Materials Science Research Horizons

MATERIALS SCIENCE RESEARCH HORIZONS

MATERIALS SCIENCE RESEARCH HORIZONS

HANS P. GLICK EDITOR

Nova Science Publishers, Inc. New York

Copyright © 2007 by Nova Science Publishers, Inc. All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. For permission to use material from this book please contact us: Telephone 631-231-7269; Fax 631-231-8175 Web Site: http://www.novapublishers.com NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material. Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS. LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA Materials science research horizons / Hans P. Glick (editor). p. cm. Includes index. ISBN-13: 978-1-60692-751-9 1. Materials science. I. Glick, Hans P. TA403.M347155 2006 620.1'1--dc22 2006032477

Published by Nova Science Publishers, Inc.

New York

CONTENTS Preface

vii

Chapter 1

Photoionization of Polyvalent Ions Doris Möncke and Doris Ehrt

Chapter 2

Growth and Characterization of δ-Bi2O3 Thin Films by Chemical Vapour Deposition under Atmospheric Pressure T. Takeyama, N. Takahashi, T. Nakamura and S. Itoh

Chapter 3

Chapter 4

Porous Materials: The Mathematical-Physical Expressions for Some Properties of Three-Dimensional Reticulated Porous Metallic Materials in the Same Analytical Model System P.S. Liu Influences of Process Parameters, Inclusion and Void in Copper Wire Drawing Somchai Norasethasopon

1

57

81

109

Chapter 5

Development of Hardfacing for Fast Breeder Reactors A. K. Bhaduri and S. K. Albert

149

Chapter 6

Tissue Engineering of Cartilage in Bioreactors Nastaran Mahmoudifar and Pauline M. Doran

171

Chapter 7

Heterogeneous Combustion Synthesis Hung-Pin Li

193

Chapter 8

Recycling of Ecocompatible Treated Red Mud and Compost from SS-MSW: Examples of Use on Sediment and Mine Soil Samples P. Massanisso, E. Nardi, R. Pacifico, L. D’Annibale, C. Cremisini and C. Alisi

217

Formation and Adjustment of Bubbles in a Polyurethane Shape Memory Polymer W.M. Huang, B. Yang, L.H. Wooi, S. Mukherjee, J. Su and Z.M. Tai

235

Chapter 9

Index

251

PREFACE Materials science includes those parts of chemistry and physics that deal with the properties of materials. It encompasses four classes of materials, the study of each of which may be considered a separate field: metals; ceramics; polymers and composites. Materials science is often referred to as materials science and engineering because it has many applications. Industrial applications of materials science include processing techniques (casting, rolling, welding, ion implantation, crystal growth, thin-film deposition, sintering, glassblowing, etc.), analytical techniques (electron microscopy, x-ray diffraction, calorimetry, nuclear microscopy (HEFIB) etc.), materials design, and cost/benefit tradeoffs in industrial production of materials. This book presents new research directions in this rapid-growing field. Chapter 1 - The effect of polyvalent dopants on photoinduced defect formation was studied in different glasses. Ionization of the glass matrix results in intrinsic defects, positively charged hole and negatively charged electron centers. Polyvalent dopants can be photooxidized or photoreduced. These extrinsic defects might replace selectively one or several intrinsic defects and / or cause an increase in the number of opposite charged defects. Photoionization can also result in unusual dopant valences otherwise not observed in glasses. The systematic comparison of different dopants and glass systems irradiated by excimer lasers helps to understand defect generation processes and might eventually help in the design of UV-resistant or UV-sensitive glasses. Defect formation occurs in the ppm range and was analyzed by optical and EPR spectroscopy. A series of polyvalent dopants such as typical trace impurities, glass or melt additives and typical dopants used for optical components like filter glasses, optical sensors, fluorophores or photochromes, were studied. Distinct melting conditions give rise to different valences of various dopants and as a consequence different photoinduced redox-reactions might be observed after irradiation. Qualitative and quantitative changes in the defect formation rates depend on the: • kind and concentration of the dopant, c was varied from 50 to 5000 cation ppm. • radiation parameters such as wavelength, or power density of the excimer lasers used. • glass matrix; (fluoride-)phosphate and borosilicate glasses give rise to different intrinsic defects of varying stability. The matrix determines also the initial incorporation like valence or coordination of the dopants and stabilizes or destabilizes photoionized dopant species.

