P h / o ~Ph CbzN_.,.~_..O Me 123

NaHMDS,-78 ~ BrCH2CO2t-Bu

Ph~o

15-crown-5,THF 71%

CbzN.,.~--. 2. EEDQ Me" ""[ 0 CO2t-Bu NH4HCO3 95% (2 steps) 124

!

Ph

H2 (60 psi)

+ O _ .a.."O PdCI2 (30 molYo) o 1. TFA, CHzCI2 Ph"~1/ " CbzN_--~ = H3N" - . ~ o ~ . 'O 2:1 EtOH Me "] THF (0.30 M) CONH2 94% 125

SCHEME 37. Asymmetric synthesis of (S)-2-methylasparagine.

Me" "[ CONH2 126

290


Ph Ph,,,lr/~O CbzN

P_h NaHMDS

- v ' ~ O BrCH2CO2t-Bu THF, -78~

(+)-1

Ph

Ph,,,~,,~O NaHMDS,-78 ~

Ph,,,l~O

CbzN~

CbzN ~ . . O Me" ""l CO2t-Bu 128

O

15-crown-5 Mel, THF

"..CO2t_Bu 63%

81%

127

SCHEME 38. Asymmetric synthesis of (S)-2-methylasparagine.

but diastereomeric at the phenyl carbons (Scheme 38). Conversion to the amide proceeded in slightly lower yield, presumably due to the steric hindrance posed by the phenyl rings on the same side of the oxazinone, and hydrogenolysis again yielded (S)-2-methylasparagine. C. SYNTHESIS OF PHENYLALANINE ANALOGS

Williams and Im reported the synthesis of phenylalanine in the original oxazinone enolate paper (Table 5, entry d). Alkylation of the oxazinones with substituted benzyl halides has become one of the most prevalent uses of the lactone enolate chemistry, due to the typically high yields, high diastereoselectivities, and ease of cleavage of the oxazinone to the corresponding phenylalanine derivative. Schow reported alkylation of oxazinone (-)-1 with the benzylic bromide 129 to give, after deprotection with TBAF, alcohol 130 (Scheme 39). Catalytic hydrogenation with Pearlman's catalyst in THF gave cleavage of the N-Cbz and O-benzyl residues, but only partially cleaved the N-benzyl bond. Acidic catalytic hydrogenation was used to complete the synthesis of azatyrosine 131. 47 Ph H2 (50 psi) OTBDPS Ph...].. Pd(OH)2/C (9 mol%) +~O1. NaHMDS " ~ " ~ O 1:1 i-PrOH/THF (0.13 M); CIH~ O_ + CbzN .~r/'-~. =CbZN.v ~ ~ a THF (60%) O H2 (50 psi) O 2. TBAF, THF ~ Pd/C (4 mol%) CI Ph phVJ,,,O

(-)-1

B

(79%)

130

~J

I

HCl/H20 (0.06 M)

N/'..~'~'OH '

64%

131

H~

SCHEME 39. Asymmetric synthesis of an azatyrosine.

Solas reacted (-)-1 with the benzylic bromide 132 to give the alkylation product 133 in good yield (Scheme 40). 48 Catalytic hydrogenation over palladium chloride proceeded quantitatively to give amino acid 134 as its hydrochloride, which was further converted to the protected phosphotyrosine isostere 135.

8

DIPHENYLOXAZlNONES

291

CI OH OH Ph H2 (50 psi) .. + - [ FmocHN_ LiHMDS Ph~ . , , PdCI2 (50 m~176176 n31'~~ 0 _ --'T" "0 --T.-~ O . 2:1. EtOH/THF . . -:" " ] ~ CbzN...~O ~ OO / F--+I:'O78%P" (0.18 M) ]/EL ~/~l~(OEt)2 '/IL-.v~-h~P(OH)2 F (OEt)2 ~ 0 quant" F["F Fi "F

Ph B r"|h Ph~ ' ~ 0 i/~ CbzN' . v ~ + ~ (-)-1

132

133

(OEt)2

134

135

F " SCHEME 40. Asymmetric synthesis of a protected phosphotyrosine isostere.

Zhang also synthesized 135 indirectly, alkylating ( - ) - 1 with 4iodobenzyl bromide to give aryl iodide 136 (Scheme 41) and converting to the same difluoromethylphosphonate 133. Auxiliary removal at lower pressure with palladium on carbon gave zwitterionic amino acid 137. 49 Ph ph~[...O

-, | Ph H2 (1 atm.) + O~ n y ' ~ - O 100 Pd/C (25 mol%) H3N~L~ O -____._~._.,, -

Ph NaHMDS,ZHFph J [ -78 ~ -"~r ~ \ o

c z=

=:,

o

80%~/"~?~!

(-)-1

136i~ I~ //J... '1

" ~

Oq u a n t "

'/K"v~"'~'P(OEt)21"F

133 ~ J ' ~ / P(OEt)2 ~ , , F F

F 137

SCHEME 41. Asymmetric synthesis of a protected phosphotyrosine isostere.

Bender and Williams conducted a rapid synthesis of m-tyrosine via alkylation of ( - ) - 1 with 3-benzyloxybenzyl bromide (Scheme 42). 5o Catalytic hydrogenation of the alkylated oxazinone 138 removed the auxiliary and protecting groups to give m-tyrosine in excellent yield. LiHMDS Ph B r / " ' Y ' S " y O B n Ph Ph....JL. ~ PhiL. O '~ "O CbZN.v~,. , u (-)-1

THF

,.

I T CbzNy'~" O

87%

~o/OBn

H2 (50 psi) PdCl2 (30 mol%) 1"1 EtOH/THF (0.09 M)

+ H3N~o ~

O

"...~fOH

99%

138

meta-tyrosine

SCHEME 42. Asymmetric synthesis of m-tyrosine.

Paquette and co-workers modified the procedure slightly to produce the protected m-tyrosine 140 (Scheme 43), which was used for the synthesis of (-)-sanglifehrin A. 51

292


NaHMDS B r ~ OB~

Ph ~ Ph O CbzNv.,,~--.0

THF / HMPA 60%

Ph H2 (50 psi) OH Phi.,.. O PdCI2 (30 mol%) CbzHN....~ CbzN~ 1"1 EtOH/THF (0.08 M); _ O -_ 0 ,IOBoc Cbz-O-Su ~~OBoc 92% (2 steps)

(-)-1

139

L ~

140

SCHEME 43. Asymmetric synthesis of protected m-tyrosine.

Seto alkylated ( - ) - 1 with the benzyl bromide 141 to give ketoester 142 (Scheme 44). 52 Auxiliary removal also reduced the benzylic ketoester to yield the corresponding ~-hydroxyester 143. Ph

Br.....~

Ph

H2 (1 atm.)

+

_

2:1 MeOH/91%THF(0.018 M)

O-

s

CbzN'-/'~'O ~ (-)-1

-78 ~ -> r.t.

O~

O

45%

2 14

141 Ot-Bu

"'"[/~

O

O

~

Ot-Bu Ot-Bu

143

OH

O SCHEME 44. Asymmetric synthesis of an ~-hydroxyester amino acid.

Roller alkylated ( - ) - 1 with the benzyl bromide 144 to give adduct 145. Hydrogenation gave amino acid phosphonate 146 (Scheme 45). 53

Br ~ P/~ Ph" 7 " O +

NaHMDS

CbzN" P~ ~'O Ph

H2 (36 Pd black (80psi) mol%)~

+ OH3N,~ O _

CbZN'v~o

(-)-1

~7 " ~ THF.78oc/HMPA %..0 .- 78% 144 (Ot-Bu)2

_ 5 14

~

MeOH (0.023 M) O ,, quant. P(OtBu)2

~

O

'/~.~x'....../P(Ot-g u)2 146

SCHEME 45. Asymmetric synthesis of an amino acid phosphonate.

Burke alkylated ( - ) - 1 with the benzyl bromide malonate 147 to give the adduct 148 in good yield and with high diastereoselectivity (Scheme 46). Auxiliary removal via hydrogenolysis and Fmoc protection gave the protected 4-(2'-malonyl)phenylalanine derivative 149. 54 The analogous enolate alkylation with the diethyl malonate derivative proceeded in only 14% yield. However, conversion of 148 to the corresponding diethyl ester allowed access to the Fmoc-protected amino acid containing the diethylmalonate side chain. 55

8

293

DIPHENYLOXAZINONES

Ph Ph

Br

Ph~o

Ph~o

+

LiHMDS, THF

CbzN _ @

0

y

CbZN.v~ O (-)-1

-78 ~ -> r.t. 66%, >97:3 dr t-BuO2

_

~

O2t-Bu 147

1. H2 (20-45 psi) Pd black (33 m o l % ) 1:1 EtOH / THF (0.28 M)

~

148

CO2t-Bu CO2t-Bu

~OH FmocHN.,_./~O_.

2. Fmoc-OSu, NaHCO3" 95% (2 steps)

~ ] ] ~ ~ C O 2

t-Bu

CO2t-Bu

149

SCHEME 46. Asymmetric synthesis of a protected malonyl phenylalanine derivative.

Garbay alkylated methyllactone 123 with benzyl bromides 150a/b to give dialkylated lactones 151 (Scheme 47); catalytic hydrogenation gave the disubstituted amino acids 152a/b. 56 Ph

Ph

Ph Br~ O + L

KHMDS

Ph~o CbzN e. . . ~ .

o

H2 10% Pd/C (6 mol%)

THF CbzN _ O -78 oC M~ c ~ ) O 1_ a:n=0,44% b:n=l 31% BuO2 150

O

H3'e"',~", O

1:1(0.025 EtOH /M) THF a:n=0,99% b: n = 1,90%

-

123

+

t-BuO2

1 151

t_BuO2C.

1 152

SCHEME 47. Asymmetric synthesis of a disubstituted phenylalanine derivative.

Burke and co-workers adapted the above procedure for the synthesis of the phosphorus-containing amino acid 156 (Scheme 48). 57 Taylor alkylated (-)-1 with the sulfonamide 157 to give 158 (Scheme 49). Auxiliary removal under a hydrogen balloon and reprotection of the free amine yielded protected amino acid 159.58 Jin and Williams alkylated (-)-2 with the benzyl bromide 160 to give 161 in good to excellent yield (Scheme 50). The benzyl ether of 161 could be hydrogenated to the free phenol 162 without cleavage of the diphenyloxazinone auxiliary. Further elaboration gave the N-methyl species 163,

294


Ph Ph Ph Br\ O ~.. /~.. KHMDS CbzN Ph O + ~ THF M CbzN O -78 ~ Me / 86%

Ph Ph 10%Pd/C(13 mol%) OH O CbzN I:IEtOH/THF BocHN ..~--e.~O ~ (0.025M) M 2. Boc20,Et3N 61% (2 steps)

I

123

153

154 I

155 O',P(OEt)2

156 O"P(OEt)2

SCHEME 48. Asymmetric synthesis of a disubstituted phosphorus-containing amino acid. Ph

~. H2 (1 atm.) Ph ~ 0 PdC,2 (50 mol%) Fm~ ?H~o Ph [/ LiHMDS CbzN~~o 11 EtOH/ diox. (0.064M) Phi,.? + ~ I:ITHF:HMPA z'-....r ~ Fmoc-O-Su,Na2CO3 ~ J CbzN v ' ~ ' O F\"~O 80% ~[~.~F99~176 F~...~ 159 (-)-1 F">"'~-N(DMB)2 158 O=S:O O:S:O O I N(DMB)2 157 >98.5:1.5er N(DMB)2 Br

_

I

SCHEME 49. Asymmetric synthesis of a sulfonamide-containing amino acid.

Ph

Ph

Ph

OMe Ph~ Ph Me~ ~ /OBn O O + "~,..~ NaHMDS,- B~ _~-"O

BooN"v~O (-)-2

"~"T H F79-92~ Br ~ 160

~OBn 161

~\OMe

Ph~o

H2 (1 atm.) Pd/C = BooN..,..~ - O EtOH95%

=.

2-~OH 16

OMe

Ph Me Me Ph~o~ 1) H2Pd012(8psi) 0 FmocMeN ~Ho_ MeNlo 1:1 EtOH/ THF : = ~OTBS :"-,rJ"~--,rZOTBS 2) Fmoc-OSu,NaHCO3 163 DMF 164 ~ "OMe "OMe 80% (2 steps) Me Me SCHEME 50. Asymmetric synthesis of a highly substituted tyrosine derivative.

which suffered loss of the auxiliary under higher pressure hydrogenation. The resultant free amino acid was soluble in ether and was thus directly protected to give the N-methyl-N-Fmoc-tyrosine derivative 164. 59 Extension of this methodology to the tetrasubstituted benzyl iodide 165 provided 166 (Scheme 51). Elaboration as before gave the N-methyl-N-

8

295

DIPHENYLOXAZINONES

OMe .O n ~"~O (+2

-78~ 88% 165

OH .

NaHMDS BocN "

+ B~

Ph O == ~ O B n l I .~1 MeO~ ~ \OMe 166 Me

0

OTBS .J.L. /)... MeO" " ~ OMe Me 167

SCHEME 51. Asymmetric synthesis of a highly substituted tyrosine derivative.

Fmoc-tyrosine derivative 167. 60 These highly substituted tyrosine derivatives have been deployed by the Williams laboratory in total syntheses of members of the saframycin/ecteinascidin family of antitumor agents; 166 was utilized for both halves of these agents. Partial auxiliary cleavage from 166 by catalytic hydrogenolysis gave the free acid, whose reduction gave diol 168 (Scheme 52). Further Pictet-Spengler elaboration gave the diol 171, whose N-bibenzyl group could be removed by catalytic hydrogenolysis in absolute ethanol. Inclusion of Boc20 in the reaction mixture led to the protected amine 172 in excellent yield. We speculate that the nitrogen of 168 is deactivated by the carbamate-protecting group and thus is not reduced, while the dialkyl nitrogen of 171 is more electron-rich and thus more readily hydrogenated. Protection of the diol of 172 and removal of the Boc group gave free amine 173, which underwent peptide coupling with the acid chloride 174 derived from 167. The Fmoc-protected coupling product 175 was converted to the corresponding Boc-protected amine, and the primary O-TBS group was selectively removed. Oxidation of the resultant primary alcohol to the aldehyde allowed formation of the hemiaminal species and removal of the phenolic O-TBS group gave diol 176. Treatment of 176 with TFA converted the hemiaminal to the iminium ion, which underwent Pictet-Spengler-type cyclization to give the pentacycle. Reductive amination of the secondary amine provided the phenol 177, which was used as a common intermediate for the synthesis of both (-)-renieramycin G (Scheme 52) and (-)-jorumycin (Scheme 53). D. GLYCINATE A L D O L CONDENSATIONS

Miller and co-workers were the first to explore the utility of boron enolates of the oxazinone for glycine-based aldol condensation reactions. 61 Thus, formation of the boron enolate of ( - ) - 1 and reaction with aldehydes

296

ROBERT M. WILLIAMSAND CAMERON M. BURNETT OMe M e a l

Ph

Ph

~)~ ..'~

OBn ~

1 Pd/C(10%), H2, MeOH

MeO e

OH

Ph?BocN,v...~O NaHMDS'THFVMeO. i O 2"IBCF' NMM' M e O ~ B~ 88o/0 ~ ~"1 then NaBH4 ' 88% 2 steps HO (-)-2

Me"~//[''OBn

MeO ~ Me

168

OMe 161 MeO H_ OTBS EtO2CCHO ,.._ Me ~

"N

ph C.3CN,,0oc Ph

HO

92%

1. DBU,THF OTBS 50 ~

THF

_

HO O%EtPh

169

MeO H M e ~ O T B

Ph

v 2"TBSCI' NEt3' 0H2CI2 88% 2 steps

MeO S Pd/C(10%), H2 M e ~ . / - . ~

89% 2steps

170 OTBS1. KHMDS MeO TBSO allyI-Br j,. M e ~

MeO'//~"r~-N _ ~ P h i_ i : HO ~ 15h OH 171 (dr 1:0.3)

EtOH,Boc20 "- ...J-L /it... : .NBoc 69O/o(+23%otherMeO"HO"~/" "~ .... 2. TMS-OTf, M e O ~ NH diastereomer) 2,6 lutidine AllylO \ 172OH 97% 2 steps 173 OAIlyl Me OMe MeO. ~ .OMe ..~. 2,6-1utidine MeO TBSO ~ \--~ 1. piperidine,CH2Cl2 HO Me CH2CI2 Me"r/,'/'/'/'/~,.., _'/'~"~'~OTBS 2. Boc20, EtOH ,.._ MeO H OH ~H "- Me\/~/~.--hl-/JL.~Boc OMe 3. HCO2H,THF, H20 TI "T H 7" Cl[O~ MeO N~NHFmoc 4. Dess-Martin[O] . . . . ~ / / J - ~ N . / . Fmoc..N'''h AllylO \ O 5. TBAF,0 ~ Meu T ]r ]7 "H 174H r ~ ~OMe OAIlyl 175 77% 5 steps AllylO t,,OAIlyl O 176 OMe TBSO" "7:" Me OMe O Me OMe HO.~ ~ ..Me 1. Bu3SnH,Pd~ O HH 93% "r"/ ~ 2.2,4,6-TCB-CI M e ~ - : 0 TFA MeO H I. It NEt3' PhCH3 1. anisoCH2CI2,M e ~ O M e O _t N MeO" O~ ~,, O 2. HCHO, M e O / ~ "~"'1~'14 O..~1 Me NaBH3CN AllylO I,,, C) " H O Me 71% 2 steps OAIlyl 177

3. DDQ 42% 3 steps

(-)-renieramycin G

Me

SCHEME 52. Asymmetric total synthesis of (-)-renieramycin G.

OMe HO~Me

M

MeO e ~

1. LiAIH4, then aq. KCN, AoOH 2. Pd~ H .H .,L\ IJ,.,.OMe 3. Bu3SnH, DDQ M ~ =

MeO" "1>" "~"- "1]" "14 AllylO ~, O OAllyl 177

4.(49%3stepS)Ac20 5. AgNO3 (78% 2 steps)

OMe O O H H \ / , J ~ ~ Me :_- : Meo~]N O

\

___.~...H OH OAc

(-)-jorumycin

SCHEME 53. Asymmetric total synthesis of (-)-jorumycin.

~e

8

297

DIPHENYLOXAZINONES TABLE 7 Selectivity and yield of lactone aldol reactions % Recrystallized yield of major diastereomer (179) 57 38 42

17: 3:1 5:1 5:1

Me n-Pr i-Pr

Ph Bu2BOTf, Et3N CH2CI2 Phi'- O -78 ~ CbzN. v . ~ o RCHO (-)-1

Ph

Ph

l

Ph~ " . ~ ~0-O. -BBu CbZN~o'" "l 178

|

Ph

O 0 H2PdCI2 + OH (40 psi)_ H3N' OO THF/EtO~ R R" "OH 179 180

= CbzN

SCHEME 54. Aldol reaction of oxazinone 1.

gave the aldol adducts 179 with moderate diastereoselectivities (Table 7, Scheme 54); recrystallization provided the pure major diastereomers. Catalytic hydrogenation gave the corresponding [3-hydroxy-~-amino acids 180. Confirmation of the anti-diastereoselectivity (erythro) was secured through the synthesis of threonine itself. The Zimmerman-Traxler chair transition state shown mandates that the aldehyde approach from the more open face of the constrained E-enolate. This aldol methodology was quickly adapted for the synthesis of more complex natural products. Williams and co-workers reported aldol reaction between (-)-1 and the oxazinone-derived aldehyde 181 to give the aldol adducts 183 and 184 in a-~1:25 ratio (Scheme 55). Barton deoxygenation, catalytic hydrogenation, and cleavage of the methyl ether gave (2S,6S)-2,6-diamino-6-(hydroxymethyl)pimelic acid (6-HMDAP), a naturally occurring DAP acid analog (see above for Baldwin's shorter entrde to this natural product). 62 This methodology was then extended by Williams and Yuan to the synthesis of three stereoisomers of DAP (Scheme 56). 63 Thus, (+)-1 was alkylated with 4-iodo-l-butene to give 188, which after ozonolysis afforded the aldehyde 189. Aldol reaction with 190 provided the adduct 191. Subsequent Barton deoxygenation gave the dimeric species 192, which was hydrogenated as usual to give (2R,6R)-DAP (193) quantitatively. Mixing and matching the oxazinone templates 1 and 2 gave access to (2S,6S)-DAP and meso-DAP, a natural constituent of Gram-negative bacterial cell walls.

298


Ph Phi--.

Ph O

3steps

Ph

_ Ph~,~? "-- CbzN

+

ph~[~. O ~

n-Bu2BOTf, NEt3 p h ~ . . . ~ O

4o,

c o, ooc

MeO/

[-..~,.H ,~C

8~

,., //0 HO H 0~_c~ ~/ ~ .-. //0 HO H 0\\ ,-, u .-" "-' u . w Ph~ ~ .....~' ~'"Ph + P h ~ ~ ~'"Ph ~,--N.N., \ N r.t. 58%

Ph

Ph

~

ooc 0o

03, MeOH Ph = I Me2S

194 ~ Ph

Ph Ph CH2CI2 O Ph,,,1 ~ O = ~ + -78 ~ C b z N / , ~ O B B u 2 56%

195 H/~.O Ph

HO

~

2. n-PhaSnH,AIBN OH

CbzN

O

Ph

toluene/ reflux

(

....V

EtOH/THF BocHN c

CbzN~ O

O -O-

....(

NH3

72%

+

198

Ph

SCHEME57. Asymmetricsynthesisof a differentiallyprotectedmeso-DAP derivative.

Employing the N-Boc-oxazinone ( - ) - 2 as one coupling partner in this protocol gave access to a differentially protected chiral form of meso-DAP (198, Scheme 57). This substance should provide access to numerous stereochemically defined peptides of DAP of biological significance. Interestingly, Williams and Yuan found that the differentially protected dilactone 197 could be selectively ring-opened with HC1 in dioxane to give, after esterification, amine 199 (Scheme 5g). 64 The selectivity stems from the protonation of the N-Boc-deprotected amine, making the adjacent lactone linkage more prone to hydrolysis. Oxidative bibenzyl cleavage furnished amino ester 200. Coupling of the free amine with either enantiomer of a protected glutamate derivative (201) yielded dipeptide precursors 202. Catalytic hydrogenation removed the chiral auxiliary as well as the glutamate-protecting groups, and HC1 in dioxane cleaved the methyl ester to give the 7-Glu-DAP dipeptide 203. DeMong and Williams conducted a stereocontrolled synthesis of (2R,3R) and (2S,3S)-[3-hydroxyornithine (205), beginning with the aldol reaction between ( - ) - 1 and protected 3-aminopropionaldehyde (Scheme 59). The resultant alcohol 204 was obtained as an 8:1 mixture of diastereomers, from which the minor diastereomer could be removed by recrystallization. High-pressure hydrogenation over palladium chloride, followed by neutralization with ammonium hydroxide, yielded

300


Ph

Ph

h

", r

O

Ph Ph*~o ~0 t-BOCN /"

Ph OH ,,~ Ph /CO2Me diox. HCI 1. conc.

BocNIv ~O+cbzN v ~ o (-)-2

C l ~ l

HN'~"(

2. CH2N2 Cbz N

(+)-1

P

Ph 197

CH2CI2,MeOH

199

Cbz.N / ~ O

77o (two steps)

~ p

(

1. Pb(OAc)4

5

O 65% 0

H2N"-,,/Co2Me

O Ph

Ph

200

o

CO2H 201

COzH Et3N O ~ \\ N31P( OPh OPh 58%

~ NyCO2Me O ( CH2CI2/MeOH \ 2. conc. HCI,dioxane Cbz .-: 3. Dowex50x8-200 'N " O 92% ~ ~ O 202 Ph /

RT +H3N NH3+ 203, y-D-GlutamyI-L-meso-DAP

Ph

SCHEME58. Asymmetric synthesis of 7-glutamyl-L-meso-DAE

Ph

Ph@o

Ph Bu2BOTf, Et3N Ph~N.~.. CH2CI2 O -78 ~

1. H2 (78 psi) OPdCI2 (2 equiv) ,. I~,~~. 1:2 EtOH / THF (0.028 M) H3 O

O NHCbz Cbz ~.~ "O C b Z N ' v ~ O H...L[v.J O (N~:~

(-)-1

69% 8:1 dr

204

z

2. NH4OH, pH6

OH

68%

(NHOI-H 9 CI 205 (2S, 3S-[3-hydroxyornithine)

SCHEME59. Asymmetric synthesis of ~-hydroxyornithine.

(2S,3S)-[3-hydroxyornithine; the enantiomer was obtained simply by beginning with (+)-1. 65 Scott and Williams synthesized two of the four possible stereoisomers of [3-hydroxypipecolic acid as part of their investigations into elucidating the relative and absolute stereochemistry of the natural antitumor antibiotic tetrazomine. 66 Using the boron enolates derived from 1, both the (2R,3R)- and (2S,3S)-stereoisomers were obtained (Scheme 60). Their total synthesis subsequently revealed that tetrazomine was constituted from the (2S,3R)-isomer. 67 DeMong and Williams also examined Mannich-type reactions. The boron enolate of ( - ) - 1 reacted with imine 211 in low yield to form

8

301

DIPHENYLOXAZINONES

Ph 1. Bu2BOTf, Et3N

CbZN.v~ O

2. O CH2CI2, -78 ~ .,.,~. NCbz H -"[~""--.~ " Ph" Ph " 206 69%

(+)-1

~ ?

O H _OH ~ "-

Ph,,,lr~O

.~

1 03, CH2CI2,-78 ~ " 2. Me2S, 25 ~ 69%

.,.~,,~,fNCbz 15h 207

Ph"

O H OH H2 (1 atm), 5%Pd/C .~

O ~ . 1/ I -

CH2CI2 66%

O H .OH H2 (50 psi), PdCl2 ~

Ph'"'~"Y" N'v'~F'h 208

O H OH ? I L l - I~ O

~

HO~-

.N..~

EtOH, THF, 25 ~

209, (2R, 3R)-3-hydroxypipecolicacid

92% (er >99.5 90.5)

Ph Ph

O

4 steps =_~

CbZNv~ O (-)-1 25% overall yield (er >99.5"0.5)

,."~OH

H H_

O OH HO/[I,,,~

~HOHH_ O

HN.v.p

N

210, (2S, 3S)-3-hydroxypipecolic a c i d

(-)-tetrazomine

SCHEME 60. Asymmetric synthesis of the two anti-stereoisomers of 13-hydroxypipecolic acid.

(-)-1

1. LHMDS 2. Me2AICI 3. ~lBn TBSO/., ~ 211 "THF, lh., -78 ~ 60%

1.7% HF in MeCN

Ph O Ph CbzNO ~ .L. -.

Ph Ph O CbzNv . ~ O

Ph J. SMe Ph BocN,.j,,. NHBoc y" O CbzN- ' ~ ' O TEA, HgCI2 ~ TBSO/"'v'~NBn DMF, rt, o/n 212 67% 213 BocHN" " NBoc 3.3:1 dr (1H-NMR) Ph p h~..Jl... O H2NJ"--.--" 1. H2, PdCI2 " O--'OH DIAD,PPh3, THF CbzN.....~ O~-115 psi, 4d. _-Ph Phi..,,~ 7" O CbzN~ - - . . O _TBSO' ' ' ' ' v ~ NHBn

-

HO.,.-..,,v...~,NBn BocHN"~NBoc 214

0 ~ 15min, rt lh. 87o/o

I~ N B n L...N/~ Boc 215

reflux, 1.5h.

~N.. ~

2 HCI N

NBoc

95%

H

216, (2S,3R)-capreomycidine

SCHEME 61. Asymmetric synthesis of (2S, 3R)-capreomycidine.

a boron chelate that was resistant to hydrolytic cleavage. It was found that ( - ) - 1 could be converted to the corresponding aluminum enolate by treatment with LHMDS and transmetallation with dimethylaluminum chloride (Scheme 61). Mannich-type reaction with the protected imine 211 gave the protected amino alcohol 212 as a 3.3:1 mixture of diastereomers. Interestingly, the diastereoselectivity of this process provides the threo or

302


I-

HPh

-1 /

Ph ~ ,: 0 r..''~OTBS |

|CBzN-~z:E~.. O AI,,,Me I

/

L

"ue/

A

an

__1

FIGURE 2. Zimmerman-Traxler transition state for the Mannich-type reaction of 1.

syn relative configuration, opposite to the selectivity seen with aldehydes. After conversion to the protected cyclic guanidine 215, high-pressure hydrogenation over palladium chloride and cleavage of the two N-Boc groups with hydrochloric acid yielded (2S,3R)-capreomycidine (216). 68 The syn-diastereoselectivity in this Mannich-type reaction can be explained by the Zimmerman-Traxler transition state (A) between the E-enolate and the E-imine as shown in Figure 2. DeMong and Williams then utilized the protected serinal species described above in Scheme 27 in conjunction with the asymmetric synthesis of capreomycidine (216, Scheme 61) to achieve a convergent and concise total synthesis of capreomycin IB, a natural tuberculostatic agent (Scheme 62). A Hofmann rearrangement was exacted on a late-stage pentapeptide in order to transform an asparagine residue into a diaminopropanoic acid residue. The serinal residue was utilized to construct the unusual 13-urea-containing unsaturated amino acid. This synthesis, which is the only modem synthesis reported since Shiba's first synthesis of NHCBz

1.1.5 M MeOH-HCI, A (97%) 2. N-methylmorpholine, CH2CI2 (92%) CBzHN _:..,~CO2H

BocHN ___y~/ONSu

:"" N 217 H2

BocHN/-...v j

O 218

3.2N NaOH, THF, M eOH, rt (99%)

I....LCO2H BocHN _ : ~ . . ~ N H

(

j

0

.

NHBoc

BocHN ..,.,..CO2H

1. EDCI, HOBt O Me Et3N, CH2CI2 (82%)BooHN _ . , ~ H'"~CO2H 219 ~CONH 2 H H2, 10% Pd-C :-...CONH2 B BnO2Cy,,.NH2"HCI 2. MeOH (98%) 220 i~leH .

+

SCHEME 62. (a) Synthesis of fragments A and B for the total synthesis of capreomycin IB.

8

DIPHENYLOXAZINONES

303 CbzHN....,fCO2H

1. TiCI4, Et3N, CH2CI2 (EtO)3C H, -78~176 (85%) 2. H2 (60 psi), Pd(OH)2 THF, EtOH (99%) 3.3N EtOH.HCI, A (96%) :.._

Ph Ph,,,l~O CbzN v . ~ O

4. A, EDCI, HOBt, NMM CH2CI2, (91%) BocHN ~ 5. H2, Pd(OH)2, EtOH (99%) _:

(+)-1

6

o0.o.t,7

.,

7. PhI(O2CCF3)2, py. MeCN, H20, DMF (70%)

;

1. H2, 10% Pd-C, EtOH 2.1N LiOH, EtOH 3. EDCl, HOAt, DMF CH2CI2, 20% (3 steps)

"O ..~ HN O BocHN _..r...---...T~N ~ .... O

4. 99% formic acid 5.2N HCI, acetone, A; urea, (50-90%)

II

HN~CO2Et

O

~"NHBoc

223

II

o

O O

.. CO Et

EDCI, HOBt DMF, THF, 0~ 86-89%

EtO..L o Et 221

, H H CbzHN ~ N...I ~ ) H,,,~. H BocHN 'rl" ~,, e

-

HN N

HBoc

H NH,.~,' " ,NH '~'NH'HCI

-

~ - NH L , ,.~---. 222 ' ~ J NH.HCl H 1 step from 216 (56%)

H2N% H , . J [ H ,,1 / -...~ N -'H'e BocHN O T ~

EtO" "OEt

O

Me h

XNH

O

NH2

H "UHN/%'O,

HN"" H

capreomycin IB

SCHEME 62. (b) Asymmetric total synthesis of capreomycin ]B.

capreomycin IB in 1977, required 27 steps and proceeded in 2% overall yield from (-)-1. Johns, Mori, and Williams have recently reported an interesting application of the diastereoselective aldol condensation reaction in an approach to quinine. 69 Aldol condensation of ketenesilyl acetal 224 with 225 produced the thermodynamic aldol adduct 226 in 76% yield. This possessed the correct relative and absolute configuration at C8 and C9 of quinine (Scheme 63). Further elaborations to the key piperidinone 233 set up a Pd-mediated SN2'-type cyclization that gave the quinuclidine having the desired relative configuration. The initial cyclization product proved to be unstable and the Pd-mediated reaction was quenched with diisobutylaluminum hydride to give 7-hydroxyquinine. This approach to quinine, which still faces the difficult deoxygenation at C7, is unique in that the C8/C9 stereogenic centers were installed in the very beginning of the

1"i21N

ROBERTM.WILLIAMSANDCAMERONM.BURNETT

304

Ph Ph -

CHO Ph NaHMDS, Ph MeO~ ...r TBSCl ..._ ' " ~ 1 + O "~ ~ ~ CbZN.v,~ O l THE,-78 ~ v CbzN/.~/L..OTBS ~ "N" 97% (+)-1 224 225

ph,,,./)~O

1. Pd-C, H2, EtOAc 2. ZnCI2, MeOH

H CO2Me B~

3. Pb(OAc)4 M e O ~ MeOH-CH2CI2 4. Pd-C, H2 (Boc)20, EtOAc 227 79% 4 steps OBn

Ph...., ~ O ~ o CbzN~OTES

1. TBAF THF, 0 ~

MeO~

2. TESOTf 76% 2 steps

226 (X-ray) H CHO DIBAL-H BocHN-~, .,OTES 1. B n O / ~ / ~ M g B r (81%; M e O ~

THF-ether,-78~ 2. Dess-Martinox. 3. DIBAL-H 50% 3 steps

228 OMs

Ac o, O.AP

OAc

BooHN--OTES 0H2012

n,,~"OC D '~''--~"

2. H2, Pd-C, MeOH MeO 3. MsCI, Et3N,CH2CI2 77% 3 steps

M e O ~

230

229 1. LAH, Et20

BzO

( H ~,'"OTMS

~. 2. TMSOTf,GH2GI2 2,6-1utidine MeO~ 3. i~'-..~-.~OBz 87% 3 steps

ES

232

1.1 M HCI, 0 ~ THF-H20 2. Swern ox. 90% 2 steps

n-Bu3SnF, tol. P(2-furyl)3 2. DIBAL-H 44% 2 steps 234

BzO~ , N ~ " ~-../ 0 ~--.,.,,/ ~ O H T ES MeO",~'~'/i I - -~----' 233

1. Pd2dba3,A MeO

OTES H

231

~OTMS

BzO TMSCI 59%

o N ' "~ ,OTES Nail BocN 'H~ THF, ~"--,, ... 99% Meu

~.

M e ~ H HO'~~N~ N~~ . ~ , , , O H

235, 7-hydroxyquinine

SCHEME63. Asymmetricsynthesisof 7-hydroxyquinine. synthesis, and features a novel C2/C3 C-C bond-forming strategy to access the quinuclidine ring system. V. Other Manipulations

The use of the oxazinones as glycine radical equivalents has been briefly explored. Substitution at either the nitrogen atom or the lactone carbonyl has also proven synthetically useful.

8

DIPHENYLOXAZINONES

305

A. OXAZINONES AS GLYCINE RADICALS

Williams showed that the bromolactone 20 could be reduced with Bu3SnD to give the corresponding syn-deuteriolactone with modest diastereoselectivity (Scheme 64). Subsequent hydrogenation gave chiral, deuterated glycine in 80:20 er. This protocol gives the stereochemistry opposite to that observed for palladium reduction of 20 under D 2 gas. z~

Ph O CbzN- ~ O 13r

1. Bu3SnD H3o NS ~O 2. H2, Pd/C

DH 80:20er

19

236

SCHEME 64. Asymmetricsynthesis of s-deuterium-labeled glycine.

Vederas attempted to synthesize meso-DAP using ~,[3-dehydrolactone 237 as an acceptor for conjugate addition of the 3,-radical species derived from the decomposition of 238 (Scheme 65). 7~ The initial adduct 239 was too sterically hindered to be reduced as expected by 1,4-cyclohexadiene, instead losing a hydrogen atom at the [3-position to give the unsaturated product 240 in low yield. O

-

Ph

Ph~o Ph \ BocN" "~O

238

~, 237

Ph

Ph

Ph~o

" Hcbz

Ph ~ " - -

CO2Me/2 =-

BocN~ < ' 0

Phil

32%

0 -

H9 BocN~ 0

k..~,,,NHCbz

K..~,,,NHCbz

CO2Me

241) CO2Me

239

SCHEME 65. Radical conjugate addition to an ~,13-dehydrolactone.

Kabat used the phenylselenolactone 241, prepared by alkylation of (+)-2, as a radical donor for conjugate addition to the ~,13-unsaturated glycine derivative 242 (Scheme 66). While the addition proceeded exclusively anti to the phenyl rings, the quench at the y-carbon exhibited poor diastereoselectivity, giving a 2:3 mixture of diastereomers 243:244. Catalytic hydrogenation of the auxiliary and protecting group removal gave the 2,4-diaminoglutamic acid hydrochlorides (245 and 246). 71

306

ROBERT M. WILLIAMS AND CAMERON M. BURNETT Ph

Ph,,,~,,~

H2,10% Pd/C ClH~ THF/EtOH,reflux; CIH~ "J O 6N HCI, reflux " 82% (2 steps) CO2H

.~ BocN AcHN~. -'~ Ph Ph , , , ~ O

LiHMDS, Ph THF;= P h , , , ~ O

~CO2Me 243 CO2Me 2 245 /NHAc 2 4 2 Ph "" CIH.H2N__/CO2H B o c N v ~ O PhSeBr B o c N . . ~ nBu3SnH,AIBN- Ph,,' ~ " 0 3 H2 10% Pd/C 84% O PhMe,80 ~ . ~ . THF/EtOH, reflux; CIH~ SePh BocN O (+)-2 74% 241 AcHN,,,~ 6N HCl, reflux 90% (2 steps) CO2H r 244 CO2Me 246

SCHEME 66. Asymmetric synthesis of 2,4-diaminoglutamic acids.

Ph Ph

O

RCHO,p-TsOH

H N . . v ~ O benzene, A

e.4.eL. s.~.4"~ ~'-O

=Ph

O

H2 / PdCI2/ EtOH

R1

002 R2

RI.~'N~o

THF, 25 ~ 40-60 psi . ., ".CO2Me R2= H MeO2C "CO2Me or 250

k__z, :

247

MeO2C

CO2Me

MeO2~ _

,

/

248, E/Z-~-endo

2. Pb(OAc)4,MeOH, CH2C[2 R2 = Me

SCHEME 67. Asymmetric synthesis of highly substituted proline derivatives.

B. OXAZINONES AS GLYCINE-BASED A Z O M E T H I N E YLIDES

Williams used the oxazinone template as a glycine azomethine ylide equivalent for the synthesis of highly substituted proline derivatives. 72 The oxazinone 247 is prepared by TFA removal of the N-Boc group from oxazinones 2. Condensation of 247 with various aldehydes proceeds under mild conditions forming the corresponding azomethine ylides 248 (Scheme 67, Table 8). Dimethyl maleate was found to be a generally reactive dipolarophile that underwent [1,3]-dipolar cycloaddition via the endo transition state to give the bicyclo[4.3.0] oxazinones 249. Formation of both E- and Z-ylides led to generally poor stereoselectivity at C7, but complete stereochemical control was observed at the other three stereogenic centers. Significantly, isobutyraldehyde gave a single diastereomer via the less sterically encumbered E-ylide. The cycloadducts resulting from saturated aldehydes could be subjected to catalytic hydrogenation to directly yield free amino acids 250 (R 2 = H).