viii

Hans P. Glick •

initial transmission of the glass sample, which also depends on the dopant (kind, valence, coordination), its concentration, and the thickness of the sample plate, d was varied from 0.5 to 2mm. Some dopants are photooxidized while others are photoreduced Some defects recombine easily or transform into more stable defects while others are stable for months or years. Chapter 2 - Bismuth oxide (Bi2O3) thin films are interesting materials within the class of oxide semiconductors, owing to a variety of physical properties determined by its many polymorphs. This semiconductor is characterized by significant values of band gap, dielectric permittivity and refractive index as well as marked photosensitivity and photoluminescence. These properties make Bi2O3 films well suited for many applications in various domains such as microelectronics, sensor technology and optical coatings. However, the characteristics of this film strongly depend on its crystal phases: its electrical conductivity may vary by over 5 orders of magnitude, while its energy gap may change from around 2 to 3.96 eV. Therefore, it is required to manufacture high-quality Bi2O3 films with a single phase. Thin films of δ-Bi2O3 were prepared on the sapphire (0001) and the yttria-stabilized zirconia (YSZ) (111) substrate by means of chemical vapour deposition under atmospheric pressure. X-ray diffraction measurement revealed the deposited δ-Bi2O3 films on the YSZ (111) substrates have good crystal quality and a flat surface. The full width at half maximum value of out-of-plane rocking curve is 0.0260˚ (93.6 arcsec.). An optical band gap of 3.28 eV was estimated by the optical transmittance measurement. Spectroscopic ellipsometry shows that the refractive index n of the single crystalline δ-Bi2O3 film at 800 ˚C is 2.4940 with 632.80nm. We believe this is the first time to investigate the optical properties of δ-Bi2O3 thin film. Chapter 3 - New developments are ceaselessly gained for the preparation, the application and the property study of porous materials. As to the theories about the structure and properties of porous materials, the famous classical model - Gibson-Ashby model has been being commonly endorsed in the field of porous materials all over the world, and is the theoretical foundation widespreadly applied by numerous investigators to their relative researches up to now. But there are some shortcomings in this classical model in fact, e.g., the impossible close-packed of pore units and the unequivalent struts. In this chapter, another model for porous materials are introduced which is put forward by the present author, and this new model can make up those shortcomings existed in Gibson-Ashby model. More importantly, the mathematical-physical expressions, which are well in agreement with the relevant experimental results, can be smoothly acquired for some properties of threedimensional reticulated foamed materials using this new model. These expressions include the relationship between electrical resistivity and porosity, the relationship between tensile strength and porosity, the relationship between relative elongation and porosity, and the relationship between biaxial loading strength and porosity. The experimental results showed that, the obtained mathematical-physical relations from this new model are obviously more excellent than that from Gibson-Ashby model when applying into the porous materials. Chapter 4 - In the copper fine wire drawing, the breakage and defect of the wire were fatal to the success of quantitative drawing operations. The first part of this paper shows how three of the main process parameters, the die half-angle, reduction of cross-sectional area and numbers of the drawing pass influenced drawing stress and internal defect by experiment. The influences of a non-central inclusion and void in the single-pass copper shaped-wire drawing were investigated by 2D FEM. The effects of the lateral and longitudinal sizes of a