8

307

DIPHENYLOXAZINONES TABLE 8 Preparation of substituted prolines (250)

Entry a b c d e f g

Rl

249 % Yield (dr)

Cleavage method (249-250)

250 % Yield (er)

H n-Pr i-Bu Ph 4-H2N-Ph 4-OzN-Ph 2-Furyl

71 32 (1.33:1) 52 (1:0) 70 (1.7:1) 71 (1"1) 71 (1" 1) 61 (1"1)

H2, PdC12 (0.04 M reaction concentration) H 2, PdC12 (0.02 M reaction concentration) H 2, PdC12 (0.02 M reaction concentration) HC1, MeOH/Pb(OAc) 4 HC1, MeOH/Pb(OAc) 4 HC1, MeOH/Pb(OAc) 4 H 2, PdC12 (0.02 M reaction concentration)

98 (>99.5:0.5) 93 (>99.5:0.5) 99 (>99.5:0.5) 57 (>99.5:0.5) 66 (>99.5:0.5) 56 (>99.5:0.5) 99 (>99.5:0.5) (R = THF)

Under these conditions, adducts with aromatic R l groups decomposed, and basic ring-opening resulted in extensive epimerization. Fortunately, an alternative oxidative cleavage protocol was employed for these substrates involving acidic methanolysis/ring-opening to the corresponding methyl esters. Subsequent oxidative cleavage of the amino alcohol auxiliary gave amino acid methyl esters 250 ( R 2 = Me). The C7 diastereomers could be separated by either column chromatography or preparative TLC. The C7 syn-diastereomers readily reacted under these conditions, while the anti-diastereomers gave the amino acid methyl esters in low yield and with attendant epimerization at the ~-stereogenic center. The furylpyrrolidine could not be obtained from 249 (R ~ = furyl); the furyl ring was saturated under catalytic hydrogenation conditions and decomposed under oxidative cleavage conditions. Wudl, Prato, and co-workers reported that N-alkylation of amine 247 and treatment of 251 with acid gave the corresponding azomethine ylide, which reacted with fullerene C60 to give the protected fulleroproline 252 in 26% yield (Scheme 68). 73 Ph

Ph~o Ph C1~OC8H17 P h ~ o 060 HN.v~ O Et3N,A = I / N v ~ o pTsOH= OC8H17 26% 247 251 Ph

PhyLo

N~'- O

252

SCHEME 68. Asymmetric synthesis of a protected fulleroproline.

308


Sebahar and Williams reported a concise asymmetric synthesis of spirotryprostatin B, a cell cycle inhibitor isolated from Aspergillus fumigatus by Osada and co-workers. TM Reaction of oxazinone 253 with the masked prenyl aldehyde 254 gave the E-ylide selectively, in accordance with the results obtained earlier with isobutyraldehyde. The unsaturated oxindole 255 reacted as the dipolarophile to give the desired cycloadduct 257 in 82% yield (Scheme 69), setting four contiguous stereocenters in a single step. X-ray crystallography of the adduct revealed that the dipolarophile reacted in the exo-orientation from the [3-face of the dipole (256). Catalytic hydrogenation of the chiral auxiliary proceeded quantitatively, and further manipulations yielded spirotryprostatin B. Ph Ph,,,l~ O + Me\ J ~ ~

Me ,,,.,~OMe Swernox. ~u Me 89%~'~ Me Ph

Ha

254

J. ~

o

3A,,o,

2s6

II

820/0

9

-

Ha

4S ,

257

(x-ray)

o

~,.[ /~=0

-'N H isatin

Me HN -.

Me

EtO2C\

253

[I

eO4"

O H C v ~ ONe

p h , , , ~O. ~

Ph MePh. -~ MeO4 "'~"" "O 02etazomethineylideMe/~. . . . . N....~.._o [,,3]-dipolar Oi....~' "~'H" cycloaddition._ " ~". ~ ',,CO2Et

(Ph)aPCHCO2Et

diglyme, A

~'abnet~i-oeX~

te

Me MeO'7( H 1. D-Pro-OBn, OH.H,,.~~ M~ \ ....,("N""*CO2H'f;,U BOP,Et3N, Me. I ,y H2, PdCl2 MeCN (74%) MeO---~--...N J. ---~ O-.~~., n -" Me O ~ O "'CO2Et 2. H2, Pd-C, etOH THE, etOH 60 psi, 36h H N ~ 3. BOP et3N, MeCN Hlki'..~...""C02 St 99% ~ 94% 2 steps 258

259 __

1. Lil, py., A

TsOH O- H"I ~ 2 equiv Me " ~ \N / = --- N tol., A Me~ 0_~"'~ " N' " ~ H 0

(70-74%) 2. DCC, DMAP

8 _80o,o

,-,o-,,,y 260 ~

34-43o,0

S BrCCI3,A 3. NaOMe, MeOH

spirotryprostatinB SCHEME 69. Asymmetric total synthesis of spirotryprostatin B.

8

DIPHENYLOXAZINONES

309

Similar methodology was used to synthesize spirotryprostatin A. 75 The unstable exo-methylene oxindole was prepared in situ from 261 and reacted as above to give the cycloadduct 262 as well as a stereochemically homologous isomer that had suffered elimination of methanol (Scheme 70). Catalytic hydrogenation and thermodynamic epimerization of the ~-methine gave acid 263. The absence of the carboethoxy group reduced steric bulk and allowed lower catalyst loading and hydrogen pressure, compared with the procedure required for the hydrogenation of 257 for spirotryprostatin B. Coupling to D-proline, hydrogenation, and pentacycle formation gave 264. Elimination of methanol under acid-catalyzed conditions yielded spirotryprostatin A in 8% overall yield along with 9-epispirotryprostatin A. Me Ph Ph OHC.v~Lo Me Me Ph. i Me Ph 254 Me aeO4 "'~'-"0 aeO~2/~ H '"1,/'~O mol. sieves, tol. Me/~. N~...~O 1. H2, PdCl2 Me~ \ N. ,,CO2H HN"v~O TMS....] "~ O~"'H THF, EtOH~ O~" /

253 MeO/''''~" ~ H ~ ~''

O

H262 N~'~

2" HOAc n-prcHO

263 HN. ~

44% OMe + -20% MeOH elimination

261

OMe

.e

1. L-Pro-OBn, MeO-_-~ "''jN~ ...."'~O BOP'Et3N'MeCN),. Me O,k/~~. "

p-TsOH, H20

2. H2, Pd-C, EtOH 3. EDCl, Et3N, MeCN

3A, tol., 100 ~

HN~ I

~--

O... Me. Y IN Me~ ' " N ;'""~O

HN~ - - - ~

/I

264 OMe

OMe spirotryprostatin A (8% overall) + 9-epi-spirotryprostatin A

SCHEME 70. Asymmetric total synthesis of spirotryprostatin A.

Sebahar and Williams also conducted an investigation into the effect of varying the aldehyde side chain on the dipolar cycloaddition with ethyl oxindolylidene acetate (255). 76 In addition to the E-~-exo products 265, other products isolated resulted from endo dipolarophile approach (266) or from the Z-ylide (267) (Scheme 71). Bulky aldehydes favored the E-ylide and gave the standard cycloadducts 265. However, isobutyraldehyde (Table 9, entry d) gave all three products with an 8.6:1 ratio of diastereomers with the desired regioselectivity. Less branched aldehydes gave

310


Ph RCHO 3,A.,PhMe

Ph

Ph,,.1S~O

Ph.

Ph

Ph

Et02C.

HNv-~ 0

@o

253

255

H

265

266

267

SCHEME 71. Dipolar cycloadditions of 253.

TABLE 9

Spirooxindole pyrrolidine cycloadducts Entry a b c d e f g h i j

R H CH2OBn CH2OBn i-Pr i-Pr i-Bu i-Bu HRCCMeR(OMe) H2CCMez(OMe ) p-MeOPh

Temperature 265 % Yield Reflux Reflux 60 ~ Reflux 60 ~ Reflux 60 ~ Reflux 60 ~ Reflux

28 44 54 43 74 84 86 29 82 60

266 % Yield

267 % Yield

dr (265:266)

ll 14 8 11 6 1 0 0 1 0

0 0 0 5

> 20:1 >20:1 8.6:1 >20:1 >20:1 >20:1 >20:1 >20:1 >20:1

trace 0 0 0 0 0

high diastereoselectivity but moderate e x o selectivity (Table 9, entries a-c), and bulkier aldehydes gave high diastereo- and facial selectivity (Table 9, entries f-j). Most reactions showed improved regio- and diastereoselectivity upon lowering of the reaction temperature from reflux to 60 ~ p-Anisaldehyde, though, required refluxing conditions to react, presumably because the electron-donating methoxy group slows formation of the azomethine ylide. Conversion of the cycloadducts 265 into the corresponding amino acids was generally straightforward, requiring high-pressure hydrogenation, and was followed by derivatization to the methyl esters 268 to aid characterization (Scheme 72, Table 10). Unfortunately, the p-anisaldehyde cycloadduct 265 (R = p-MeO-Ph) gave only a small amount of product under these conditions, and elevated temperatures or pressures gave decomposition. Pearlman's catalyst aided somewhat, while acidic conditions were found to give the desired amino acid in tolerable yield.

8

311

DIPHENYLOXAZINONES

Ph H Ph,, O 1. H2(70psi) R,, N...,CO2Me o R ~ O Peel2(1.0equiv.) O~ ''C ,, ~. O2Et [ '-. "'C02Et 11 MeOH/THF(0.05M) HN HN,~ 2. TMSCHN2 "~ 265

268

SCHEME 72. Cycloadduct auxiliary removal and methyl ester installation. TABLE 10 Conversion of cycloadducts 265 into methyl esters 268 Entry a b c d e f

Substrate

Method

% Yield

265a 265f 265h 265j 265j 265j

H 2, PdC12 H 2, PdCI 2 H 2, PdC12 H 2, PdC12 H 2, Pd(OH) 2 H 2, Pd/C, 1 M HC1

93 89 85 5 25 59

Attempts to cleave the bibenzyl auxiliary from 265 via oxidative conditions resulted in decomposition, presumably due to competing oxidation of the electron-rich oxindole. Schreiber and co-workers have extensively adapted this methodology to a solid-phase combinatorial platform as part of their diversity-oriented synthesis program. 77 Oxazinone 253 (or its antipode) is first reacted with resin-bound aldehydes with magnesium perchlorate in the presence of trimethyl orthoformate (Scheme 73). This is followed by Pd-catalyzed alkynylation, amine coupling, and N-acylation to give the tagged library pool 270. Cleavage of the resin from 270 allows direct assays, and some compounds, including 271, were found to have promoter activity for the actin polymerization inhibitor latrunculin B. Ahrendt and Williams used the oxazinone azomethine ylide in an approach to the manzamine-type alkaloid nakadomarin A (Figure 3). 78 The E-ylide formed from oxazinone- 253 and mannitol-derived aldehyde 272 reacts with the unstable alkene 273 in a E-~-endo fashion to give the cycloadduct 275 in moderate yield (Scheme 74). Catalytic hydrogenation removed the biphenyl auxiliary in excellent yield, and further elaboration gave ester 280, containing the ADE-ring core of the hexacyclic natural product.

312


Tag-1 ~

Ph 1. amine PyBOP N-acylating i-Pr,, ,i-Pr + Ph'"l'~O 1 Mg(CIO4)2 1. alkyne Si'o__~__CHO HN'v~ O HC(OMe)3'PY. Pd(ll), Cu(I) i-Pr2NEt reagent 2.Tag-3 2.Tag-4 2.Tag-2 269 253 Ph Tag-1 Ph Ph,,,W-o TAG-. ~ Ph 2 y ~ % i-Pr\ i-Pr "'T/-\O .... , N 0 T A G ~ ~ ~ / ~ / si[ O-~,./N ,'~, O HO/~/O ~ ....i,,fO TAG-4 O HO O'~ N ' ~ H N v ~ 270

44

~/;'~,~,,,. a1

MeO~O

HO--~--CHO CHO O~ CHO + ~~/~]/CHO ~ . OHN~O OH OH " OH "[" CHO OH L,,,./O L,.,,./O L,,,/o I

271

o../~ o

SCHEME 73. Library of 3,520 spirooxindole cycloadducts.

FIGURE 3. Structure of nakadomarin A.

C. D I R E C T N I T R O G E N S U B S T I T U T I O N

Recent reports have explored substitution of the oxazinone nitrogen prior to auxiliary removal to give heterocyclic products. Funk alkylated the sodium enolate of ( - ) - 2 with his bromomethylvinyl ketone equivalent 281 to give 282 (Scheme 75). Thermal unmasking of the vinyl ketone 283 was tbllowed by removal of the N-Boc group to yield the free amine, which suffered conjugate addition in situ to give the 6,6bicyclic system 284. Stereoselective reduction of the ketone, followed by catalytic hydrogenation, yielded (2S,4R)-4-hydroxypipecolic acid 2 8 5 . 79

8

DIPHENYLOXAZINONES

o

313

o

Ph' "O "O ~ M e +

r~h,,,

~_

Ph"I"~'-"N/HH +

I'

13oc

253

272

o--o

[.

Ph,,, .O.T ,/_O . - - , Boc

]~ -'~O~=(~-~N-B~

H20"--/,/iPh'"''" N/'~M'~l---/~//O_.~ eO 1_

Me

~

1 44-60% PO.... O ' -

Me'~Me 275 Boc 20 mol%Pd(OH)2/C / " ' - - - ~ O /N--,~ 1. TMSCHN2 _k./ MeOH H2 (1 atm) ~ C l H. /---~'"\\ = MeOH/EtOAc(1:1) O NEt3' CH2CI2 O , ~ ~ " ,~, O 2.SnCI2 92% Me.C) "H ~ "CO2H 0~ - r.t. . ~__/"H ? "CO2H CH3NO2/H20 Me 276 75% Me L) v v -.~ 80% Boc Boc 277 Boc 'N '~'"--~k ~l , l Mes 'NA ~....~ MeslCN"'r/N" H /--~'"\\ PPh3, imidazole,12-~ H '_~/.." ~,. "O 5mo1% .Ru=~ HO~...~,.. 0 toluene,80 ~ ///'- "1',!" "'r,~ ,,I~ Cl 2Me "CO~e 60O~o l . L .... '0~2 h O ~ /// v v "O CH2CI2,reflux 278 279 280 65% \

273

h

274

. Boc

.~,.. ],

/-. ~'~

SCHEME 74. Dipolar cycloaddition approach to nakadomarin A.

Ph

P~"~O

NaHMDS HMPA, THF, -78 ~

B~ (-)-200

~ B r 281 86%

-v-

Ph BH3, THF TMSOTf, P h ~ o -78 ~ lutidine; ~ L____.~ MeOH " 59%

_ "[~:

0

O 284

83%

Ph Ph~o BocN _,~L.. O 150 ~ [ ' f ' ~ / " 282 v

Ph Ph ~JL.. 0 BocN _ @

PhMe 95%

283

O

o.. ,.o

H2 (50 psi) 5% Pd(OH)2 (20% w/w) EtOAc (0.05 M) 97%

0

~ H

-

OH

OH

.....~o 285

SCHEME 75. Asymmetric synthesis of 4-hydroxypipecolic acid.

Lee and co-workers alkylated (-)-2 with 5-chloro-l-iodopentane to give chloride 286 (Scheme 76). The N-Boc group was removed and cyclized to give the 7,6-bicyclic system 287. Catalytic hydrogenation and reprotection of the amine gave the protected (2S)-2-azepanecarboxylic acid 288. 80

314


Ph

NaHMDS

o BocN v

Ph

h@'o "O HMPA/THF"-

73%

(-)-2

BocN

O

,~

C~

87%

45%

286

Ph

Ph~

H2 (50 psi) PdCI2 (30%w/w)

O

O 2:1 EtOH:THF (0.015 M) 96%

Boc20

Boc..,/\'?H N A,,.O

60% 288

287

SCHEME 76. Asymmetric synthesis of protected 2-azepanecarboxylic acid.

O O Ph [ ~ ' . ~ O E t Ph~ " ' O CH2 N2289 ,. r/N-,..~oCu(acac)2,.

Ph

Ph~o

eto cu':

/

HN-..~ O 247

Et3N 79%

N~O 290

Ph Ph~./]L...O EtO2C,,,~ ~::::O O~N----/

293

quant. ""/'/~CO2Et O 291

OEt

O/ ~/ 292

Ph THF= 97%

~N'[

.... z~OH

294 H

SCHEME 77. Asymmetric synthesis of a hydroxylated pyrrolidine.

Saba used the oxazinone in a highly unusual way as a template for synthesis of hydroxylated pyrrolidines. Conjugate addition of amine 247 to vinyl ketone 289 gave diazo species 290 (Scheme 77). Decomposition of 290 with copper(II)acetylacetonate yielded, after a Stevens rearrangement (291-292), the 7,5-bicyclo product 293. Reduction with LAH gave the tetraol 294 in excellent yield for the sequence. 8~ D. C A R B O N Y L M A N I P U L A T I O N A N D P E P T I D E ISOSTERES

Homologation at the oxazinone lactone carbonyl group has significantly expanded the scope and utility of these glycine templates and has

8

315

DIPHENYLOXAZINONES

Ph Ph ~ : O

NaN(TMS)2

~O HH IPh Ph -

THF, -78 ~ Me OTf

CbzN~ O _ Me j

Me./~75O/o

"~Me

Ph OTBS 1. DIBAL-H, CH2CI2 Ph.d.... H? J]... -78 ~ OMe ~ CbzN~ / ~ O A c ZnBr2, CH2CI2 2. Ac20, DMAP ; 58% Et3N' CH2CI2 Me.,~ ,,,.._

CbzN - . ~''~'O

(-)-1 Ph Ph ~ - O

+

295

Me 296

82%

Ph Ph ~ : O

OH O

1 KOH, dioxane MeOH

CbzN..~, ..../CO2Me Me\.r,"/: 297

CbzN_.~,.~~CO2Me H Me,,,~ 298 - 1 : 4, SYNANTI Me Me

Me.

2. Li~ NH3, THF n-BuOH 69 %

H

Me statine

SCHEME 78. Asymmetric synthesis of statine.

provided access to a number of peptide isosteres and related compounds. The original example of Williams, Colson, and Zhai began with alkylation of ( - ) - 1 with 1-isobutyl triflate to give lactone 295 (Scheme 78). Reduction to the lactol and acetylation gave the hemiacetal 296 in good yield as an equimolar mixture of diastereomers. Treatment of this mixture with the ketenesilyl acetal of ethyl acetate gave the masked 13-hydroxy esters 297:298 in moderate yield as a 1:4 mixture, favoring the desired relative and absolute stereochemistry. Hydrolysis of the methyl ester allowed dissolving-metal reduction of the auxiliary, giving the [3-hydroxy,/-amino acid statine. 82 Further investigation of the lactone carbonyl homologation revealed that Sakurai coupling of allyltrimethylsilane with lactol hemiacetals 300 proceeded in good yield with a slight level of stereoselectivity (Scheme 79). 83 _Ph Ph,,,.~O f CbzN O ~ '.L. '~

P_h

NaHMDS RX

1) DIBAL-H C H2CI2 -780C

Ph,,,~,~ 0 D,

THF_78 / HMPAoc

CbzN DMA_ Et3N' CH2CI2

R

(+)-2

299

Ph Ph,,,~jk.. O

CbzN.]~~-~

Ph Ph, "',t'/~O : == C b z N - - ~

R 301, R = H: 98% R = Me: 61% >99:1 d.r.

Ph

Ph,,,1~

R

302

R

BF3 9 OEt2, MeCN ~ SiMe3 -15 ~ (R = H) OAc -40 -> 0 ~ (R = Me)

300, R = H: 78% 68:32 d.r. R = Me: 76% 3:2 d.r.

0

Li/NH 3

+ OH O H 3 N ~ OOH EtOH/THF" R 303, R = H: 81% R = Me: 80% >99:1 d.r.

SCHEME 79. Asymmetric synthesis of hydroxymethylene peptide isosteres.

316


The alkene was ozonized and oxidized to the [3-hydroxyacid 302, allowing dissolving-metal removal of the bibenzyl auxiliary to give the hydroxymethylene peptide isosters 303. Alternatively, a hydroboration-oxidation protocol yielded protected 7-hydroxy-]3-aminobutyric acid 302, whose dissolving-metal reduction gave hydroxyethylene peptide isostere 303 (Scheme 80). Formation of the unsubstituted hemiacetal and displacement with allyltrimethylsilane as previously gave alkene 306, which was converted to the acid 307 (Scheme 81). Hydrogenation at high temperature and pressure gave (S)-7-hydroxy-]3-aminobutyric acid ((S)-GABOB, (S)-308a),

Ph

Ph ~

Me 301

Li/NH3

' CbzN...]/~.'~OH Me O 302

~ H3N 80% >99:1 e.r.

OMe 303

SCHEME 80. Asymmetric synthesis of hydroxyethylene peptide isosteres.

Ph CbzN~.,r, ,'Ph

.o

L~ O

Ph CbzN--~,, ,Ph L..../,

RLi/CeCl3 THF,-78~ (45-73%)

(+)-1

-

~/--SiMe3 TiCI 4, 0H2012

.._ " -

R'

-78~ or BF3-Et20, MeCN

Ph ~

_Ph Me3Si"" I (~ /~

'"r CbzN~

304a R = H, R'= Ac -15 to -20 ~

R

305

304b, R = Me, R'= H

Ph

304c, R = n-Bu, R'= H

Ph C b z N ' ~ l ,,''Ph 1.03, MeOH/CH2CI 2, ~ Ph 1. 120 psi H2, PdCI2, O -78 ~ then M e 2 S CbzN "'t"" THF/H20, 75-80 ~ -

2. PDC, DMF, rt, 77-91% 306a, R = H (98%) 306b, R = Me (65%) 306r R = n-Bu (67%) H2NHO~ R ~.. v ~"OH

308a, R = H 308b, R = Me 308c, R = n-Bu

HO2C

2. DOWEX 50WX2-100 93-99%

307a, R = H 307b, R = Me 307c, R = n-Bu

t. aq. CH20, 10% Pd/C 80 psi H2, rt (97-99 %) 2. Mel, DMF, rt 3. Amberlist IRA-400 (HO-) (68-85%)

+ u HIiOR ,O, Me3N ~ ~ \ c ~

309a, R = H ((S)-(+)-carnitine) 309b, R = Me 309c, R = n-Bu

SCHEME 81. Asymmetric synthesis of (S)-(+)-carnitine and congeners.

O

8

DIPHENYLOXAZINONES

317

whose N-permethylation gave (S)-(+)-carnitine (309a). 84 Application of this protocol to substituted lactols (304b,c) produced the tertiary alcohol derivatives 308b,c and 3 0 9 b , c . Alternatively, ketenesilyl acetal addition directly to the oxazinone (+)-1 produced O-silyl lacto1311 (Scheme 82). 85Treatment with boron trifluoride effected elimination to 312, which was diastereoselectively hydrogenated to produce 313 (dr = 94:6). Hydrogenation gave (R)-GABOB 314, whose N-permethylation gave (R)-(-)-camitine. Ph Cbz- N ~.~,.,Ph L~o

OSiMe2t-Bu F EtO O -7 /~oet ,H H ~,._.2 SiMe2t-BuI "~ /Ph.... I ~ - '' O ~) (..~I

/

O

-20 ~

(+)-1

CH2CI2 0 ~

O O

312

Os, e

85%

t-Bu

bz" N L

Ph BF3.Et20 CbZ.N.~r,,,Ph

95%

I

Ph Cbz-. N/~1,,,, Ph

EtO""~O 310

Ph H2, PdCl2, EtOH ,,.. HCI.HN~r,,Ph HCI, rt, 99%

dr=94:6

311

_J

L~o

O~

313

1.1M NaOH, THF, rt, 12 h 2.1M HCI (till neutral) 3. H2, PdCI2, 75-80 ~ 3 h 4. DOWEX 50WX2-100 91% 4 steps

OEt

H2N

OEt aq. CH20 OH O 1. CH31,DMF, rt OH O 10%Pd/C,, M e 2 N ~ = OH OH 2. Amberlist 80 psi H2, rt IRA-400 (OH) 314 99 % 315 83% 2 steps

Q OH O Me3N~oQ (R)-(-)-carnitine

SCHEME 82. Asymmetric synthesis of (R)-(-)-carnitine and congeners.

The reaction of hemiacetal 316 with crotylsilanes was examined by Aoyagi and Williams (Scheme 83). 86 Yields of the coupling products 317 and 318 ranged from moderate to good (Table 11). The stereochemistry at the former carbonyl carbon was assigned on the basis of NOE experiments. The methyl-substituted stereocenter was not so easily assigned, and further derivatization was required. Ph

1"

Ph

~

CbzN ~ - ~ r , ^c,.,,.,

_ _ H

conditions

-15 ~ 316

Ph

"~ C b z N ~ anti

317

lVle

~

~

+ CbzN

syn

318

SCHEME 83. Coupling reaction of 316 with crotylsilanes.

Me

318

ROBERT M. WILLIAMS AND CAMERON M. BURNETT TABLE 11

Lewis acid mediated coupling reactions of hemiacetal 316 with crotylsilanes Entry

Crotylsilane SiMe3

Lewis acid

% Yield

317:318

B F3.OEt 2

96

74:26

48

74:26

a

Me~

b

Me~SiMe

3

TIC14

c

Me~ S i P h

d

Me

3

BF3"OEt2

54

84:16

/---SIMe3

BF3.OEt 2

93

38:62

e

Me~

SiPh3

B F3.OEt 2

70

37:63

f

Me

BF3"OEt2

81

39:61

/~TBS

An ozonolysis-reduction-mesylation protocol converted alkene 319 to the mesylate 320 (Scheme 84). Hydrogenation and intramolecular ringclosure yielded a highly polar pyrrolidine, whose isolation was aided by protection of the amine to give pyrrolidine 321. Surprisingly, the catalytic hydrogenation had cleaved both the Cbz and bibenzyl groups from nitrogen, while the C-O benzylic bond was untouched. Dissolving-metal conditions yielded the protected 3-hydroxypyrrolidine 322. _Ph

Ph,,
95:5) CO2H 337 ! H2, 120 psi | PdCI2, 75 ~ THF:H20 OH HCI-H2N_ ~ C O 2 H

338

IR

SCHEME 87. Asymmetric synthesis of isothreonine and nor-C-statine.

5 mol% CbzN v-"/~" O

(+)-1

TBAF / alumina (1 mol~ 5 equiv PMHS / toluene 92%

2

-

. CbzN

OH

339

SCHEME 88. Titanium-based reduction of oxazinone 1.

Buchwald reported a catalytic titanium-based method for reduction of lactones to lactols and applied it to (+)-1 to give the lactol 339 in excellent yield as a mixture of diastereomers (Scheme 88). 88 The procedure offers the advantage of using cheap polymethylhydrosiloxane as a stoichiometric reductant, but has not found synthetic application in this area. Interestingly, the procedure did not epimerize lactones with ~-stereogenic centers, suggesting possible application to carbonyl manipulation of the wide variety of ~-substituted lactones, though steric hindrance at the position appears to reduce the procedure's effectiveness. 89

8

DIPHENYLOXAZINONES

321

Baldwin reported that ( - ) - 2 could be substituted at the carbonyl with fluoride-activated TMS-CF 3 to give the substituted lactol 340 (Scheme 89). 9o Unfortunately, cleavage to the desired ~-trifluoroketone 341 failed under a variety of conditions. Additionally, ~-substitution with even a methyl group reduced the yield to 35%, and ~-substitution with a phenyl group completely blocked the addition. Ph ~ . ~ Ph 0 BocN.v~o

TMS-CF3 Ph CsFor -~ P h ~ TBAF/THF B~ 'O"OH-v~ 82% CF3 340

(-)-2

Ph ~r X = HO "'Pho B~

a

341

SCHEME 89. Trifluoromethyl substitution of oxazinone 2.

Williams and co-workers explored the stabilized Wittig reaction for homologation of the lactone carbonyl. 9~ Reaction of (+)-1 with (triphenylphosphoranylidene) acetonitrile in xylenes at 210 ~ gave the trisubstituted alkene 342 quantitatively, via tautomerization of the initial exo olefin (Scheme 90). High-pressure hydrogenation removed the Cbz group and reduced both the alkene and nitrile to give oxazine 343 as a single diastereomer in quantitative yield. Addition of hydrogen from the less hindered face of the olefin results in a syn-relationship between the phenyl rings and the side chain. Selective protection of the primary amine Ph Ph . . . < O (+)-1

Ph3P%/CN

Ph Ph...1~ O

xylenes, 210 ~ >99%

342

+

H N v ~ ....

/

344 1. Na~ NH3(0, EtOH

Ph ~ [O Obz"N- ~ O

~" NHCbz O

"~ 2. H2(80 psi), PdCI2, 80 ~

CN

120 psi, rt >99%

343

L

NH2-HCI

Ph

_Ph Cbz20, 2N NaOHO ) P,h ._. . ~ THF, rt 70%

Ph Ph.. ' ~ O _

H2, PdC12,HCl

H

DIPEA~-

Ph.. ~

O,~... ,,.0

120~ 75-85%

Ph

N ' ~ Cbz

OH

H O ' ~ ....~ ~ / N ~ N H NH2 hypusine

PhN ~

346

Ph

"" NHCbz

2

83% (two steps) SCHEME 90. Asymmetric synthesis of hypusine.

322


allowed alkylation of the secondary amine 344 with known iodide 345, yielding dilactone species 346. Hydrogenation at elevated temperature and pressure allowed removal of both bibenzyl groups to give (+)-hypusine. Wittig reaction of (+)-1 with methyl (triphenylphosphoranylidene)acetate gave the trisubstituted alkene 348 quantitatively (Scheme 91). Catalytic hydrogenation removed the N-Cbz group and saturated the enamine. N-protection gave ester 349, which was reduced to the aldehyde and converted to the benzylimine 350. A chelation-controlled allylation (via 351) of the imine yielded alkene 352; the alkene was converted to the acid 353 and coupled to give hydrazide dipeptide 354. Global deprotection via hydrogenation gave (+)-negamycin. 92

Ph Ph'"l'~O

CO2Me PhaP=/

I

CbZN.v~ O xylenes, 210 o~ (+)-1 ph,,, / ~ O ' r Ph OMe __ CbzN "v~"' "/'~O 349, dr = 94:6

2. BnNH2, AI203 CH2CI2, 0 ~

HN-Bn CbZNv,J,,,, ~ 9 352, dr= 4.4:1 Ph Ph,,, .Bn CbZN'v~ ...." ~ H

~/

_Ph Ph'"f~O

Ph

Ph,,~ ~Bn

CbzN'/"'~"~O

"'

~

~

I~~~R"

Cbza

THF, -40 ~ ZnBr 351

Ph Ph,, f . ~ Cbz .Bn WSC,Et3N, HOBt, CH2CI2 "'I'" "? N" Me O CbzN..,..~ ....~ C O 2 H - v "JJ"oB N- n H2N" 353 80%

MeOH,H20, HOAc

SCHEME 91.

H

CeCl 3

1. 115 psi H2, PdCi2 MeOH-HCI,rt 2. CbzCI, EtsN DMAP, CH2CI2 96% 2 steps

350

40 psi H2, 10% Pd-C

"OBn

OMe

348

Cbz

2. 03, MeOH CH2CI2 3. PDC, DMF 56% 3 steps

MeO . E L

O/"N"

J 100%

83% 2 steps 1. Cbz-CI, NaOH

r

~

347

~

Ph',,~2"-0

] OMe

3bZN~o

1. DIBAL-H, CH2CI2

Ph

354

Ph 3h' " ~ "-O

OH N_H2 0 (+)-negamycin

Me O H

75%

Asymmetricsynthesisof negamycin.

VI. Conditions to Remove the Chiral Auxiliary A primary consideration when using the diphenyloxazinone template is the choice of method and specific conditions for removal of the bibenzyl moiety. As shown above, dissolving-metal conditions are appropriate when the target amino acid contains unsaturation, as well as for the direct deprotection of the N-Boc-lactones to the corresponding N-Boc-protected amino acids. The ring-opening/oxidative cleavage protocols are usually

8

DIPHENYLOXAZINONES

323

required when an aromatic side chain is introduced at the 0t-carbon, producing a new and potentially reducible benzylic C-N bond. Hydrogenation, while the most-used method for auxiliary cleavage, suffers from a lack of such clear-cut guidelines, especially in the choice of hydrogen pressure or catalyst. The report by Kagan that inspired our work is an example of the ambiguity inherent in choosing hydrogenolysis conditions: the alkene is saturated using Raney nickel in dioxane, while the auxiliary is cleaved with Pearlman's catalyst in ethanol, and an atmospheric pressure of hydrogen is sufficient for both reactions. In this case, the general rule that protic solvents increase catalyst activity appears to account for the differential reactivity, though the choice of catalysts remains unclear. Our general rule has been to first attempt the catalytic hydrogenation with 30 mol% of the catalyst and atmospheric pressure of hydrogen in either a protic solvent (MeOH, EtOH) or, if necessary for solubility, a mixture of protic and nonprotic solvents (e.g., EtOH-THF). The monosubstituted lactones acquired from either the electrophilic or enolate manifolds are often susceptible to catalytic hydrogenation at 14-20 psi of hydrogen. However, the procedure appears highly substratespecific, with similarly substituted adducts requiting varying pressure and catalyst loading. In extreme cases, deprotection of the same substrate is reported under different conditions (cf Schemes 40 and 41). The disubstituted lactones accessible via the enolate manifold require higher hydrogen pressure, presumably due to increased steric hindrance of the proximal benzylic C-N bond, which is typically the last residue to be reduced. In all cases, though, ambient temperature appears sufficient for the hydrogenolysis. Products of carbonyl homologation generally require higher hydrogen pressures and often elevated temperatures for successful hydrogenolysis. The absence of the lactone carbonyl group makes the ring oxygen more electron-rich, suggesting that the reduction should be more difficult. This, coupled with the increased steric hindrance presented by the 13-substituent, hinders reduction of the benzylic C-O bond, which usually is the first to be cleaved in lactone-containing systems. Dipolar cycloaddition products in which the pyrrolidine ring was substituted at every position required elevated hydrogen pressures for successful bibenzyl removal, though an unsubstituted methylene on the proline allowed for the use of atmospheric pressure. The limited examples of six- and seven-membered rings on nitrogen show that 50 psi of hydrogen allows catalytic hydrogenation of the auxiliary. Probably the most interesting case comes from Williams' recent synthesis of reineramycin and jorumycin. Hydrogenation in a mixture of

324


protic and aprotic solvents selectively cleaves the bibenzyl C-O bond, while hydrogenation in neat ethanol is required for cleavage of the C-N bond. Also, while dialkyl substitution of the nitrogen appears to promote C-N cleavage, the N-methyl substrate in the same synthesis requires elevated pressure for auxiliary removal. In this case the TBS-protected phenol may impart sufficient steric hindrance to require the higher pressure. VII. Conclusion

The diphenyloxazinone template has been shown to possess a wide variety of reactivity manifolds that provides access to the synthesis of structurally diverse amino acids, their derivatives, and nitrogen-containing natural products. Current methods include glycine electrophilic reactivity, conversion to the phosphonate, enolate alkylation, radical chemistry, azomethine ylide dipolar cycloadditions, formation of heterocycles at nitrogen, and homologation at the lactone carbonyl, which has provided access to peptide isosteres and other nitrogen-containing substances in optically pure form. While some guidelines are given here for selecting hydrogenation conditions, the auxiliary removal procedure remains highly substrate-dependent and necessarily empirical. Efforts to further extend the breadth of reaction manifolds in these simple heterocyclic templates and the practical utility of these agents are an ongoing and active area of research in our laboratory. Current natural products whose syntheses are being pursued in the author's laboratory include palau' amine, axinellamine, quinine, tuberostemoninol, MPC-1001, ecteinascidin 743, lemonomycin, renieramycin I, bioxalomycin ~2, microsclerodermin H, quinine, nakadomarin A, zetekitoxin, and numerous isotopically labeled amino acids and their derivatives. Acknowledgments The authors are grateful to the National Institutes of Health (GM068011) and the National Science Foundation for financial support. Prof. David A. Evans of Harvard University is acknowledged for providing useful discussions at the earliest phases of this work.

References and Footnotes 1. Herbert, R. A., The Biosynthesis of Secondary Metabolites; Chapman and Hall: London, 1981. 2. Coppola, G. M., Schuster, H. E, Asymmetric Synthesis: Construction of Chiral Molecules Using Amino Acids, Wiley-Interscience: New York, 1987.

8

DIPHENYLOXAZINONES

3 25

3. Williams, R. M., Synthesis of Optically Active a-Amino Acids, Pergamon Press: Oxford, 1989. Vol. 7. 4. An earlier review on this topic has appeared: Williams, R. M., Aldrichim. Acta 1992, 25, 11. 5. Williams, R. M., Armstrong, R. W., Dung, J.-S., J. Am. Chem. Soc. 1984, 106, 5748. 6. Vigneron, J. P., Kagan, H., Horeau, A., Tetrahedron Lett. 1968, 54, 568 I. 7. Both enantiomers of 5 are available from Aldrich: (-)-5, catalog #331899; (+)-5, catalog #331880. 8. Dastlik, K. A., Sundermeier, U., Johns, D. M., Chen, Y., Williams, R. M., Synlett 2005, 693. 9. All four lactones are available from Aldrich: (+)-Cbz, catalog #331856; (-)-Cbz, catalog #331872; (+)-Boc, catalog #331813; (-)-Boc, catalog #331848. The (+)Boc-lactone is also available under catalog #331821. 10. (a) Weijlard, J., Pfister, K., Swanezy, E. E, Robinson, C. A., Tishler, M., J. Am. Chem. Soc. 1951, 73, 1216. (b) Williams, R. M., Sinclair, P. J., DeMong, D. E., Chen, D., Zhai, D., Org. Synth. 2003, 80, 18. 11. Aoyagi, Y., Agata, N., Shibata, N., Horiguchi, M., Williams, R. M., Tetrahedron Lett. 2000, 41, 10159. 12. Chang, H.-T., Sharpless, K. B., Tetrahedron Lett. 1996, 37, 3219. 13. van den Nieuwendijk, A. M. C. H., Warmerdam, E. G. J. C., Brussee, J., van der Gen, A., Tetrahedron: Asymmetry 1995, 6, 801. 14. Sinclair, E J., Zhai, D., Reibenspies, J., Williams, R. M., J. Am. Chem. Soc. 1986, 108, 1103. 15. Williams, R. M., Sinclair, E J., Zhai, D., Chen, D., J. Am. Chem. Soc. 1988, 110, 1547. 16. Williams, R. M., Sinclair, P. J., Zhai, W., J. Am. Chem. Soc. 1988, 110, 482. 17. Zhai, D., Zhai, W., Williams, R. M., J. Am. Chem. Soc. 1988, 110, 2501. 18. Williams, R. M., Zhai, W., Tetrahedron 1988, 44, 5425. 19. Williams, R. M., Hendrix, J. A., J. Org. Chem. 1990, 55, 3723. 20. Williams, R. M., Zhai, D., Sinclair, P. J., J. Org. Chem. 1986, 51, 5021. 21. (a) Ramer, S. E., Cheng, H., Palcic, M. M., Vederas, J. C., J. Am. Chem. Soc. 1988, 110, 8526. (b) Ramer, S. E., Cheng, H., Vederas, J. C., Pure Appl. Chem. 1989, 61,489. 22. Williams, R. M., Fegley, G. J., J. Am. Chem. Soc. 1991, 113, 8796. 23. Williams, R. M., Fegley, G. J., J. Org. Chem. 1993, 58, 6933. 24. Williams, R. M., Fegley, G. J., Tetrahedron Lett. 1992, 33, 6755. 25. Williams, R. M., Im, M.-N., Tetrahedron Lett. 1988, 29, 6075. 26. Williams, R. M., Im, M.-N., J. Am. Chem. Soc. 1991, 113, 9276. 27. Baldwin, J. E., Lee, V., Schofield, C. J., Synlett 1992, 249. 28. Williams, R. M., Sinclair, E J., DeMong, D. E., Org. Synth. 2003, 80, 31. 29. Williams, R. M., Liu, J., J. Org. Chem. 1998, 63, 2130. 30. Aoyagi, Y., Iijima, A., Williams, R. M., J. Org. Chem. 2001, 66, 8010. 31. Dong, Z., Tetrahedron Lett. 1992, 33, 7725. 32. Schuerman, M. A., Keverline, K. I., Hiskey, R. G., Tetrahedron Lett. 1995, 36, 825. 33. (a) Looper, R. E., Williams, R. M., Tetrahedron Lett. 2001, 42, 769. (b) Looper, R. E., Williams, R. M., Angew. Chem. Int. Ed. 2004, 43, 2930. 34. (a) Looper, R. E., Runnegar, M. T. C., Williams, R. M., Angew. Chem. Int. Ed. Engl. 2005, 44, 3879. (b) Looper, R. E., Runnegar, M. T. C., Williams, R. M., Tetrahedron 2006, 62, 4549.

326 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72.