Preface

ix

central inclusion in the multi-pass copper shaped-wire drawing were also investigated. Based on the experimental data of the optimal die half angle, wire deformation, plastic strain, hydrostatic stress and drawing stress of the copper shaped-wire containing a non-central inclusion and void were calculated. The copper shaped-wire that contained a central inclusion and void was also calculated. During drawing a wire containing a non-central inclusion, necking, bending and misalignment occurred. However, only necking occurred in the case of the central inclusion wire. In the case of the non-central inclusion wire, inclusion rotation occurred. For the same inclusion size, the inclusion size strongly influenced drawing stress but the eccentric distance slightly influenced drawing stress. The drawing stress of the copper shaped-wire that contained a central inclusion was greater than the case of the wire that contained a non-central inclusion. The drawing stress decrement due to a void and the opposite deformation behaviour between the wire that contained a central void and inclusion were found. The effects of the lateral and longitudinal sizes of a central inclusion and void on the drawing and the maximum hydrostatic tensile stress during the multi-pass copper shapedwire drawing were also carried out. The present paper also shows how two of the inclusion parameters, the size and aspect ratio of the elliptical inclusion, influenced drawing stress and maximum hydrostatic stress of the copper shaped-wire during drawing. It was found that the maximum drawing stress increased as the longitudinal inclusion size and aspect ratio increased. Both longitudinal inclusion size and aspect ratio influenced the inclusion leading edge location where the maximum hydrostatic tensile stress was induced. The necking due to a central inclusion in copper shaped-wire drawing occurred on some parts of the wire surface in front of and nearby the inclusion and the lateral neck size decreased when the longitudinal and lateral inclusion sizes increased as the inclusion passed through the die. The maximum hydrostatic tensile stress directly increased as the inclusion aspect ratio increased for the small and medium inclusions but it inversely increased for the large inclusion. It was mostly found where the inclusion leading edge was located in the drawn zone. The influences of a central inclusion on the plastic deformation, hydrostatic stress and drawing stress in the round-to-round copper wire drawing were also investigated by 3D FEM. Chapter 5 - Various components of the Fast Breeder Reactors encounter wear of adhesive or abrasive nature and sometimes erosion. Hardfacing by weld deposition have to be used to improve the resistance to high temperature wear, especially galling, of mating surfaces in sodium. Based on radiation dose rate and shielding considerations during maintenance, handling and decommissioning, nickel-base E NiCr-B hardfacing alloy was chosen to replace the traditionally used cobalt-base Stellite alloys. Studies, on the effect of long term ageing of NiCr hardface deposits on austenitic stainless steel substrate, demonstrated that E NiCr-B deposits after exposure at service temperatures up to 823 K would retain adequate hardness well above RC 40 at end of the components’ designed service-life of up to 40 years. Further, based on detailed metallurgical studies, including residual stress measurements after thermal cycling, the more versatile plasma transferred arc welding (PTAW) process was chosen for deposition of the E NiCr-B hardfacing alloy, so that the width of the dilution zone could be controlled by optimising the deposition parameters. This paper outlines the adaptation of technology for hardfacing with the E NiCr-B alloy using the selected PTAW process, through collaborative efforts with industries, for development of hardfacing technology for the various components of PFBR.

x

Hans P. Glick

Chapter 6 - The main goal of cartilage tissue engineering is to generate three-dimensional cartilage and osteochondral tissues for use in repair of large cartilage injuries. Cartilage constructs are generated by seeding and culturing viable cells in biodegradable polymer scaffolds under conditions suitable for tissue formation. In this chapter, current developments in cartilage tissue engineering are reviewed, focusing on the source of cells, the polymer scaffolds, seeding systems, bioreactors and application of mechanical stimulation for cell differentiation and tissue production. The generation of cartilage tissue constructs in the laboratory using a bioreactor system is also described. Chondrocytes were isolated from human foetal epiphyseal cartilage, expanded in monolayer, dynamically seeded into poly(glycolic acid) (PGA) polymer scaffolds and cultured in recirculation bioreactors. Composite scaffolds were used to improve the initial distribution of cells within the scaffolds and to develop cartilage constructs that were homogeneously cartilaginous throughout their thickness. The quality of the engineered cartilage was assessed after 5 weeks of bioreactor culture in terms of tissue wet weight, cell, glycosaminoglycan (GAG), total collagen and collagen type II contents, histological analysis of cell, GAG and collagen distributions, immunohistochemical analysis of collagen types I and II, and ultrastructural analysis using transmission electron microscopy. Chapter 7 - Many exothermic non-catalytic solid-solid or solid-gas reactions, after being ignited locally, can release enough heat to sustain the self-propagating combustion front throughout the specimen without additional energy. Since the 1970’s, this kind of exothermic reaction has been used in the process of synthesizing refractory compounds in the former Soviet Union. This novel technique, so-called Combustion / Micropyretic synthesis or Selfpropagating High-temperature Synthesis(SHS), has been intensively studied for process implication. This technique employs exothermic reaction processing, which circumvents difficulties associated with conventional methods of time and energy-intensive sintering processing. The advantages of combustion synthesis also include the rapid net shape processing and clean products. In addition, the combustion-synthesized products have been reported to possess better mechanical and physical properties. Heterogeneous distributions of reactants, diluents, and pores are common during combustion synthesis when powders are mixed, and this directly leads to the variations of the thermophysical / chemical parameters of the unreacted compacts. Since combustion synthesis is sustained by the sequences of the local chemical reactions, the propagation manner is strongly dependent on the parameters of each portion of the reactants. Thus, the variation of thermophysical / chemical parameters of reactants caused by heterogeneities in composition and porosity is thought to significantly change the processing parameters, such as combustion temperature and propagation velocity; and further affect the product properties. This chapter systematically introduces the impact of heterogeneities during combustion synthesis with Ni + Al. Correlations of heterogeneities in the reactants and a diluent with the propagation velocity and combustion temperature are discussed. In addition, a map, considering concurrent heterogeneities in the composition and porosity, has been generated to provide a better understanding of the change in propagation velocity on account of the heterogeneous combustion synthesis. Chapter 8 - Ecological restoration of polluted areas is an increasing necessity for many countries around the world. Current technologies used to recover polluted soil and sediment are in general too costly. Recently, on-site approaches such as metal trapping and phytoremediation have attracted attention for their ability to meet criteria of economicity.

Preface

xi

Metal trapping is based on the diminution of metal mobility and availability as a result of applying soil amendments, for example particular industrial residues. Phytoremediation is an appealing environmental cleanup technology but a deeper understanding of the complex interactions in the soil-plant system is still needed. In this study, the effect of adding treated red mud (BauxsolTM - material with the potential to immobilise metal) on mine soil and on sediment (from a volcanic coastal lagoon in Southern Italy) and of adding both red mud and compost (produced from Source-Separated Municipal Solid Waste) on trace elements fractionation and mobility, have been investigated. Barley (Hordeum vulgare) was used as a plant model to follow any change in matrices phytotoxicity: seedlings were transplanted in pots containing the contaminated mine soil or sediment and a mixture of the investigated matrices with different percentages of treated red mud and compost. Plant growth was studied also by controlling the total protein content, biomass and enzyme activity. The knowledge of trace elements mobility and “speciation” in contaminated soils and sediments is an important requisite for any further environmental evaluation and these features can be evaluated through leaching tests or by "sequential extraction procedure". In this work, total concentration of selected trace elements, their fractionation by sequential extraction procedure (BCR standardised) and leaching batch tests using a kinetic approach, were studied. The most evident result in the soil trials was that the utilization of amendments, used both separately and in a mixture, always improved the growth of barley plants. In particular, barley seedlings were practically not able to grow on the polluted mine soil and the simple adding of red mud resulted in a significant improvement in plant development. An even more drastic improvement was obtained with the addition of compost and compost plus treated red mud. In the sediment trials, the best yield in plant growth was obtained in the pot with the addition of treated red mud alone. The necessity of a delicate compromise between the maintaining of an acceptable plant viability and the control of metal mobility seems to be achievable through a careful balancing of the percentages of compost and red mud utilized as amendments. Chapter 9 - Two approaches are proposed for realizing porous polyurethane shape memory polymers using water as a non-harm foam agent. We show that it is possible to control the bubbles by varying the moisture ratio and heating procedure. We demonstrate that one can further modify the size of bubbles by further heat treatment. As such, one can make resizable micro bubbles and even channels.

In: Materials Science Research Horizons Editor: Hans P. Glick pp. 1-56

ISBN 978-1-60021-481-3 © 2007 Nova Science Publishers, Inc.