ROBERT M. WILLIAMSAND CAMERONM. BURNETT Nolen, E. G., Watts, M. M., Fowler, D. J., Org. Lett. 2002, 4, 3963. DeMong, D. E., Williams, R. M., Tetrahedron Lett. 2002, 43, 2355. DeMong, D. E., Williams, R. M., J. Am. Chem. Soc. 2003, 125, 8561. Lee, S.-H., Nam, S.-W., Bull. Korean Chem. Soc. 1998, 19, 613. Lee, S.-H., Lee, E.-K., Jeun, S.-M., Bull. Korean Chem. Soc. 2002, 23, 931. van den Nieuwendijk, A. M. C. H., Kriek, N. M. A. J., Brussee, J., van Boom, J. H., van der Gen, A., Eur. J. Org. Chem. 2000, 3683. Allevi, P., Anastasia, M., Tetrahedron: Asymmetry 2004, 15, 2091. Singh, S., Pennington, M. W., Tetrahedron Lett. 2003, 44, 2683. Kele, P., Sui, G., Huo, Q., Leblanc, R. M., Tetrahedron: Asymmetry 2000, 11, 4959. Williams, R. M., Fegley, G. J., Pruess, D. L., Schaeffer, E, Tetrahedron 1996, 52, 1149. Baldwin, J. E., Lee, V., Schofield, C. J., Heterocycles 1992, 34, 903. Aoyagi, Y., Williams, R. M., Synlett 1998, 1099. Schow, S. R., DeJoy, S. Q., Wick, M. M., Kerwar, S. S., J. Org. Chem. 1994, 59, 6850. Solas, D., Hale, R. L., Patel, D. V., J. Org. Chem. 1996, 61, 1537. Shen, K., Keng, Y.-E, Wu, L., Guo, X.-L., Lawrence, D. S., Zhang, Z.-Y., J. Biol. Chem. 2001, 276, 47311. Bender, D. M., Williams, R. M., J. Org. Chem. 1997, 62, 6690. Paquette, L. A., Duan, M., Konetzki, I., Kempmann, C., J. Am. Chem. Soc. 2002, 124, 4257. Chen, Y. T., Xie, J., Seto, C. T., J. Org. Chem. 2003, 68, 4123. Li, P., Zhang, M., Peach, M. L., Liu, H., Yang, D., Roller, P. R., Org. Lett. 2003, 5, 3095. Gao, Y., Burke, T. R., Jr., Synlett 2000, 134. Kang, S.-U., Worthy, K. M., Bindu, L. K., Zhang, M., Yang, D., Fisher, R. J., Burke, T. R., Jr., J. Med. Chem. 2005, 48, 5369. Liu, W.-Q., Vidal, M., Gresh, N., Roques, B. P., Garbay, C., J. Med. Chem. 1999, 42, 3737. Oishi, S., Kang, S.-U., Liu, H., Zhang, M., Yang, D., Deschamps, J. R., Burke, T. R., Jr., Tetrahedron 2004, 60, 2971. Hill, B., Ahmed, V., Bates, D., Taylor, S. D., J. Org. Chem. 2006, 71, 8190. Jin, W., Williams, R. M., Tetrahedron Lett. 2003, 44, 4635. Lane, J. W., Chen, Y., Williams, R. M., J. Am. Chem. Soc. 2005, 127, 12684. Reno, D. S., Lotz, B. T., Miller, M. J., Tetrahedron Lett. 1990, 31,827. Williams, R. M., Im, M.-N., Cao, J., J. Am. Chem. Soc. 1991, 113, 6976. Williams, R. M., Yuan, C., J. Org. Chem. 1992, 57, 6519. Williams, R. M., Yuan, C., J. Org. Chem. 1994, 59, 6190. DeMong, D. E., Williams, R. M., Tetrahedron Lett. 2001, 42, 183. Scott, J. D., Williams, R. M., Tetrahedron Lett. 2000, 41, 8413. Scott, J. D., Williams, R. M., Angew. Chem. Int. Ed. Engl. 2001, 40, 1463. DeMong, D. E., Williams, R. M., Tetrahedron Lett. 2001, 42, 3529. Johns, D. M., Mori, M., Williams, R. M., Org. Lett. 2006, 8, 4051. Sutherland, A., Vederas, J. C., J. Chem. Soc., Chem. Commun. 2002, 224. Kabat, M. M., Tetrahedron Lett. 2001, 42, 7521. Williams, R. M., Zhai, W., Aldous, D. J., Aldous, S. C., J. Org. Chem. 1992, 57, 6527.

8

DIPHENYLOXAZINONES

327

73. Maggini,M., Scorrano, G., Bianco, A., Toniolo, C., Sijbesma, R. E, Wudl, E, Prato, M., J. Chem. Soc., Chem. Commun. 1994, 305. 74. (a) Sebahar, P. R., Williams, R. M., J. Am. Chem. Soc. 2000, 122, 5666. (b) Sebahar, E R., Osada, H., Usui, T., Williams, R. M., Tetrahedron 2002, 58, 6311. 75. (a) Onishi, T., Sebahar, E R., Williams, R. M., Org. Lett. 2003, 5, 3135. (b) Onishi, T., Sebahar, E R., Williams, R. M., Tetrahedron 2004, 60, 9503. 76. Sebahar, E R., Williams, R. M., Heterocycles 2002, 58, 563. 77. Lo, M. M-C., Neuman, C. S., Nagayama, S., Perlstein, E. O., Schreiber, S. L., J. Am. Chem. Soc. 2004, 126, 16077. 78. (a) Ahrendt, K. A., Williams, R. M., Org. Lett. 2004, 6, 4539. (b) Ahrendt, K. A., Williams, R. M., Org. Lett. 2005, 7, 957. 79. Greshock, T. J., Funk, R. L., J. Am. Chem. Soc. 2002, 124, 754. 80. Dutton, E E., Lee, B. H., Johnson, S. S., Coscarelli, E. M., Lee, E H., J. Med. Chem. 2003, 46, 2057. 81. Saba, A., Tetrahedron Lett. 2003, 44, 2895. 82. Williams, R. M., Colson, P.-J., Zhai, W., Tetrahedron Lett. 1994, 35, 9371. 83. Aoyagi, Y., Williams, R. M., Tetrahedron 1998, 54, 10419. 84. Jain, R. E, Williams, R. M., Tetrahedron 2001, 57, 6505. 85. Jain, R. E, Williams, R. M., Tetrahedron Lett. 2001, 42, 4437. 86. Aoyagi, Y., Williams, R. M., Tetrahedron 1998, 54, 13045. 87. Aoyagi, Y., Jain, R. E, Williams, R. M., J. Am. Chem. Soc. 2001, 123, 3472. 88. Verdaguer, X., Berk, S. C., Buchwald, S. L., J. Am. Chem. Soc. 1995, 117, 12641. 89. Verdaguer, X., Hansen, M. C., Berk, S. C., Buchwald, S. L., J. Org. Chem. 1997, 62, 8522. 90. Walter, M. W., Thaker, N., Baldwin, J. E., Muller, M., Schofield, C. J., J. Chem. Res. (S) 2000, 310. 91. Jain, R. P., Albrecht, B. K., DeMong, D. E., Williams, R. M., Org. Lett. 2001, 3, 4287. 92. Jain, R. P., Williams, R. M., J. Org. Chem. 2002, 67, 6361.


Chapter 9 RHODIUM-CATALYZED CYCLOISOMERIZATION REACTIONS OF ALLENES IN DIVERSITYORIENTED SYNTHESIS Kay M. Brummond and Branko Mitasev Department of Chemistry University of Pittsburgh Pittsburgh, PA 15260, USA

I. II. III.

Introduction Designand Synthesis of the Pivotal Allenic-Amino Ester Intermediates AllenicAlder-ene Reaction Affording Amino-Ester Tethered Cross-Conjugated Trienes IV. Diversificationof Cross-Conjugated Trienes via Diels-Alder Reactions: First Generation Triene V. Design and Synthesis of a Second Generation Triene VI. Synthesis and Stereoselective Diels-Alder Reaction of Novel Bicyclic Trienes VII. Summary Acknowledgments References and Footnotes

328 336 342 346 349 353 361 362 362

I. Introduction

Modem drug discovery efforts are based upon a deep understanding of the cellular pathways responsible for a particular disease. The extreme complexity of the cellular circuitry has made the goal of understanding all of its aspects a formidable challenge. Significant progress in this area has been made by using modem biochemical tools such as mutagenesis, which is commonly used to irreversibly modify proteins in order to understand their cellular function. However, there are many aspects of the cell's biochemistry that remain a mystery. These include important signaling pathways based on dynamic protein-protein interactions that are particularly difficult to study. ~ Therefore, there is a tremendous need for innovative strategies to discover new cellular pathways and probe the function of

9

RHODIUM-CATALYZED CYCI~OISOMERIZATION REACTIONS OF ALLENES

329

various protein targets. Small organic molecules are useful research probes for studying cellular pathways. Small molecules are capable of interacting with macromolecules such as proteins generally in a reversible manner, thereby modulating their function. 2 Observing the effects that result from such interactions in designed assays can often lead to an understanding of the role a particular target plays in the cell. Additionally, if the protein is a known therapeutically relevant target, lead compounds for drug discovery can be identified. The great number of relevant protein targets and their immense structural diversity dictates that the small molecules screened against these targets must also be structurally and functionally diverse. 3,4 Nature is one of the greatest sources of diverse small molecules with a broad bioactivity profile. However, a major limitation of screening natural products is their limited availability. 5 Therefore, synthetic small molecules constitute a major portion of the modem screening palette. Advances in the field of synthetic organic chemistry have led to development of many methodologies for efficient assembly of small molecules. The field of combinatorial synthesis has evolved in the past two decades as a consequence of the demand for diverse small molecules for biomedical research. 6 The synthetic strategy that is most commonly utilized in the combinatorial approach involves appending different building blocks around a common structural core. This approach has been greatly facilitated by the development of practical technologies that allow the streamlined parallel synthesis of large numbers of compounds (thousands) in a short period of time. Although this approach will undoubtedly continue to lead to identification of additional biological agents, researchers have questioned whether the level of structural diversity that is achieved is sufficient to complement the wide variety of modem biomedical targets. 7 The appendage diversity that is achieved by varying substituents around a common core is thought to limit the compounds to a narrow chemical space. Very often, and particularly in the pharmaceutical company setting, the molecules accessed in this manner are designed to fall within defined physico-chemical parameters that increase their chances of becoming drug candidates. 8 For example, the well-known Lipinski rules for drug-like molecules consider properties aimed at increasing bio-availability (molecular weight, solubility, number of hydrogen-bond donors and acceptors, etc.). 9 The need for novel, chemically diverse small molecules has been made clear in the National Institutes of Health (NIH) roadmap for medical research (http://nihroadmap.nih.gov/). According to these guidelines, the goal of biomedical research in the future is to identify a comprehensive set of small molecules that are capable of selectively modifying the function

330

KAY M. BRUMMOND AND BRANKO MITASEV

of the majority of biological targets in the human cell. Since the number of relevant biological targets continues to grow as a result of intense research, this important goal can only be achieved by effectively integrating the development of new synthetic technologies to generate novel chemically diverse entities, with assays against a broad range of biological targets. New synthetic technologies encompass methods for more efficient synthesis of small molecules, their purification, isolation and characterization. In recent years, organic chemists have become increasingly aware of these issues and have begun addressing them by innovative combinatorial strategies and diversity-oriented synthesis (DOS). ~~DOS aims to develop new and adapt existing synthetic methodologies for generating structurally diverse molecules specifically for biological screening. Because synthetic chemists generate novel compounds continuously, the term DOS is carefully assigned only to the designed and deliberate synthesis of collections of small molecules populating novel chemical space. 5 Contrary to the classical combinatorial approach, modern DOS efforts put their main emphasis on the diversity of the molecular scaffolds that are accessed and not on the numbers of compounds. ~ Therefore, the synthesis of libraries of compounds is generally limited to between tens to hundreds of compounds, and not thousands (or more) as in the classical combinatorial approach. ~2 Moreover, the compounds are accessed in milligram quantities, which allows a thorough assessment of their biological activity in a wide variety of biological assays. ~3 Characterizing the diversity of molecular libraries is another important aspect of D O S . 14 This is commonly done by using computational methods to predict various physico-chemical properties (i.e., molecular descriptors), which are then compared to those of existing libraries. 9 These can include not only the more common parameters such as number of rotatable bonds, H-bond donor and acceptor groups, solvent-accessible surface area and clogP (distribution coefficient for octanol/water), but also pharmacologically related ones such as the predicted affinity for serum-protein binding, intestinal permeability, metabolizable groups in the molecule, etc. Computational prediction of parameters like these allows the practicing DOS chemist to design libraries that possess a broad diversity profile. In designing such discovery libraries of novel compounds, at least three forms of structural diversity have been considered and include appendage, stereochemical and skeletal diversity. An ideal DOS strategy incorporates all three forms. Skeletal diversity is arguably the most important, but relatively difficult to achieve in an efficient manner. There are at least two conceptually distinct ways it can be accomplished. The first one involves

9

RHODIUM-CATALYZED CYCLOISOMERIZATION REACTIONS OF ALLENES

331

designing structurally similar substrates that under common reaction conditions undergo diverging transformations affording skeletally distinct products. The chemical information that leads to different products is therefore encoded in the substrates (substrate-based control). This is the most commonly utilized synthetic approach to skeletal diversity. ~5 One example from Schreiber involves a 1,3-dipolar cycloaddition between an indole dipolarophile and an in situ-generated cyclic oxonium ylide to give skeletally different products by varying the placement of the reactive functional groups around a common pyridone core (Scheme 1).~6

OMe

Me

S ~ )----N' OMe

O ; CO ~~~V~'N Me

"~,

"~0 OSi

1

4

0"~ ~-N20 ,0, ~ O S i Rh(ll)

MeH / ~ 0 " N '"H

H N ~ O S

Me

~ 5

TsN--~---~

i

0

Me HN /.~e H',~ ~ O N H tBu

o

o" N.

OSi

"%. ";0

L v j "lOSi!4 6

SCHEME 1

Despite the fact that a high degree of skeletal diversity can be accessed via the substrate-based approach, the efficiency of the overall process is compromised, since all of the precursors need to be synthesized independently. Therefore, a more efficient approach to skeletal diversity involves subjecting a common synthetic precursor to different reaction conditions to give skeletally unique products (reagent-based control). Although there are some applications of this concept, it remains the most difficult to achieve, particularly in the context of cyclic skeleton synthesis. ~7

332


Accessing all three forms of diversity requires DOS strategies with branching reaction pathways available to common synthetic precursors. In this manner, structurally distinct scaffolds can be obtained from a small pool of reactants, thereby increasing the overall efficiency of the process. Therefore, there is a strong incentive to develop new chemical transformations and design strategies toward this goal. The development of many useful transition metal-catalyzed reactions in the last decade has opened the door for their application to DOS. ~8 Transition metal-catalyzed reactions are generally environmentally benign and economic synthetic processes, proceeding with high levels of selectivity (chemo-, regio-, and/or stereoselectivity) and minimize the use of raw materials and generation of byproducts. Among these, Pd-catalyzed coupling reactions (Heck, 19 Stille, 2~ Sonogashira, 2~ Suzuki, 22 etc.) and allylic substitution reactions 23 are regarded as some of the most important with the highest impact on the field of modern organic synthesis. Not surprisingly, these reactions have seen extensive application in combinatorial and DOS. The majority of classical combinatorial strategies in the past have been largely limited to utilizing transition metal-catalyzed coupling reactions only introducing appendage diversity (e.g., see Pd-catalyzed Suzuki, 24 Heck 25 and Stille 26 coupling reactions). More recently, intramolecular versions of these and related reactions have been used to create skeletal diversity in the synthesis of small- and medium-sized rings. 27 Related transformations such as the coupling of arylboronic acids and amides recently developed by Buchwald 28 are also becoming increasingly popular in the generation of cyclic skeletons. 29Another very important transformation that is widely used in combinatorial synthesis and DOS is the Ru-catalyzed ring-closing metathesis developed by Grubbs. ~5'3~ Transition metal-catalyzed reactions that transform relatively simple acyclic starting materials to cyclic (or polycyclic) products via a carbocyclization process are another important class of reactions that has received attention in the past decade. Carbocyclization, in general, refers to a cyclization process that involves carbon-carbon bond formation via a carbometalation, wherein a C*-M (Carbon-Metal) species delivers the carbon and metal component across an unsaturated bond (C=C) thereby affording a C * - C - C - M species. 31 In particular, carbocyclization reactions of precursors containing unsaturated functional groups (e.g., alkenes, alkynes) have been very useful in the syntheses of carbocyclic and heterocyclic molecules. 32 Examples of such reactions include transition metal-catalyzed enetype cycloisomerizations, 33-35 [4 + 2] and [5 + 2] cycloadditions 36 (7 to 10 and 7 to 11, respectively; Scheme 2) and [2 + 2 + 1] cyclocarbonylation 37

9

333


reactions of enynes (7 to 12). Aside from increasing molecular complexity, an important aspect of these reactions is that metal catalysis often allows for bond formation that would be difficult or impossible using conventional methods, to readily occur under mild conditions. A typical example is the intramolecular [4 + 2] cycloaddition of electronically unactivated dieneynes proceeding under Rh(I) catalysis (e.g., 7-10). 38 The vast potential for increasing molecular complexity and achieving skeletal diversity via metal catalyzed carbocyclization reactions (e.g., cycloadditon, cycloisomerization, cyclocarbonylation, etc.) remains an untapped resource for DOS.

C( 8 R1 ~ R 1 enyne ~.~esis

13

TR1= Hor alkyl R1= alkyl ~

R2 9

co

[2+2+1]~

~ ~[4+2]

7

R1

[5+2]

10

12

11 SCHEME 2

Although olefins and acetylenes are most commonly utilized in carbocyclization reactions, use of allenes as ~-components is becoming increasingly prevalent. For a relatively long time since their first synthesis, 39 allenes were considered no more than a chemical curiosity and remained underutilized. 4~ Intense research in the past decades, however, has resulted in many useful synthetic methods involving allenes. 4~-43The two cumulated double bonds of the allene display high reactivity toward a range of transition metals, and have been exploited in a variety of ways.

334


Using allenes as olefin components in transition metal-catalyzed reactions often has the advantage of increased reactivity. This is largely due to the strain associated with having two cumulated double bonds, which is estimated at 10 kcal/mol. 44 Despite this fact, transition metal-catalyzed carbocyclizations of allenes (e.g., cycloaddition and cycloisomerization reactions) remain largely unexplored and underutilized in synthesis, presumably because there are no known control elements for effecting double bond selectivity other than substrate modification. 45,46 Recent studies by Brummond and coworkers have resulted in some of the first examples of reagent-based control of olefin selectivity in the allenic cyclocarbonylation (Pauson-Khand reaction) and cycloisomerization reactions. 47 Reagent-based control of double bond-selectivity in transition metalcatalyzed carbocyclization reactions of allenes is ideally suited for application to DOS since skeletally different products can be obtained. For example, selective engagement of the proximal olefin of allenyne 14 in a cyclocarbonylation reaction under Mo(CO)6-mediated conditions leads to an ot-alkylidene cyclopentenone 15 (Scheme 3). 48 Alternatively, the same transformation of the distal double bond under Rh(I) catalysis leads to a 4-alkylidene cyclopentenone 16. 49 Furthermore, a Rh(I)-catalyzed cycloisomerization reaction involving the distal double bond of the allene can lead to a cross-conjugated triene 17. 5o R2

Mo(CO)6

2R1

15 R2

R2 Rh(I), CO

CH2 R1

CH2R 1

16

14

R2 Rh(I)

-R 1 17 SCHEME 3

9

335

RHODIUM-CATALYZEDCYCLOISOMERIZATIONREACTIONS OF ALLENES

Each reaction results in increase of molecular complexity since relatively simple acyclic precursors are transformed to mono- or bicyclic skeletons. Furthermore, a novel, reactive moiety is generated (enone, cross-conjugated triene) that can be further exploited in diversity generating transformations. Therefore, we became interested in implementing a DOS strategy based on transition metal-catalyzed cyclocarbonylation and cycloisomerization reactions of allenes. From the beginning of this study in 2001, we have been interested in developing novel chemical transformations that can be applied to the efficient assembly of functionalized small molecules. As illustrated in Scheme 4, these studies have led us to diverse arrays of functionality. 51 Herein, we focus our attention on the design and synthesis of the pivotal allenic aminoester intermediates 18 and the details of their transformation into cross-conjugated trienes 19 via a Rh(I)-catalyzed cycloisomerization reaction. Furthermore, the diversification of these interesting molecules by using subsequent complexity-generating transformations is described. The ultimate goal of the study was twofold: (1) development of new synthetic methodologies for efficient assembly of complex small molecules; and (2) synthesis of collections of these compounds specifically for use as biological probes. R3 P~N'~ MeO20 R1 II 19A

P~N,.'~.~~ R3 MeO2C R1

R2 23

R.N,,~, ~ ~MeO2C~ R3 R1R2 22

Rh(I)

law

"~

O R3 P \ N - ' ~ "j MeO2C---~~ Rh(I),. R1 I[

R3

~

P-N

.K o

19B

18

Mo(CO)6 R3 P - N / ~ O

Me02C-~ -"~\ R1 ~-,.R2 21 SCHEME 4

MeO2C~O R1 2O

~2

336


II. Design and Synthesis of the Pivotal Allenic-Amino Ester Intermediates

In designing a pivotal allene intermediate for our DOS strategy, three aspects were considered: (1) incorporation of a higher number of N and O heteroatoms was important since compounds containing them are more likely to have a desirable pharmacological profile and exhibit interesting biological effects via specific interactions with proteins; (2) potential for diversification of the molecular scaffolds by employing both front-end and back-end appendage diversity strategies; and (3) ease of preparation and availability of the precursors. Incorporating nitrogen- and oxygen-containing functional groups in the pivotal allenes is also important because they would allow for rapid attachment of pre-functionalized alkynes and alkenes used in the carbocyclization reactions, thereby incorporatingfront-end diversity into the scaffolds. Additionally, the reactivity of nitrogen and oxygen containing functional groups can be exploited in back-end functionalization of the scaffolds subsequent to their formation. With this in mind, several known methods for preparation of functionalized allenes were considered. 42~The Claisen rearrangement of propargyl ethers and esters is a versatile method for preparation of allenes. 52 Work by Castelhano and Krantz demonstrated that mild dehydrative conditions (Et3N, CC14, PPh3) effect the rearrangement of benzoyl-protected amino ester 24 to 4-allenyl-5-oxazolone 25, which is transformed to methyl ester 26 when treated with MeOH (Scheme 5). 53 Furthermore, Kazmaier reported an ester-enolate Claisen rearrangement of a variety of propargyl amino-esters 27 using LDA/ZnC12 affording allenic amino acids 29 with diastereoselectivities greater than 93%. 54 H R I ~ I ~ R2 PPh3,MeC 0014, N Et3N ~

BzHN/-..~O ~ O I

H -".,R2

--~~ R I ~ ~O O

I

MeOH,HCI ~

/,,~

R1 MeO2C NHBz

Ph 24

25

a3

26

H - R2

a3 R2

PHN~'" 0

LDA, ZnCI2, THF

R I ~ I ' ~ R2 PN

..P-~O

~Zn.----O 27

28

[3,3] sigmatropic rearrancjement ,,HO2C

,,,~R1 *NHP

29

P = Boc, Cbz, Tos SCHEME 5

9


337

The allenic amino acid derivatives obtained via these two methods appeared ideal for adapting to our DOS strategy Notably, the propargyl esters are easily obtained in one step by coupling the corresponding Nprotected amino acid and a propargyl alcohol, allowing for multiple points of diversity to be introduced. Therefore, the allenic amino esters obtained in this manner were elected as pivotal intermediates for development of a DOS strategy based on transition metal-catalyzed cycloisomerization and cyclocarbonylation reactions of allenes. Using Castelhano's protocol, we first prepared the phenylalaninederived allene 34 as outlined in Scheme 6. Esterification of N-benzoyl phenylalanine 30 with 3-butyne-2-ol by using N,N'-dicyclohexylcarbodiimide (DCC) and 4-dimethylaminopyridine (DMAP) gave ester 32 in 78% yield. The reaction proceeds via the intermediacy of oxazolone 31, which is then subjected to nucleophilic attack by the alcohol. Treatment of ester 32 with CCl4, PPh 3 and Et3N in acetonitrile affords the 4-allenyl-2-oxazolin-5-one intermediate 33, which is treated with MeOH/HC1 to give allene 34 in 74% yield as a 1.7:1 mixture of diastereomers (as originally reported). 53

Bn BzHN/~.OH O

OH DCC, DMAP 0H2CI2,rt, 78%

Bn)~/O N~.O

BzHN'~ O O

Ph 31

30

32 PPh3, CCI4, Et3N

MeCN, rt

H

H

MeOH, HCI, rt

Bn,~~

74% (2 steps)

MeO20

Ph

34

33

SCHEME 6

A serine-derived allene (38) was also synthesized following the route outlined in Scheme 7. Esterification of the known acid 35 with 3-butyn-2-ol in 73% yield was followed by removal of the Boc protecting group

338

KAY M. BRUMMONDAND BRANKO MITASEV

in 36 with TFA. Coupling of the primary amine of 36 with benzoyl chloride to afford amidoester 37 in 60% yield for the two steps. Claisen rearrangement of 37 afforded allene 38 in 87% yield as a --~2:1 mixture of diastereomers (determined by integration of the allenic methyl group resonances in the ~H NMR spectrum).

TBSO~ B~

L H 35

OH

TBSO.. ~"~. /

DCC, DMAP B O O , N]..~, ]O/ COOH 'CH2Cl2, 73% H II O

TBSO.. 1. TFA, CH2CI2, rt 1"'['/ 2. BzCI, Et3N, CHCI3 B Z ~ N ~ O 60% (2 steps)

36

H

n O 37 1. PPh3, CCI4, Et3N MeCN 2. MeOH, HCI, rt. 87%, dr = 2 : 1 H

.,~NHBz MeO2C" ~---OTBS

38 SCHEME 7

The alternative protocol for preparing allenic amino acids with high diastereoselectivity is the ester-enolate Claisen rearrangement reported by Kazmaier. All examples reported by Kazmaier involved internal alkynes (27, R 3 = alkyl, in Scheme 5) and consequently, all of the allenes contained an alkyl group at the proximal double bond of the allene (i.e., trisubstituted allenes). Therefore, our initial efforts focused on reproducing Kazmaier's protocol by preparing trisubstituted allenes. Propargylic esters 41a and 41b were obtained by coupling the corresponding acids 39a and 39b with alcohols 40a and 40b using DCC and DMAP in 88 and 77% yield, respectively (Scheme 8). Claisen rearrangement of 41a using the reported conditions (LDA, ZnC12, THE - 7 8 ~ to room temperature) proceeded to give the intermediate allenic acid, which was converted to the methyl ester 42a in 22% overall yield by treatment with MeI and KHCO 3. This low yield was surprising since neither of these reactions revealed byproducts by TLC. Indeed, when we performed the Claisen rearrangement reactions without purifying the intermediate acids, the

9

339


yields improved. Applying this strategy to the preparation of 42b resulted in 73% yield after purification. H ='=R2

R3 R1

1.LDA,ZnCI2,THF,-78~ to rt

4 0 a or 4 0 b

PHN.~OH O

DOC, DMAP,.. P H N " ~ O CH2CI2, rt

2.KHOO3,MeI,DMF,rt.

R3 MeO2' C ~ N l l p

O

39a P = Boc, R 1 = Bn 39b P = Cbz, R 1 = Me

41a

4 2 a - 2 2 % : P = Boc, R 1 = Bn,

- 88%

41b - 77% 42b

'~~C6H

14

R2= n-Hex, R3= Me - 73%: P = Cbz, R 1 = Me, R2= i-Pr, R3= n-Bu

04H1~~ ' ~ ~ ~

OH

OH

40a

40b

SCHEME 8

We were also interested in preparing 1,3-disubstituted allenes using this protocol. To this end, propargylic ester 41e was prepared in 97% yield from N-Cbz-alanine 39b and 3-butyn-2-ol (Scheme 9). Applying the two-step reaction sequence to this substrate resulted in the corresponding allenyl-amino ester 42e in 70% yield as a mixture of diastereomers in a 1:1 ratio as determined by ~H NMR. This result was disappointing since the high diastereoselectivity is considered the major advantage of utilizing Kazmaier's protocol. Since all examples of the Claisen rearrangement proceeding with high diastereoselectivity reported by Kazmaier contain an internal alkyne, the lack of diastereoselectivity in the case of 42e is attributed to the terminal alkyne. Nevertheless, the origin of this effect is not clear.

H Q

DCC, DMAP

CbzHN

~.,.. /

..OH O

39b

1.LDA, ZnCI2,THF,-78~ to rt 2.KHCO 3, MeI,DMF,rt

CH2012, rt, 97% C b z H N ~ O I I O 41c

SCHEME 9

70%,dr=1 9 1

MeO2C

NHCbz 42c

340

KAY M. BRUMMONDAND BRANKO MITASEV

To circumvent this problem, disubstituted allenes were prepared diastereoselectively by utilizing a trimethylsilyl (TMS) group to temporarily functionalize the alkyne terminus (Scheme 10). Following the original protocol, a solution of 41d in THF (kept at room temperature) was added to a solution of LDA at - 7 8 ~ followed by addition of ZnC12 (0.5 M in THF). The resulting Zn-enolate was then warmed to room temperature affording the intermediate carboxylic acid after aqueous work-up, which was converted to the methyl ester by addition of MeI and KHCO 3. Removal of the allenyl TMS group was accomplished by treatment of 43d with tetra-n-butylammonium fluoride (TBAF) in presence of a phosphate buffer (pH = 7.0) to give the disubstituted allene 42d in 49% yield for the three steps. 55 This yield was reproducibly obtained when the three steps were performed without purification of the carboxylic acid and allenyl-TMS intermediate 43.

TMS\ ~I~R2

H .-- R2 1. LDA, ZnCI2,THF, -78~ to rt 2. KHCO 3, Mel, DMF,rt ,- MTeM:c~,e.~NR~ TBAF' pH = 7.0 b uffe r, ~ THF

PHN@ 0 0

H .-- R2 H ~.e. ,,~R1 MeO2C ~'NHP

43d 43e 43f

41d P = Cbz, R 1 = Me, R 2 = Me

41e P-- Cbz, R1 = Me, R2= i-Pr 41 f P = Boc, R1 = Bn, R2= Me

42d - 49% (three steps) 42e - 48% (three steps) 42f - 60% (three steps)

SCHEME l0

Allene 42d was obtained as nearly a single diastereomer (diastereomer ratio of --~95:5 was determined by 1H NMR). The relative stereochemistry of the major diastereomer was assigned as syn, in accordance with Kazmaier's results. 54 This route was then applied to the synthesis of allene 42e (48%), which contains an isopropyl group at the terminal position of the allene. Attempts to prepare the Boc-protected allenic aminoester 42f using this protocol led to the formation of byproducts. It was reasoned that formation of unidentified byproducts in the Claisen rearrangement step could be minimized by keeping the temperature of the reaction at - 7 8 ~ during the addition of the propargyl ester 411' to LDA. Indeed, it was found that by cooling the THF solution of 411' to - 7 8 ~ and adding it to a solution of LDA simultaneously with ZnC12 (0.5 M in THF), the yield increased from ---45 to 60% for the three-step sequence (the protocols for the formation of the methyl ester and TMS removal were kept identical to that for the preparation of 42d and 42e).

9

341


Presumably, simultaneous addition of ZnCI 2 and the substrate to LDA is advantageous due to immediate formation of the stabilized Zn-enolate, which minimizes side reactions resulting from exposure of the propargyl ester to excess LDA. The next goal was to develop a general procedure for N-alkylation of the allenic aminoesters that would introduce the alkyne component of the precursors for transition metal-catalyzed carbocyclization reactions. It was quickly found that treatment of the amides or carbamates with Nail in DMF at room temperature for 2-5 min, followed by addition of the corresponding propargylic bromide, resulted in clean N-propargylation. Using this protocol on the Bz-protected substrate 34 and 1-bromo-2-butyne gave allenyne 44a in 83% yield (Scheme 11). We were interested in

a5

R4 PHN / ~ /~ . "*~...,,R3 MeO2C R1 "~" R2

Br\ R5 Nail, DMF, rt

PN MeO2C R

~R2

44a-I

a5 BzN MeO2CB'/X Bnn~e~,,H

/

R5

BzN M eTBO SO2 C ' ~ ~ |

CbzN ,,H

MeO2C'"~,I,-, \ ' ~ '"~

44a : R5 = Me, 83% 44b :R 5 = H, 73% 44c :R5= TMS, 75%a 44d :R 5 = Ph, 89% CbzN MeO2C',"M ' ~ e ~ e ~ H 44j: 86%

44f : R5= Me, 84% 44g: R5= H, 77% 44h : R5= TMS, 86%a 44i: R 5 = Ph, 68%

44e :94%

C bzN,v..~n-Bu MeO2C"~Vle " ~ 44k: 78%

a Conditions: KH, THF. SCHEME 11

,H L

B~ "

~

MeO2C'" ~Bn " ' ~ , ,H 06H13 441: 73%

342


varying the substitution pattern of the alkyne in order to study the scope of the carbocyclization reactions. Therefore, a terminal alkyne was incorporated via N-alkylation with propargyl bromide affording 44b in 73% yield. Alkylation with 3-phenyl-l-bromopropyne gave a phenyl-substituted allenyne 44d in 89% yield. Attempts to prepare precursor 44e with a TMS group on the terminus of the alkyne led to desilylation, which was attributed to the presence of NaOH in the bulk Nail. This problem was circumvented by utilizing KH (in mineral oil) as a base and THF as a solvent, which gave 44e in 75% yield. The same protocols were applied to synthesize the Cbz-protected, alanine-derived substrates 44f-i in yields ranging from 68 to 86%. These two sets of allenynes were envisioned to serve as main model systems for studying the transition metal-catalyzed reactions and subsequent diversification of the scaffolds. We were also interested in examining the effect of allene substitution on the carbocyclization reactions so allenynes 44e, 44j, 44k and 441 were synthesized in 73-94% yields. III. Allenic AIder-ene Reaction Affording Amino-Ester Tethered Cross-Conjugated Trienes

Our synthetic investigations started with the Rh(I)-catalyzed allenic Alder-ene reaction. When allenyne 44a was submitted to the optimized reaction conditions (5 mol% [Rh(CO)2C1]2, toluene), cycloisomerization proceeded at room temperature in less than 10 min to give the expected cross-conjugated triene 45a in 80% yield (Scheme 12).5~"The structure of triene 45a was assigned based on the characteristic olefin resonances in the ~H NMR spectrum. Triene 45a was obtained as a single isomer of the exocyclic olefin, which is assigned Z-geometry in accordance with previous examples and mechanistic studies we reported. 5~ The scope of this transformation was next investigated by subjecting allenynes 44b-441 to the same reaction conditions. First, only allenynes substituted with a methyl group on the terminal allenic position were tested to avoid formation of E/Z isomers of the appending olefin. In all cases, the reaction proceeded in 10 min to afford the cross-conjugated trienes 45b-45i (Table 1). The reaction conditions were compatible with either a Bz- (entries 1-4) or Cbz- (entries 5-8) protected amine. Allenynes 44b and 44g with a terminal alkyne reacted to give 45b and 45g in 74 and 84% yield, respectively (entries 1 and 6, Table 1). Substitution of the alkyne terminus with either a TMS (entries 2 and 7, Table 1) or phenyl group (entries 3 and 8) resulted in the corresponding trienes in yields

9

343


//,.,••"f

5 mol% [Rh(CO)2CI]2

BzN

toluene, rt, 10 min, 80%

Bz. N ~ [ . - . MeO2C~'~~

Ha Hb

MeO2CB~nn~ i ' ~ , ,H 44a

45a He

Hb

Lo assigned resonance

chemicalshift ( p p m )

Hb He Ha Hd Hc

splitting/coupling constant

6.36 5.75 5.63 5.50 5.27

dd, J= 17.2, 10.7Hz s q, J = 7.0 Hz dd, J = 17.2, 1.5 Hz dd, J= 10.7, 1.5 Hz

SCHEME 12 TABLE l Scope study of the Rh(I)-catalyzed allenyl Alder-ene reaction

.R2

R2 5 mol% [ah(CO)2CI]2

P-NR~ ~o ~ MeO2C ,,H

toluene,rt,10 min

P\N'~ ~

MeO2C. _ ~ ~ , R1

44b-i Entry l 2 3 4 5 6 7 8

II

45b-i

Allenyne

P

R~

R2

Triene

Yield (%)

44b 44c 44d 44e 44f 44g 44h 44i

Bz Bz Bz Bz Cbz Cbz Cbz Cbz

Bn Bn Bn -CH2OTBS Me Me Me Me

H TMS Ph Me Me H TMS Ph

45b 45c 45d 45e 45f 45g 45h 45i

74 92 81 89 81 84 87 95

344


ranging between 81 and 95%. Allenynes with a methyl (entries 5-8), benzy! (entries 1-3) and silyloxymethylene (entry 4) group in the amino acid side chain R ~also reacted without event. In all of the examples studied, the triene is the only product observed in the reaction and can be easily separated from the Rh-containing impurities by filtering the reaction mixture over a short plug of silica gel. The Bz-protected allenynes (44a-44e, entries 1-4, Table 1 and Scheme 12) were subjected to the reaction conditions as mixtures of diastereomers (dr --~1.7:1) while the Cbz-protected allenynes (entries 5-8) were nearly single diastereomers (all compounds are racemic). Nevertheless, all reactions resulted in the corresponding crossconjugated triene as a single isomer indicating that the exocyclic olefin geometry is not related to the relative stereochemistry of the allenyne, but is a result of the last reductive-elimination step in the mechanism of the reaction. The rate at which the reaction of these aminoester-tethered substrates proceeded was noticeably higher when compared to previously reported examples, some of which required up to 6 h for completion. 5~ The increased reactivity of the aminoester substrates is likely a result of a Thorpe-Ingold effect imposed by the quaternary center adjacent to the allene. 56 In addition, the carbomethoxy group may play an activating role by reversibly coordinating to the metal center. 5v Next, allenynes possessing different substituents on the allene moiety were tested (Scheme 13). Allenyne 44j, substituted with an isopropyl group at the terminal allenic position, reacted to afford triene 45j possessing a trisubstituted appended alkene in 95% yield. In this case, there is only one hydrogen atom that can undergo [3-hydride elimination. This does not affect the rate or the yield of the reaction, since 45j was produced in 95% yield after 10 min. Trisubstituted allenes 44k and 441 underwent the Alder-ene reaction affording 45k and 451 in 78 and 80% yield, respectively. In the case of 451, only the E isomer of the appended alkene was observed, characterized by a coupling constant of 16.0 Hz for the vinyl hydrogens in the ~H NMR spectrum. This stereoselectivity is in contrast to earlier examples from the Brummond group, where Rh(I)-catalyzed reaction of alkyl allenes resulted in mixture of appended alkene isomers in E/Z ratios in the range of 3-6:1. The geometry of the appended alkene is determined during the 13-hydride elimination step of the reaction, which requires a coplanar arrangement of the Rh-C bond and the C-H bond that is being

9

RHODIUM-CATALYZEDCYCLOISOMERIZATION REACI'IONS OF ALLENES

5 mol % [Rh(CO)2CI]2

345

P\N -

toluene,rt, 10 min

MeO2C

R R11

R1

L 2 Na

Ra

44j (P = Cbz, R 1 = Me, R2 = R3 = Me, R4 = H) 44k (P = Cbz, R 1 = Me, R2 = R3 = Me, R4 = n-Bu) 441 (P = Boc, R 1 = Bn, R2 = H, R3 = C5H11, R4 = Me)

BOC~N.~

CbZ-.N.~ M e O ~ . ~ l

MeO2~

45j 95%

45k 78%

MeO2C-~.....~ 13nl 1.L...~~ 451 80%

(E isomer only) SCHEME 13

broken. Therefore, [3-hydride elimination in the formation of 451 can occur via rotamers A and B as shown in Scheme 14. Rotamer B, which would lead to the Z-olefin isomer, posseses an unfavorable steric interaction between the appending alkyl group and the methyl substituent on the ring and is strongly disfavored. Therefore, 13-hydride elimination via rotamer A leads to the E-isomer 451, exclusively. In summary, pivotal .allenynes possessing nitrogen and oxygen heteroatoms have been prepared using either the Castelhano or Kazmaier Claisen rearrangement protocols. The former provides benzamide-protected allenynes in high yields as mixture of diastcreomers. The latter gives N-Cbz and N-Boc protected allenynes in moderate yield as single diastereomers. All pivotal allenynes 44a-1 underwent the Rh(I)-catalyzed cycloisomerization reaction in minutes to give trienes 45a-1 in high yield. Although the triene products themselves may serve as useful biological probes, we were concerned about the reactivity of the triene moiety in biological systems. Instead, it was reasoned that this reactivity can be exploited toward the efficient assembly of more complex molecular scaffolds. Some of these efforts are described next.

KAYM. BRUMMONDANDBRANKOMITASEV

346

/

Boc-N

4 Bn "~ "H CHaHbR 441 I Rh(I) BocN/"~Rh L

BocN'~'T~Rh Ln R

A - favored

B - disfavored

-

BocN"~H b

BocNAx~"~Ha

MeO2C~H

MeO2C~H

BnlHbt/J~C5H11 (Z)-451

(E)-451

notobserved SCHEME 14

IV. Diversification of Cross-Conjugated Trienes via Diels-Alder Reactions: First Generation Triene

Sequential Diels-Alder reactions of acyclic cross-conjugated trienes 46 in order to give functionalized decalin systems 48 were initially studied by Tsuge, 58 utilizing bis-silylenolether 49 as the triene, and Fallis, using monosubstituted triene 50 (Scheme 15). 59 However, tandem Diels-Alder reactions with these acyclic trienes are difficult to control and typically afford complex mixtures of regioisomers. In addition, the synthesis of the acyclic triene starting material is not straightforward. This may, in part, be due to their instability and tendency to polymerize. 6~ Thus, despite their potential, synthetic applications of these compounds have been limited. 62 In a related example, Sherburn and coworkers recently reported that an acyclic cross-conjugated tetraene ([4]dendralene-51) can participate in a tandem Diels-Alder reaction. 63 For example, the reaction of 51 with excess N-methylmaleimide affords a mixture of mono-, di-, and tri-cycloaddition

9

347


E E

[4+2]

[4+2] "

"

E

~

E

E E

E 46

E

47

48

so# Ph~OTMS

R

49

50

51

SCHEME 15

products. Nevertheless, this example underscores the fact that rapid increase in molecular complexity can be obtained via tandem cycloaddition reactions of cross-conjugated polyenes. It was reasoned that the cyclic trienes 45a-i would not pose regioselectivity issues in these tandem cycloaddition reactions because one diene is locked in an unreactive s-trans conformation. Therefore, we became interested in exploring the feasibility of the cycloaddition pathway illustrated in Scheme 16. Furthermore, exploring these pathways offered an opportunity to study the reactivity of cyclic cross-conjugated trienes as novel chemical entities. Finally, it was reasoned that these rigid and conformationally-defined polycycles would serve as interesting biological probes. 64 R2

R3

~

P\N

R3

R1 MeO2C

[4+2]

II 45

./.~

P'N

R2

O~.,,. 2l~'~ ~ ~'~ R1 Me a3" " ~ R3 52

P = C bz or Bz SCHEME 16

]~ R4 R4 ,, [4+2]

R2

P.~ R1 1 MeO2 R3

R4 R4

R3 53

348

KAY M. BRUMMONDAND BRANKOMITASEV

Our investigations began with the reaction of triene 451' with Nphenylmaleimide (Scheme 17). Intermediate Diels-Alder cycloadduct 54 was not isolated and, instead, immediately underwent a second Diels-Alder reaction to afford a 83% yield of pentacycle 55 as a 5:2:1 mixture of diastereomers. The ratio of diastereomers was determined by HPLC, which allowed for their complete separation. 0

~ MiboZ~//2C

0

m

_

N-Ph

o

CbZ..N~J/

0 II toluene,reflux,2hr M e 0 2 ~ 83%

0

45f

. Me 0 2 ~

....~0

~ -

54

55 5 92 91 mixture of diastereomers

-

(major diastereomer shown) S .....

4

,,~.

,%

....~

.::

.....~

~

.,~. :,~.~.,-........ 9 .

.

X-raY;55ctu re

SCHEME 17. (See color insert.)

The relative stereochemistry of the major diastereomer, as determined by X-ray crystallography, results from endo approach of the dienophile from the same face of the triene as the methyl group in the first cycloaddition, while the second equivalent of dienophile approaches in endo mode from the less-hindered convex face of the newly formed diene (Scheme 17). All attempts to isolate cycloadduct 54 by reaction of 45t' with an equimolar amount of the dienophile, gave tandem Diels-Alder cycloadduct 55 and recovered triene. This result can be attributed to the higher reactivity of the diene of 54, compared to the starting triene,

9


349

since 54 is locked in an s-cis conformation. Other dienophiles (maleic anhydride and 4-phenyl-[1,2,4]-triazole-3,5-dione) also reacted with 45f to give mixtures of diastereomeric products similar to 55. Although the cycloaddition reaction of 45f affords a complex molecular scaffold in a rapid manner, obtaining the product as a mixture of diastereomers was discouraging. Furthermore, biological testing of compounds as diastereomeric mixtures is not ideal due to variability in the assay concentrations. Even though separation of the diastereomers by HPLC was feasible, it would be costly and time consuming when preparing a larger-scale library. Therefore, controlling the chemo- and diastereoselectivity of the Diels-Alder reaction of the triene was important, and a new strategy for tandem intermolecular cycloaddition was considered. V.

Design and Synthesis of a Second Generation Triene

Our efforts to control the selectivity of the Diels-Alder reactions focused on designing a new triene. It was reasoned that constraining the appended ester as part of a ring would reduce the steric bulk from the C2 position and, therefore, increase the reactivity of the diene involved in the first Diels-Alder reaction. Moreover, the cyclic constraint of the ester may block one face of the sterically biased triene with the R ~ group. Finally, the rate of the second Diels-Alder reaction could be slowed by placing an electron withdrawing carbonyl group at the C6 position. Structures such as the novel imidazo-pyridinone triene 58 address all of these issues (Scheme 18). Putting the synthesis of 58 into practice required examination of the Rh(I)-catalyzed cycloisomerization of amide-tethered allenyne 56 to form 8-1actam triene 57. Traditionally, lactams are synthesized via carbon-nitrogen bond formation. For example, lactams are formed via dehydration of amino acids, 65 by cyclization of an amide onto an alkene, 66 alkyne 67 or an allene, 68 and intramolecular vinylation of amides. 69 Alternatively, lactams can be synthesized from ketones by a Schmidt or Beckmann rearrangement, v~ There are very few examples of lactam syntheses via transition metal catalyzed carbon-carbon bond formation, and most involve a ring-closing metathesis. 7~ Synthesis of lactams via cycloisomerization reaction appears particularly attractive, since additional functionality is generated in the course of the reaction (a cross-conjugated triene in this case). There are a few examples of lactam formation via cycloisomerization reactions, and they are strictly limited to preparing y-lactams. For example, in 1999,

350


carbonyl slows the rate of the second cycloaddition

control of facial

selectivity ~ J ~

O

l

~

improved reactivity of the diene due to reduced steric hinderance at C2

0

0

HN

....

56

-~

R2

0 - - - "~

57

R3 -

N

i

R2

l

58

SCHEME 18

Lu reported a Pd(0)-catalyzed tandem cyclization/amination of dienyne 59 leading to ~-alkylidene-~,-lactam 60 (Scheme 1 9 ) . 72 More recently, Zhang reported an enantioselective Rh(I)-catalyzed cycloisomerization of amide-tethered enyne 61, affording y-lactam 62 with >99% ee. 73 Notably, both reports used a benzyl protected amide, and Zhang reported that the reaction did not proceed with the unprotected amide. To test the feasibility of a Rh(I)-catalyzed formation of ~i-lactams, amides 56a-56d were synthesized by Boc-deprotection of amine 42f to give 63 in 85% yield, followed by coupling with alkynoic acids 64a-d (Scheme 20). Employing a DCC/DMAP coupling protocol (conditions A) proved useful in preparing amides 56b and 56d in sufficient amounts for testing the subsequent cycloisomerization reaction (~100mg). Nevertheless, this protocol resulted in the formation of byproducts including dicyclohexylurea, which made the purification difficult and the yields irreproducible. To circumvent this issue, an alternative protocol was applied for the preparation of 56a and 56e. Treatment of the alkynoic acid with isobutyl chloroformate and N-methylmorpholine afforded a mixed anhydride, which was treated in situ with amine 63 to give the

9

~

O.

H 10 mol% Pd(OAc)2

B

351

RHODIUM-CATALYZEDCYCLOISOMERIZATIONREACTIONSOF ALLENES

n

-

N

+

Phi

+

Bn-N.~

"

Ph

20 mol% PPh3 MeCN, 80~

59

60

(•

O [Rh(cod)CI]2,(R)-BINAP

Bn-N

,..

"~

Bn--N

AgSbFs,rt,91%,>99%ee

61

(-)-62

SCHEME 19

corresponding amides. This proved a robust protocol that reproducibly gave 55-85% yield of the amides. O H O ~ BocH N. /..~ Meo~c,,y -.~H Bn

--

CF3COOH, CH2CI2

/111 10 min, 85%

H2N~, " MeO2C"B~n'~,H

o .),~.... R

R

64a-d

HN =" Me02C B n ~ e ~ ,'H conditions A or B

63

42f

entry

R

56a-d conditions

product

yield

1

Me

B

56a

85%

2

H

A

56b

82%

3

TMS

B

56c

72%

4

Ph

A

56d

55%

conditions A: DCC,DMAP,CH2CI2, rt. conditions B: /-BuOCOCI,NMM,-10oC, then amine 63.

SCHEME 20

With allenynes 56a-d in hand, the cycloisomerization reaction was tested (Table 2). 74 When 56a was subjected to the optimized conditions for triene formation (5 mol% [Rh(CO)=C1] 2, toluene, 0.3 M), reaction did not occur at room temperature (entry 1). Instead, triene formation was effected by heating 56a to 90 ~ With this information in mind, the

352

KAY M. BRUMMOND AND BRANKO MITASEV TABLE 2 Optimization and scope study of the Rh(I)-catalyzed Alder-ene reaction of propiolamides affording 6-1actams

~ R [Rh(CO)2CI]2 MeO20~o HNI~n ~i "H

toluene(0.03M)' 90~ '-

56a-d Entry 1 2 3 4 5 6 7

0 R HN"J~ "j MeO2C _ @ ~ Bn II 57a-d

Allenyne

R

Catalyst (tool%)

Time (rain)

Triene

Yield (%)

56a 56a 56a 56a 56b 56c 56d

Me Me Me Me H TMS Ph

5 2.5 5 l0 10 l0 l0

90 45 30 120 15 30

57a 57a 57a 57a 57b 57c 57d

0~ 18b 47 b 92 45 77 66

~Condition" 5 tool% [Rh(CO)2C1], toluene, rt. bIncomplete reaction" starting material was recovered.

catalyst loading was varied between 2.5 and 10 mol% (entries 2-4) to establish its effect on the efficiency and yield of the reaction at 90 ~ All reactions were performed in toluene at 0.03 M concentration since it was found that increasing the concentration led to the formation of by-products and a lower yield of the triene. TM With 2.5 mol% of the catalyst, the reaction was relatively sluggish and consumption of the starting material was incomplete after 90 rain, affording triene 57a in only 18% yield (along with ---50% recovered starting material). Increasing the catalyst loading to 5 mol% resulted in a shorter reaction time of 50 min and increased the yield of the triene (47%). A further increase in catalyst loading to 10 mol% resulted in complete consumption of the starting material in less than 30 min, and 92% yield of the triene 57a (this reaction was performed on --~1 g of 56a, demonstrating the scalability of the reaction). Next, the cycloisomerization of propynamide 56b was tested with 10 mol% of catalyst. Consumption of the starting material occurred after 2 h and resulted in the isolation of 57b in only 45% yield (entry 5). Trimethylsilyl (56e) and phenyl (56d) substituted propynamides also underwent the cycloisomcrization reaction to afford the corresponding trienes 57e and 57d in 77 and 66% yield, respectively (entries 6 and 7). Notably, the reaction of the TMS-substituted alkyne required only 15 min.

9

353


The higher temperature required to effect the cycloisomerization of these unprotected amides can be attributed to a preferred trans-conformation of the secondary amide, placing the reactive termini away from each other (Scheme 21). 75 To confirm this, N-benzoyl protected allenyne 56e was synthesized in 69% yield by treatment of 56a with BzC1 at 70 ~ (Scheme 22). The cycloisomerization reaction of this precursor using 10 mol% of [Rh(CO)2C1]2 occurred in less than 1 h at room temperature, affording cross-conjugated triene 57e in 75% yield.

__~

~--NH ...~\

O,~

Rh(I),90~

HN

,~ ../S \;~., 'v'~"J2L' Bn

'H

MeO2C Bn "~,' 'H

MeO2CBn~

|

56a trans-amide

56a cis-amide

favored

disfavored

57a

SCHEME 21

~. / / / .

BzCI,Etmm,DMAP

0 Ph~ / ~ ~ ' / ~

Ph 0 I 0 mol o/o[RH(CO)2CI]2,. 0..~ m / ~ , ,

moecuarseves

Bn 56a

'

Bn

69% 56e

II

57e

SCHEME 22

VI. Synthesis and Stereoselective Diels-Alder Reaction of Novel Bicyclic Trienes

With a synthetic route to lactam 57a, we next moved toward the synthesis of a facially differentiated precursor for the Diels-Alder reaction. Saponification of the methyl ester in 57a occurred within 5 min upon treatment with LiOH. After aqueous work-up, the acid was immediately coupled with glycinemethyl ester using EDCI, HOBt and DMAP to give diamide 65a (Scheme 23). In order to complete the synthesis of hydantoin 58a, amide 65a was reacted with phosgene (COC12). Unexpectedly, imino-oxazolidinone 66a was the only product isolated (55% yield for three steps) but was expected to provide the same steric and electronic control elements as the hypothetical hydantoin 58a. The structural assignment of 66a was made later based on an X-ray crystal structure of the

354


Diels-Alder product 67a, vide infra. An additional substrate functionalized as an isobutyl amide 66b was prepared using the same protocol in 41% yield over three steps. To our knowledge, oxazolidinones with this substitution pattern have not been reported. 76

1.LiOH,THF/H20,rt. 2.RCH2NH2,EDCI,DMAP HOBt,CH2CI2,rt.

O HN MeO2C Bn 57a

II

, ~ ~ H HN ( R

Q ~,~ ,,-~'~'N

phosgene,Et3N CH2CI2,_10oC. ~

"

/,

"~

O R

65a,R =-CO2Me 65b,R = i-Pr

66a, R =-CO2Me 55% yield (3 steps) 66b, R = i-Pr 41% yield (3 steps)

O

O

=eO2J 58a or 58b not observed SCHEME 23

We suspected that the imino-oxazolidinone 66a resulting from Oacylation of the appending amide was a kinetic product of the reaction, and a consequence of the preferred conformation of the amide side chain in the precursor 65a. To confirm this, computational modeling of 65a was performed using Cache. 7v The energy-minimized model placed the amide oxygen 02 and lactam nitrogen N1 of 65a in the same direction, confirming the observed reactivity (Figure 1). Finally, it should be noted that the newly obtained imino oxazolidinone ring in 66a and 66b proved relatively sensitive to silica gel; prolonged exposure during chromatography led to ring opening giving the diamides 65a and 65b, respectively. The newly obtained bicyclic triene 66a was also modeled in order to visualize its three-dimensional structure, vv As illustrated in Figure 1, the fused bicyclic structure is relatively planar due to the presence of multiple spZ-hybridized atoms and the bottom face is blocked by the benzyl substituent. Therefore, cycloaddition reaction with the diene was expected to occur selectively with the top face (all compounds are racemic). To test this hypothesis, triene 66a was reacted with N-phenylmaleimide (1.3 equiv.) in toluene. Reaction occurred in less then 1 h at 90 ~ to afford the cycloadduct 67a in 73% yield as a single diastereomer

9

355


o'

o

4

/--N ~"' 002 Me

II

M e02C

65a

66a

;

.

..~

FIGURE 1. Cache minimizations of 65a and 66a. (See color insert.)

(Scheme 24). An X-ray crystal structure of 67a confirmed that the [4 + 2] cycloaddition occurred with endo-selectivity and the diene approached from the face opposite the benzyl group. Attempts to effect a tandem cycloaddition reaction of 66a by using excess dienophile still gave only 67a. Similarly, cycloaddition of 66b with N-methylmaleimide resulted in formation of 67b in 95% yield as a single diastereomer. Additional experimentation demonstrated that this cycloaddition is limited mainly to using maleimides. 78 Other dienophiles (diethylfumarate, p-benzoquinone and dimethylacetylene dicarboxylate) generally gave low yields of the cycloadduct (~50%). The X-ray crystal structure of 67a revealed several interesting features. As a result of the endo-cycloaddition occurring from the concave face of 66a, the product adopts a folded shape with the N-phenylpyrrolidinone moiety projecting directly above the new diene (Figure 2). Moreover, the two double bonds of the 1,3-diene are twisted out of planarity as evidenced by a dihedral angle of--~40 ~ In addition, the ~,13-unsaturated amide is also twisted in the opposite direction by a dihedral angle of ~42 ~ These steric and electronic features of the new diene significantly lower its reactivity toward electron-poor dienophiles. For example, the second cycloaddition could not be effected using electron deficient

356


a2 i

-

O.

N/R 2 -

o_ ~ _ ~ o R1 o ~ ~.;_;#~

1.3 equiv.

\N

toluene, 1h,90~

;--',-

67a, R 1 = CO2Me,R 2 = Ph, 73% 67b, R 1 = i-Pr, R2= Me, 95%

66a, R 1 =-CO2Me 66b,Rl= i-Pr

SCHEME 24

......,i) .."~ ......3

....

a.

....

-

ph/N

"iD

67a

"--~',-----",:

torsional angle of diene = 39.9 ~

,

torsional angle of o~,13-unsaturated amide = 42.4 ~

FIGURE 2. X-ray crystal structure of 67a. (See color insert.)

dienophiles, except with diethyl fumarate. 78 Therefore, ethyl vinyl ether was examined as a small, electron rich dienophile to better match the character of the diene. Heating 67a in a mixture of toluene/ethyl vinyl ether at 90 ~ afforded 70% yield of pyran 68a as a single diastereomer. Few inverse electron demand hetero-Diels-Alder reactions of ~,13unsaturated amides have been documented, and they generally result in the formation of an aromatic compound (e.g., indole, thiazole, pyrazole). 79 Since lanthanide Lewis acids (e.g., Eu(fod) 3) can been used to catalyze hetero-Diels-Alder reactions, we tested this reagent on the cycloaddition of 67a and ethyl vinyl ether, s~ With 10 mol% of Eu(fod) 3, the reaction proceeded at room temperature, giving 68a in 95% yield as a single diastereomer (Scheme 25). Hydrolysis and isomerization of pyran 68a afforded aldehyde 69a.

9

357


OEt

.o

t

0

,

HO,

Me02C ph/N---'~0 67a

10 mol%Eu(fod)3 M 1,2-dichloroethane rt,95%

Ph/N'~;

ph/N%

68a

69a

SCHEME 25

Next, hydrolysis of the oxazolidinone moiety in 67a to the parent diamide was explored as a means to introduce structural diversity in the products and increase the number of hydrogen-bond donors. It was anticipated that this transformation would increase the water solubility of these compounds, and improve their pharmacological profile. 8 When 67a was heated to 70 ~ in 1 M HC1/dioxane (1:1) for 1 h, only starting material was recovered in ---80% yield. The stability of this imino oxazolidinone to acidic conditions is in sharp contrast to the bicyclic triene 66a, which is readily hydrolyzed in presence of aqueous acid. Next, we attempted to cleave the imino oxazolidinone under Lewis acid conditions (BF3-OEt 2 and Me2S), 81 which also gave recovered starting material in 86% yield. On the contrary, treatment of 67a with LiOH in THF/H20 caused complete decomposition. 82 Because strongly basic conditions caused decomposition of the starting material, primary amines were examined as weaker bases/nucleophiles. To this end, a solution of 67a in CDC13 was treated with benzyl amine, and the reaction was followed by ~H NMR. Although cleavage of the oxazolidinone moiety was evident by the appearance of new amide and urea N-H resonances in the downfield region (8-9 ppm), gradual disappearance of both olefinic peaks of the diene was also observed. Based on these observations, it was speculated that the final product of the reaction was 71a resulting from cleavage of the imino oxazolidinone and 1,4-addition of benzylamine to the ~,[3unsaturated amide, and subsequent isomerization of the remaining olefin into conjugation within the ~-lactam ring (Scheme 26, via intermediacy of 70a). Unfortunately, 71a was obtained as a mixture of diastereomers. Since this 1,4-addition side process was not desired, reduction of the diene in 67a and 67b was attempted using Pd/C and H 2 (1 atm). Interestingly, reduction of either substrate after 4 h at room temperature,

358

KAY M. BRUMMOND AND BRANKO MITASEV m

O

m

Bn"NH O

O

-NL/

BnNH2 (10 equiv.) CDCI3, rt

/--N

olefin

MeO2

NHBn

isomeriza%

O n

"

/

N-'"-~ _H II

NHBn

~

MeO2C

67a

70a

71a

SCHEME 26

gave the 0~,[3-unsaturated amides 72a and 72b in 80% and 95% yield, respectively (Scheme 27). This result was attributed to the steric hindrance around the diene. Presumably, reduction of the more accessible exocyclic double bond led to an intermediate [3,7-unsaturated amide followed by isomerization of the remaining olefin into conjugation. Next, solutions of 72a and 72b in CDC13 were treated with a primary amine (isobutylamine and allylamine, respectively), which clearly effected opening of the imino-oxazolidinone to ureas 73a and 73b. In addition to isobutylamine and allylamine shown in Scheme 27, benzylamine and 2-methoxyethylamine were also used to afford ring opening products in > 80% yield. a3

0

0

03\

Oii

NH

0

,o 67a, R 1 = CO2Me, R 2 = Ph 6 7 b R 1 = i-Pr, R 2 = Me

72a, 80% 72b, 95%

73a, R 1 = CO2Me, R 2 = Ph, R3= i-Bu, 75% 73b, R 1 = i-Pr, R 2 = Me, R 3 = Allyl, 92%

SCHEME 27

The structure of 73a was assigned by ~H NMR (in CDC13) and was based on the presence of two downfield resonances at 9.02 ppm (t, J - 5.8 Hz, 1H) assigned to the urea N-H proton and 8.46 ppm (dd, J - 7.1, 4.6 Hz, 1H)

9


359

assigned to the amide proton. These N-H resonances were unusually sharp, suggesting an ordered secondary structure of the urea and amide side chains. The downfield chemical shifts of the amide and urea protons support the notion of intramolecular hydrogen bonding. 83 To examine this computationally, 73a was modeled using Cache. 77 The minimized structure of 73a resulted in arrangement of the side chains as shown in Figure 3, with two potential hydrogen bonds: (a) between the urea N-H and lactam carbonyl oxygen (distance 2.287 /k); and (b) between the appending amide N-H and the adjacent pyrrolidine-dione carbonyl oxygen (distance 2.140 ,~). As a result of this secondary bonding, the two side chains are presumably held rigidly, which accounts for the sharp N-H resonances in the NMR spectrum. It was reasoned that this feature may be useful in designing biological probes that project functional groups in specific three-dimensional space. For example, many potent protease inhibitors are small peptide-like molecules that possess a defined secondary structure, resulting in strong interaction with the enzyme. 84 We anticipate that the rigid amido-ureas may also prove as useful biological probes to study protein function.

2.287,~,

"~NH

0

Me02C O " ~ n

-

o

Ph'

2.140A

73a

energy-minimized model of 73a; the two phenyl groups were replaced with a methyl group for clarity purposes FIGURE 3. Energy-minimized model of 73a; the two phenyl groups were replaced with a methyl group for clarity purposes. (See color insert.)

360


Our focus then shifted to designing a library of these polycyclic scaffolds using the synthetic protocols described above. As outlined in Scheme 28, this synthetic pathway offers at least five points of diversity to be introduced gradually as the complexity of the scaffold increases. The imino-oxazolidinone moiety was envisioned as a crucial part of the triene precursor 76 because it enables a highly stereo- and chemoselective Diels-Alder reaction with a number of maleimides. Furthermore, this moiety is used as a key diversity element, because the transformation of 77 to 78 results in the conversion of a molecule rich in hydrogen-bond

200 compounds of various scaffolds synthesized in solution phase 0

R\ NH

zR2

O

R2

R1 = Me, Bn R2 = Me R3 = various primary amines R4 = H, Me, Et, Ph, -CH2CO2Me R5 = various primary amines

HN

MeO2C---Xr~o. i~1 "~,,,H 74

78

Rh(I) l

t

a4

O R2 1. LiOH. ~..~ 2" R3NH2' EDCI' HN DMAP, HOBt. ,.

O O~

O ...~ N

R20.....~/1~1~..O

O

O

R2

3. phosgene, Eta N. Ra_ 76

75

77 R3 =

a5 _

H2N"~-.~

H2N

H2N'~'-~

H2N~CO2Me

H 2 N ~

H2N~CO2Me

H2N~

OMe

H2N~

SCHEME 28

OH

H2N N..J f"'O H2N~N'v

~

9


361

acceptors to one that contains two hydrogen-bond donor groups. This transformation is expected to bring about significant differences in the physico-chemical properties and potentially the biological activity of the compounds. Since both classes of compounds were envisioned as library members, a broad range of diversity was accessed via a relatively simple set of transformations. The library synthesis was put into practice by the staff at the University of Pittsburgh Center for Chemical Methodologies and Library Development (UPCMLD, http://ccc.chem.pitt.edu/). Using four points of diversity (allenic amino-ester, amine for oxazolidinone formation, N-alkyl maleimide and amine for the oxazolidinone opening reaction), the center synthesized 200 library members in quantities of 5-100 mg each. 85 These compounds are continuously being sent out for biological testing and are available to the academic and industrial researchers at no cost (see http://ccc.chem.pitt.edu/). VII.

Summary

In summary, we have successfully applied allenic Rh(I)-catalyzed cycloisomerization reactions to DOS. The overall goal of this study was to design robust and efficient methods for the synthesis of complex small molecules to be used as biological probes. As part of our broader DOS strategy, we have developed a synthesis of pivotal allenic amino-ester intermediates by utilizing a Claisen rearrangement of amino-acid propargyl esters. The scope of the allenic cycloisomerization reaction was thoroughly studied with respect to the substitution of the amino acid side chain, amine protecting group and the alkyne and allene reactive moieties. Consequently, we have gained efficient access to various polysubstituted cross-conjugated trienes that would be difficult to access via existing methods. Moreover, a cycloisomerization of amide-tethered allenynes was utilized to prepare novel 8-1actams. This achievement is important, since previously, only y-lactams have been available via transition metalcatalyzed cycloisomerization reactions. The resulting trienes that are accessible via these methods represent a novel class of compounds. Here it was demonstrated that these trienes can be utilized in sequential Diels-Alder reactions to gain rapid access to non-aromatic polyheterocyclic skeletons. The first generation of sterically and electronically undifferentiated trienes underwent the reaction without chemo- or stereoselectivity, thus affording products as mixtures of diastereomers. This obstacle was overcome by designing second generation bicyclic triene precursors that underwent a stereoselective Diels-Alder reaction to afford

362


tetracyclic skeletons. The synthetic route was then applied to the synthesis of a library of compounds, thus fulfilling the goal of DOS. Acknowledgments

We are grateful to the National Institute of General Medical Sciences (NIGMS P50GM067982) for generous support of this project. B.M. would like to thank the University of Pittsburgh for an Andrew Mellon Fellowship. We also thank Dr. Donald A. Probst and Dr. Bingli Yan for their contribution to this project.

References and Footnotes

1. Arkin, M. R., Wells, J. A., Nat. Rev. Drug. Discov. 2004, 3, 301. 2. (a) Crews, C. M., Mohan, R., Curt. Opin. Chem. Biol. 2000, 4, 47. (b) Crews, C. M., Splittgerber, U., Trends. Biochem. Sci. 1999, 24, 317. 3. (a) Lokey, R. S., Curt. Opin. Chem. Biol. 2003, 7, 91. (b) Schreiber, S. L., Science 2000, 287, 1964. (c) Schreiber, S. L., Bioorg. Med. Chem. 1998, 6, 1127. 4. (a) Lee, M. L., Schneider, G., J. Comb. Chem. 2001, 3, 284. (b) Kim, Y. K., Arai, M. A., Arai, T., Lamenzo, J. O., Dean, E. E, Patterson, N., Clemons, E A., Schreiber, S. A., J. Am. Chem. Soc. 2004, 126, 14740. 5. Fergus, S., Bender, A., Spring, D. R., Curt. Opin. Chem. Biol. 2005, 9, 304. 6. (a) Bunin, B. A., Ellman, J. A., J. Am. Chem. Soc. 1992, 114, 10997. (b) DeWitt, S. H., Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 6909. 7. Wess, G., Urmann, M., Sickenberger, B., Angew. Chem. Int. Ed. 2001, 40, 3341. 8. Lipinski, C. A., Lombardo, F., Dominy, B. W., Feeney, E J., Adv. Drug. Deliv. Rev. 2001, 46, 3. 9. For computational methods for predicting physico-chemical properties, see (a) van de Waterbeemd, H., Gifford, E., Nature Rev. Drug. Disc. 2003, 2, 192. (b) Jorgensen, W. L. Science 2004, 303, 1813. 10. (a) Burke, M. D., Schreiber, S. L., Angew. Chem. Int. Ed. 2004, 43, 46. (b) Arya, E, Joseph, R., Gan, Z., Rakic, B. Chem. Biol. 2005, 12, 163. (c) Burke, M. D., Bergen E. M., Schreiber, S. L., Science, 2003, 302, 613. (d) Beeler, A. B., Schaus, S. E., Porco, J. A. Jr., Curr. Opin. Chem. Biol. 2005, 9, 277. 11. (a) Wipf, P., Coleman, C. M., Janjic, J. M., Iyer, P. S., Fodor, M. D., Shafer, Y. A., Stephenson, C. R. J., Kendal, C., Day, B. W., J. Comb. Chem. 2005, 7, 322. (b) Hotha, S., Tripathi, A., J. Comb. Chem. 2005, 7, 968-976. (c) Hanessian, S., Kothakonda K. K., J. Comb. Chem. 2005, 7, 837-842. (d) Lei, X., Zaarur, N., Sherman, M. Y., Porco, J. A., Jr., J. Org. Chem. 2005, 70, 6474-6483. (e) Oikawa, M., Ikoma, M., Sasaki, M., Tetrahedron Lett. 2005, 46, 415-418. (f) Simon, R. A., Schuresko, L., Dendukuri, N., Goers, E., Murphy, B., Lokey, R. S., J. Comb. Chem. 2005, 7, 697-702. 12. For examples see the comprehensive survey of combinatorial library synthesis for 2005: (a) Dolle, R. E., Bourdonnec, B. L., Morales, G. A., Moriarty, K. J., Salvino, J. M., J. Comb. Chem. 2006, 8, 597. 13. Stockwell, B. R., Nature, 2004, 432, 846.

9


363

14. (a) Blaney, J. M., Martin, E. J., Curt. Opin. Chem. Biol. 1997, 1, 54. (b) Willett, E, Curt. Opin. Biotechnol. 2000, 11, 85. 15. For a recent example, see Spiegel, D. A., Schroeder, E C., Duvall, J. R., Schreiber, S. L., J. Am. Chem. Soc. 2006, 128, 14766. 16. Oguri, H., Schreiber, S. L., Org. Lett. 2005, 7, 47. 17. For recent examples, see (a) Cordeiro, A., Quesada, E., Bonache, M. C., Velazquez, S., Camarasa, M. J., San-Felix, A., J. Org. Chem. 2006, 71, 7224. (b) Wipf, E, Stephenson, C. R. J., Walczak, M. A. A., Org. Lett. 2004, 6, 3009. (c) Chiara, J. L., Garcia, A., Sesmilo, E., Vacas, T., Org. Lett. 2006, 8, 3935. (c) Aurrecoechea, J. M., Suero, R., de Torres, E., J. Org. Chem. 2006, 71, 8767. (d) Micalizio, G. C., Schreiber, S. L., Angew. Chem. Int. Ed. 2002, 41, 3272. 18. For comprehensive summaries of transition metal catalysis in the past decade, see (a) Hegedus, L. S., Coord. Chem. Rev. 1996, 147, 44. (b) Hegedus, L. S., Coord. Chem. Rev. 1997, 161, 129. (c) Hegedus, L. S., Coord. Chem. Rev. 1998, 168, 49. (d) Hegedus, L. S., Coord. Chem. Rev. 1998, 175, 159. (e) Hegedus, L. S., Coord. Chem. Rev. 2000, 204, 199. (f) S6derberg, B. C. G., Coord. Chem. Rev. 2006, 250, 300. (g) S6derberg, B. C. G., Coord. Chem. Rev. 2006, 250, 2411. 19. (a) Beletskaya, I. E, Cheprakov, A. V., Chem. Rev. 2000, 100, 3009. (b) Dounay, A. B., Overman, L. E., Chem. Rev. 2003, 103, 2945. 20. (a) Farina, V., Krishnamurphy, V., Scott, W., J. Org. React. 1997, 50, 1. (b) Duncton, M. A. J., Patterden, G., J. Chem. Soc., Perkin. Trans. 1 1999, 1235. 21. Negishi, E., Anastasia, L., Chem. Rev. 2003, 103, 1979. 22. Stanforth, S. E, Tetrahedron 1998, 54, 263. 23. Trost, B. M., Crawley, M. L., Chem. Rev. 2003, 103, 2921. 24. (a) Frenette, R., Friesen, R. W., Tetrahedron Lett. 1994, 35, 9177. (b) Xie, E, Cheng, G., Hu, Y., J. Comb. Chem. 2006, 8, 286. (c) Ma, Y., Margarida, L., Brookes, J., Makara, G. M., Berk, S. C., J. Comb. Chem. 2004, 6, 426. 25. (a) Hiroshige, M., Hauske, J. R., Zhou, E, J. Am. Chem. Soc. 1995, 117, 11590. (b) Kulkarni, B. A., Ganesan, A., J. Comb. Chem. 1999, 1, 373. (c) Coelho, A., Sotelo, E., J. Comb. Chem. 2006, 8, 388. 26. (a) Plunkett, M. J., Ellman, J. A., J. Am. Chem. Soc. 1995, 117, 3306. (b) Yun, W., Li, S., Wang, B., Chert, L., Tetrahedron Lett. 2001, 42, 175-177. 27. (a) Bolton, G. L., Hodges, J. C., J. Comb. Chem. 1999, 1, 130. (b) Yun, W., Mohan, R., Tetrahedron Lett. 1996, 37, 7189. (c) Zhang, H. C., Maryanoff, B. E., J. Org. Chem. 1997, 62, 1804. (d) Arumugam, V., Routledge, A., Abell, C., Balasubramanian, S., Tetrahedron Lett. 1997, 38, 6473. (e) Yu, Y., Ostresh, J. M., Houghten, R. A., Tetrahedron Lett. 2003, 44, 2569. (f) Krishnan, S., Schreiber, S. L., Org. Lett. 2004, 6, 4021. (g) Xiang, Z., Luo, T., Lu, K., Cui, J., Shi, X., Fathi, R., Chen, J., Yang, Z., Org. Lett. 2004, 6, 3155. (h) Cuny, G., Bois-Choussy, M., Zhu, J., J. Am. Chem. Soc. 2004, 126, 14475. 28. (a) Antilla, J. C., Buchwald, S. L., Org. Lett. 2001, 3, 2077. (b) Kwong, E Y., Klapars, A., Buchwald, S. L., 2002, 4, 581. (c) Yang, B. H., Buchwald, S. L., J. Organomet. Chem. 1999, 576, 125. 29. Masse, C. E., Ng, E Y., Fukase, Y., Sanchez-Rosselo, M., Shay, J. T., J. Comb. Chem. 2006, 8, 293. 30. Kim, Y., Arai, M. A., Arai, T., Lamenzo, J. O., Dean, E. E III, Petterson, N., Clemons, E A., Schreiber, S. L., J. Am. Chem. Soc. 2004, 126, 14740.

364


31. (a) Negishi, E. In Comprehensive Organic Synthesis, Trost, B. M. (Ed.), Pergamon: Oxford, 1991, Vol. 5, pp. 1163-1184. (b) Schore, N. E., Chem. Rev. 1988, 88, 1081. (c) Trost, B. M., Angew. Chem., Int. Ed. Engl. 1986, 25, 1. (d) Tamao, K., Kobayashi, K., Ito, Y., Synlett 1992, 539. (e) Lautens, M., Klute, W., Tam, W., Chem. Rev. 1996, 96, 49. 32. For a review on synthesis of heterocycles, see Nakamura, I., Yamamoto, Y., Chem. Rev. 2004, 104, 2127. 33. (a) Trost, B. M., Angew. Chem. Int. Ed. 1995, 34, 259. (b) Trost, B. M., Science, 1991, 254, 1471. 34. Trost, B. M., Krische, M. J., Synlett, 1998, 1. 35. Aubert, C., Buisine, O., Malacria, M., Chem. Rev. 2002, 102, 813. 36. (a) Wender, E A., Jenkins, T. E., J. Am. Chem. Soc. 1989, 111, 6432. (b) Wang, B., Cao., P., Zhang, X., Tetrahedron Lett. 2000, 41,8041. (c) Murakami, M., Ubukata, M., Itami, K., Ito, Y., Angew. Chem. Int. Ed. 1998, 37, 2248. 37. (a) Brummond, K. M., Kent, J. L., Tetrahedron, 2000, 56, 3263. (b) Gibson, S. E., Stevanazzi, A., Angew. Chem. Int. Ed. 2003, 42, 1800. (c) Rivero, M. R., Adrio, J., Carretero, J. C, Eur. J. Org. Chem. 2002, 2881. 38. Jolly, R. S., Luedtke, G., Sheehan, D., Livinghouse, T., J. Am. Chem. Soc. 1990, 112, 4965. 39. Burton, B. S., Pechman, H. V., Chem. Ber. 1887, 20, 145. 40. Hendrickson, J. B., Cram. D. J., Hammond, G. B., Organic Chemistry, 3rd edn., McGraw-Hill: New York, 1970, pp. 104-105. 41. For the most recent review, see Ma, S., Chem. Rev. 2005, 105, 2829. 42. (a) Modern Allene Chemistry, Krause, N., Hashmi, A. S. K. (Eds.), Wiley-VCH: Weinheim, 2004. (b) The Chemistry of Ketenes, Allenes, and Related Compounds Part 1, Patai, S. (Ed.), Wiley: New York, 1980. (c) Allenes in Organic Synthesis, Schuster, H. F., Coppola, G. M. (Eds.), Wiley: New York, 1984. 43. For reviews on reactions of allenes, see (a) Hashmi, A. S. K., Angew. Chem., Int. Ed. 2000, 39, 3590. (b) Marshall, J., Chem. Rev. 2000, 100, 3163. (c) Zimmer, R., Dinesh, C., Nandanan, E., Khan, E, Chem. Rev. 2000, 100, 3067 (d) Bates, R., Satcharoen, V., Chem. Soc. Rev. 2002, 31, 12. (e) Ma, S., Topics in Organometallic Chemistry, Tsuji, J. (Ed.), Springer-Verlag: Heidelberg, 2005, pp. 183-210. (f) Sydnes, L. Chem. Rev. 2003, 103, 1133. (g) Brandsma, L., Nedolya, N. A., Synthesis 2004, 735. (i) Tius, M., Acc. Chem. Res. 2003, 36, 284. (h) Wei, L. L., Xiong, H., Hsung, R. P., Acc. Chem. Res. 2003, 36, 773. (i) Lu, X., Zhang, C., Xu, Z., Acc. Chem. Res. 2001, 34, 535. (j) Wang, K. K., Chem. Rev. 1996, 96, 207. 44. Padwa, A., Filipkowski, M. A., Meske, M., Murphree, S. S., Watterson, S. H., Ni, Z., J. Org. Chem. 1994, 59, 591. 45. For examples of 4+ 2. reaction, see (a) Wender, E A., Jenkins, T. E., Suzuki, S., J. Am. Chem. Soc. 1995, 117, 1843. For examples of 5 + 2. reaction, see (b) Wender, E A., B i, E C., Gamber, G. G., Gosselin, F., Hubbard, R. D., Scanio, M. J. C., Sun, R., Williams, T. J., Zhang, L., Pure Appl. Chem. 211112,74, 25. (c) Wender, E A., Glorious, E, Husfield, C. O., Langkopf, E., Love, J. A., J. Am. Chem. Soc. 1999, 121, 5348. For an example of a 2 + 2 + 2. reaction, see (d) Aubert, C., Llerena, D., Malacria, M., Tetrahedron Lett. 1994, 35, 2341. 46. (a) Wender, P. A., Fuji, M., Husfield, C. O., Love, J. A., Org. Lett. 1999, 1, 137. (b) Wender, E A., Zhang, L., Org. Lett. 2000, 2, 2323. 47. Please see reference 45a for an additional example of catalyst-based control of double bond selectivity in a 4 + 2. reaction.