Chapter 1

PHOTOIONIZATION OF POLYVALENT IONS Doris Möncke∗ and Doris Ehrt Otto-Schott-Institut für Glasschemie, Friedrich-Schiller-Universität, Fraunhoferstr.6, D-07743 Jena, Germany

ABSTRACT The effect of polyvalent dopants on photoinduced defect formation was studied in different glasses. Ionization of the glass matrix results in intrinsic defects, positively charged hole and negatively charged electron centers. Polyvalent dopants can be photooxidized or photoreduced. These extrinsic defects might replace selectively one or several intrinsic defects and / or cause an increase in the number of opposite charged defects. Photoionization can also result in unusual dopant valences otherwise not observed in glasses. The systematic comparison of different dopants and glass systems irradiated by excimer lasers helps to understand defect generation processes and might eventually help in the design of UV-resistant or UV-sensitive glasses. Defect formation occurs in the ppm range and was analyzed by optical and EPR spectroscopy. A series of polyvalent dopants such as typical trace impurities, glass or melt additives and typical dopants used for optical components like filter glasses, optical sensors, fluorophores or photochromes, were studied. Distinct melting conditions give rise to different valences of various dopants and as a consequence different photoinduced redox-reactions might be observed after irradiation. Qualitative and quantitative changes in the defect formation rates depend on the: − − −

∗

kind and concentration of the dopant, c was varied from 50 to 5000 cation ppm. radiation parameters such as wavelength, or power density of the excimer lasers used. glass matrix; (fluoride-)phosphate and borosilicate glasses give rise to different intrinsic defects of varying stability. The matrix determines also the initial incorporation like valence or coordination of the dopants and stabilizes or destabilizes photoionized dopant species.

Tel.: +49-3641-948511 / 948506; fax: +49-3641-948502, [email protected] or [email protected].

2

Doris Möncke and Doris Ehrt −

initial transmission of the glass sample, which also depends on the dopant (kind, valence, coordination), its concentration, and the thickness of the sample plate, d was varied from 0.5 to 2mm.

Some dopants are photooxidized while others are photoreduced Some defects recombine easily or transform into more stable defects while others are stable for months or years.

INTRODUCTION Solarization in glasses was first described by Faraday in 1825. The effect of irradiation induced transmission changes has been investigated since for its scientific and technological significance [1-31]. Pelouze recognized already in 1867 that a change in the oxidation state of the typical glass impurities Fe and Mn can cause strong solarization effects [2]. UV-radiation excites valence electrons in the irradiated material and complicated photoreaction processes lead subsequently to the formation of irradiation induced defects. Defects are generated in ppm concentrations and occur in pairs of negative electron centers (EC) and positive hole centers (HC). While intrinsic defects arise from the glass matrix itself are extrinsic defects connected to dopants or impurities. The formation of defects may result in transmission changes but also in changes of the refractive properties of the material. Studying the processes and mechanisms that govern defect formation requires more attention as stronger lamps and lasers, which work at increasingly shorter wavelengths, are more and more utilized. This knowledge can than be exploited in the development of photosensitive or photoresistant appliances. Because of their strong electronic transitions in the ultraviolet (UV) and visible (VIS) spectral range were, in analogy to similar absorbances in crystals, defects initially called color centers. Optical spectroscopy is the method of choice when studying defect formation; however, optical spectra of doped glasses are often dominated by transitions of the dopants that overlay the defects bands [3-18]. Complementary information on these defects, even regarding their detailed structure, can be derived from EPR-spectroscopy as many defects are paramagnetic [7-23, 3-18] The addition of polyvalent ions often initiates or enhances considerably the formation of defects in a glass sample. Extrinsic defects can form in addition to intrinsic defects and thus cause the increased generation of reversibly charged intrinsic defects. On the other hand can extrinsic defects substitute selectively one or more intrinsic defects of like charge [3, 19-23]. Irradiation induced defects can further be classified according to their stability in transient or stable defects. Some initially formed defects transform rapidly into more stable defects, sometimes even during the irradiation process. The transformation of defects in thermodynamically more stable centers can then again be hindered kinetically. Defect formation is a dynamic process and the kind and rate of defect development depends on many factors, e.g. the glass matrix, the concentration and species of any dopants, the initial transmission of the sample, or on the radiation parameters. For example excites UV-radiation only valence electrons while X-ray radiation detaches even the inner electrons in the material. Accordingly are different defects initiated by different radiation sources [24].

Photoionization of Polyvalent Ions

3

This chapter intends to compare the role of a wide range of polyvalent metals in the formation of irradiation induced defects. All glass samples were irradiated with excimer lasers in the UV and the laser induced defects were characterized by optical and EPR spectroscopy. Only defects stable at room temperature are discussed. Even these relative stable defects show transformation and recombination reactions when the samples were stored at room temperature in the dark. The glass types studied were selected for their high transparency in the deep ultraviolet (λ0~160-185 nm) and consequently their application in high performance optics [25-31, 35].