9


365

48. (a)Kent, J. L., Wan, H., Brummond, K. M., Tetrahedron Lett. 1995, 36, 2407. (b) Brummond, K. M. In Advances in Cycloaddition Chemistry, Vol. 6, Harmata, M. (Ed.), JAI Press: Stamford, CT, 1999, pp. 211-237. (c) Brummond K. M., Wan, H., Tetrahedron Lett. 1998, 39, 931. (d) Brummond, K. M., Wan, H., Kent, J. L., J. Org. Chem. 1998, 63, 6535. (e) Brummond, K. M., Lu, J., Petersen, J., J. Am. Chem. Soc. 2000, 122, 4915. (f) Brummond, K. M., Lu, J., Petersen, J., J. Am. Chem. Soc. 2000, 122, 4915. (g) Brummond, K. M., Kerekes, A. D., Wan, H., J. Org. Chem. 2002, 67, 5156. 49. (a) Brummond, K. M., Chen, H., Fisher, K. D., Kerekes, A. D., Rickards, B., Sill, E C., Geib, S. J., Org. Lett. 2002, 4, 1931. (b) Brummond, K. M., Gao, D., Org. Lett. 2003, 5, 3491. 50. Brummond, K. M., Chen, H., Sill, E, You, L., J. Am. Chem. Soc. 2002, 124, 15186. 51. (a) Brummond, K. M., Mitasev, B. M., Org. Lett. 2004, 6, 2245. (b) Brummond, K. M., Curran, D. E, Mitasev, B., Fischer, S., J. Org. Chem. 2005, 70, 1745. (c) Brummond, K. M., Chen, H., Mitasev, B., Casarez, A., Org. Lett. 2004, 6, 2161. (d) Brummond, K. M., Chen, D., Org. Lett. 2005, 7, 3473. 52. (a) Spry, D. O., Bhala, A. R., Heterocycles, 1986, 24, 4641. (b) Ley, S. V., Gutteridge, C. E., Pape, A. R., Spilling, C. D., Zumbrunn, C., Synlett, 1999, 1295. (b) Aoki, Y., Kuwajima, I., Tetrahedron Lett. 1990, 51, 7457. (c) Henderson, M. A., Heathcock, C. H., J. Org. Chem. 1988, 53, 4736. 53. Castelhano, A., Home, S., Taylor, G., Billedeau, R., Krantz, A., Tetrahedron 1988, 44, 5451. 54. Kazmaier, U., G6rbitz, C. H., Synthesis 1996, 1489. 55. (a) Oda, H., Sato, M., Morizawa, Y., Oshima, K., Nozaki, H., Tetrahedron 1985, 41, 3257. (b) Oda, H., Sato, M., Morizawa, Y., Oshima, K., Nozaki, H., Tetrahedron Lett. 1983, 24, 2877. 56. For examples of rate acceleration of cyclization reactions as a result of Thorpe-Ingold effect, see (a) Grubbs, R. H., Chang, S., Tetrahedron 1998, 54, 4413. (b) Alexander, J. B., La, D. S., Cefalo, D. R., Hoveyda, A. H., Schrock, R. R., J. Am. Chem. Soc. 1998, 120, 4041. (c) Yamamoto, Y., Nakagai, Y., Ohkoshi, N., Itoh, K, J. Am. Chem. Soc. 2001, 123, 6372. (d) Okamoto, S., Livinghouse, T. Organometallics 2000, 19, 1449. (e) Buchwald, S. L., Hicks, E A., J. Am. Chem. Soc. 1999, 121, 7026. 57. Rh or Ir coordination to an alcohol or ester is proposed as a reason for stereoselectivity in hydrogenation reactions (a) Crabtree, R. H., Davis, M. W., J. Org. Chem. 1986, 51, 2655. (b) McCloskey, E J., Schultz, A. G., ,/. Org. Chem. 1988, 53, 1380. 58. (a) Tsuge, O., Wada, E., Kanemasa, S., Chem. Lett. 1983, 12, 239. (b) Tsuge, O., Wada, E., Kanemasa, S., Chem. Lett., 1983, 12, 1525. 59. (a) Woo, S., Squires, N., Fallis A. G., Org. Lett. 1999, 1,573. (b) Woo, S., Legoupy, S., Parra, S., Fallis, A. G., Org. Lett. 1999, 1, 1013. 60. For synthetic methods for preparation of acyclic trienes, see (a) Arisawa, M., Sugihara, T., Yamaguchi, M., Chem. Commun. 1998, (23), 2615. (b) Trahanovsky, W. S., Koeplinger, K. A., J. Org. Chem. 1992, 57, 4711. (c) Shi, M., Shao, L. X., Synlett 2004, 807. (d) Moriya, T., Furuuchi, T., Miyaura, N., Tetrahedron 1994, 50, 7961. 61. For a review on cross-conjugated polyenes, see Hopf, H., Angew. Chem. Int., Ed. Engl. 1984, 23, 948. 62. For example, Fallis' methodology was applied by Schreiber and coworkers to the diversity-oriented synthesis of a library of polycyclic small molecules: (a) Kwon, O., Park, S. B., Schreiber, S. L., J. Am. Chem. Soc. 2002, 124, 13402.

366

63. 64.

65. 66. 67. 68. 69. 70. 71.

72. 73. 74. 75. 76. 77.

78. 79.

80.

KAY M. BRUMMONDAND BRANKOMITASEV Forother applications of heteroatom-containing, cross-conjugated trienes, see (b) Spino, C., Liu, G., Tu, N., Girard, S., J. Org. Chem. 1994, 59, 5596. (c) Dion, A., Dub6, E, Spino, C., Org. Lett. 2005, 7, 5601. Payne A. D., Willis A. C., Sherburn M. S., J. Am. Chem. Soc. 2005, 127, 12188. Conformationally defined polycyclic small molecules often exhibit higher potency and specificity of binding to biological targets compared to their acyclic analogues, since they do not have to undergo conformational changes in order to adapt to a binding site. King, F. D., Strategy and Tactics in Drug Discovery. In Medicinal Chemistry: Principles and Practice, 2rid edn., King E D. (Ed.), The Royal Society of Chemistry, Cambridge, UK, 2002, pp. 342-346. See Richard, C. L. Comprehensive Organic Transformations. A Guide to Functional Group Preparations, 2nd edn., Elsevier: Oxford, 1870 pp. Yeung, Y.-Y., Hong, S., Corey, E. J., J. Am. Chem. Soc. 2006, 128, 6310. Serna, S., Tellitu, I., Dominguez, A., Moreno, I., SanMartin, R., Org. Lett. 2005, 7, 3073. Grimaldi, J., Cormons, A., Tetrahedron Lett. 1986, 27, 5089. Hu, T., Li, C., Org. Lett. 2005, 7, 2035. (a) Golden, J. E., Aub6, J., Angew. Chem. Int. Ed. 2002, 41, 4316. (b) Sharghi, H., Hosseini, M., Synthesis, 2002, 1057. For examples of ~,-lactam formation via ring-closing metathesis, see (a) Clayden, J., Tunbull, A., Pinto, I., Tetrahedron: Asymm. 2005, 16, 2235. (b) Badorrey, Cativiela, C., Diaz-de-Villegaz, M. D., Diez, R., Galvez, J. A., Tetrahedron Lett. 2004, 45, 719. For an example of 5-1actam formation via ring-closing metathesis, see (c) Niida, A., Tomita, K., Mizumoto, M., Tanigaki, H., Terada, T., Oishi, S., Otaka, A., Inui, K., Fujii, N., Org. Lett. 2006, 8, 613. (a) Xie, X., Lu, X., Tetrahedron Lett. 1999, 40, 8415. (b) Xie, X., Lu X., Liu, Y., Xu, W., J. Org. Chem. 2001, 66, 6545. Lei, A., Waldkirch, J. P., He, M., Zhang, X., Angew. Chem. Int., Ed. 2002, 41, 4526. Brummond, K. M., Painter, T. O., Probst, D. A., Mitasev, B., Org. Lett. 2007, 8, 347. Stewart, W. E. III, Siddall, T. H., Chem. Rev. 1970, 70, 517. For the unexpected formation of a similar oxazolidinone, see Granier, T., Vasella, A., Helv. Chim. Acta. 1998, 81,865. Geometry optimization was performed with Cache Worksystem Pro. version 6.1.10. First, global minimum search was performed using MM3 parameters, followed by optimization of the side chain conformation with AM I parameters. Then, energy minimization of the sample was performed using AM 1 parameters. Mitasev, B., Yan. B., Brummond, M, Heterocycles, 2006, 70, 367. (a) Tacconi, G., Iadarola, P., Marinone, E, Righetti, E P., Desimoni, G., Tetrahedron 1975, 31, 1179. (b) Rudnichenko, A. V., Timoshenko, V. M., Chernega, A. N., Nesterenko, A. M., Shermolovich, Y. G., J. Fluorine Chem. 2004, 125, 1351. (c) Burdisso, M., Desimoni, G., Faita, G., Righetti, E, Tacconi, G., J. Chem. Soc. Perkin Trans. 2, 1989, 7, 845. For use of Eu(fod) 3 to catalyze hetero Diels-Alder reactions, see (a) Bednarski, M., Danishefsky, S., J. Am. Chem. Soc. 1983, 105, 3716. (b) Dujardin, G., Leconte, S., Coutable, L., Brown, E., Tetrahedron Lett. 2001, 42, 8849. (c) Hughes, K. D., Nguyen, T. L. N., Dyckman, D., Dulay, D., Boyko, W. J., Giuliano, R. M., Tetrahedron: Asymm. 2005, 16, 273. (d) Dujardin, G., Rossignol, S., Brown, E., Synthesis 1998, 763.

9


367

81. Sanchez, I. H., L6pez, E J., Soria, J. J., Larazza, M. I., Flores, H. J., J. Am. Chem. Soc. 1983, 105, 7640. 82. For an example of hydrolysis of oxazolidinones using LiOH, see Davies, S. G., Hermann, G. J., Sweet, M. J., Smith, A. D., Chem. Commun. 2004, (9), 1128. 83. (a) Lagenhan, J. M., Fisk, J. D., Gellman, S. H., Org. Lett. 2001, 3, 2559. (b) Gardner, R., Liang, G. B., Gellman, S. H., J. Am. Chem. Soc. 1999, 121, 1806. (c) Dado, G., Gellman, S. H., J. Am. Chem. Soc. 1993, 115, 4228. (d) Gellman, S. H., Dado, G., Liang, G. B., Adams, B.,J. Am. Chem. Soc. 1991, 113, 1164. 84. (a) Han, Y., Giroux, A., Colucci, J., Bayly, C. I., McKay, D. J., Roy, S., Xanthoudakis, S., Vaillancourt, J., Rasper, D. M., Tam, J., Tawa, E, Nicholson, D. W., Zamboni, R. J., Bioorg. Med. Chem. Lett. 2005, 15, 1173. (b) Quibell, M., Benn, A., Flinn, N., Monk, T., Ramjee, M., Ray, E, Wang, Y., Watts, J., Bioorg. Med. Chem. 2005, 13, 609. (c) Sperka, T., Pitlik, J., Bagossi, P., Toezser, J., Bioorg. Med. Chem. Lett. 2005, 15, 3086. (d) Verhelst, S. H. L., Bogyo, M., ChemBioChem 2005, 6, 824. (e) Grimm, E. L., Roy, B., Aspiotis, R., Bayly, C. I., Nicholson, D. W., Rasper, D. M., Renaud, J., Roy, S., Tam, J., Tawa, P., Vaillancourt, J. P., Xanthoudakis, S., Zamboni, R. J., Bioorg. Med. Chem. 2004, 12, 845. (f) Helal, C. J., Sanner, M. A., Cooper, C. B., Gant, T., Adam, M., Lucas, J. C., Kang, Z., Kupchinsky, S., Ahlijanian, M. K., Tate, B., Mennitti, E S., Kelley, K., Peterson, M., Bioorg. Med. Chem. Lett. 2004, 14, 5521. (g) Johansson, E O., Chert, Y., Belfrage, A. K., Blackman, M. J., Kvarnstroem, I., Jansson, K., Vrang, L., Hamelink, E., Hallberg, A., Rosenquist, A., Samuelsson, B., J. Med. Chem. 2004, 47, 3353. (h) Reid, R. C., Pattenden, L. K., Tyndall, J. D. A., Martin, J. L., Walsh, T., Fairlie, D. E, J. Med. Chem. 2004, 47, 1641. 85. Chambers, P., Turner, D., Werner, S., Mitasev, B., Brummond, K. M. Synthesis of a 200-member Library of Polycyclic Oxazolines and Amides. Manuscript in preparation.

515

COLOR PLATE SECTION

0

..~

2 / N ~

/ C02Me

Me02C

65a

66a

X,

Please refer to Figure 1 of Chapter 9 in text for figure legend.

0

0

F_~O~N.,H .4 Me02C

0==~ P 67a

torsional angle of diene = 39.9 ~

-

torsional angle of o~,l]-unsaturated amide= 42.4 ~

Please refer to Figure 2 of Chapter 9 in text for figure legend.

516

COLOR PLATE SECTION

2.287A

"~NH

O

73a

energy-minimized model of 73a; the two phenyl groups were replaced with a methyl group for clarity purposes Please refer to Figure 3 of Chapter 9 in text for figure legend.

c~~

0

~ MiboZ~2c

N-Ph

0 II toluene,reflux,2hr MeO2(~ ,~ 83%

45f

0

~oN_ph c0z.N~ .... . MeO2~

J

....~-Ph

~

-

54

55 5:2:1 mixture of diastereomers (major diastereomer shown)

-

/

X-ray structure of 55

~

..~

Please refer to Scheme 17 of Chapter 9.


Chapter 10

HYDROGEN-BOND-MEDIATED SYNTHESIS

IN THE SOLID

ORGANIC STATE

Leonard R. MacGillivray Department of Chemistry University of Iowa Iowa City, IA 52242, USA

I. II.

III.

Introduction and Background The Problem of Crystal Packing Template-Controlled Solid-State Reactivity Using Resorcinol A. Prior Work B. Resorcinol as a Template

IV.

V.

Target-Oriented Synthesis in the Solid State A. [2+2] Paracyclophane B. Ladderanes Other Templates A. 1,8-NAP B. 2,3-NAP C. Reb-Im

Conclusion Acknowledgments References and Footnotes VI.

368 369 371 371 373 374 375 376 378 378 379 380 381 381 381

I. Introduction and Background

Covalent bond-forming reactions (e.g., carbon-carbon bond) lie at the heart of organic synthetic chemistry. ~ Such reactions are used to construct molecules of remarkable complexities. Methods that form covalent bonds in the most efficient manner possible - in high yields, with limited byproducts, and with minimal waste- are highly valued. 2 Organic chemists continue to search and develop improved ways to control the formation of covalent bonds. In this context, the organic solid state represents an intriguing medium within which to control the formation of covalent bonds. 3 The solid state is sufficiently flexible to allow atoms to move and react, yet

10

HYDROGEN-BOND-MEDIATED ORGANIC SYNTHESIS IN THE SOLID STATE

369

sufficiently rigid to enable reactions to proceed with remarkable stereocontrol. The homogeneous nature (i.e., 3D regularity) of the solid state means that covalent-bond-forming reactions that occur can proceed in up to quantitative yield. The benefit of controlling the formation of covalent bonds in such a solvent-free environment is apparent. 4 The solid state also enables molecules to adopt geometries impossible to achieve in the liquid phase. This means that opportunities exist to form molecules in the solid state that may be inaccessible in solution. Although covalent bonds have been known to form in the organic solid state for more than a century, 3 it has only been within the last decade that chemists have developed tools and concepts that are enabling the solid state to be used as a medium to construct molecules by design. 5 Rapid advances in the field of X-ray crystallography (e.g., CCD X-ray diffraction technology) have permitted structural data to be collected and analyzed on timescales of minutes-to-hours as opposed to days-to-months using more traditional instruments (e.g., single-point detectors). Moreover, conceptual advances in the field of crystal engineering 6 and the related field of supramolecular chemistry 7 have resulted in marked improvements of our understanding of how intermolecular forces influence the organization of molecules such that the design and construction of solids with predetermined physical properties (e.g., reactivity, optical activity) is now becoming a reality. It is with these ideas in mind that work performed in our laboratory during the past 5 years that has been aimed to achieve control of a carboncarbon-bond-forming reaction in the solid state; namely, the [2+2] photodimerization, 8 will be discussed. It will be shown that control of this cycloaddition reaction can be achieved by using small organic molecules, termed linear templates, 9 which are used to assemble and preorganize alkenes in geometries suitable for the photoreaction. The templates operate by assembling the alkenes via hydrogen bonds. It will be demonstrated that this supramolecular approach to covalent bond formation can be used to code the construction of carbon-carbon bonds with virtually perfect control of regiochemistry. Whereas our initial focus was to expand the nature of the templates, we are beginning to learn how this method can be employed to construct molecules with architecturally rich frameworks. II. The Problem of Crystal Packing

The field of crystal engineering finds its origins in the early work of Gerhardt Schmidt who, through a large number of crystallographic investigations, determined specific geometry criteria for a [2 + 2] photodimerization

370

LEONARD R. MACGILLIVRAY

to occur in the solid state. 8 By systematically studying a series of cinnamic acids, Schmidt determined that two carbon-carbon double (C =C) bonds should be aligned parallel and separated by less than 4.2 A to react to form a cyclobutane product. Thus, UV-irradiation of ~-trans-cinnamic acid, a polymorph of cinnamic acid with olefins arranged head-to-tail, was shown to produce ~-truxillic acid while irradiation of ~-trans-cinnamic acid, with olefins arranged head-to-head, produced [3-truxinic acid (Scheme 1). Cinnamic acid itself undergoes cis-trans isomerization in the liquid phase; thus, the fact that the acid underwent a dimerization in the solid state represented an intriguing departure from solution-phase chemistry. Schmidt also showed that the rigid environment of the solid state enables the dimerization to occur stereospecifically such that the molecular structure of the product is consistent with the geometric relation of the olefinic reactants in the crystal lattice. From a purely topological standpoint, the photodimerization provides a simple way to crosslink two stacked molecules.

/---cO2H hv ~ CO2H solid 13-cinnamicacid HO2C

~

truxinicacid hv

CO2H

CO2H CO2H

solid

HO2C ~~'--

o~-cinnamicacid

CO---~

truxillicacid SCHEME 1

Whereas the work of Schmidt provided criteria for a [2+2] photodimerization to occur, Schmidt also showed how the extreme sensitivity of solid-state organization to subtle changes to molecular structure severely hampered the synthetic value of the covalent-bond-forming process. ~~In particular, Schmidt showed that members of a homologous series of cinnamic acids did not exhibit homologous reactivities in the solid state. Such differences in reactivity were shown to result from the steric and electronic influences of the organic functional groups on the crystal packing. Thus, whereas p-methylcinnamic acid was found to form a solidstate structure that was photoactive, o-methylcinnamic acid formed a structure that was photostable. These observations lie in stark contrast to the

10


371

liquid phase, where such subtle changes to the structures of reactants typically do not have such adverse effects on geometry and reactivity. III. Template-Controlled

Solid-State Reactivity Using Resorcinol

In 2000, our group embarked on studies to determine whether it is possible to separate the [2+2] photodimerization from the effects of crystal packing. We hypothesized that by effectively decoupling reactivity from packing, we could broaden the synthetic scope of the solid-state reaction. In particular, we expected that the photoreaction could be decoupled from packing using small, ditopic molecules that function as linear templates. 9 In this design, a template would assemble and preorganize, via relatively strong directional forces such as hydrogen bonds, two functionalized olefins in a geometry that is both suitable for the photoreaction and independent of crystal packing (Scheme 2). In the minimalist case, a single template could assemble two monofunctionalized olefins for the photoreaction. Additional copies of the template could also be integrated into the proposed reactivity scheme so as to exploit the process of selfassembly. Owing to the fact that the template would assemble along the exterior of two stacked olefins, we expected that the approach could be eventually used to modify the structure of the olefin (e.g., addition of a functional group) so as to apply the templates to a variety of reactants and, therefore, construct a variety of products.

2 E

+2R--R

crystallize

--R--R--

E r

-R--R-

]

~

--R--R--

E, ,3 - -"

dissolve

-

"~

2

E

+ I I "R'--R -"

= linear template, R - - R = reactant

SCHEME 2

A. P R I O R W O R K

Prior to our proposed work, there had been two studies that suggested that the [2+2] photodimerization could be separated from the effects of long-range packing. First, Campbell and Feldman had described an olefinic 'J'-shaped dicarboxylic acid that self-assembled in the solid state to form a dimer held together by four O - H - - - O hydrogen bonds (Scheme 3). ~ The C = C bonds

372

L E O N A R D R. M A C G I L L I V R A Y

of the monomer units conformed to Schmidt's geometry criteria for a photodimerization within the dimer. Moreover, that the carboxylic acid formed a 0D hydrogen-bonded assembly (i.e., where the strongest intermolecular forces occurred within the dimer) meant that the arrangement of the olefins, and reactivity, would be largely independent of the effects of packing. When irradiated with UV radiation, the solid produced the expected cyclobutane product in quantitative yield.

"H-O x

O'H"O~V hv

"

~.0""H'O.

lsolid state

SCHEME

3

Second, Ito and Scheffer described a series of crystalline diammonium cinnamate salts in which the ammonium groups participated in N + - H - . - O - hydrogen bonds with the cinnamates (Scheme 4). 12 A number of salts were photoactive, which was accounted for by the ammonium ions adopting gauche conformations (e.g., ethylenediammonium) that forced the cinnamates to stack in positions to react. The ammonium ions also adopted anti conformations that forced the cinnamates into unreactive geometries. In some cases, the ammonium ions participated in N + - H . . . O - hydrogen bonds with other components of the lattice so as to give infinite hydrogen-bonded polymers that were photostable. Nevertheless, the studies involving the salts demonstrated that olefins could be forced into reactive geometries using a ditopic system.

NH; "OOC ,~~Ar (CH2)n'rr"~"NH~"OOC ~.,,A r

"OOC'~"Ar state "OOC.,,~~Ar hv

solid

SCHEME 4

10


B. R E S O R C I N O L

373

AS A TEMPLATE

Our first studies involved developing 1,3-dihydroxybenzene, or resorcinol, as a linear template. ~2 Specifically, we anticipated that cocrystallization of resorcinol with trans-l,2-bis(4-pyridyl)ethylene (4,4'-bpe) would produce a four-component complex, of composition 2(resorcinol).2(4,4'-bpe), held together by four O - H . . - N hydrogen bonds wherein two molecules of 4,4'-bpe would be positioned for a [2+2] photodimerization (Scheme 5). In line with our strategy, cocrystallization of resorcinol with 4,4'-bpe produced the 0D complex 2(resorcinol)-2(4,4'-bpe). The two C =C bonds of the two olefins were oriented parallel and separated by 3.65 A, an ideal position for a photodimerization. UV-irradiation of the solid produced the anticipated photoproduct, rctt-tetrakis(4-pyridyl)cyclobutane (4,4'-tpcb). The product formed stereospecifically and in 100% yield. The structure of the product was determined using single-crystal X-ray analysis in which 4,4'-tpcb was shown to assemble with resorcinol in the three-component complex 2(resorcinol).(4,4'-tpcb). We also determined that the template-controlled solid-state reaction could be conducted in gram-scale amounts.

1

o...

.....

.......

template

~0

hv

-H..... N ~ " ~ X

-H..... N -,........J

solid state

template

J,~'~tN ..... H-O

~~-~N

..... H-O

SCHEME 5

Following our initial report, we demonstrated that resorcinol could be used to assemble an olefin with a single 4-pyridyl group. ~4 In particular, resorcinol was used to assemble two stilbazoles in (resorcinol).2(transl-(4-pyridyl)-2-(4-chlorophenyl)ethylene) in a head-to-head geometry for a regiocontrolled photodimerization that produced the corresponding

LEONARDR. MACGILLIVRAY

374

head-to-head product rctt-l,2-bis(4-pyridyl)-3,4-bis(4-chlorophenyl) cyclobutane in quantitative yield (Scheme 6). C1 ... C1 interactions also formed between the hydrogen-bonded complexes such that the nearestneighbor assemblies constituted six-component assemblies held together by both O - H . . . N and C1...C1 forces. As a consequence of these interactions, the olefins were organized in close proximity, the two C = C bonds being separated by 3.98 A. Unlike 2(resorcinol).2(4,4'-bpe), however, the olefins adopted an antiparallel orientation, which was expected to render the C =C bonds photostable.

O'H ..... N

o.,

~

CI

c,

..... ..,

hvlS~ state

_!-. .....

c'

O-H ..... N r

~

Cl

SCHEME6 We attributed the generation of the photoproduct to the olefins undergoing a pedal-like change in conformation in the solid state that enabled the C = C bonds to adopt a parallel orientation suitable to react and form the product. IV. Target-Oriented Synthesis in the Solid State

That resorcinol served as a template to direct the [2+2] photodimerization within crystalline assemblies prompted us to explore whether the method could be applied to more structurally diverse olefins. In particular, we decided to study di- and tri-olefins as reactants. We hypothesized that the self-assembly process involving the templates could adapt to a lengthening of the reactants with the placement of additional C - C bonds between the pyridyl groups.

10


375

A. [2 + 2] P A R A C Y C L O P H A N E

Specifically, we expected that we could use a resorcinol to construct a lengthened molecule in the form of a [2.2]paracyclophane. The paracyclophane framework, introduced by Cram, bears relevance to physical organic chemistry and materials science owing to its reactive, photochemical, and electrochemical properties. 15 Despite these features, however, the synthesis of molecules based on [2.2]paracyclophane remains a synthetic challenge. A retrosynthetic analysis of the targeted [2.2]paracyclophane suggested that the molecule could be constructed from the bifunctional diene trans,trans-l,4-bis[2-(4-pyridyl)ethenyl]benzene (1,4-bpeb). Cocrystallization of resorcinol, or a derivative, with 1,4-bpeb would yield the four-component complex, 2(resorcinol).2(1,4-bpeb), with the two dienes positioned for a double [2+2] photodimerization. UV-irradiation of the solid would generate the cyclophane target tetrakis(4-pyridyl)-l,2,9,10diethano[2.2]paracyclophane (4,4'-tppcp). In the ideal case, the target would form in the solid state in quantitative yield and gram quantities. We determined that co-crystallization of 5-methoxyresorcinol (5OMe-res) with 1,4-bpeb produced the four-component assembly 2(5OMe-res).2(1,4-bpeb) (Scheme 7). z3 Similar to 2(resorcinol).2(4,4'-bpe), the components were held together by four O - H . . . N hydrogen bonds with the C=C bonds being separated by 3.70 A. The assemblies packed

O ' H ..... MeO

N..r~-'~

~~-~t~N ...... H-O

.--~

--/?OMe b'H ..... N.,':'--~

~,~,

~,~ ....... ,,,.4 n'u

hvlS~ state O-H.....

~l..... H-O OMe

SCHEME 7

376

LEONARD R. MACGILLIVRAY Q

-0

O-

FIGURE 1. ORTEP perspective of the targeted [2.2]paracyclophane.

with C =C bonds of adjacent assemblies parallel and separated by 3.95 A, a distance also suitable for a photodimerization. As determined by ~H NMR spectroscopy, UV-irradiation of 2(5-OMe-res).2(1,4-bpeb) produced 4,4'-tppcp in 60% yield. In addition to the paracyclophane, a monocyclized dimer and indefinable products formed. We attributed the formation of the side products to cross-reactions between nearestneighbor hydrogen-bonded structures. Later, we showed that 4-benzylresorcinol (4-bn-res) yields 4,4'-tppcp stereospecifically in 100% yield. 16 Similar to 2(5-OMe-res).2(1,4-bpeb), co-crystals of 2(4-bn-res).2(1,4-bpeb) consisted of four-component assemblies with the two dienes positioned for a double [2 + 2] photodimerization. However, in contrast to 2(4-bn-res).2(1,4-bpeb), the closest separation between C - C bonds of adjacent assemblies was 5.4 *. This longer distance was outside the distance criterion of Schmidt for a photoreaction. We attributed the different packing of the hydrogen-bonded assemblies to steric and electronic influences of the benzyl group attached to the resorcinol template. UV-irradiation of 2(4-bn-res).2(1,4-bpeb) produced 4,4'tppcp in 100% yield. The structure of the [2.2]paracyclophane target was confirmed via single-crystal X-ray diffraction (Figure 1). B. L A D D E R A N E S

In addition to a [2.2]paracyclophane, we anticipated that a resorcinol could be used to construct [n]-ladderanes (where n = 3 or 5) in the solid state. Near the time of our proposed work, ladderanes had been discovered in intracellular membrane lipids of anaerobic ammonium-oxidizing, or anammox, bacteria. 17The ladderanes were shown to rigidify the membrane of an organelle that allows the bacteria to participate in the oceanic nitrogen cycle. A retrosynthetic analysis of a ladderane suggested that a template-controlled [2+2] photodimerization of the conjugated diene

10


377

trans, trans- 1,4-bis(4-pyridyl)- 1,3-butadiene (1,4-bpbd) would yield the corresponding [3]-ladderane, all-trans-tetrakis(4-pyridyl)-[3]-ladderane (4-tp-3-1ad). Likewise, a templated [2+2] photodimerization of the triene trans, trans, trans- 1,6-bis(4-pyridyl)- 1,3,5-hexatriene (1,6-bpht) would produce the corresponding [5]-ladderane all-trans-tetrakis(4pyridyl)-[5]-ladderane (4-tp-5-1ad) (Scheme 8). O-H ..... N ~" ' ~ ~ ' ' ~ ~ ~

~

9

Q.H

~

H0

. . . . . . .

..... N

....... H . ~

hvlS~ state

c -H ..... O-H ..... N ~,,,,,~

~ N - -

-H-O

SCHEME 8

We determined 5-OMe-res to be a template for the quantitative construction of both the [3]- and [5]-ladderanes. ~8In particular, co-crystallization of 5-OMe-res with either 1,4-bpbd or 1,6-bpht produced the four-component assemblies 2(5-OMe-res).2(1,4-bpbd) and 2(5-OMe-res).2(1,6-bpht), respectively. In each case, the C =C bonds of the polyenes were organized in appropriate positions for the photoreaction. As established via 1H NMR spectroscopy, UV-irradiation of each solid produced the targeted [3]- and [5]-ladderane in gram amounts and 100% yield. Each product was characterized via single-crystal X-ray analysis (Figure 2). The template-controlled

w n

wD..

..

..

' ,

0

w

.

n

n

n

..

o

" 2- L......... ! 1 1 I1-,,5 .~7 ~ :, ~-,r,%.,a FIGURE 2. ORTEP perspective of the [5]-ladderane.

.

378

LEONARD R. MACGILLIVRAY

solid-state [2+2] photodimerization of 1,4-bpbd in 2(5-OMe-res).2(1,4bpbd) provided the first example of a high-yielding solid-state synthesis of a [3]-ladderane while the reaction within 2(5-OMe-res)-2(1,6-bpht) represented the first example of a solid-state synthesis of a [5]-ladderane. V. Other Templates

That we could use templates based on resorcinol to construct a [2.2]paracyclophane and [n]-ladderanes in the solid state prompted us to further explore the scope of the method. We reasoned that having additional templates on hand could increase the synthetic flexibility of the method by providing a means, for example, to increase a product yield where a resorcinol may be less effective. Additional templates could also be used with recognition sites other than 4-pyridyl groups attached to the olefins. In addition to expanding the method in terms of the supramolecular synthesis, different recognition sites could be used to increase the structural diversity of the photoproducts. Since our initial studies involving resorcinol, we have determined that 1,8-naphthalenedicarboxylic acid (1,8-nap), ~9 2,3-bis(4-methylenethiopyridyl)naphthalene (2,3-nap), 2~and Rebek's imide (Reb-im) 21 function as linear templates in the solid state. A. 1,8-NAP

Our first study to extend linear templates beyond resorcinol involved 1,8-naphthalene dicarboxylic acid (1,8-nap, Scheme 9). 19 Similar to 2(resorcinol).2(4,4'-bpe), co-crystallization of 1,8-nap with 4,4'-bpe was expected to produce the four-component assembly 2(1,8-nap).2(4,4'-bpe). That the 1,8-naphthalene framework had been previously used to enforce face-to-face stacking of aromatics provided a measure of support for our hypothesis. 22 That 1,8-nap was known to form a hydrogen-bonded dimer in the solid state also supported our hypothesis. 23 We expected that each carboxylic acid group of each diacid would interact with the 4-pyridyl group of 4,4'-bpe via an O - H . . . N hydrogen bond. As expected, cocrystallization of the two components produced a four-component complex with two C = C bonds organized parallel and separated by 3.73 ,~. UV-irradiation of the solid produced 4,4'-tpcb stereospecifically in 100% yield.

10

379


O'H ...... N ~ ' ~ / ~ '

~

-~.

:.o O-H ...... N I

I

~

~ ~,~

~N ...... H ' O

o.~

~N .....

I

-H-O 9

I

hv 1 solid

1,8-nap

1,8-nap !

state

~--%..

........N , ; ~ ~ ,~Z~~ ...... o,

,~ ~/~O-H

--.-o\ ........N

N.........H-O SCHEME 9

B. 2,3-NAP Following our work involving resorcinol and 1,8-nap, it occurred to us that the hydrogen-bonding, or code, involving the templates and reactants could be reversed. Thus, whereas resorcinol and 1,8-nap function as hydrogen-bond-donors, we aimed to develop a template that functioned as a bifunctional hydrogen-bond acceptor. Following a study involving agentophilic forces, 24 we determined that code reversal could be achieved using 2,3-bis(4-methylenethiopyridyl)naphthalene (2,3-nap) (Scheme 10). 2~ In particular, co-crystallization of 2,3-nap with fumaric acid (fum) produced the four-component assembly 2(2,3-nap).2(fum) held together by four O - H . . - N hydrogen-bonds. Two acid molecules were juxtaposed by the

S

[ ~ [ ~

"

O

~

0

S ---~"~s~N ....... H'O -"~~2N

O'H ...... N~/'~-'~---/S

...... H"O~"

~""~r

...... N:~'~/~)'--" S

m

I

m

I

hv i

solid ~state

2,3-nap S ~"-~'~s~N ..... H

098% de and 97% ee, although the direct C-H insertion was a competing reaction. High stereoselectivity was also exhibited in the reaction with the unsaturated lactone 27 (>98% de, 99% ee) but in this case, some direct C-H insertion and cyclopropanation also occurred as competing reactions. During the exploratory studies, Qihui Jin identified that dihydronaphthalenes were excellent substrates for the combined C-H activation/Cope rearrangement. This was seen in the Rhz(S-DOSP)4 catalyzed reaction of vinyldiazoacetate 29 with dihydronaphthalene 28 (Scheme 8). 34 The C-H activation/Cope rearrangement product 30 was obtained in >98% de and 98% ee. One of the signature features of the donor/acceptor-substituted

11

391

TOTAL SYNTHESES OF NATURAL PRODUCTS

Me [~

Ph Rh2(SDOSP)4 Me "~~_.~ PhCF3,0 ~ CO2Me CO2Me 68%

N2~ +

26

Ph

25

97% ee, >98% de

Me Ph Ph N2, ~ oO~ Rh2(S-DOSP)4 ~ e~-~\ ~ O + PhCF3,0 ~ ""CO2Me CO2Me 87% O 27 25 99%ee, >98% de SCHEME 7

carbenoids is that the trajectory of attack of reagents to the carbenoids appears to be very specific. Even though the actual trajectory is not known with certainty, an excellent predictive stereochemical model has been developed for these reactions. 11a,35 Applying this model to the C-H activation/Cope rearrangement reaction of vinyldiazoacetates with dihydronaphthalene 28 correctly predicts the stereochemistry observed in product 30 (Scheme 8). The substrate approaches the vinylcarbenoid (S-catalyst) from the front face and the constricted trajectory leads to excellent control of the two stereogenic centers.

CO2Me + N2=~/ Rh2(S'DOSP)4. "Me 23 ~ t Me Et 28 29 MeO2C 30

[~~

Me flip E "CO2Me

>98% de, 98% ee

r ueO.Gi i Etl lJ . ,Tj

9. 180~ [ ~ ~ MeO2C"~v "~MeH

(S-DOSP) Catalyst substrate approaching from front SCHEME 8

180~ ,. H ~ C O 2 M e Me

392

HUW M.L. DAVIES AND XING DAI

C. E N A N T I O D I F F E R E N T I A T I O N " M O D E L S T U D Y

Inspired by the results of the combined C-H activation/Cope rearrangement shown in Scheme 8, we recognized that this methodology could be a wonderful strategic reaction for the synthesis of the marine natural products shown in Figure 1. Presumably, the C-H activation/Cope rearrangement product 32, which contains the three common stereocenters, could be a universal intermediate to these natural products (Scheme 9). Initially we thought it may be necessary to begin the synthesis with enantiomerically pure dihydronaphthalenes. During the analysis of this synthetic problem, however, we recognized that the racemic dihydronaphthalene (+_)-31 could be used as starting material. Applying the predictive models to the Rhz(R-DOSP)4-catalyzed reaction of ( +)-31 with 18 suggested that the two enantiomers would lead to different products. Only (S)-31 would be capable of a matched, combined C-H activation/Cope rearrangement to form 32, whereas (R)-31 would be matched for a cyclopropanation to form 33 (Scheme 9). This would be a very exciting outcome because the dihydronaphthalene (+)-31 could potentially be used as the limiting agent, as both enantiomers would be consumed but would form different products.

Me

Me

R . , E ~ . f ~ ~,H S-enantiomer ~ " H

Me

,14' Racemic

> OMe

)

Starting +Material Rh2(R-DOSP) (, ,

CO2Me

\

18 Me

~

,-ha "od" e'ge t ransitionState ($ubs*-"tes approaching from front)

.,J... I~',, Me" v 9

Me 32

R-enantiomer

r._ f

Rh I (R-DOSP) Catalyst SCHEME 9

le

Me" v

O2Me ~ "~Me 33

393

l 1 TOTALSYNTHESES OF NATURALPRODUCTS

As this was a bold hypothesis, Abbas Walji conducted model studies to test the predictions. Rh2(S-DOSP)4-catalyzed reaction of dihydronaphthalene 34 with the phenylvinyldiazoacetate 35 gave a very promising result (Scheme 10). 36 A 1:1 mixture of the combined C-H activation/Cope rearrangement product 36 and the cyclopropane 37 were formed in a combined yield of 80%. Both products were produced in 98% ee and essentially as single diastereomers. The relative and absolute stereochemistry of 36 was determined by conversion of 36 into the crystalline p-bromobenzoate 38, whose configuration was confirmed by X-ray crystallography.

Me

Me -

N2 35

""H

---

2% Rh2(S-DOSP)4 34

Me

h

80%

O,~'

+

,~Ph

(1:1)

36 98% ee

CO2Me 37

98% ee

/

OMe i, H2, Pd/C ii, LiAIH4 iii, DCC, DMAP, p-BrC6H4CO2H Me _

9 ,1 a

Br\~...~

"Ph 0

SCHEME l0 IV. Total Synthesis of (+)-Erogorgiaene

Having successfully completed the model studies, attention was then directed toward the total synthesis of (+)-erogorgiaene (2). 36 The key step is the rhodium-catalyzed reaction between the vinyldiazoacetate 18 and the dihydronaphthalene (+)-39 (Scheme 11). The Rhz(R-DOSP)4 catalyzed reaction of vinyldiazoacetate 18 with (+)-39 resulted in enantiodifferentiation to form 41 with 90% ee. The other enantiomer of the dihydronaphthalene 39 preferentially formed the cyclopropane 40. Due to the high diastereoselectivity of the combined C-H activation/Cope rearrangement, a single diastereomer of 41 was formed with the correct configuration for (+)-erogorgiaene (2). Owing to the

394

HUW M.L. DAVIESAND XING DAI

tendency of 41 to undergo a retro-Cope rearrangement, the combined mixture of 40 and 41 was globally hydrogenated, and the ester was reduced to the alcohol 42, which was isolated in 31% overall yield from the dihydronapthalene (+_)-39 (62% yield from the matched enantiomer (S)-39). Completion of the total synthesis of (+)-erogorgiaene (2) was achieved by oxidation of 42 to the aldehyde with pyridinium chlorochromate (PCC) followed by a Wittig reaction. Most impressively, Abbas Walji, the graduate student who conducted this project, completed the whole synthesis of (+)-erogorgiaene, including the sixstep synthesis of the dihydronaphthalene substrate, in just 10 days!

Me

. ~ ~ Me

Me

MeSH

2%Rh2(R-DOSP)4

2,2-DMB, rt, 2 h ~ Me~

39 +

N2 Me02C1 8 . ~ M e

,,H H

s

Me+ Me (1:1)

]~ "Me

4 1 0 " ~ ~ 90% OMe ee

40

i,H 2,Pd/C ii, LiAIH4

Me

[ 31% (62%) ~ over 3 steps Me

i, PCC (89%)

S

e

ii, Ph3P=CMe2(82%) OH

(+)-erogorgiaene SCHEME l l V. Total Syntheses of (-)-Colombiasin A and (-)-Elisapterosin B A. RETRO-SYNTHESIS OF (-)-COLOMBIASIN A

(-)-Colombiasin A (3) is one of the most well-recognized members of these diterpene marine natural products. It was envisioned that the combined C-H activation/Cope rearrangement would again be effective as illustrated in the retrosynthetic analysis outlined in Scheme 12. There was no reason to diverge from the published [4 + 2] cycloaddition approach for the formation of the bridged ring. 16b'16c The precursor diene 43 was

11

395

TOTALSYNTHESESOF NATURALPRODUCTS

simplified to aldehyde 44. The aldehyde 44 could be accessed with relative ease from the ester 45. The pivotal step of the synthesis, of course, would involve a combined enantioselective C-H activation/Cope rearrangement between the 1,2-dihydronaphthalene 46 and vinyl diazoacetate 18. During this step, the C6-C7 bond would be formed and the stereogenic centers at C3, C6 and C7 would be installed with the desired absolute and relative stereochemistry.