2. EXPERIMENTAL SECTION Fluoridephosphate (FP) and metaphosphate glasses (NSP) were selected as primary matrix glasses for the irradiation experiments. The generated extrinsic defects were characterized by EPR and optical spectroscopy and when possible identifified yb type and charge. Additional experiments using borosilicate type samples were added later in order to study the dopants effect on defect formation in an entirely different glass matrix. All samples were prepared and irradiated under defined and comparable conditions.

2.1. Sample Preparation The preparation of the different high purity glasses has been described in detail before [718, 25-31]. Only high purity reagents were used for all glasses. The iron content of the duran type borosilicate glass was < 1 ppm, of NSP ~ 5ppm and of FP10 < 10 ppm. The total iron content was analyzed by wet-chemical analysis. The Fe3+ content was also determined from the absorption of its CT-transition at 250 nm in the optical spectra [9, 27, 28-31]. The high purity dopant components were added in various amounts between 50 and 5000 ppm (cation %). The fluoroaluminate FP10 has the synthetic composition [10 P2O5 · 90 (AlF3, CaF2, SrF2, MgF2) mol%] and was melted at 1100°C under air in platinum crucibles. In order to obtain reduced dopant species were some samples also remelted under reducing melting conditions under argon atmosphere in glass-carbon crucibles. The metaphosphate glass NSP [10 Na2O · 40 SrO·50 P2O5 mol%] was melted under air at 1300°C in SiO2-crucibles. The dopants were reduced by the addition of 0.2 to 1 wt% carbon to the batch. Low alkaline borosilicate samples of the duran type [82 SiO2·12 B2O3·5 (K/Na)2O·1 Al2O3 mol%] were prepared under air at 1650°C. 250 to 1000 g batches were processed for 35 hours. For some samples were oxidizing or reducing conditions established by using the corresponding nitrate or tartrate salts of the reagents. All melts were cast in preheated graphite moulds and annealed from 500 or 550 °C to room temperature with a cooling rate of 30 K / h.

4

Doris Möncke and Doris Ehrt

2.2. Radiation Sources Polished samples plates of the dimensions 10 x 20 mm and a thickness of either 0.5, 1, or 2 mm were irradiated with excimer lasers. The sample thickness was chosen for each sample in accordance to the initial absorbance at the irradiation wavelength. Excimer lasers working at 193 nm (ArF-laser), 248 nm (KrF-laser), and 351 nm (XeFlaser) were used. The power density of the laser was 200 mJ/cm² per pulse by a pulse duration of ~ 20 ns. The optical spectra were taken with increasing accumulated pulse numbers (at 10, 100, 1000, 5000, and 10000 pulses). The final pulse number normally suffices to reach the saturation level. EPR spectra of the samples were taken once after the final irradiation. Further optical spectra were obtained at increasing time intervals in which the irradiated samples were stored in the dark at room temperature. Some samples were irradiated by a high pressure mercury lamp or HOK lamp with a spectral power density of 1 kW with a wide continuos spectrum from 190 nm in the UV to the NIR.