O Me

O intramolecular Diels-Alder

M MeI

~"~ ~v

"Me

(-)-colombiasinA

OR Me Grignard addition elimination

"~'H

O/~M

Nic~176 strategy

3

Me

~

e

Me" ~ "I~H O~//%M

e

43

II

0

44

OR Me

OR Me M e O ~

OR Me , M e O ~
95% ee) - - ]/ "Me 34% (68%.) y ~-I over3steps OH 57

~

Me OTBS

OMe

f H. OTBSMe

HO

Me

58

SCHEME15 C. TOTALSYNTHESESOF (-)-COLOMBIASINA AND (-)-ELISAPTEROSINB The completion of the synthesis would require access to the quinone diene 62 and this was achieved using very standard steps (Scheme 16).37 PCC oxidation of 57 afforded aldehyde 59, followed by a Grignard addition, which generated the allylic alcohol 60. Conversion of 60 to the triflate followed by elimination generated the diene 61, which was readily desilylated and air oxidized to the quinone-diene 62. The final steps of the synthesis are well established as Kim and Rychnovsky have previously shown that the diene 62 can be converted to (-)-colombiasin A by an intramolecular Diels-Alder reaction, while treatment of 62 with boron trifluoride etherate generates (-)-elisapterosin B by means of a [5 + 2] cycloaddition (Scheme 17). 16d'37 When diene 62 was heated to 180 ~ in toluene, followed by demethylation of 63 with

HUWM.L.DAVIESANDXINGDAI

398

TBSO Me MgBr MeO~ MeO~ PCC,DCM, M e O ~ (1.2 eq) - Me" -,T,- -~, H 87% S ~ "~H "~Me Me"T B"~ Me"T B"(" S O ~~I" /,~

Me 33

Me

0

34

SCHEME 17

a formal Diels-Alder adduct between cyclopentadiene and a simple enone dienophile. Furthermore, the stereochemistry of this transformation fell beautifully into place in two important ways. First of all, such a Diels-Alder adduct would certainly afford the desired relative stereochemistry of each stereocenter thanks to the Alder endo rule. Also, the 1990s were a time of much activity in the area of asymmetric Diels-Alder catalysis, and the go-to diene for nearly every paper in this area was cyclopentadiene. Thus, it seemed very likely that we would be able to obtain a fully stereocontrolled route to 32 using some variation of this overall plan. Despite the length of this section, most of these ideas arose and were solidified in the matter of only about an hour total, with probably another hour over the next few days taking a look at some of the relevant literature. Satisfied that we had a shot at a synthesis of alkaloid 251F that served the key intramolecular Schmidt step and would also permit an attractive route to the key enone 32, it was time to get to work.

430

KEVIN J. FRANKOWSKI, AARON WROBLESKI, AND JEFFREY AUBI~

IV. First Generation Work: Synthesis of the Key Enone and a Total Synthesis of Racemic Desmethyl 251F

I asked Dr. Kiran Sahasrabudhe, an excellent postdoctoral colleague who had joined my laboratory after completing a Ph.D. program at Penn State with Professor Ken Feldman, to begin work on the 251F project. Kiran had been working with us on a problem related to asymmetric ring expansions and was itching to get involved in total synthesis, so she viewed this as an attractive opportunity. We decided to initially carry out a model study of a racemic 251F derivative lacking the 3-methyl group as our first subgoal (Figure 6). The main advantages were that working out the chemistry pertaining to Me

Me

....?H

/OH ....

MeaN ~

Me

Me

(_+)-3-desmethyl251F(3)

(-)-251F(2)

Me

Me

. . . . . . .

/

. . . . . . .

H"~

/

Me

N3

N3 35

U

X H

30

..../O''

+

N3

S

95% diastereomeric purity. Reductive removal of the chiral auxiliary with LAH afforded the enantiopure alcohol 43 that was immediately benzylated to the ether 69. Finally, oxidative cleavage of the olefin afforded the desired aldehyde 70 in 36% yield over the five-step sequence. Conversion of 64 to 65 was effected by treating a mixture of 64 and cyclopentadiene in methylene chloride at - 1 0 0 ~ with diethylaluminum chloride. The Diels-Alder adduct 65 was isolated in ca. 93-95% diastereomeric purity on a 2-g scale. With protocols in hand for the asymmetric synthesis of acid 41 and the provisions met for the stereocontrolled installation of the 3-methyl group, we began to think seriously about re-tooling the ozonolysis/aldol route toward the bicyclic enone intermediate 31, given the length of this route and our concerns that in particular the lactone olefination and the aldol steps were insufficiently robust for use with the more hard-won enantiomerically pure materials. Aaron was discussing this problem at a group meeting when I had a "V8 moment", referring to an ad campaign for the popular tomato juice concoction: "We could use a metathesis reaction for that!" The thought that ruthenium-catalyzed ring-opening/ring-closing metathesis (RO/RCM) could be utilized in the desired [2.2.1] --, [3.3.0] skeletal rearrangement had very strong literature precedent. Robert Grubbs, in 1990 (i.e., before ring-closing metathesis was the force of chemistry that it has become), published a route to the triquinane natural product capnellene using titanocene alkylidene complexes (Scheme 30). 44This synthesis, which

Me

Me

CP2Ti=CH2=

Me

Me

.

MeO

M 71

M e ~

Cp2

Me01~.O TiCP2

72

Me O ~ Cp2,,,- Me

_

M

~ OMe-

73

steps H capnellene

MeO 74

SCHEME30

446


is one of very few to come out of the Grubbs laboratory, uses the Stembach Diels-Alder/ring-shuffling strategy, but with a twist. Thus, the Tebbe reagent 35 served as a source of titanaethylene (CP2Ti=CH2), which cycloadded onto the endocyclic olefin 71 and converted to the titanocene 73. In standard Tebbe fashion, the titanocene 73 olefinated the ester carbonyl with loss of O=TiCP2 to provide the skeletally rearranged enol ether 74. It seemed very attractive to update this approach for the synthesis of enone 31 using the more recently developed ring-opening metathesis catalysts based on ruthenium. 45At the time, there were already a few applications of combining ringopening metatheses with the ring-closing variety but at about the time that our work came out, many other related examples of this chemistry were starting to emerge. 46 The student of total synthesis will especially want to study some very elegant examples of this strategy used by Andrew Phillips and coworkers working at the University of Colorado. 47 So inspired, it seemed plausible that with proper manipulation of the [2.2.1]bicycloheptane skeleton of acid 41, its transformation to the desired [3.3.0]bicyclooctane enone intermediate would be possible. To accomplish this, we would need to convert the methyl ketone used for the aldol transformation into an appropriate olefin. In building this substrate, the Diels-Alder adduct 65 was cleaved to the enantiopure acid 41 with LiOH/H202 .48 Reduction of the acid to the alcohol 75 (LAH), oxidation to the aldehyde, and subsequent vinyl lithium addition provided the first potential metathesis substrate 76 as a 1:1 mixture of diastereomers (Scheme 31). Under a variety of conditions

2. 1. LAH LiOH, H202,.

OM.M.M~e~~_( O

7

Me

. oxidation . vinyllithium

.~~ .

Me I

%-~'~OH

H Ph 65

75

76 Cl,,, P(CY)3

~ u--Nph

Grubbs (77)

HO

Me

CI P(Cy)3 77

78 not observed .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.4

SCHEME 31

12

447

THE TOTAL SYNTHESIS OF AMPHIBIAN ALKALOIDS

examined, however, the desired RO/RCM did not provide 78 and only the starting allylic alcohol 76 could be detected. This was not especially discouraging as it seemed possible that the alcohol portion of 76 would have complexed the catalyst, thus inhibiting the desired reaction pathway. In addition, an additional oxidation step would have been necessary to convert 72 to the desired enone. A more economical strategy would be carrying out the tandem metathesis reaction on a substrate already at the enone oxidation state, thus avoiding the additional reduction and oxidation steps shown in Scheme 31. Thus, the conversion of 41 to the Weinreb amide 49 followed by addition of vinyl Grignard afforded 79 in 85% yield for the two-step sequence (Scheme 32). The enone 79 represented a substrate that was both free of a potentially troublesome alcohol and in the desired oxidation state of the targeted bicyclic enone 80.

1. HNMe(OMe).HCI, Me

2. vinyl Grignard

CH2Ct2

9.

85%

'"

ca. 30% ~"'~-'~/%O

41

=_

79

O

,\\ \\

+ oligomerized by-products

Me 80

SCHEME 32

At this stage it became possible to explore the use of Grubbs's ruthenium catalysts to effect the desired RO/RCM reactions. Standard reaction conditions 46a afforded the desired enone 80, albeit in poor yield. Under a variety of modified conditions examined, it appeared that a good deal of the substrate had been converted to oligomerized material. Though only a 30% yield of the metathesis product was initially isolated, we felt that this yield could be increased with the identification of more fitting conditions. Upon further examination of the literature 46b and a very productive breakfast conversation with Professor Grubbs, we soon learned that certain additives to metathesis reactions can have drastic effects on their overall efficiencies. In particular, it was recommended that we attempt the reaction in the presence of another olefin (3-hexene or ethylene) to allow for redirection of unproductive pathways back into useful chemistry (this point is discussed in detail below). As it happened, the reaction of 79 was

448


attempted using the Grubbs first generation catalyst 77 in methylene chloride with 5 equiv, of 3-hexene. The resulting product was a derivative of 80 (containing a 2-butenyl group in place of the vinyl substituent), however, the yields remained in the 30-40% range. Remarkable increases in reaction efficiencies were realized, however, when the tandem RO/RCM was carried out in ethylene-saturated methylene chloride. With a minimal amount of optimization, enone 80 was isolated in 93% yield using 5 mol% of ruthenium catalyst 77 on multigram scale reactions. Scheme 33 presents a likely mechanism for production of the desired enone 80.

79

5 mol% 77, C2H4, 0H2CI2

LnR

~

~ LnRu . ~ . . ~ . ~

~ O

~ O

81

82

_ LnRU~.,.~ "t~O 83

93% 80 SCHEME 33

Metallocycloaddition of 77 onto the endocyclic olefin of 79 generates metallocyclobutane 81, which upon retro-cycloaddition provides alkylidene 82. Typically, the norbornyl olefin is more reactive than less strained acyclic ones, and in this case the fact that the other double bond is conjugated with the ketone is expected to make it less reactive still. A second intramolecular metallocycloaddition onto the enone olefin provides metallocyclobutane 83, which again undergoes retro-cycloaddition to provide the targeted enone 80. The yield of 80 was better than tripled upon adding ethylene. As suggested earlier, an excess of ethylene present in the reaction allows for the recycling of unproductive intermediates (those leading to polymers and other byproducts) back into the desired pathway leading to enone 80. Scheme 34 represents one possible unproductive pathway (amongst n u m e r o u s others) that can lead to an alternative product. To retard polymer production, dimeric or oligomeric materials can also be reverted back to monomeric alkylidene intermediates capable of entering the pathway depicted in Scheme 33. This complex equilibrium of

12

449


u/•Me

LnR

_ -

LnRU~...[~... Me

Me

- LnR~'~. undesired

1LC=H4 %~/Me

LnRu/~ [~.... Me

[Ru] %[~....f Me _ LnRu~ . ~ ] /

~o

~o 82

~~.O

Me LnRu ~./"-~O

80 desired

83

SCHEME 34

intermediates ultimately proceeds toward what is likely the most thermodynamically stable component, enone 80. The successful incorporation of a tandem RO/RCM in place of an ozonolysis/aldol route proved to increase the efficiency and ease of bicyclic enone synthesis tremendously. In only three steps, 41 was converted to 80 in c a . 80% overall yield (compared to seven steps and 27% overall yield using the ozonolysis/aldol strategy). Utilizing this sequence, multigram quantities of 80 were produced with relative ease. As developed in the model study, it was envisioned that a one-pot conjugate addition/aldol reaction would serve to install the e x o methyl group and ultimately the azido side chain. Advancing toward the key Schmidt precursor, the enone 80 was reacted with MezCuLi to afford the intermediate enolate 84, which was quenched with the aldehyde 70 to provide the dehydrated aldol adduct 85 (Scheme 35). The aldol adduct was isolated as a single olefinic isomer in 65% yield. The geometry of the newly formed olefin was investigated using 2D NMR techniques. The use of COSY, HMQC, and HMBC allowed for the assignment of values to all the protons and carbons of 85. With a NOESY experiment, it was found that an

450

KEVIN J. FRANKOWSKI, AARON WROBLESKI, AND JEFFREY AUBt~

NOE f-h

Me

.... ,,Me2cu.i O

Me

,o

'%

Li

Me

H Me

65% (from 80)

....

0

80

84

Me

85

SCHEME 35

NOE existed between the newly installed exo methyl group and the allylic enone protons (Scheme 35). For these groups of protons to be within NOE proximity requires that the exocyclic enone olefin have an E configuration. Apparently, the additional steric interactions encountered by placing the allylic methylene group close to the C-9 methyl group on the ring were overcome by the electronic benefit of placing the alkyl chain anti to the cyclopentyl ketone. Following the protocol developed previously, 85 was converted to the azido ketone Schmidt precursor 86 (Scheme 36). The multipurpose Na/NH 3 reduction converted 85 to the corresponding reduced alcohol (not shown) as an approximately 4:1 mixture of inseparable diastereomers epimeric at the ~-position. The major isomer was presumed to be that placing the side chain in an endo orientation. A subsequent Mitsunobu reaction 39 transformed the primary alcohol to the corresponding azide 86, also as an inseparable 4:1 mixture of diastereomers (50% yield, two Me

H Me

0

1. Na/NH3 2. Zn(N3)2~ pyr, Me Me DEAD'PPh3 = N3~,,, ~ ....% 50% (2 steps) Me 0

85

TfOH or

TiCl4 ....% Me

86

Me

0

87 SCHEME 36

Me

86 plus decomposed products

12

451

THE TOTAL SYNTHESIS OF AMPHIBIANALKALOIDS

steps). With the necessary functional groups in place, the key intramolecular Schmidt reaction was then examined. Treating the azidoketone 86 with either TfOH or TiC14 resulted only in re-isolation of starting material or apparent decomposition of the olefin functionality. None of the desired cyclized lactam product 87 could be detected. This was a momentary cause for concern, but this particular problem was overcome in fairly short order. We hypothesized that the olefin was providing another site for protonation and possibly even interaction with the azide in the acidic medium; such chemistry has been extensively explored by the group of William Pearson and coworkers at the University of Michigan. 5~ Because the olefin was serving as a masked hydroxymethylene equivalent, Aaron decided to convert this compound to its oxidized analog before re-attempting the Schmidt reaction. Initial experiments focused on a one-pot conversion of the olefin to the primary alcohol using an ozonolysis/borohyride reduction sequence (Scheme 37). This method yielded the desired alcohol 88; however,

M ~ ~ k Me

\

/ Me

1.0 3

Me ....

0

....

O

/

2"NaBH4" ~

Me

Me

Me 87

Me

86

Me

1.0 3 2. excess NaBH4

Me

Me

Me

....

0

....

HO

Me

86

90 SCHEME 37

Me

\ OH

OH

452

KEVIN J. FRANKOWSKI, AARON WROBLESKI, AND JEFFREY AUBt~

incomplete reduction afforded mixtures of the desired alcohol 88 and the ozonide 89. Even more alarming was that the alcohol 88 was only isolated in 20-30% yield. When adding a larger excess of reducing reagent (NaBH4), over-reduction took place to give the diol 90 (stereochemistry neither shown nor determined). An alternative route involved carrying out the above sequence in a two-pot protocol. Because of difficulties in trying to reduce the ozonide 89 in acceptable yields, direct conversion of 86 to the corresponding aldehyde 91 was investigated (Scheme 38). Toward that end, olefin 86 was ozonized and treated with DMS to afford aldehyde 91 in c a . 70% yield. At this juncture a few experiments were conducted on 91 to determine whether this substrate was a viable Schmidt substrate. Treating the azido ketoaldehyde 91 with TfOH resulted in the detection of a lactam product (as ascertained by ~H and ~3C NMR) among a variety of unidentified impurities. Accordingly, attention was then redirected toward carrying out the Schmidt reaction on the alcohol 88. Caution was exercised with this reduction, though, as once 91 was produced, three functional groups capable of undergoing reduction were present. Reduction of either the azide or ketone eliminates one of the functional groups necessary for the Schmidt reaction. Fortunately, addition of 1 equiv, of NaBH 4 chemoselectively reduced the aldehyde in the presence of both the ketone and the azide to cleanly afford the alcohol 88 in 50-55% yield over two steps. Conversion of the 4:1 diastereomeric mixture of 86 to the alcohol 88 allowed for facile chromatographic separation of the diastereomers at this stage.

Me

Me

....

1.0 3

=

M~~k, ' M e

~' O

N3

Me

86

1. NaBH4 2. separation Me of isomers

91

Me \

D,

50- 55% (2 steps)

O O

0

Me 88

SCHEME 38

OH

Me

12

453


The key intramolecular Schmidt reaction was now tried on the alcohol 88. In this case, treatment of 88 with excess TfOH proceeded smoothly and yielded the ring-expanded Schmidt product 92 in 79% yield (Scheme 39). X-ray analysis of the crystalline lactam 92, which was to our knowledge the first carried out on any analog of 251F, nicely confirmed the tricyclic structure and orientation of all seven stereogenic centers in this penultimate intermediate. Me

Me ,||1

Me '||'k

N3

~

OH 0

79%

Me

"|'

0

88

\ OH

Me Me

92 Me

LAH iiii k

86-100~176 Me

OH Me

251F (2) SCHEME 39

As expected, LAH reduction generated alkaloid 251F 2 in 86-100% yield. Synthetic 251F was identical in all respects to published spectra (see Table 1 for comparison of 13C values). 24,26 X-ray crystallography and 2D NMR analysis are immensely powerful methods for determining molecular structure and stereochemistry as has been demonstrated throughout this chapter. Far less prevalent, though, is the use of IR spectroscopy to help determine the stereochemistry of natural products. Figure 9 provides the IR spectra of both the lactam 92 and alkaloid 251F. Excluding the lack of a carbonyl absorption for 251F (as expected after LAH reduction), there is a subtle difference between the two IR spectra apparent in the C-H stretching region. Noticeable in the IR spectrum of 251F are two small bands at 2756 and 2800 cm-~; clearly lacking in the lactam spectrum, these are known as Bohlmann bands. 19 Such stretches are observed when a nitrogen lone pair is antiperiplanar to three C-H bonds as is the case in 251F. The nitrogen lone pair places electron density into the antibonding orbitals of the C-H bonds. Consequently, the C-H bonds assume a different stretching frequency,

454

KEVIN J. FRANKOWSKI, AARON WROBLESKI, AND JEFFREY AUBI~ TABLE 1 Experimentally determined and reported ~3C values for alkaloid 251F Entry

Experimental J3C NMR for 251F (125 MHz) (ppm)

Daly's reported 13C NMR for 251F (ppm)

14.7 15.6 17.5 27.1 28.3 30.2 31.4 35.3 37.6 41.2 44.3 47.3 53.1 61.3 64.8 67.3

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

14.8 15.7 17.5 27.1 28.4 30.2 31.4 35.4 37.7 41.3 44.4 47.4 53.2 61.4 64.8 67.4

ME

I 60

.... \

55

Me 0

Me

OH

92 aO00

;r162 wt~e-,u,~b~t O::m.t)

ID 84

Me

10

i"

~o

MeJ~~~N~

\ .... OH Me

251F (2) w

~

{r

FIGURE 9. The IR spectra showing the characteristic Bohlmann bands. Note the two small absorptions at 2800 and 2756 cm-1.

12


455

and it is only when the combined effects of three C-H bonds vibrating at a new frequency occur that these bands are observed. Bohlmann bands were absent in the lactam because a carbonyl occupied the position where a third antiperiplanar C-H bond was needed. In fact, the observation of Bohlmann bands in the racemic 3-desmethyl 251F synthesis was the first piece of evidence obtained, indicating that the azido side chain occupied the desired e n d o position. It was not until later that sophisticated 2D NMR techniques and X-ray crystallography confirmed what the presence of Bohlmann bands had already indicated. VII. Final Remarks

Overall, the asymmetric total synthesis of 251F proceeded in 13 steps with an overall yield of ca. 5-8%. Approximately 100 mg of the natural product was produced in this way. In Daly's report on the isolation and structural elucidation of 251F, it was noted that 100 frog skins were required to isolate only ca. 300 ~g of the alkaloid (or ca. 3 lag per frog). 25 Accordingly, 100 mg of 251F corresponds to the amount of alkaloid that would be harvested from in excess of 30,000 frog skins! The potential of natural sources to provide fascinating new compounds remains unchallenged. However, it is clear that the thoughtful use of synthesis to provide rare compounds in quantity is not only valuable for feeding downstream biological applications, but also can play a role in conserving precious natural resources. The synthetic materials have been submitted to various programs to determine their pharmacological profiles, but the results of these studies will have to wait for future publications. One interesting element of this project arises from the fact that we made synthetic 251F in enantiomerically pure form. However, the absolute stereochemistry of the naturally occurring material is unknown. This leads to the possibility that our enantiomerically pure (-)-251F would correspond to the less biologically potent version of the molecule (assuming that 251F is biologically active in the first place). This point can only be settled through experiment, but the availability of r a c e m i c desmethyl-251F could be valuable for estimating the biological potential of the series enantiomeric to our (-)-251F preparation. In conclusion, a study ending in a synthesis of racemic 3-desmethyl 251F allowed us to develop chemistry that ultimately culminated in an asymmetric total synthesis of alkaloid 251F. Key to both syntheses was the use of an intramolecular Schmidt reaction on advanced intermediates, which delivered the core tricyclic skeleton of the natural product. In addition,

456

KEVIN J. FRANKOWSKI,AARON WROBLESKI,AND JEFFREY AUBt~

a second-generation approach toward an important bicyclic enone was developed that greatly streamlined the synthetic sequence toward 251F. At seven steps and 27% overall yield, the ozonolysis/aldol strategy toward enone 31 proved satisfactory. However, replacing this route with a domino metathesis strategy better than halved the number of steps (total three) and tripled the yield (80%) e n r o u t e to the analogous enone 80. And of course, these syntheses demonstrated the use of the intramolecular Schmidt reaction in the synthesis of increasingly complex alkaloids. This was an extremely enjoyable project from beginning to end. In retrospect, the Schmidt chemistry was not really the highlight from the perspective of the principal investigator, but rather the recognition that we could explore other chemistry, not so familiar to us, in ways that would really make the Schmidt steps shine. For example, I had been impressed with the Sternbach approach to quinane natural products for years, and was finally able to utilize this elegant scheme to procure a product that was important to our program. In addition, we were able to plug into the RO/RCM whirlwind that has become such an everyday part of fine organic synthesis. The synthesis of 251F was an important milestone in the group in that it demonstrated that the intramolecular Schmidt reaction would not necessarily be relegated to the synthesis of simple bicyclic compounds. Acknowledgments The three authors would like to thank, first and foremost, the other members of the laboratory who played decisive roles in the chemistry described in this document. They are, in chronological order, Greg Milligan, Pat Rafferty, and Kiran Sahasrabudhe. None of this work could have been done without their contributions. We have all benefited from a most stimulating work environment at the University of Kansas provided by other members of the group and would like to acknowledge all of the group members who contributed to this positive atmosphere. Other important technical contributions made by Douglas Powell (Xray) and David Vander Velde (NMR) are also acknowledged, and we are grateful to all of the scientists, named throughout the chapter, who shared their insights with us along with the way and provided useful suggestions. We are also grateful to Dr. John Daly, who provided both insights and valuable comments on our publications at various times throughout this project. Dr. Daly is to be congratulated for also providing the worldwide community of organic chemists with a lifetime of challenges and inspiration through his singular research program. Of course none of this work could have been accomplished without the generous support of the United States taxpayers as administered by the National Institutes of Health. We are specifically grateful to the National Institute of General Medical Sciences for its continuous support of our alkaloid synthesis program through GM-49093.

12

THE TOTALSYNTHESISOF AMPHIBIANALKALOIDS

457

References and Footnotes

1. Aub6, J., Chem. Soc. Rev. 1997, 26, 269. 2. (a) Aub6, J., Milligan, G. L., J. Am. Chem. Soc. 1991, 113, 8965. (b) Aub6, J., Milligan, G. L., J. Am. Chem. Soc. 1995, 117, 10449. 3. Daly, J. W., J. Nat. Prod. 1998, 61, 162. 4. Albuquerque, E. X., Daly, J. W., Witkop, B., Science 1971, 172, 995. 5. Daly, J. W., Proc. Natl. Acad. Sci. USA. 1995, 92, 9. 6. Daly, J. W., Spande, T. E, Garraffo, H. M., J. Nat. Prod. 2005, 68, 1556. 7. Daly, J. W., J. Med. Chem. 2003, 46, 445. 8. LaLonde, R. T., Muhammad, N., Wong, C. F., Sturiale, E. R., J. Org. Chem. 1980, 45, 3664. 9. Le Dr6au, M.-A., Desma6le, D., Dumas, E, d'Angelo, J., J. Org. Chem. 1993, 58, 2933. 10. (a) Smith, A. L., Williams, S. E, Holmes, A. B., Hughes, L. R., Lidert, Z., Swithenbank, C., J. Am. Chem. Soc. 1988, 110, 8696. (b) Gnecco, D., Marazano, C., Das, B. C., J. Chem. Soc., Chem. Commun. 1991, 625. (c) Holmes, A. B., Smith, A. L., Williams, S. E, Hughes, L. R., Lidert, Z., Sithenbank, C., J. Org. Chem. 1991, 56, 1393. (d) Shishido, Y., Kibayashi, C., J. Org. Chem. 1992, 57, 2876. 11. (a) Yamazaki, N., Kibayashi, C., Tetrahedron Lett. 1988, 29, 5767. (b) Nagao, Y., Dai, W.-M., Ochiai, M., Tsukagoshi, S., Fujita, E., J. Am. Chem. Soc. 1988, 110, 289. (c) Yamazaki, N., Kibayashi, C., J. Am. Chem. Soc. 1989, 111, 1396. (d) Nagao, Y., Dai, W.-M., Ochiai, M., Tsukagoshi, S., Fujita, E., J. Org. Chem. 1990, 55, 1148. (e) Polniaszek, R. P., Belmont, S. E., J. Org. Chem. 1991, 56, 4868. (f) Polniaszek, R. P., Belmont, S. E., J. Org. Chem. 1990, 55, 4688. (g) Taber, D. E, Deker, P. B., Silverberg, L. J., J. Org. Chem. 1992, 57, 5990. 12. Yao, L., Smith, B. T., Aub6, J., J. Org. Chem. 2004, 69, 1720. 13. Marx, J. N., Norman, L. R., J. Org. Chem. 1975, 40, 1602. 14. Mirrington, R. N., Schmalzl, K., J., J. Org. Chem. 1972, 37, 2871. 15. (a) Loibner, H., Zbiral, E., Helv. Chim. Acta 1977, 60, 417. (b) Chen, C.-P., Prasad, K., Repic, O., Tetrahedron Lett. 1991, 32, 7175. 16. Mossman, C. J., Aub6, J., Tetrahedron 1995, 52, 3403. 17. Lipshutz, B. H., Harvey, D.F., Synth. Commun. 1982, 12, 267. 18. Stevens, R. V., On the stereochemistry of nucleophilic additions to tetrahydropyridinium salts: A powerful heuristic principle for the stereorationale design of alkaloid syntheses. In: Strategies and Tactics in Organic Synthesis, vol. 1, Lindberg, T. (Ed.), Academic Press: Orlando, FL, 1984, pp. 275-298. 19. (a) Bohlmann, E, Angew. Chem. 1957, 69, 541. (b) Bohlmann, E, Chem. Ber. 1958, 91, 2157. (c) Gribble, G. W., Nelson, R. B., J. Org. Chem. 1973, 38, 2831. (d) Carson, J. R., Carmosin, R. J., Vaught, J. L., Gradocki, J. J., Costanzo, M. J., Raffo, R. B., Almond, H. R. J., J. Med. Chem. 1992, 35, 2855. (e) Pearson, W. H., Gallagher, B. M., Tetrahedron 1996, 52, 12039. (f) For a review, see: Crabb, T. A., Newton, R. E, Jackson, D., Chem. Rev. 1971, 71, 109. 20. Daly, J. W., Spande, T. E Amphibian alkaloids: Chemistry, pharmacology and biology. In: Alkaloids: Chemical and Biological Perspectives, vol. 4, Pelletier, S. W. (Ed.), Wiley: New York, 1986, pp. 1-274. 21. Aub6, J., Rafferty, P. S., Milligan, G. L., Heterocycles 1993, 35, 1141.

458

KEVIN J. FRANKOWSKI,AARONWROBLESKI,AND JEFFREYAUBt~

22. (a) Momose, T., Toyooka, N., J. Org. Chem. 1994, 59, 943. (b) Aehman, J., Somfai, E, Tetrahedron 1995, 51, 9747. (c) Jefford, C. W., Sienkiewicz, K., Thornton, S. R., Helv. Chim. Acta 1995, 78, 1511. (d) Michael, J. E, Gravestock, D., Synlett 1996, 981. (e) Bardou, A., Celerier, J.-E, Lhommet, G., Tetrahedron Lett. 1998, 39, 5189. (f) Michael, J. E, Gravestock, D., J. Chem. Soc., Perkin 1 2000, 1919. (g) Back, T. G., Nakajima, K., J. Org. Chem. 2000, 65, 4543. (h) Michel, P., Rassat, A., Daly, J. W., Spande, T. E, J. Org. Chem. 2000, 65, 8908. (i) Shu, C., Alcudia, A., Yin, J., Liebeskind, L. S., J. Am. Chem. Soc. 2001, 123, 12477. (j) Song, Y., Okamoto, S., Sato, E, Tetrahedron Lett. 2002, 43, 8635. (k) Ma, D., Pu, X., Wang, J., Tetrahedron: Asymmetry 2002, 13, 2257. (1) Davis, E, Yang, B., Org. Lett. 2003, 5, 5011. (m) Toyooka, N., Dejun, Z., Nemoto, H., Garraffo, H. M., Spande, T. E, Daly, J. W., Tetrahedron Lett. 2006, 47, 577. (n) Davis, E A., Yang, B., Deng, J., Zhang, J., ARKIVOC 2006, 120. 23. Myers, C. W., Daly, J. W., Am. Mus. Novit. 1980, 1. 24. Myers, C. W., Papdis Avulsos Zool. S~to Paolo 1987, 36, 301. 25. Spande, T. F., Garraffo, H. M., Yeh, H. J. C., Pu, Q.-L., Pannell, L. K., Daly, J. W., J. Nat. Prod. 1992, 55, 707. 26. Daly, J. W., Garraffo, H. M., Spande, T. E, Amphibian alkaloids. In: Alkaloids, vol. 43, Cordell, G. A. (Ed.), Academic Press: New York, 1993, pp. 185-288. 27. Taber, D. F., You, K. K., J. Am. Chem. Soc. 1995, 117, 5757. 28. Wender, E A., Ternansky R. J., Tetrahedon Lett. 1985, 26, 2625. 29. Gracias, V., Zeng, Y., Desai, E, Aub6, J., Org. Lett. 2003, 5, 4999. 30. Sternbach, D. D., Hughes, J. W., Burdi, D. E, Banks, B. A., J. Am. Chem. Soc. 1985, 107, 2149. 31. Sternbach, D. D., Ensinger, C. L., J. Org. Chem. 1990, 55, 2725. 32. Sternbach, D. D., A new strategy for the synthesis of polyquinanes. In: Strategies and Tactics in Organic Synthesis, vol. 2, Lindberg, T. (Ed.), Academic Press: San Diego, CA, 1989, pp. 415-438. 33. (a) Hoffmann, H. M. R., Vathke-Ernst, H., Chem. Ber. 1981, 114, 2208. (b) Parlar, H., Baumann, R., Korte, E, Z. Naturforsch. B: Anorg. Chem., Org. Chem. 1981, 36B, 898. 34. (a) Burke, S. D., Murtiashaw, C. W., Dike, M. S., J. Org. Chem. 1982, 47, 1349. (b) Burke, S. D., Murtiashaw, C. W., Saunders, J. O., Oplinger, J. A., Dike, M. S., J. Am. Chem. Soc. 1984, 106, 4558. 35. Tebbe, E N., Parshall, G. W., Reddy, G. S., J. Am. Chem. Soc. 1978, 100, 3611. 36. Petasis, N. A., Bzowej, E. I., J. Am. Chem. Soc. 1990, 112, 6392. 37. Dollinger, L. M., Ndakala, A. J., Hashemzadeh, M., Wang, G., Wang, Y., Martinez, I., Arcari, J. T., Galluzzo, D. J., Howell, A. R., J. Org. Chem. 1999, 64, 7074. 38. Wrobleski, A., Aub6, J., J. Org. Chem. 2001, 66, 886. 39. Viaud, M. C., Rollin, E, Synthesis 1990, 130. 40. (a) Vandewalle, M., Van der Eycken, J., Oppolzer, W., Vullioud, C., Tetrahedron 1986, 42, 4035. (b) Fukuzawa, S., Matsuzawa, H., Metoki, K., Synlett 2001, 709. (c) Krotz, A., Helmchen, G., Tetrahedron: Asymmetry 1990, 1,537. (d) Evans, D. A., Barnes, D. M., Johnson, J. S., Lectka, T., yon Matt, E, Miller, S. J., Murry, J. A., Norcross, N. D., Shaughnessy, E. A., Campos, K. R., J. Am. Chem. Soc. 1999, 121, 7582. (e) Evans, D. A., Chapman, K. T., Bisaha, J., J. Am. Chem. Soc. 1988, 110, 1238. (f) Evans, D. A., Miller, S. J., Lectka, T., J. Am. Chem. Soc. 1993, 115, 6460. (g) Evans, D. A., Miller, S. J., Lectka, T., yon Matt, E, J. Am. Chem. Soc. 1999, 121, 7559.

12

THE TOTALSYNTHESISOF AMPHIBIANALKALOIDS

459

41. (a)Evans, D. A., Miller, S. J., Lectka, T., J. Am. Chem. Soc. 1993, 115, 6460. (b) Desimoni, G., Faita, G., JCrgensen, K. A., Chem. Rev. 2006, 106, 3561. 42. (a) Evans, D. A., Chapman, K. T., Bisaha, J., J. Am. Chem. Soc. 1988, 110, 1238. (b) Poll, T., Abdel Hady, A. F., Karge, R., Linz, G., Weetman, J., Helmchen, G., Tetrahedron 1989, 30, 5595. 43. Evans, D. A., Bender, S. L., Morris, J., J. Am. Chem. Soc. 1988, 110, 2506. 44. (a) Stille, J. R., Grubbs, R. H., J. Am. Chem. Soc. 1986, 108, 855. (b) Stille, J. R., Santarsiero, B. D., Grubbs, R. H., J. Org. Chem. 1990, 55, 843. 45. For reviews on ruthenium metathesis catalysts see: (a) Sanford, M. S., Love, J. A., Grubbs, R. H., J. Am. Chem. Soc. 2001, 123, 6543. (b) Trnka, T. M., Grubbs, R. H., Acc. Chem. Res. 2001, 34, 18. 46. Selected examples include: (a) Zuercher, W. J., Hashimoto, M., Grubbs, R. H., J. Am. Chem. Soc. 1996, 118, 6634. (b) Stragies, R., Blechert, S., Synlett 1998, 169. (c) Arjona, O., Csaky, A. G., Medel, R., Plumet, J., J. Org. Chem. 2002, 67, 1380. (d) Hagiwara, H., Katsumi, T., Endou, S., Hoshi, T., Suzuki, T., Tetrahedron 2002, 58, 6651. (e) Weatherhead, G. S., Ford, J. G., Alexanian, E. J., Schrock, R. R., Hoveyda, A. H., J. Am. Chem. Soc. 2000, 122, 1828. (f) Zuercher, W. J., Scholl, M., Grubbs, R. H., J. Org. Chem. 1998, 63, 4291. (g) Lee, D., Stello, J. K., Schreiber, S. L., Org. Lett. 2000, 2, 709. 47. (a) Minger, T. L., Phillips, A. J., Tetrahedron Lett. 2002, 43, 5357. (b) Pfeiffer, M. W. B., Phillips, A. J., J. Am. Chem. Soc. 2005, 127, 5334. (c) Chandler, C. L., Phillips, A. J., Org. Lett. 2005, 7, 3493. (d) Hart, A. C., Phillips, A. J., J. Am. Chem. Soc. 2006, 128, 1094. (e) Phillips, A. J., Hart, A. C., Henderson, J. A., Tetrahedron Lett. 211116,47, 3743. 48. Evans, D. A., Britton, T. C., Ellman, J. A., Tetrahedron Lett. 1987, 28, 6141. 49. Nahm, S., Weinreb, S. M., Tetrahedron Lett. 1981, 22, 3815. 50. For recent examples see: (a) Pearson, W. H., Hutta, D. A., Fang, W.-K., J. Org. Chem. 2000, 65, 8326. (b) Pearson, W. H., Hines, J. V., J. Org. Chem. 2000, 65, 5785. (c) Schkeryantz, J. M., Pearson, W. H., Tetrahedron 1996, 52, 3107.


Chapter 13 SEDUCED BY A SIREN'S CALL" EXPANDING APPLICATIONS FOR AROMATIC COMPOUNDS AND THE SYNTHESIS OF (+)-RISHIRILIDE B Thomas Pettus and Todd Wenderski Department of Chemistry and Biochemistry University of California at Santa Barbara Santa Barbara, CA 93106-9510, USA

Burgeoning Interest in Aromatic Compounds Why Choose a Particular Phenol Skeleton for Oxidation? II. Developing Easy Access to Various Resorcinol Systems III. Total Synthesis of (_+)-Epoxysorbicillinol IV. Plans for Chiral Auxiliaries Disintegrate V. Stereochemistry Describes an Unexpected Transition State VI. VII. Beating the Clock, Tuning the Reaction and Improving Yields VIII. Controlling B-Dione Tautomerization and Protection IX. A Lactic Acid-Derived Directing-Protecting Directing Group X. New Methods for Cleavage of the Directing-Protecting Group XI. The Total Synthesis of (+)-Rishirilide B XII. New Horizons for This Method and Oxidative Dearomatizations Acknowledgments References and Footnotes I~

460 464 466 471 472 473 474 476 478 482 483 484 488 488

I. Burgeoning Interest in Aromatic Compounds

It is often tough to tell when a story truly begins. I will start at Virginia Tech, where I was working in the laboratories of Professor Tom,is Hudlicky, (T'Hud). My organic chemistry college professor, Dr. Maurice Maxwell (Max), was visiting VPI for a summer research program and had brought me along. I was assigned to the bench of Gustavo Seoane, who served as my immediate mentor and handler for the summer of 1988. Gustavo's mission, which no doubt doubled his workload, was to prevent me from hurting myself, others, and taking out a wall while carrying out an experiment. T'Hud had recently recognized the importance of a Gibson paper reporting

13

461

SEDUCED BY A SIREN'S CALL

the formation of (+)-cis-2,3-dihydroxy- 1-methyl-cyclohexadiene from toluene by a mutant strain (39D) of the soil microbe Pseudomonas putida, and he and his group were just beginning to finish enantioselective syntheses of terpenes and prostanoids from arene diols (Scheme 1). 1 T'Hud had not yet been mellowed by the birth of his son and in those days he ran about the lab with a cigarette in hand whipping his students into line.

I

HO

Pseudomonas putida

succ,n,c

HO"

o

Gibson 1970

anhydride

S C H E M E 1. Gibson's observation with PP-39D. 2

T'Hud publicized a list of broken glassware along with its past owner. The idea was that the list, posted above the sink at the first bench found when entering the lab, would enable glassware to be efficiently reordered and returned to its original owner. In reality, the list indicated those with the worst technique and I seemed to be its most frequent contributor. After adding some rather expensive separatory funnels, I was banished to the microbiology lab. The microbiology lab at that time was an old broom closet in the bowels of Davidson laboratory. Here, surrounded by plastic petri dishes, I was assigned the task of generating the group's supply of arene diol from styrene (Scheme 2). The process began by growing agar plates of the soil microbes for about three days, followed by incubation of the microbe in the presence of the styrene, and subsequent feeding of the styrene in correct proportions with oxygen to a sugary aqueous suspension of microbes. The feeding apparatus had been cobbled together from plywood, Erlenmeyer flasks, two fish tank pumps, and two airflow control devices. By mid-afternoon in the sweltering heat of Blacksburg, Virginia, the small non-air-conditioned microbiology laboratory reeked of unconsumed styrene from the past twenty-four hours of force-feeding. The microbes, now dead, stewed in their own arene diol excrement. After centrifuging

Pseudomonas. putida

o.

A, oBz zeylena

HO,,

steps = O : : ~ / I ~ L ~/

HO"

IMDA

"i --" Ph

1989

S C H E M E 2. Total synthesis of (-)-zeylena from styrene?

462

THOMAS PETTUS AND TODD WENDERSKI

2 L of PP-broth, and a labor-intensive emulsion prone extraction, Mother Nature rewarded me with 500 mg of the corresponding arene diol. In all likelihood, this material would polymerize, forming a beautiful pink solution if not immediately purified. By the end of the summer, I had learned many things, but three stood out. First, I hated the smell of styrene. Second, unless you wanted it to be shredded, never show T'Hud a spectrum containing ethyl acetate. Thirdly, an aromatic ring could be an extremely useful six-carbon building block, but the bugs could provide only one enantiomer of it. By the end of the summer, I inquired about my graduate school options. T'Hud provided me with two choices: "Work for me or for Dick Schlessinger at UR." From what little that I can recall of my graduate school experience beyond the truth of the adage "out of the frying pan and into the fire," my fascination with the economy and elegance of aromatic intermediates continued to grow. While working in the Schlessinger group on the 4th floor of Hutchinson Hall, I discovered my talent for the reverse process of my work in T'Hud's group and converted many delicate chiral building materials into useless aromatic compounds (Scheme 3). Needless to say, Schlessinger shared his trademark expression of disgust and utter reprehension by flicking a lit cigarette into my hood in the midst of an extraction with ether. O Br

a) KHMDS c) dimethylfumarate

TMSO CO2Me BC, " ~~~

OH Br~CO2Me H~

/r,. OMe

"SOMe

SCHEME 3. Destruction of chirality. 4

While collecting references in Carlson Library for an upcoming NIH proposal, I encountered some beautiful and inspirational syntheses starting from aromatic building blocks that employed oxidative dearomatization for the introduction of chirality. These syntheses included Delongchamp's ryanodol, 5 Danishefsky's calicheamicinone, 6 and Corey's ovalicin (Figure 1).7 All had chosen to begin with an aromatic compound and arrived at an ornate albeit racemic natural product. With Schlessinger's NIH grant on asymmetric Diels-Alder reactions of chiral amino furans renewed, I escaped the University of Rochester. I married my bench mate, Dr. Liping Hong, whose earlier defection from the

13

463

SEDUCEDBY A SIREN'S CALL

NO .~ ~ Me Me "~Me~~

OH I OH Me

Me

O

/~NHCO2Me /O Me Me H O(,/ " HO ~ H .~. I ?H!~O .~lMe

~o o. L-..,,NH (+)-ryanodol

MeSSS (-)-calicheamicinone

ovalicin

FIGURE 1. Structures built by oxidative dearomatization.

Schlessinger group had often served as a point of inspiration. Liping completed her Ph.D. with Rick Borch at University of Rochester and began work with Art Schultz at Rensselaer Polytechnic Institute. Professor Schultz coincidently was among Schlessinger's first Ph.D. students and his research focused on the reductive counterpart to oxidative dearomatization. In 1996, I moved to NYC and began work in Havemeyer Hall at Columbia University as an NSF fellow on a terpene called tricycloillicinone. Shocked by my survival of graduate school, and thankful that my new advisor did not smoke, so that I might now quit, I busied myself making many aromatic compounds with a Nicoderm CQ | patch attached to each of my arms. The initial [5+2] dearomatization strategy failed. Instead, we devised a radical cyclization (Scheme 4). However, neither Danishefsky, group members, nor I could imagine a straightforward nonracemic entry leading to the key o-quinol building block.

OTBDPS 50~ O O

OTBDPS ~

O

1) Snider cyclization

~

-

O~oo~O 2)carbonylerasure " ~

(_+)-tricycloilicinone1998

SCHEME 4. Total synthesis of tricycloillicinone.

After these diverse experiences with aromatic compounds, most would have wisely chosen to avoid them further. Sure, it may seem that aromatic materials offer attractive starting points for synthesis, but in reality the yields for most aromatic transformations are exceedingly low, certainly not as high as those found for most aldol and polyketide chemistry, pervasive throughout the chemical literature these days. Moreover, despite a hundred years of research, the control of regiochemistry during aromatic

464


substitution remains difficult, and distinguishing between the phenols of a poly-hydroxylated aromatic compound is still problematic. Although metalmediated reactions have overcome some of the problems, the issue of regiocontrol is often pushed upstream in a strategy requiting selective access to a halogenated aromatic compound. Therefore, chemists still resort to chemically beating an aromatic compound to afford a mixture containing some of the desired substitution pattern and leave the rest up to a mechanical separation. Like a moth to the flame, I was seduced by the malleability embedded within the aromatic skeleton, despite intimate knowledge of these problems, and decided to find some new procedures for transforming them into useful building blocks. II. Why Choose a Particular Phenol Skeleton for Oxidation?

No general method for enantioselective oxidative dearomatization existed, but I was not alone in the pursuit of a procedure. From careful inspection of past literature, 8 four intractable obstacles seemed to have barred the development of enantioselective oxidative dearomatization processes from offering genuine utility for asymmetric synthesis. First, oxidative dearomatizations proceed in low yield and, as all synthetic chemists know, a low yield dooms a synthesis to inefficiency and anonymity. Therefore, the few notable examples of oxidative dearomatization in synthesis have applied the transformation at a very early stage when starting material was abundant or near the end when the target could be reached without much further exploration. Many have claimed improvements for oxidative dearomatization involving additives and/or solvent effects. However, upon critical examination and comparison, it appears that only kinetically favored intramolecular oxidative dearomatizations, such as the Wesley oxidation, consistently provide good yields (>65%); the rest is mostly rubbish. Secondly, oxidative dearomatization results in an oxygenated derivative, be it a p-quinol, o-quinol or an arene diol. Therefore, subsequent steps are required to protect the acidic hydroxy residue before most subsequent reactions. Since the protection step had to be followed at some later point by a deprotection step, the efficiency of the method is again lowered. Exceptions to this problem are application of the method at the conclusion of a synthesis or inclusion of some useful functionality such as the epoxide generated by the Wesley oxidation. Thirdly, oxidative dearomatization usually leads to derivatives of p-quinol, o-quinol, and arene diols that are very unstable and susceptible to re-aromatization. This problem is usually nucleus dependent. For example,

465

13 SEDUCEDBYA SIREN'SCALL

o-quinols are susceptible to dimerization, whereas p-quinols succumb to single electron transfer reduction and rearomatization; arene diols undergo elimination and/or reoxidation. None of these problems can be avoided altogether. However, electronic and structural features can be introduced to dissuade undesired reactivity. Fourthly, control of regiochemistry and stereochemistry in the subsequent reactions of the products afforded by oxidative dearomatization often proves challenging. The problem stems from the nature of the product, which as a rule contains at least four or five trigonal carbon atoms. Without many tetrahedral atoms or other steric control elements, the planar products undergo reactions from both faces of their rt-systems; the stereocenter usually exerts a sizable influence only on vicinal tfigonal atoms. Given these four problems, most organic chemists have wisely avoided the enchanting siren's call of the functionality embedded within aromatic starting materials. Those who have pressed-on have cobbled together a series of fixes to sidestep these impediments. For example, one of the first methods aimed at chemically producing non-racemic building blocks from aromatic compounds carried out the finicky dearomatization reaction early, thereby producing a less precious achiral starting material (Scheme 5). This material is then trans-ketalized with a chiral diol. In principle, several diastereomers could arise. However, positioning the larger substituent in the equatorial position of the newly created 1,3-dioxane is favored. Unfortunately, chirality is far removed from subsequent reaction sites. Thus, the chemoselectivity and stereoselectivity of subsequent reactions suffer. Nevertheless, the method has been applied in the synthesis of miroestrol, 9 LL-C10037R, ~~and (-)-jesterone. ~

R~~

OMe

oxidize,, yield R2 10-65%

OH O RI'"~.

O R2

R OH

MeO OMe R ~ O achiral

transketalization =

R2

partially diastereoselective reaction, some / chemoselectivity

-._1,2-addition of R3deketalization

RZ~.. R2 O

RI~~o R2 0

SCHEME5. Trans-ketalizationof a quinonemonoketalwith a chiral diol.

466

THOMAS PETFUS AND TODD WENDERSKI

In a related strategy, the phenol is subjected to dearomatization affording a symmetric p-quinol. The non-precious symmetric system is then subjected to a desymmetrization (Scheme 6). The newly added functionality is then incorporated into the final target. This method has been used in the synthesis of manumycin A. ~2

OH ~

0 0 oxidation10_40% ~ ~1~. Desymmetrization ~J~.b"pO=

R3

R3 "OP achiral

chiral

SCHEME 6. Desymmetrization after oxidation.

In a similar approach, the achiral p-quinol is temporarily desymmetrized. After some further selective reactions, the blocking group, which also serves to direct the stereochemical outcome of succeeding reactions, is then removed. This modification has been employed in the synthesis of diepoxin. J3 IH.

Developing Easy Access to Various Resorcinoi Systems

While some advances have been reported, it was evident to us that much work remained in order to develop a versatile method that would provide genuine applicability to a diverse range of molecules. Our initial plan aimed at addressing this problem seemed simple enough (Scheme 7)--construct a 4-alklyated resorcinol (C) with differentially protected phenols and then submit a mono-deprotected variant of C to chemical oxidation. In principle, the structure that emerged (A or B) O

~,~.. / oxidation P" . R A,P"= -H ~ OP" P"= -H

OP"

P'~ c

B, P'=-H SCHEME 7. o- and p-quinols via selective protection.

13

467


should reflect which phenol oxygen atom was unprotected at the time of oxidation as well as the atom ortho or para to the unprotected phenol displaying the greatest cationic coefficient in the intermediate. After reviewing the chemical literature, we chose to target p-quinols (A), with the notion that this platform would be more easily adaptable to enantioselective processes because of the proximity of the - O P ' residue, as well as the perceived controllable and divergent reactivity of A. Our next step was to secure a route to differentially protected 4-alkyl resorcinols such as C. This notion proved to be harder in practice than we had ever imagined (Scheme 8). All attempts to cause the more accessible phenol in 1 to undergo a selective protection failed. The reactions produced a mixture of 2, 3, and 4 that could not be readily separated. In hindsight, this problem should have been anticipated. First, there is a solubility issue. The mono-protected resorcinols 2 and 3 are vastly more soluble than the starting resorcinol 1. Secondly, the A-values that are normally associated with steric size and often used to predict the outcome of a reaction apply to sp 3 systems. In an aryl ring, the steric effects are only felt in two planes. Careful inspection of Protecting Groups by Greene and Wuts reveals only one example of selective protection of 4-t-butyl resorcinol with pivaloyl chloride and a bulky amine. ~5 It is clearly a loaded example.

~ HO 1

OH

OP"

largeprotecting ~ group oragent amine HO transfer TBDPSCi-IPr2NEt , orPivCI,i-Pr2NEt 2

~

OH

P'"O

3

OP""

~ P"'O 4

SCHEME 8. Selective protection proves impossible. 14

Obviously, a better entry into the resorcinol starting material was needed for us to continue. 4-Bromo-resorcinol was commercially available, and for a time we considered metal-mediated coupling reactions. However, 4-bromo-resorcinol has rarely been utilized in a palladium mediated coupling because of the electron rich nature of the C-Br bond due to donation from the ortho and para C-O residues. Therefore, we concluded that development of new chemistry was necessary to address differentially protected 4-alkyl resorcinols in an efficient manner. Based on McLoughlin's observations regarding the sodium borohydride reductions of o-O-acetyl ketones, j6 we invented some simple procedures

468


that could afford a vast range of differentially protected 4-alkylated resorcinols. McLoughlin speculated that the corresponding reduction involves an o-quinone methide intermediate. Our amended sequence proceeds from an o-O-acylated 2,4-dihydroxybenzaldehyde 5 to the elaborated material 7 in one or two-pot processes (Scheme 9). 17 The latter involves addition of a Grignard reagent to compound 6. The o-OBoc group was selected after considerable experimentation and best enables use of a wide range of alkyl magnesium and lithium reagents. OBoc

OBoc

BocO 5

1 pot

H R2

0/~"-- H

BHs.DMS~

R2MgB/ / in Et20 OBoc

BocO 6

R2= H Me etc R1= H Me Br

HO j

SCHEME 9. New access to 4-alkylated resorcinols via o-QMs.

With easy access to a range of 4-alkylated resorcinols, we next examined oxidative conditions intended to afford a p-quinol derivative in some protected form. In 1997, Pelter reported that exposure of a phenol to [bis(trifluoroacetoxy)iodo]benzene (PIFA) resulted in an ortho cyclization with the attached alcohol becoming embedded within the product (Scheme 10). At the time, it was the only example of an ortho cyclization. All of the other reports of intramolecular oxidative dearomatizations involved an ipso cyclization, in which the nucleophile attacked the aryl carbon atom to which it was attached. Although the order of events could not be known with certainty, ortho cyclization seemed assured, given the gem-dimethyl and cis olefin residues. OH

O

PIFA MeOH

~O

1997 Me

SCHEME 10. First example of ortho oxidative cyclization. ~s

13

469


Pelter's loaded example inspired our first idea: oxidation of the carbamate 8. Nontoxic hypervalent iodide reagents have become the reagents of choice for oxidative dearomatization. In principle, exposure of 8 to PIFA or [bis(acetoxy)iodo]benzene (PIDA) would lead to a cation intermediate and a cyclization would ensue, whereupon addition of water to the iminium intermediate should provide the desired carbonate 9 (Scheme 11). PIFA is believed to involve cation intermediates, whereas PIDA tends to involve single electron transfer. Despite a Herculean effort on the part of then graduate student Ryan Van De Water, the reaction failed to produce any of the carbonate 9. Instead, paltry amounts of products were isolated, which indicated addition of a ligand from the oxidant to the ethylated carbon of the starting phenol.

OH

0

0 .

8

.

.

.

o, '

Et.N0_...0 ~ Et

.

0

l~t

o L.

Et-N,Et

0 9

S C H E M E 11. First idea fai|s. 19

Dreiding models suggested that this outcome reflects the failure of the carbonyl to adopt an alignment necessary for cyclization. Therefore, the chain was extended by one carbon atom and the corresponding amide was next examined. To produce this material, the differentially protected resorcinol 10 was coupled with the ~-chloroamide 11 and the remaining OBoc residue was cleaved to afford phenol 12 (Scheme 12).2oAn extraordinary number of conditions were examined for the oxidation of 12, which included an examination of additives, the use of solid supported oxidants, as well as phenol silylation. The yields of products, particularly the desired product, were meager. Thallium nitrate afforded 14 and 18, arising from ortho and para nitro addition to the intermediate cation. Koser's reagent [PhI+OH-OTs] gave the tosylate 15, corresponding to ortho addition. Lead tetraacetate gave a mixture of the bis-ortho acetate 16 and the para acetate 17. The latter product was also produced as the major product with PIDA. Based on the notion that introduction of water halfway through the reaction might lower the energy barrier by quenching the iminium intermediate, we explored the use of 2-iodoxybenzoic acid (IBX) as an oxidant, because it would generate an equivalent of water as Iv

470

THOMAS PE'I'TUS AND TODD WENDERSKI

N 11 1.

OH

O _--

oxidation;H20 O

OH

OH

~

~OTs

PO" "~

PO" ""~

14

15

El

Et

O

PO

O

O

.,,~OAc PO" "I>" 16

Ac

Et

20

21

O CF3 "~ 0

PO

,

18

O

~ "

O

~l~O PO

17

O

PO

~~O PO

Et

O

~/~ ~ PO

N.-0

19

O

OHn-lI ,,~'~

Fs PO

PO H

CF 3

22

Et O OF3

23

Et

24

SCHEME 12. The oxidation proves capricious.

was reduced to I III. However, instead of the desired product 13, Van De Water isolated the o-quinone 20. Derek Magdziak and Andy Rodriguez would subsequently develop this discovery into a general method for the synthesis of o-quinones and catechols from substituted phenols. 2~ Yaodong Huang and Dr. Jinsong Zhang later applied this discovery in the synthesis of ( +)-brazilin. 22 The best outcome was obtained using PIFA. However, the reaction gave highly variable amounts of the lactone 13, along with the biphenyl 24 and the trifluoroacetate 19. Addition of PIFA to 12 in tetrahydrofuran (THF) afforded adduct 21 in a 72% yield and showed participation by THE When PIFA was used in hexafluoroisopropanol or acetonitrile, similar solvent adducts 22 and 23 formed in respective yields of 11 and 20%. The latter required an aqueous work-up to saponify a trifluoroacetoxy intermediate. Of the conditions surveyed, a 42% yield of 13 proved to be the best and it occurred only once from addition of I. 1 equivalents of PIFA to 0.1 M solution of the 12 in CH3NO 2 at 0 ~ However, inconsistent amounts of compounds 19, 23, and 24 and 13 emerged from these conditions. In spite of the poor performance of the reaction, the adduct 13 had displayed great synthetic promise by partaking in a myriad of stereoselective and chemoselective reactions (Scheme 13).

13

471


0

Ph.. OH

II

,.-,~"./C02Me

__~

O~v~~ 25101 ~0~ I I 0~'---/' [J

t,....6~ u

27

dia?en~Ytll. ~L,~~ ~N~MgBr KHMDS/ ~ 50%~ O /70%

HR/~PPc~31CoVh ~ f~ / , jl Ij ~ 0/ ~ 72% l: ",w-~

C[O2Me

26

O

OMe

Danishefsk3fs diene JJ..... 87% BORSM= O" ~ v [-..0 "r"- ]~

/ ~ 13 X MeOH ~ O ~ ~ NBS Jl. .~ Rh(PPh3)3CI/ 63%~ 0 II~ "1 " H2/CH2CI2'2h .... N /JJ~ 0 _ ~' ~ 61~ Me~.~ 0

"0

28

o

0 SCHEME 13. Versatility of glycolic derived cyclohexa-2,5-dienone adducts.

IV. Total Synthesis of (+_)-Epoxysorbicillinol

Therefore, we focused this emerging methodology toward epoxysorbicillino136. 23 Sorbicillin 31 was prepared in three pots and smoothly transformed into 32 by a Mitsunobu coupling. One of the phenols was engaged in hydrogen bonding with the sorbate carbonyl and could not participate (Scheme 14). However, the remaining phenol partook in oxidation when exposed to PIFA. The desired cyclization proceeded in a surprisingly good 65% yield and was followed by an unexpected diastereoselective epoxidation when two equivalents of PIFA were used. The product was unequivocally established by X-ray analysis. The masked carbonyI was freed by two different procedures. As suggested by Kirby's treaty describing the effects of secondary orbital interaction upon rates of SN2 reaction, 24 the lactone 33 was first opened to afford the amide 37 by Weinreb's conditions. This material undergoes a rapid reaction with tin tetrachloride and affords epoxysorbicillinol 36, whereas 33 does not. In addition, 33 smoothly opens with cesium hydroxide and the resulting carboxylate 34 undergoes lactonization to afford 35 upon treatment with acid. Another saponification affords 36. A manuscript describing this six-pot synthesis was submitted in October 2000 to Angew. Chem. It was rejected and a twelve-step racemic synthesis submitted in November appeared in J. Am. Chem. Soc. in February. 25 Ours was published in Org. Lett. in January 2001. 26 C'est la vie.

472


/

oJ coo "o0

I

II

_

?-!e~

> 90% yield -

"

]

"

OH sorbicillin (31) pots

inthree

I

9

g.,:o

H ]I r../

~o.,.~o

' "-" over three steps epoxysorbicillinol(36) one pot

33

o

CsO. / L H=O \ ' ~ /

L "~ -'-

0 ~ ' " T :-~ H4 . . . . / 0

"Phi=O"

32 ~

0.~ 0 0

-78 ~

/0 [" ...~ ~ 0 I~" ">

\

0%

~ ~ ' , , , o: ~176 .~o oJ ~ ;

0 .....

'''~

o..;

0 :

o~Oo ~ 35

"~OGs

34 / //~/"

~~oc.,

I

/

AIMe2

\~ ~-~

"T 11

o\ o- 2 >>>>>> 0 /4"..... oxidant: PIFA - PhlO+TMSOTf >>>> PIDA A/B ratio optimized solvent: CH3NO2 > CH2CI2 >> any other solvent additives showing little effect: propylene oxide, K2003, phenol silylation, Bu4N+BF4-, fluorinated solvent

SCHEME 17. Rapid optimization by measuring A/B ratio.

The identity of the R ~ substituent, the length of the tether (n), as well as the shape and electronics of the amine contributing to the nucleophilicity of the amide were all examined in combination with different oxidants, solvents, additives, and other tricks that would supposedly improve the yield. Many of the results were expected. For example, electron-rich amines with little steric encumbrance afforded a greater A/B ratio than amides derived from electron-poor amines or amides derived from congested amines. The most surprising result, however, was the effect of the R 1 substituent. The presence of this residue had a far greater influence on the success of cyclization than any other criterion. For example, oxidations of electron poor oxazolidinone amides afforded B exclusively for R 1 = -H, but gave an almost 3:1 A/B ratio when R ~ = - B r or -Me. This result most likely reflects fewer degrees of freedom in the transition state, which is caused by reorientation of the amide residue into an arrangement prone to the desired cyclization. VIII. Controlling B-Dione Tautomerization and Protection

Meanwhile, Dr. Wang's progression toward rishirilide B had stalled. Our plan entailed a very risky gambit of controlling the formation of a particular B-dione tautomer in order to manage the subsequent addition of an anion to one to the carbonyl residues. Our appreciation of this problem had been dimmed by our recent involvement with epoxysorbicillinol (36) - a tautomeric natural product that favors the tautomer displaying hydrogen bonding. Therefore, we naively expected 61, which was the hydrogenbonded and fully conjugated tautomer, to predominate. We further suspected

477

13 SEDUCEDBY A SIREN'S CALL

that its predominance would be reflected in the subsequent reactivity of this mixture (Scheme 18). _

~~

equilibrium solvent dependent OH O 1~

O

OMe J 60

1LiOH then HCI

i

then LiOH

OO~ I

~

_ R1 OH

HCI viathen ~'-Iactone

OMe.... ,r O-H

O

i

61 CH2N

i NCOCI

O

\O

62

O Et2N O/

OMe~I

R1 ...

H 63

0

'

SCHEME 18. Controllingtautomer formation and protection.31

Over time, the mixture of pseudo-acids 61-62 equilibrates to structure 61 (for R ~= - M e and-Br). However, the reaction of the predominating tautomer 61 with diazomethane affords a 1:1 mixture of the vinylogous esters 63 and 65. After considerable experimentation, Dr. Wang discovered that the carbamate vinylogous ester 64 could be formed selectively. We attribute this discovery to the notion that the less stable carbamate corresponding to acylation of 62 reverts to the more stable product 64 under the reaction conditions. Next, Dr. Wang found that addition of an excess of lithiated ethyl vinyl ether to 64 exclusively produced the desired diastereomer 66 upon work-up (Scheme 19). This stereochemical outcome had been expected, because we knew that the unprotected hydroxyl residue would direct the 1,2-addition to the adjacent carbonyl while the pseudoaxial isopentyl residue would control the stereochemistry of the methyl residue. Unfortunately, Dr. Wang was unable to cause the enol ether 66 to smoothly convert into the ethyl ester 67. The best yield was obtained using RuO 4 under biphasic conditions and it afforded a 30% yield of the desired ethyl ester 67 along with a considerable amount of product corresponding to oxidation of the electron rich naphthalene. A bis-~-hydroxy ketone, which was easily produced by epoxidation and ring opening of the enol ether 66, returned the tautomeric mixture of 61 and 62 on treatment with

478

THOMAS PETrUS AND TODD WENDERSKI

o

Et2N'JJ'O

OEt

0

~'~Li ..

O

69%yield 64

RuO4

.....

th e n NH4CI

" OH~o~E

\

.

"'-...-~

t

less than

30%yield

~ OMe

66

87

DMAP Et2NCOCl

o

....OH

/~OHo~E: --~

m-CPBA

tautomeric mixture of

pseduoacids61+62

HIO 4

OH0 Rq,,,%O H bis-*x-hydroxyketone R2

SCHEME 19. Stereoselective formation of the stereotriad.

periodic acid. This outcome was quite surprising, given the considerable number of reports involving hydroxycortisone, which had shown that cleavage occurred between the primary hydroxyl and hydrated carbonyl residues to give the corresponding ~-hydroxy aldehyde. 32 While the ethyl ester 67 may seem to have still been a viable intermediate, from the earlier Danishefsky account of failed saponification, we knew that we had reached an impasse. Therefore, we set the project aside and tackled the asymmetry. IX. A L a c t i c A c i d - D e r i v e d D i r e c t i n g - P r o t e c t i n g D i r e c t i n g

Group

The realization of the importance of the R ~ substituent (R~4: -H) had reinvigorated the project. Lupe Mejorado began to systematically examine chiral auxiliaries (Scheme 20). Though the yields had improved, the enantiomeric excess remained low. Only 60% ee was obtained with 68 and those for chiral oxazolidone derivative 69 hovered around 20%. Of some note was the surprising stability of the oxazolidinone-derived intermediate 70. This compound proved stable to work-up and chromatography. Upon stirring it in an ethereal solution over a slightly basic aqueous layer, the desired product 46 formed. Given the disappointment afforded by our auxiliary plan, we next investigated the effects of substituents among the intervening atoms of the protecting group. Since the glycolic protecting group would be jeuisoned, a stereocenter on the intervening carbon atom might prove useful, provided its source was inexpensive and the unit was easy to install and remove. We considered a variety of non-racemic a-hydroxy acid derivatives,

13

479


0

poRI~R Phl(OeOCF3) H20"= 2;~ OH

R 1 = Me-, B r -

2 68-69

R1

2

- chiral amine e.r.

2:1-4:1

0

46 0

o

RI

.

,Pr ....

69

0

~ i OC(O:~)CF~ o.~R

70

SCHEME 20. Auxiliaries fail for glycolic derivatives.

which were attached to 71 by a stereochemical inversion using the Mitsunobu process. The stereointegrity of these couplings was carefully checked against the corresponding racemic standard. Mandelic derivatives were ruled out because of their propensity to undergo oxidation rather than inversion under Mitsunobu conditions. We setfled upon simple lactic derivatives and further suspected that the corresponding amides would prove more selective, because of the pseudo allylic strain incurred during the progression into products. Indeed, the ester derivative 72 afforded a diastereomeric ratio of 3:1 (Scheme 21). On the other hand, the pyrrolidine derivative 73 proceeded to the lactone 77 in good yield and diastereoselectivity. 33Only a trace amount of the undesired diastereomer was evident in the 400 MHZ ~H-NMR of crude product and it was easily separated by chromatography. The methoxyamine 74 afforded a slightly improved yield of 77. It is interesting to note that 75, which was prepared from the ester 72, could have afforded the quinamine corresponding to 77. However, the oxidation instead produced the oxime ester 76. The oxime ester 75 smoothly converted to the lactone under acidic conditions or upon treatment with Koser's hypervalent iodide reagent. Wenderski found that the enone within 77 smoothly reduced with potassium azodicarboxylate (PADA). The resulting product 81 underwent further hydrogenation and hydrogenolysis to afford the trans-fused decalin 79 (Scheme 22). The modification, which was borrowed from Corey's ovalicin synthesis, was a significant advancement for us. Earlier, we had employed a single step reduction with the glyco|ic derivative 13 using the Wilkinson catalyst. However, this and other metal-mediated reductions

480

THOMAS PETTUSAND TODDWENDERSKI OBoc

RI~

71 Me2AINHOMe

HO Mitsunobu i

I MeO.~ O

OH

R2

OH

~R2 ~ N ~ 72 o

~R2 I 73

OH

OH

oMe: oMe: O//

LR2

74

0

I

I

i~,,

good d.r.>l0:1 good yield > 65%

good d.r.>10:l better yields>70%

75

good d.r.>10:1 better yields>70%

l

0 R1= -Me, -Br

R ~

0 PhlOTsOH R2

o

.....~ 0 0

77

>70% yield

o .....i ~ 0 N

MeO"

a~ 76

SCHEME 21. A diastereoselective dearomatization with lactate derivatives.

proved to be more capricious for the corresponding lactic derivatives because of competing single electron transfer and rearomatization. Since the equatorial positioning of the [~-alkyl hydroxyl residue, compound 79 proved surprisingly robust toward most bases. However, upon saponification of the lactone with potassium trimethylsilanolate (KOTMS), the system underwent a ring flip and ~-elimination, producing the corresponding enone 80 on work-up. Another interesting stereoselective transformation, discovered by Hoarau for lactic derivatives, was the product afforded by the addition of primary amines. The lactone proceeded first to the corresponding secondary amide, and then underwent cyclization to afford the lactam 87, presumably via a cis addition of N-H across the olefin of the vinylogous ester. On the other hand, reduction of the cyclohexadienone with MAD/L-Selectride afforded the enone 86 as a single diastereomer. Further indications of significant differences between the chemical reactivity of lactic and glycolic backbones came upon prolonged exposure of 77 to tertiary amine bases or 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU).

13 SEDUCEDBYASIREN'SCALL O

R~

~ ~ )

O

].,..~O,,,,/ R2

O KOTMS R2

O

78

O 77

Bu4NOH

ifR 1: -Br / //// or-C(O)R

J R3 OH ,., R11--I...~u

H 2 Pd/C if R1 =-Br

O R

cso ,-,

O

R2 OM O 81 \ pyrrolidine

O . O

R1

R1

O ~2 O

86

,i, ,

/

"""

MAD,"....~ ,1. O ~

O

H+

H20 O ~ O " Hd '"~. "..~R 2

I

L-selectride

80

O

~

HO ()/ 88 R2 R3NH2--J/

O

O

: DBU R ~

O

.

481

82

O

83

,~KOH SnCl 4

OH

O "~ N

85

84 SCHEME22. Lactic derivatives behave differently than glycolic

derivatives.

The equatorial methyl residue adjacent to the lactone carbonyl proved surprisingly easy to epimerize into an axial position, as shown by exclusive equilibration into adduct 78. At first, we suspected that perhaps the amine caused the lactone to open by an SN2' displacement and the system then re-closed to produce 78. However, after several experiments we attributed the complete epimerization to ~-deprotonation due to the presence of an attached electron-poor oxygen substituent and the release of torsional strain. When the R ~ substituent was electron-deficient and no proton source was available, a stereospecific electrocyclization occured during enolization to afford the x-lactone 88. Lactone 81 also underwent saponification and furnished the corresponding metal carboxylate 82. As opposed to an efficient lactonization of its glycolic counterpart, treatment of 82 with acid proceeded to the diastereomeric mixture of the x-lactones 83 in poor yield. Moreover, saponification of this mixture failed to produce a significant amount of the pseudoacid 84 and returned instead the metal carboxylate 82, which can afford the starting 8-1actone 81 during chromatography. While the lactone 81 could be opened with pyrrolidine to afford the amide 85, this lactic acid derivative failed to cleave with tin tetrachloride as opposed to its glycolic counterpart. Therefore, the chemistry prevented retention of

482


the ketone oxidation state, making the non-racemic method unsuitable for both rishirilide B (45) and epoxysorbicillinol (36). X. New Methods for Cleavage of the Directing-Protecting Group

A solution to our cleavage problems was revealed while progressing toward 89, a structure reported by Cordell and to which anti-cancer activity had been attributed. 34At the time of isolation, the enone 89 was proposed as the biosynthetic precursor for 90 (Scheme 23), which suggested an unknown biological dienone rearrangement. Malik had reported another natural product, 91, which appeared to display a similar core ring system. 35 Our nonracemic method seemed applicable to either chiral enone. Hoarau built the core lactic derivative 92 and converted it into the lactol 93. All attempts to cause a [3-elimination resulted in an epimeric mixture of 89, protected as its corresponding silyl ether. We imagined that opening of the lactol and formation of the corresponding hydrazone would enable a ring-flip under milder conditions, and thus lead to the desired enone via [3-elimination. Surprisingly, the alkoxy residue 0~to the aldehyde functionality cleaved when forming the hydrazone and afforded the secondary alcohol 94. Hoarau converted this alcohol into its corresponding mesylate, and after elimination and deprotection isolated the target structure 89. However, its spectra did not match that reported for the natural product. On closer inspection, we determined that both 89 and the related structure 91 had both been misassigned. Our revision for 91 had been reported elsewhere as the natural product acremine A. 37 We requested that the authors, who had isolated acremine A, transfer their sample to the deuterated

O

0

OH ,,OH

-b-i~

Hd 89 reportedstructure 0 B r ~ O

R2

Ov ~

R2

OH

93

o

92

HO/I

90

0 steps ~ , , , O T B S

0 H2NNMe2" TMSC..~,,,OTBS HO" ~

91 reportedstructure steps

=

R2

He

94

S C H E M E 23. Accidental cleavage of the directing-protecting group. 36

89

13

483


solvent employed with 91. The ~H-NMR spectra agreed. The actual structure for the natural products misassigned as 89 and 91 is the corresponding regioisomer of the enone functionality. Thus, we had completed a rather sophisticated non-racemic synthesis of 89. XI. The Total Synthesis of (+)-Rishirilide B

Despite this setback, these cleavage conditions, in combination with a solution conjured for the endgame of rishirilide B (45), indicated that the synthesis was nearly complete. The remaining question was the choice of a protecting group for the naphthol. After methoxymethyl (MOM) and silyl ethers failed, we resorted to a benzyl residue and prepared a modified dimethide precursor. In this regard, the Comins method for regioselective alkylation of 3-alkoxybenzaldehydes again proved most useful in providing compound 96 (Scheme 24). 38 Subsequent application of the DfirstCharlton chemistry affords the benzylated compound 97. Upon heating, the dimethide was liberated and underwent a regioselective cycloaddition. The reaction occured at slightly elevated temperatures (150 v s . 110 ~ when compared to the earlier union between the methoxy and glycolic counterparts. At this elevated temperature, the initial adduct succumbed to immediate [3-elimination. Subsequent oxidation with DDQ afforded the chiral naphthalene derivative 99. The remainder of the synthesis passed without incident. Addition of the aluminum amide of dimethylhydrazine to 99 produced the 13-dione 100

.CHO

95 OBn

I Li/N~N

I/ ,

3 equiv PhLi then Mel

1. SO2, 450W Hg-hv ~ 2. M e O H , ~

~ _ . . . C HO

OBn 96 0

SO2

1. (-)-98,ZnO, 155 ~ 9h 2. DDQ (680/otwo steps)

OBn 97

OBn reo,,~ (-)-99 " ~ ~) Me

SCHEME 24. Rapid access to a viable intermediate.

484


99 Me2AINH_NMe2

~o

",,

~

, - ~ Me 0

RI

n~

oA~"'~ =

Me

-

(o

HN ' N \

~

o ... M e

-..

0

100 Et2NC(O)CI HiJnig'sbase /

0

" OH OH

102

-78 ~ Li 1. EtO~-,,, 9m-CPBA

NEt2

0"'~"'0 ~ M e ,, "~

1) Na(OAc)3BH

0

i

9,'

0

101 i

0

1. NaOCI , ~ , , , . ' H 2. H2,Pd/C 90%~ L v~"O4Hi[o]/ -

2) Nal04on silica 75%overall

,,,

103 _~

SCHEME 25. Completion of the first synthesis of (+)-rishirilide B.

(Scheme 25). It should be noted that a direct cleavage from lactone derivatives proceeded only with the electron-deficient vinylogous ester functionality. However, the alkoxy cleavage process has been effective with all functional variations of the corresponding lactol derivatives, such as compound 93. The desired vinylogous ester 101 was obtained as before, by carbamylation of dione 100 and the remaining carbonyl in 101 underwent 1,2-addition with the lithium species shown to produce a vinyl ether on work-up. Treatment of the enol ether intermediate with peracid gave the bis-~-hydroxyketone 102. To circumvent the earlier cleavage problem, 39 the bis-~-hydroxy carbonyl was reduced with sodium tri-acetoxyborohydride, whereupon the least hindered diol succumbed to cleavage with periodic acid to produce the ~-hydroxy-aldehyde 103. Lindgren oxidation and debenzylation of the naphthol completed the first total synthesis of the (+)-rishirilide B (45); 40 the enantiomer that is found in nature. 41 XII. New Horizons for This Method and Oxidative Dearomatizations

When we began, our original goal was to devise a general enantioselective dearomatization technique that would prove useful for a variety of

13

485

SEDUCED BY A SIREN'S CAI.L

synthetic applications. Given our synthesis of (_+)-epoxysorbicillinol and (+)-rishirilide B, and our manipulation of these core ring systems, we would like to think that we have succeeded to some degree. We have fashioned a protected chiral cyclohexa-2,5-dienone adduct capable of serving in a myriad of subsequent chemoselective and stereoselective transformations. Schemes 26 and 27 should give readers a clear indication as to the breadth and scope of the chemistry available with these chiral adducts. A surprising interplay of reactivity has been revealed while examining reductions of the cyclohexadienone 92. As described earlier, reduction with diimide generated from PADA in CH3NO 2 proceeded as expected with the electron deficient enone to produce 117. On the other hand, 1,4-reduction of the vinylogous ester proceeded smoothly with MAD and L-Selectride | to give the ~-bromoketone 104, in which the bromine atom is positioned in an equatorial arrangement with respect to the six-membered cyclohexenone. These results were anticipated. However, the results of other reduction conditions have been quite surprising. O

1~

O

R3

? ~

Me2AIN-NMe2 71%

PADA CH3NO2 85% \ OH

:

~28~% ~''~

NaBH3CN ZnCl2, 80% O_H 3....,,,,~,,, Br

R3 a

0

124 OH

O.~

107~

R30"v")'" 106 OH .,,,,"~/"OH Me2N--NH2 _

--n5

O

~

",

O

i::o -. . . . .

N"

i~3 OH 123

NH2

R3

O

"" O.v ,,,j..... 122 (3H

.....

125 SCHEME 26.

O ri ] .'.,~.p,,,o,

..~

CHCI3, >95%

92 O

/NaBH4 CeCl~ Br MeOR 75%

O ' v " J .....

r

~

NaBH4 THF, H20 75% /

O \~O

105 O B~AL~,,,~ I)DI 2) MnO2 60% 2-steps

MAD

R\Xo O

Br

R30..~ .....

L-Selectride 75%

O /~~

118 ~ . . ~ B r

O

TMSCI Nal 90%

R30.~) ..... O

1 1 7 0 - ~ ) .....

O

Br

Reductions and subsequent removal of the chiral tether.

486


9

O /~J'~,Br

OSiR3 Zn~ / ~ R3SiOTf

o

PADA CH3NO2 90%

'"o

111 O

R30..~ ..... 114 O DBU DMDO1 R3SiOTf / oOSi~R3

112~

O

O Pd/C,H2 >95%

o

R30,,~ ..... 104 O

I

"'9 = 2-J .....

o 1130 ~ O

'

o

_

Et2Zn,

/+

O

105

OSiR3

"?

.,,,.# R OH 115

116 '"O R30,~ ..... O

SCHEME 27. Enolate trapping, bicyclic and ring expanded products.