2.3. UV-VIS-Spectroscopy UV-VIS-NIR spectra were taken in the range from 190 to 3000 nm. A double beam spectrophotometer (UV-3102, Shimadzu) recorded the absorbance Aλ=lg(T0/T) with an error 1.0. Wire breakage occurred when Di/Do = 0.4 in the fifth pass drawing and Li/Do = 0.3 and 0.4 in the fourth and fifth pass drawing. The drawing pass numbers strongly influenced maximum hydrostatic stresses in the fourth and fifth pass drawing. They inversely and directly influenced drawing stress in the first-to-fourth and fifth pass drawing, respectively. Central Void Effects: Multi-Pass Copper Shaped-Wire Drawing. Necking occurred on the copper shaped-wire surface at the wire portion that contained a void and the lateral neck size decreased as the void size increased. The maximum hydrostatic tensile stress was not found at the wire centreline but was found on the wire surface at the neck. The maximum hydrostatic tensile stress directly increased as the void size increased. The void was transformed to be an almond-shaped void and the sharp-edge of a transformed void point in the opposite direction of the metal flow and also was transformed to be a linear crack or "pipe". The wire breakage due to the high hydrostatic tensile stress and large plastic deformation of the copper matrix in the neck was found when the void size was very large. The maximum drawing stress was normally equal to the drawing stress of the copper shaped-wire drawing without void. The minimum drawing stress of copper shaped-wire that contained the largest void, the lateral void size was equal to 1.0 (lateral wire size equal to lateral void size), was obtained and was equal to 0.0. Inclusion Size and Aspect Ratio Effects: Single-Pass Copper Shaped-Wire Drawing. The necking behavior is the same as described above. The lateral neck size decreased in accordance with the increase in longitudinal and lateral inclusion sizes while the inclusion passed through the die. The medium inclusion (a/h ratio was approximately 0.2 to 0.5) strongly influenced the maximum hydrostatic tensile stress and it rapidly increased as longitudinal inclusion size increased in this range. It directly increased as the inclusion aspect ratio increased (elliptical inclusion approach to be a circular inclusion) for the small (a/h ratio was approximately 0.0 to 0.2) and medium inclusion. It inversely increased as the inclusion aspect ratio increased for the large (a/h was approximately 0.5 to 0.8) inclusion. The maximum hydrostatic tensile stress was found where the inclusion leading edge was located in the drawn zone and was far away from the die exit. It was not found in the case of a/h = 0.4 where b/a = 0.6, a/h = 0.2 where b/a = 0.8 and a/h = 0.2 where b/a = 1.0. The maximum drawing stress occurred when the inclusion passed through the die and increased as the longitudinal inclusion size and aspect ratio increased. It was found when the inclusion leading edge was located at the inclusion displacement ratio equal to 0.55 in the reduction zone. Inclusion Size Effects: Single-Pass Round-to-Round Copper Wire Drawing. The inclusion size directly strongly influenced necking and maximum hydrostatic tensile stress of the copper wire. The maximum hydrostatic tensile stress was found when the inclusion leading edge was located around the die exit. But the maximum drawing stress was found when the inclusion leading edge was located around the reduction zone center.

146

Somchai Norasethasopon

ACKNOWLEDGMENT The author wishes to express his appreciation to the Director of the National Metal and Materials Technology Center (MTEC), National Science and Technology Development Agency, Thailand, for his support and assistance in many details of the finite element program "MSC.MARC" for this problem simulation. The author would like to thank Prof. Dr. Yoshida, K., Department of Precision Mechanics, School of Engineering, Tokai University, Japan, and Nissapakul, P., Tangsri, T., and Pramaphant, P., Department of Mechanical Engineering, Faculty of Engineering, King Mongkut’s Institute of Technology Ladkrabang, Thailand, for giving him valuable discussion and comment.

REFERENCES [1] [2]

[3] [4] [5] [6] [7] [8] [9]

[10] [11]

[12]

[13]