For example, exposure of 92 to diisobutylaluminum hydride (DIBAL) alone resulted in reduction of both the alkene of the bromo vinylogous ester and the carbonyl of the lactone to afford 122. While the double reduction of 92 with NaBH 4 in THF/H20, which afforded the allylic alcohol 108, may seem reasonable given the DIBAL result, the formation of the lactol 124 by Luche conditions is difficult to rationalize. Moreover, the reduction of 92 with NaBHsCN and ZnC12 to give the alcohol 125 seems stranger still. An explanation for this divergent reactivity begins to emerge upon examining the carbonyl frequencies in the infrared spectrum from compound 92. The ketone carbonyl appears at 1668 cm-~, whereas the lactone carbonyl resides at 1759 cm -~. The bromine atom in 104 cleaved on exposure to trimethylsilyl iodide (TMSI) to afford 105, whereupon both carbonyls were reduced upon treatment with excess DIBAL. Regioselective allylic oxidation with MnO 2 yielded the lactol 106. The directing-protecting group could be cleaved at this stage by treatment with dimethylhydrazine to produce the diol 107, or in the case of lactol 122, the anti-epoxy alcohol 123. Alternatively, as shown in our synthesis of (+)-rishirilide B, the lactic acid tether can be cleaved directly from lactone 117 upon treatment with

13

487


the corresponding aluminum amide to give the vinylogous acid 118. Compound 104 is an important branching point for many other transformations as well (Scheme 27). For example, it led to the silyl enol ether 112 as shown, as well as to its regiomeric enol counterpart 114. The latter resulted from the regioselective enolization of ketone 110, which occurred parallel to the trans-decalin-like ring junction. The Baeyer-Villiger oxidation of 110 also proved entirely regioselective and afforded the seven-membered lactone 116. In addition, these systems, while protected with the lactic acid group that had been used to direct stereochemistry, underwent a variety of other subsequent diastereoselective transformations such as the epoxidation of 114 that produced 113 and the cyclopropanation of 104 that afforded 109. However, in order to remove the directing-protecting group by a 13-elimination to produce 115, the lactone 110 must first be opened so that the underlying cyclohexanone can undergo a ring flip. While many future applications can be imagined, the simplest are syntheses of natural products displaying a chiral six-membered ring containing a tertiary alcohol- compounds that are not easily procured by other strategies, such as cleorindicin D, illudin J1, and oxysporidinone shown in Figure 3. However, one need not limit applications to ornate six-membered rings. Given the capacity to rupture the core ring system by Baeyer-Villiger and ozonolysis reactions, and the power to expand or contract the core cyclohexyl ring system, applications to acyclic arrays such as phoslactomycin and other fused cyclic natural products such as the trans-fused 7,5-ring system found in (-)-kessane are equally plausible. 0

0 oH

(+)-cleroindicin D

OH (-)-illudin J1

H3C (-)-kessane

HO 0 HO" \ 0

H2N

OH

i~

...~...)~..5~~..~J OH O~)_..,

oxysporidinone

FIGURE 3. Future targets.

phoslactomycin B

488


Some could argue that a catalytic enantioselective chemical oxidative dearomatization process would be superior to that which we have developed. From serious contemplation of strategies leading to catalytic processes, we beg to differ. We speculate that as far as oxidative dearomatization processes are concerned, a catalytic process, if it were ever developed, would be fairly limited in scope and applicable to a small assortment of phenolic nuclei. The method that we have developed permits oxidative dearomatization of an assortment of resorcinols and affords a chiral p-quinol nucleus that is both protected and fortified as well as discriminant in its subsequent reactivity. When considering that new fermentation and separation processes have reduced the cost of R- and S-lactic acid directing group to less than $0.50 per pound as provided from renewable biosources such as corn, the issue of cost seems mute. However, others will likely succumb to the siren's call and with some luck may develop more efficient methods for unlocking the synthetic potential embedded within six carbons of simple aromatic compounds. Acknowledgments The development of this synthetic method and the completion of (+_)-epoxysorbicillinol and (+)-rishirilide B are all testaments to the well-known saying "any hard experience that one survives leads to greater strength and wisdom." This work is a testimony to the resiliency and ingenuity of many graduate and undergraduate students, as well as postdocs who nobly fought to make a round peg fit into a square hole. Postdocs Dr. Junhua Wang and Dr. Jinsong Zhang; and graduate students Ryan Van De Water, Lupe Mejorado, Christophe Hoarau, Todd Wenderski, Yaodong Huang, and Maurice Marsini; and undergraduate students Andy Rodriguez, Kevin McQuaid, Dave Freeman, and Simon Meek all contributed to aspects of this work. This narrative is as much their story as mine. However, I would like to bestow a special thanks to my many mentors including Joe Shelor, Maurice Maxwell, Tom,is Hudlicky, Dick Schlessinger, and Samuel Danishefsky and my dear wife Dr. Liping H. Pettus. These people played a large role in shaping my life. To all of these unusual individuals, and to other inspirational folks such as Peter Wipf, I pass on my heartfelt thanks and gratitude.

References and Footnotes 1. 2. 3. 4. 5.

Hudlicky, T., Luna, H., Barbieri, G., Kwar, L. D.,J. Am. Chem. Soc. 1998, 110, 4735. Gibson, D. T., Hensley, M., Yoshioka, H., Mabry, T. J. Biochemistry 1970, 9, 1626. Hudlicky, T., Seoane, G., Pettus, T. R. R., J. Org. Chem. 1989, 54, 4239. Schlessinger, R. H., Pettus, T. R. R., J. Org. Chem. 1994, 59, 3246. (a) Deslongchamps, P., Belanger, A., Berney, D. J. F., Borschberg, H. J., Brousseau, R., Doutheau, A., Durand, R., Katayama, H., Lapalme, R., Leture, D. M., Liao, C. C.,

13

6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.


489

MacLachlan, E N., Maffrand, J. R, Marazza, R., Martino, R., Moreau, D., Ruest, L., Saint-Laurent, L., Santonge, R., Soucy, E, Can. J. Chem. 1990, 68, 127. (b) Deslongchamps, E, Belanger, A., Berney, D. J. E, Borschberg, H. J., Brousseau, R., Doutheau, A., Durand, R., Katayama, H., Lapalme, R., Leture, D. M., Liao, C. C., MacLachlan, E N., Maffrand, J. E, Marazza, R., Martino, R., Moreau, D., Ruest, L., Saint-Laurent, L., Santonge, R., Soucy, E, Can. J. Chem. 1990, 68, 115. Cabal, M. P., Coleman, R. S., Danishefsky, S. J., J. Am. Chem. Soc. 1990, 112, 3253. Corey, E. J., Dittami, J. P., J. Am. Chem. Soc. 1985, 107, 256. Magdziak, D., Meek, S. J., Pettus, T. R. R., Chem. Rev. 2004, 104, 1383. Corey, E. J., Wu, L. I., J. Am. Chem. Soc. 1993, 115, 9327. Wipf, P., Kim, Y., Jahn, H., Synthesis 1995, 1549. Hu, Y., Li, C., Kularni, B. A., Stobel, G., Lobkovsky, E., Torczynski, R. M., Porco, J. A., Jr., Org. Lett. 2001, 3, 1649. MacDonald, G., Alcaraz, L., Lewis, N. J., Taylor, R. J. K., Tetrahedron Lett. 1998, 39, 5433. Wipf, E, Jung, J. K., Org. Lett. 2000, 65, 6319. Unpublished results from these laboratories. Greene, T. W., Wuts, E G. M. Protecting Groups in Organic Synthesis, 3rd edition, John Wiley and Sons, Inc.: New York, NY, 1999, p. 265. McLoughlin, B. J., J. Chem. Soc. Chem. Commun. 1969, 540. Van De Water, R. W., Magdziak, D. J., Chau, J. N., Pettus, T. R. R. J. Am. Chem. Soc. 2000, 122, 6502. Pelter, A., Hussain, A., Smith, G., Ward, R. S., Tetrahedron 1997, 53, 3879. Unpublished results from these laboratories. Van De Water, R. W., Hoarau, C., Pettus, T. R. R., Tetrahedron Lett. 2003, 44, 5109. Magdziak, D., Rodriguez, A. A., Van De Water, R. W., Pettus, T. R. R., Org. Lett. 2002, 4, 285. Huang, Y., Zhang, J., Pettus, T. R. R., Org. Lett. 2005, 5, 5841. Sperry, S., Samuels, G. J., Crews, E, J. Org. Chem. 1998, 63, 10011. Kirby, A. J. Stereoelectronic Effects, 1st edition, Oxford University Press, Inc: New York, NY, 1996, p. 40. Wood, J. L., Thompson, B. D., Yusuff, N., Pflum, D. A., Matth~ius, M. S. E, J. Am. Chem. Soc. 2001, 123, 2097. Pettus, L. H., Van De Water, R. W., Pettus, T. R. R., Org. Lett. 2001, 3, 905. McQuaid, K., Pettus, T. R. R., Synlett 2004, 2403. Najdi, S., Reichlin, D., Kurth, M. J., J. Org. Chem. 1990, 55, 6241. (a) Allen, J. G., Danishefsky, S. J., J. Am. Chem. Soc. 2001, 123, 351. (b) Yamomoto, K., Hentemann, M. E, Allen, J. G., Danishefsky, S. J., Chem. Eur. J. 2003, 9, 3242. Durst, T., Kozma, E. C., Charlton, J. L., J. Org. Chem. 1985, 50, 4829. Wang, J., Pettus, T. R. R., Tetrahedron Lett. 2004, 45, 5859. (a) Kertesz, D. J., Marx, M. J., J. Org. Chem. 1986, 51, 2315. (b) Ashton, M. J., Lawerence, C., Karlsson, J. A., Stuttle, K. A. J., J. Med. Chem. 1996, 39, 4888. Mejorado, L., Hoarau, C., Pettus, T. R. R., Org. Lett. 2004, 6, 1535. David, J. M., Chfivez, J. E, Chai, H. B., Cordell, G. A., J. Natural Prod. 1998, 61,287. Aziz-ur-Rehman, Malik, A., Riaz, N., Nawaz, H. R., Ahmad, H., Nawaz, S. A., Choudhary, M. I., J. Nat. Prod. 2004, 67, 1450. Hoarau, C., Pettus, T. R. R., Org. Lett. 2006, 8, 2843.

490

THOMAS PETTUSAND TODDWENDERSKI

37. Assante, G., Dallavalle, S., Malpezzi, L., Nasini, G., Burruano, S., Torta, L., Tetrahedron 2005, 61, 7686. 38. Comins, D. L., Brown, J. D., J. Org. Chem. 1985, 50, 4829. 39. Mejorado, L., Pettus, T. R. R., Synthesis 2006, 3209. 40. Mejorado, L., Pettus, T. R. R., J. Am. Chem. Soc. 2006, 128, 15625. 41. Iwaki, H., Nakayama, Y., Takahashi, M., Uetsuki, S., Kido, M., Fukuyama, Y., J. Antibiot. 1984, 37, 1091.

INDEX

ab initio methods, 127 acetic anhydride(Ac20 ), 60, 106 acetonitrile, 470 acetylation, 60, 76, 81,315 acid-catalyzed cyclizations, 246 acremine A, 482 activation, 384 activation barrier, 131, 134 acyl anion equivalent, 248, 264 acylation, 52, 311 acyloin ring closure, 222 1,4-addition, 357 AD-mix-[3, 85, 244 Ag20, 94 aglaforbesin, 222, 228 Aglaia forbesii, 222 aglain, 220, 226 aglaroxin C, 219 AIBN, 192 Alder end() rule, 429 aldol, 8, 434, 436-437, 445, 449, 456 aldol addition, 16, 20 aldol condensation, 295, 233 aldol methodology, 297 aldol reaction, 388 alkenylsilane, 22, 25, 30 alkyl azides, 440 alkyl magnesium and lithium reagents, 468 4-alkylresorcinols, 467 allenes, 334 allenic amino acids, 338 allenic aminoester, 340

4-allenyl-2-oxazolin-5-one, 337 4-allenyl-5-oxazolone, 336 allenynes, 342, 351 allyl cuprate, 166, 195 allyl ether, 176 allyl magnesium bromide, 166, 196 allyl magnesium chloride, 158, 165 allylic chloride, 248 ~-allylic palladium complex, 78 allylic rearrangements, 260 allylic substitution reactions, 332 allyltrimethylsilane, 316 aluminum ate complex, 182 aluminum enolate, 301 aluminum reagents, 168 Amberlyst-15, 28 0~-amino acids, 268 3-aminopropionaldehyde, 299 (1R,2S)-2-amino-l,2-diphenylethanol, 270 ammonium hydroxide, 299 amphibian alkaloids, 408 amphilectane, 35 amphotericin B, 3 amphotericins, 1 anhydrous HC1, 47 anti conformation, 372 1,5-anti aldol addition, 30 anticancer, 220, 385 antifertility activity, 59 antifungal activity, 244 antifungal agents, 2 antimalarial, 244 antimicrobial, 244 antlproliferative and antileukemic activities, 220

491

492 1,5-anti stereoinduction, 8, 21, 23, 31 antitubercular, 385 an tiviral, 244 arene diols, 464 arene-arene interactions, 99 arginine, 146 aromatic C-H functionalization, 384 arsine ligands, 27 arylboronic acids, 332 3-aryl-3-hydroxy-l,2-indandione, 230 aspergillusfumigatus, 308 asymmetric Diels-Alder catalysis, 429 asymmetric dihydroxylation, 271 asymmetric photocycloaddition, 233 asymmetric synthesis, 268, 442, 444 asymmetric synthesis of spirotryprostatin B, 308 auxiliaries, 443 A-values, 467 axinellamine, 324 azatyrosine, 290 (2S)-2-azepanecarboxylicacid, 313 azides, 28 l 4-azidobutyraldehyde, 441 azidohydrin, 439 azido-Schmidt reaction, 412 azomethine ylide, 306--307, 310

Baeyer-Villiger, 73, 75 Baeyer-Villiger oxidation, 487 bands, 453 Barton deoxygenation, 297 Barton-McCombie protocol, 54, 192 base-catalyzed isomerizations, 43 base-induced fragmentation, 264 bathochromic shift, 116 batrachotoxin, 410 9-BBN-H, 73, 75, 87, 91 Beckmann or Schmidt chemistry, 409 Beckmann rearrangements, 412 benzo [b]cyclobutapyran-8-one, 221-222, 228, 230 benzo[b]oxepine, 220 benzo[k]fluoranthene, 116 benzocyclobutene, 107 benzoin oxime, 271 ( + )-p-benzoquinone, 385

INDEX benzoyl chloride, 338 benzyl bromide, 87 benzyl chloroformate, 270 benzyl ether, 76 benzylation, 68 3-benzyldimethylsilylpropynol, 31 benzyldimethylsilyl group, 31 3-benzyloxybenzyl bromide, 291 4-benzyloxybutanal, 437 4-benzylresorcinol, 376 benzylsilane, 27 betaine, 87 BF3 etherate, 262 BF3-Et~O, 44, 319, 357 BH.~, 90 BiBr.~, 200 BiBr3, Et~SiH, 201 [3.3.0]bicyclooctane, 423, 446 bicyclo[4.3.0] oxazinones, 306 bicyclomycin, 269 (S)-BINOL, 169, 233 biogenetic pathway, 220 biological membranes, 232 biomimetic conversion, 255 bioxalomycin K~Fe(CN)6, 83 I~2OsO~(OH) 4, 83 K3PO4, 91 kaempferol, 230, 232 kalihinane, 35 karenia, 154 KCtt~SOCH> 62 (-)-kessane, 487 ketal cyclization, 202 ketalization, 76 ketene acetal, 11 ketene actal, 187 ketenesilyl acetal, 303, 315 keto-enol isomerization, 106 keto-enol tautomers, 228-229 13-ketoesters, 56 338, 340 KHMDS, 284, 319, 435 kinetic resolution, 6 KOH, 319, 434 Koser's reagent, 469 Kosower salt, 125 KOTMS, 26, 480

INDEX

L-(+)-diethyl tartrate, 79 lactam, 349, 451,453 lactams via cycloisomerization reaction, 349 lactol, 315 [3]-ladderane, 376-378 [5]-ladderane, 377-378 [n]-ladderanes, 378, 381 LAH, 314, 440, 445-446, 453 lanthanide Lewis acids, 356 latrunculin B, 311 LDA, 50, 64, 336, 340 lead tetraacetate, 279, 469 lemonomycin, 324 Lewis acid, 272, 357 Lewis acid mediated rearrangement, 44 Lewis base catalysis, 13 Lewis-base-catalyzed aldol addition, 31 LHMDS, 301 LizCuCl4, 65 LiA1H4, 41, 44, 68, 76, 87, 92, 94 Librium, 428 LiHMDS, 28, 435 Lindgren oxidation, 484 Li-NH3 reduction, 40 LiOH, 28, 54, 138, 253, 353, 357, 446 Lipinski rules, 329 5-1ithio-2,3-dihydrofuran, 69 lithium, 40, 44, 47, 275, 280 lithium ammonia reduction, 41 lithium enolate, 21 lithium ethanethiolate, 399 LL-CI0037R, 465 longitblene, 49 L-Selectride, 46, 480, 485 Luche conditions, 486 2,6-1utidine, 52, 81 lysine, 146

M

macrolactonization, 6 MAD, 480, 485 magnesium perchlorate, 311 magnetic anisotropy, 117 (2'-malonyl)phenylalanine derivative, 292

499 manganese(IV) oxide, 16 Mannich reaction, 300, 302, 388 manumycin A, 466 marine sponges, 35 mass spectroscopy, 135 m-CPBA, 65, 69, 71, 75, 90, 176, 285 Me2A1C1, 68 MezCuLi, 435-436, 449 M%S, 357 M%M, 168, 169 MeI, 338, 340 MEM ether, 255, 260 membrane lipids, 376 MeMgBr, 69 MeMgC1, 61 menthol, 232 (-)-menthyl 3,4-bis(dibromomethyl) benzoate ortho-quinodimethane, 112 mercury(II)-acetate, 17 mesylation, 68, 71, 318 metal-ammonia reduction, 39, 56 metal-assisted [2 + 2] cycloaddition, 222 (R)-2-methyl-1,4-butanediol, 260 metal-catalyzed carbocyclizations of allenes, 334 metallocycloaddition, 448 metallocyclobutane, 448 metalloenamine, 207 metathesis, 157, 196 metathesis reaction, 169, 445 methanephosphonic acid dichloride, 138 methoxymethyl chloride, 78 methoxymethyl ether, 92 methoxymethylene triphenylphosphorane, 45 1-methoxy-l-trifluoromethylphenylacetyl (MTPA) chloride, 21 2-methyl-1,4-butanediol, 260 3-methyl-2-butenyllithium, 67 2-methyl-6-methylene-(E)-2, 7-octadien-l-ol, 64 2-methylpent-4-enoic acid, 87 2-methylpropenylmagnesiumchloride, 184 methyl cinnamate, 227, 231 methyl cyanoformate, 41, 50 (+)-O-methyl-elisabethadione, 401-402 methyl iodide, 106 methyl lithium, 432 methyl magnesium bromide, 199

500 (4R,5 &- ( + )-4-me thyl-5-phenyl2-oxazolidinone, 444 methyl triphenylphosphanylideneacetate, 10 methyl (triphenylphosphoranylidene)acetate, 322 methylalumination, 67-68 methylation, 137 methyldimethylphosphonate, 92 methylmagnesium bromide, 13 1,2-methyl migrations, 256 methyltriphenylphosphonium iodide, 73 methyltriphenylphosphonium ylide, 90 MgClz, 196 micelles, 232 Michael reaction, 39, 41, 49, 52, 56, 220, 228, 282, 388 Micromonospora chalcea, 289 microsclerodermin H, 324 Minyobates bombetes, 421 miroestrol, 465 Mitsunobu reaction, 177, 255,260, 418, 438, 441,450, 471 mixed cuprate, 249 MnOz, 486 Mo(CO)6-mediated, 334 molecular, 149 molecular complexity, 333 molecular LEGO, 107 molecular recognition, 99, 122, 138 molecular scaffold, 349 molecular tweezers, 100, 102 molecular tweezers and clips, 112 molluscicidal, 244 MOM group, 44 Montanoa tomentosa, 59 montanol, 59 Morita-Baylis-Hillman reaction, 233 Mosher analysis, 170 Mosher ester, 21, 23 MPC-1001, 324 MTPA ester, 22-23 Mukaiyama aldol coupling, 207 Mukaiyama-type addition reaction, 187 mycoticin A, 3 mycoticins, 1 myrcene, 65

INDEX N

n PrSLi, 54 N,N-dimethylaniline, 138 N~CHCOzEt, 92 Na, 450 NaBH~CN, 486 NaBH 4, 452, 486 NaCIO z oxidation, 40 N-acyl iminium, 272 N-acyl sulfinamides, 232 N-acyloxazolidinones, 232 NAD +, 142, 145 Nail, 73, 92, 341-342 NaHMDS, 87, 287, 435, 444 NaI, 109 NaIO4, 62 nakadomarin A, 311,324 N-alkyl maleimide, 361 N-alkylation, 341 NaN(TMS)~, 255 1,8-nap, 378 2,3-nap, 379 1-naphthaleneboronic acid, 111 1,8-naphthalenedicarboxylicacid, 378 natural products, 268 natural receptors, 112 Nazarov cyclization, 258, 262, 265 N-Boc-allylglycine, 280 N-Boc-crotylglycine, 282 N-Boc-oxazinone, 287 N-Boc-protected amino acids, 274 N-bromosuccinimide, 255, 272 NBS, 284 n-Bu4NE 62 n-BuLi, 68, 78, 87 (+)-negamycin, 322 neodolabellanes, 247, 264 a-neodolabellenol, 244, 248, 250 NF-~.B activity, 220 Ni (0)-catalyzed coupling, 69 N-methyl-4-cyanopyridinium iodide, 137 N-methylmaleimide, 346, 355 N-methylmorpholine, 350 N-methylnicotinamide iodide, 141 NMNA, 142 NMO, 87 NMR, 145 NMR titration experiments, 119

501

INDEX NOE, 317, 439, 450 NOE correlation experiments, 45 NOE difference spectra, 48 NOESY, 439, 449 noncovalent conformers, 129 non-covalent interactions, 100 norbornadiene, 107 (2S,3R)-nor-C-statine, 319 N-phenyl maleic imide, 106 N-phenylmaleimide, 348, 354 N-propargylation, 341 a-sulfonyl allyl carbanion, 250 n-tetrabutylammonium iodide, 87, 92 nystatin, 1

0-di (dibromomethyl) benzene, 109 odorine, 222, 232 olefin metathesis, carbonyl olefination, 161 o-naphthoquinodimethane, 109 ophiobolins, 258 0-quinone methide, 468 organoborane, 85, 87, 91 organotin acetylides, 275 ortho ester Claisen rearrangement, 260 OsO4, 87 ovalicin, 462, 479 oxaphosphetane, 87 oxazinone, 270, 272, 311 oxazolidinone, 354, 357 oxepane, 60, 62, 64, 67, 69, 71, 76, 79, 83, 87, 90, 158, 196, 198, 211-212 oxidation, 316 oxidative addition, 384 oxidative cleavage, 14, 62, 220, 226, 247, 279, 307, 445 oxidative cyclization, 222 oxidative dearomatization, 463, 464, 472 oxidative hydroboration, 90 oxidative hydrolysis, 202 oxidopyrylium, 219, 223-224, 226-227 232-233, 237, 239 oxindole, 311 oxocarbenium ion, 165, 174, 186, 188 oxonium species, 255 oxy-Cope rearrangement, 7, 250

oxynitrilase-mediated asymmetric cyanohydrin-forming reaction, 272 oxysporidinone, 487 ozonolysis, 260, 287, 297, 318, 423, 431, 445, 451,456, 487

paclitaxel, 220 PADA, 485 palau'amine, 324 palladium chloride, 274, 282, 290, 299, 302 palladium mediated coupling, 467 palladium on carbon, 291 palladium-catalyzed coupling, 26 palladium-catalyzed cross-coupling, 9, 31 palladium-catalyzed reductive coupling, 396 palladium-catalyzed Suzuki-Heck type couplings, 111 p-anisaldehyde, 310 [2.2] paracyclophane, 375-376, 378, 381 parallel synthesis, 329 Pauson-Khand reaction, 334 Pb (OAc) 4, 226, 228 PbC12, 160, 161 p-benzoquinone, 103, 355 PCC, 394, 399 PCC on alumina, 69 Pd (0)-catalyzed tandem cyclization/ amination, 350 Pd(dba) 2, 28 Pd (PPh~) 4, 87, 91 PDC, 92 Pd-catalyzed alkynylation, 311 Pd-catalyzed coupling reactions, 332 PdClz, 81 p-DCNB, 126, 129 Pd-mediated reaction, 303 Pd-mediated SN2'-type cyclization, 303 Pearlman's catalyst, 270, 290, 310, 323 peptide coupling, 295 peptide isosteres, 268, 314 peptide-like molecules, 359 periodate cleavage, 276 periodic acid, 478, 484 Petasis reagent, 158, 204, 433 Ph2MeSiH, 200 phenanthrene ring, 236

502 3-phenyl-l-bromopropyne, 342 4-phenyl-[ 1,2,4]-triazole-3,5-dione, 349 phenylalanine analogues, 290 phenylmagnesium chloride, 187 phenylselenation, 263 PhI+OH-OTs, 469 phorbol, 224 phosgene (COC12), 353 phoslactomycin, 487 phosphines, 17 phosphine oxides, 17 phosphoramide catalysis, 15 phosphorus tribromide, 28 photoannulation chemistry, 425 photochemical apparatus, 235 photochemical cycloaddition, 230, 232, 233 photocycloaddition, 221,222, 228, 231 [3 + 2 ] photocycloaddition, 237 [2 + 2] photodimerization, 369-371,374, 375, 380, 381 photoreaction, 377 phytoalexin, 37 phytotoxic effects, 244 Pictet-Spengler, 295 PIDA, 469 PIFA, 468-471 pinacol rearrangement, 165 pinacolic coupling, 222 pKa values, 223 PKC, 384 Plasmodium falciparum, 36 PMB ether, 171, 177 :PMB group, 25 ]~methylcinnamic acid, 370 POCI~, 138 poison darts, 410 polyepoxide cascades, 156 polyketide, 463 polymethylhydrosiloxane, 320 1,3-polyols, 8 potassium acetate, 65 potassium azodicarboxylate (PADA), 479 potassium hexamethyldisilazane, 287 potassium hydroxide, 107, 419 p-quinol, 466 preamphilectane, 35 prenyl chloride, 82

INDEX prenyllithium, 69, 95 primary amine, 321 propanal, 15 propargyl ethers, 336 propargylic bromide, 341 (Z)-propenyllithium, 260 propenyl magnesium chloride, 171,173, 175 propiophenone, 15 propynylmagnesium chloride, 186 protease inhibitors, 359 protected 2,7-diaminosuberic acid, 280 protecting group, 305 pseudo allylic strain, 479 Pseudomonas putida, 461 Pseudopterogorgia elisabethae, 384 pseudopterosins, 385 p-toluenesulfonohydrazide, 255 pulegone, 417 pumiliotoxins, 411 pyridine, 195 pyridine alkaloids, 411 pyridinium chlorochromate, 394 pyrrolidine, 481

Q quadrone, 431 quinane natural products, 456 quinine, 303, 324, 470 quinuclidine, 303

R

radical anion, 47 Ramberg-B/icklund reaction, 206 Raney nickel, 270, 323 RCM (ring-closing metathesis), 83, 177, 182, 184, 190, 195-196, 200, 202, 204, 211, 215, 349, 447 rctt-l ,2,3,4cyclobutanetetracarboxylic acid, 380 rctt-l ,2-bis ( 4-pyridyl )-3,4-bis ( 4-chlorophe nyl ) cyclobutane, 374 reaction, 387, 394, 431,450 reactive conformer, 189 reagent-based control, 331

503

INDEX Rebek's imide, 378 red tide, 154 Red-N, 10 1,4-reduction, 314-315, 485 reductive amination, 295 reductive cleavage, 76 reductive elimination, 85 regioselective allylic oxidation, 486 regioselective oxidation, 78 regioselectivity, 309 renieramycin I, 323-324 (-)-renieramycin, 295 (+)-resiniferatoxin, 224 resorcinol, 371,373, 375, 378, 380 retro-aldol, 41 retro-aldol reaction, 43-44 retro-cycloaddition, 448 retro-Dieckmann fragmentation, 417 retro-Michael decomposition, 287 retrosynthetic analysis, 83, 90, 375, 415 retrosynthetic conversions, 423 reverse-micelles, 232 R-GABOB, 317 Rh (I) catalysis, 333 Rh ( I)-catalyzed cycloisomerization reaction, 334-335, 345, 349, 350, 361 Rh (I)-catalyzed formation of 5-1actams, 350 Rh (I)-catalyzed reaction, 344 Rh 2 (R-DOSP)4, 396 Rh2 (DOSP)4, 388

Rh2(S-DOSP)4-catalyzed reaction, 393 (-)-rhazinilam, 384 rhodium on alumina, 261 rhodium-catalyzed cycloisomerization, 328 rhodium-catalyzed insertion, 92 ring strain, 250 ring-opening, 447 (+)-rishirilide B, 460, 474, 476, 482-483 RK-397, 1 rocaglamide, 219 roflamycoin, 3 rotation, 131 roxaticins, 1, 3 Ru-catalyzed ring-closing metathesis, 332 RuO4, 477 ruthenium, 446 ruthenium catalyst, 400, 448 (R)-~ethyl aspartate, 281

S

S-(-)-cucurbitine, 279 (S)-2-methylasparagine, 289 (S)-GABOB, 316 (S)-q-hydroxy-~-aminobutyric acid, 316 sadenosylmethionine, 146 Saegusa oxidation, 48, 194, 197, 199, 253 safety concerns, 440 saframycin, 295 Sakurai coupling, 315 SAM, 146 (-)-sanglifehrin A, 291 saponification, 69, 353, 478 scaffolds, 336, 342 Schmidt or Beckmann rearrangement, 349 Schmidt reaction, 409, 423, 440, 451-452 Schrock Mo catalyst, 173, 190, 197, 211 secodolastanes, 247 secondary metabolites, 245 secondary orbital, 471 secondary orbital overlap, 474 secondary structure, 359 selective, 284 selective cleavage, 87 selective cleavage of the benzyl ether, 69 selective protection, 321 self-assembly, 99, 138, 146, 381 SEMCI, 289 SeO2, 65, 79 (+)-sertraline (Zoloft), 390 Sharpless asymmetric dihydroxylation, 83, 90, 92, 95 Sharpless asymmetric epoxidation, 3, 78, 423 SiCI4, 17, 20 SiF4, 215 sigmatropic 1,5-hydrogen shift, 106 silanol, 22, 27 silicon-based aldol, 32 silver(I) oxide, 69 silvestrol, 219-220 silyl dienol ether, 31 silyl ether, 252, 482 silylene, 183 Simmons-Smith cyclopropanation, 45 Simmons-Smith reagent, 51 single electron transfer reduction, 465 singlet oxygen, 264 (+)-rishirilide B, 485

504 six-component assemblies, 374 six-membered transition, 474 skeletal rearrangement, 220 SN2 displacement, 273 SN2 reaction, 471 SN2' displacement, 249, 481 SnC14, 69, 71 sodium, 280, 285, 467 sodium azide, 442 sodium borohydride, 22, 106, 419 sodium channels, 411 sodium chlorite oxidation, 200 sodium cyanoborohydride, 53 sodium hydrogen sulfite, 264 sodium hydrosulfite, 403 sodium iodide, 106-107 sodium naphthalide, 263 sodium silanolate, 27 sodium tert-amylate, 261 solid state, 369 solid-state reactivity, 381 solvent, 146 solvent-free environment, 369 Sonogashira, 332 sorbicillin, 471 spectrofluorimetric titrations, 119, 122 spirotryprostatin B, 309 S-proline, 473 stabilized ylide, 261 re-stacking interactions, 231 stereocontrolled, 253 stereocontrolled synthesis of (2R,3R), 299 stereoselective and chemoselective reactions, 470 stereoselective reduction, 214, 312 stereoselective vinylogous aldol addition, 30 stereoselectivity, 71,199, 231 stereospecific 1,2-shift, 245 stereospecific electrocyclization, 481 stereospecific sodium borohydride reduction, 272 steric hindrance, 358 steric strain, 125 stevens rearrangement, 314 stilbazoles, 373 Stiles carboxylation, 222 Stille, 332 Stille coupling, 6, 31 stoichiometric conversion of pentane to 1-pentene, 384

INDEX structural diversity, 329 Strycker reagent [Ph~PCuH] 6, 82 sulcatone, 423 sulfides, 17 sulfones, 250 a-sulfonyl ketones, 261 supermolecules, 99 supramolecular capsules, 100 supramolecular chemistry, 369, 381 Suzuki, 332 Suzuki-Heck type coupling, 111 Suzuki-Miyaura cross-coupling, 85, 87, 90-91, 95 Swern oxidation, 71, 86, 195, 261 syn elimination, 255

syn-3-hydroxy-4-methyl-pyrrolidine, 319 syn-diastereoselectivity, 302 1,5-syn stereoinduction, 17, 32 synthetic receptors, 138 (S)-methyl aspartate, 270

T TADDOL, 233, 236-237, 239 Takai cyclization, 206 Takai protocol, 158, 179, 196, 208 Takai reagent, 160 Takai-Utimoto reagent, 173, 191-192, 208, 211 Takeda reagent, 206 tandem intermolecular cycloaddition, 349 tautomer, 476 tautomerization, 262 taxusin, 226 TBAF, 27, 87, 182, 290, 340 TBDMS ether, 185, 189 TBDPSC1, 87 TBS ethers, 195, 203, 400 TBS-enol ethers, 396 TBSOTf, 2,6-1utidine, 71 TCNB, tropylium tetrafluoroborate, 119, 126, 130 TCNQ, 122, 126 Tebbe, 204 Tebbe reagent, 158, 161,433, 446 temperature, 234 temperature of coalescence, 134 temperature-dependent spectra, 131

505

INDEX templates, 379 tetrabromonaphthalene, 110 tetrabromo-0-quinodimethane, 110 tetrabromo-0xylene, 106 tetrabutylammonium hydroxide, 76 1,2,4,5-tetracyanobenzene, 119 tetrakis (4-pyridyl) 1,2,9,10diethano [2.2] paracyclophane, 375 tetramethylammonium, 24 tetramethylethylenediamineTMEDA, 78 tetrazomine, 300 TFA, 284, 295 TfOH, 438-439, 451-453 thallium nitrate, 469 thermodynamic control, 39 thermodynamic protonation, 437 thiocarbonyldiimidazole, 255 Thorpe-Ingold effect, 344 Ti (OiPr)4, 78 TiC14, 160, 427, 439, 451 TIPS ether, 182, 188 tltanacyclobutanes, 208 utanium enolate, 284 utanium methylidene, 208 titanium tetrachloride, 110 titanium(IV) enolates, 18 utanium-promoted regioselective opening, 79 titanocene alkylidene complexes, 445 TMEDA, 160 TMS ether hydrolysis, 192 TMS iodide, 276 TMS triflate, 52 TMSBr, 92 TMS-CF~, 321 TMSCN, 319 TMSI, 486 TMSOK-promoted coupling, 9 TMSOTf, 200 TMSOTf, Et~SiH, 201 TNE 122 TNFe, 220 toluenesulfonyl hydrazone, 53 tomentanol, 59 tomentol, 59 torsional strain, 212, 481 tosyl chloride, 106 tosylhydrazone, 73 TPAP, 199-200, 203, 212 4,4'-tppcp, 375

trans-l,2-bis(4-pyridyl)ethylene (4,4'bpe), 373 trans-conformation of secondary amide, 353 trans-crotyl chloride, 282 trans-2,5-dimethyl-pyrrolidine, 473 trans-diaxial couplings, 42 trans-[3.3.0]octane, 425 trans-stilbene, 271 tra ns, tra ns- l , 4-b is ( 4-p yri dyl ) - l ,

3-butadiene, 377 trans, trans, trans- 1,6-bis (4-pyridyl)-1,3,

5-hexatriene (1,6-bpht), 377 transamination, 272 transannular cyclizations, 252 transannular olefin-enone cyclization, 255 transannular reactions, 246, 264 transition metal-catalyzed cyclocarbonylation and cycloisomerization reactions, 335, 361 transition metal-catalyzed reactions, 332, 334 transition state, 21,252, 473 transmetalation, 85 1,2-transposition, 76 triacetoxyborohydride, 24 triallylaluminum, 174 triallylborane, 174 triazoline, 416 2,4,6-trichlorobenzoyl chloride, 5, 28 trichlorosilyl enolate, 15, 17, 20 tncycloillicinone, 463 triethyl orthoformate, 284 triethylamine, 103, 106 triethylphosphite, 28 triethylphosphonoacetate, 94 triflate, 194 triflic acid, 438 trifluoroacetic acid, 51 trimethyl orthoformate, 284, 311 trimethylaluminium, 68 trimethylphosphonoacetate, 64 ~trimethylsilyl acrolein, 11 triphenylphosphine, 61, 68, 336-337 2-triphenylphosphoranylidenebutanedioate, 68 triphenylsilanol, 81 triquinanes, 428 tris (dibenzylide neacetone)-dipalladium (0)-chloroform, 27 trypticene, 116 tuberostemoninol, 324 m-tyrosine, 291

506

0~,[3-unsaturated amide, 355 unsaturated amide, 355 unnatural amino acids, 269 ureas, 17

valium, 428 vanadium-catalyzed Sharpless epoxidation, 255 vanadyl bis-acetoacetate [VO(acac)2], 62 van-der-Waals or dispersion forces, 99 vibsane, 245 Vilsmeier reaction, 40 vinyl epoxide, 78, 81 vinyl magnesium bromide, 187 vinyl triflate, 396 vinyldiazoacetate, 393 vinyldiazoacetates, 388, 391 vinylogous aldol addition, 11-12 vinylogous ester, 477, 480, 484-485 vinylstannane, 81

W

Wacker's conditions, 81 Weinreb amide, 13, 94-95, 431,447 Wesley oxidation, 464 Wieland-Miescher ketone, 73, 78 Wilkinson catalyst, 479 Williamson cyclization, 78 Williamson reaction, 78

INDEX Wittig olefination, 16, 182, 261 Wittig reaction, 45, 321,399, 402 Wittig reagent, 90 Wolff-Kishner reaction, 53

X-ray analysis, 471 X-ray crystal structure, 355 X-ray crystal structure analysis, 236 X-ray crystallography, 44, 308, 348, 369, 453 X-ray diffraction, 247, 260, 264 X-ray diffraction analyses, 256

Yamaguchi method, 5 Yamaguchi protocol, 204 ylide, 306, 308

zetekitoxin, 324 Zimmerman-Traxler chair transition state, 297, 302 zinc, 110, 255 zinc chloride, 275 Zn, 160 Zn (Cu) amalgam, 215 Zn(N3) 2, 441 Zn(OTf) 2, 214 ZnC12, 336, 338, 340-341,486 (+)-zoapatanol, 59

Strategies and Tactics in Organic Synthesis, Volume 7 (Strategies and Tactics in Organic Synthesis)

Strategies and Tactics in Organic Synthesis, Volume 4 (Strategies and Tactics in Organic Synthesis)

Strategies and Tactics in Organic Synthesis, Volume 6 (Strategies and Tactics in Organic Synthesis)

Strategies and Tactics in Organic Synthesis, Vol. 6

Strategies for Organic Drug Synthesis

Strategies for Organic Drug Synthesis and Design

Palladium in Organic Synthesis

Tellurium in Organic Synthesis

Ruthenium in Organic Synthesis

Introduction to Strategies for Organic Synthesis

Ruthenium in Organic Synthesis

Radicals in Organic Synthesis

Microwaves in Organic Synthesis

Palladium In Organic Synthesis

Microwaves in organic synthesis

Organic Synthesis

Photochemistry and Organic Synthesis

Organic Synthesis

Organic Synthesis

Reagents for Organic Synthesis, Volume 7

Reagents for Organic Synthesis, Volume 7

Nitrile Oxides, Nitrones and Nitronates in Organic Synthesis: Novel Strategies in Synthesis

Hydrogen Bonding in Organic Synthesis

Domino Reactions in Organic Synthesis

Cycloaddition Reactions in Organic Synthesis

Cycloaddition Reactions in Organic Synthesis

Iridium Complexes in Organic Synthesis

Unit Processes In Organic Synthesis

Domino Reactions in Organic Synthesis

Microwave Methods in Organic Synthesis

Metal Carbenes in Organic Synthesis

Strategies and Tactics in Organic Synthesis, Volume 7 (Strategies and Tactics in Organic Synthesis)

Strategies and Tactics in Organic Synthesis, Volume 4 (Strategies and Tactics in Organic Synthesis)

Strategies and Tactics in Organic Synthesis, Volume 6 (Strategies and Tactics in Organic Synthesis)

Strategies and Tactics in Organic Synthesis, Vol. 6

Strategies for Organic Drug Synthesis

Strategies for Organic Drug Synthesis and Design

Palladium in Organic Synthesis

Tellurium in Organic Synthesis

Ruthenium in Organic Synthesis

Introduction to Strategies for Organic Synthesis

Ruthenium in Organic Synthesis

Radicals in Organic Synthesis

Microwaves in Organic Synthesis

Palladium In Organic Synthesis

Microwaves in organic synthesis

Organic Synthesis

Photochemistry and Organic Synthesis

Organic Synthesis

Organic Synthesis

Reagents for Organic Synthesis, Volume 7

Reagents for Organic Synthesis, Volume 7

Nitrile Oxides, Nitrones and Nitronates in Organic Synthesis: Novel Strategies in Synthesis

Hydrogen Bonding in Organic Synthesis

Domino Reactions in Organic Synthesis

Cycloaddition Reactions in Organic Synthesis

Cycloaddition Reactions in Organic Synthesis

Iridium Complexes in Organic Synthesis

Unit Processes In Organic Synthesis

Domino Reactions in Organic Synthesis

Microwave Methods in Organic Synthesis

Metal Carbenes in Organic Synthesis

Recommend Documents