Morton, J., (1999). Thomas Bolton and Sons and the rise of the electrical industry. Engineering Science and Education Journal, 5-12. Miyashita, K., Sugiyama, K., Moriai, H., Kamata, K., Tachikawa, K., and Fukuda, K., (1999). Electromagnetic Properties of Bronze Processed Nb3Sn Superconducting Wires and Multi-strand Cables for A.C. Use with a Cu-Sn-X (X,Ge,Ni,Mn,Si) Matrix and a Nb-Ta Core. IEEE Transactions on Applied Superconductivity, Vol. 9, No. 2, 709-712. Mielnik, E. M., Metalworking Science and Engineering; McGraw-Hill, Inc.; New York, 1991, pp 397-462. Raskin, C., (1997). Proceedings of the WAI International Technical Conference. Italy. Amstead, B. H., Ostwald, P. F., and Begeman, M. L., Manufacturing Processes; John Wiley and Sons, Inc.; Singapore, 1987, pp 1-687. Johnson, H. V., Manufacturing Processes; Bennett and McKnight; USA, 1984, pp 14581. Kutz, M., Mechanical Engineers’ Handbook; John Wiley and Sons, Inc.; New York, 1998, pp 3-1205 Colangelo, V. J., and Heiser, F. A., Analysis of Metallurgical Failures; John Wiley and Sons, Inc.; Singapore, 1989, pp 240-322. Avitzur, B., Metal Forming: Processes and Analysis; McGraw-Hill; New York, 1968, pp. 153-258. Revised edition reprinted by Robert Krieger Publishing Co., Inc.; Huntington, N.Y., 1979. Avitzur, B., Study of Flow Through Conocal Converging Dies; Metal Forming; A. L. Hoffmanner (ed.), Plenum Press, New York, 1971, pp 1-46. Campos, H. B., and Cetlin, P. R., (1998). The influence of die semi-angle and to the coefficient of friction on the uniform tensile elongation of drawn copper bars. Journal of Materials Processing Technology, Vol. 80-81, 388-391. Campos, H. B., Castro, A. L. R., and Cetlin, P. R., (1996). Influence of die semi-angle on mechanical properties of single and multiple pass drawn copper. Journal of Materials Processing Technology, Vol. 60, 179-182. Norasethasopon, S., and Tangsri, T., (2001). Experimental Study of the Effect of a Half-Die Angle on Drawing Stress during Wire Drawing. Ladkrabang Engineering Journal, Vol. 18, 134-139.

Influences of Process Parameters, Inclusion, and Void in Copper Wire Drawing

147

[14] Norasethasopon, S., and Yoshida K., (2003). Influence of an Inclusion on Multi-Pass Copper Shaped-Wire Drawing by 2D Finite Element Analysis. International Journal of Engineering, I. R. Iran, Vol. 16, No. 3, 279-292. [15] Norasethasopon, S., and Yoshida, K., (2006). Influences of inclusion shape and size in drawing of copper shaped-wire. Journal of Materials Processing Technology, Vol. 172, No. 3, 400-406. [16] Norasethasopon, S., and Yoshida, K., (2006). Finite-element simulation of inclusion size effects on copper shaped-wire drawing. Materials Science and Engineering: A, Vol. 422, No. 1-2, 252-258.

In: Materials Science Research Horizons Editor: Hans P. Glick pp. 149-169

ISBN 978-1-60021-481-3 © 2007 Nova Science Publishers, Inc.

Chapter 5

DEVELOPMENT OF HARDFACING FOR FAST BREEDER REACTORS A. K. Bhaduri and S. K. Albert Materials Joining Section, Materials Technology Division, Indira Gandhi Centre for Atomic Research, Kalpakkam 603102, India

ABSTRACT Various components of the Fast Breeder Reactors encounter wear of adhesive or abrasive nature and sometimes erosion. Hardfacing by weld deposition have to be used to improve the resistance to high temperature wear, especially galling, of mating surfaces in sodium. Based on radiation dose rate and shielding considerations during maintenance, handling and decommissioning, nickel-base E NiCr-B hardfacing alloy was chosen to replace the traditionally used cobalt-base Stellite alloys. Studies, on the effect of long term ageing of NiCr hardface deposits on austenitic stainless steel substrate, demonstrated that E NiCr-B deposits after exposure at service temperatures up to 823 K would retain adequate hardness well above RC 40 at end of the components’ designed service-life of up to 40 years. Further, based on detailed metallurgical studies, including residual stress measurements after thermal cycling, the more versatile plasma transferred arc welding (PTAW) process was chosen for deposition of the E NiCr-B hardfacing alloy, so that the width of the dilution zone could be controlled by optimising the deposition parameters. This paper outlines the adaptation of technology for hardfacing with the E NiCr-B alloy using the selected PTAW process, through collaborative efforts with industries, for development of hardfacing technology for the various components of PFBR.

1. INTRODUCTION The Indian 500 MWe Prototype Fast Breeder Reactor (PFBR) is a pool-type liquidsodium-cooled reactor having two separate sodium circuits with the intermediate heat exchanger (IHX) providing thermal contact between the primary pool and the secondary circuit. The secondary sodium circuits transfer heat from the IHX to the steam generator, the

150

A. K. Bhaduri and S. K. Albert

steam from which drives the conventional steam turbines. The minimum sodium temperature in the primary pool during normal operation is 673 K while the mean above-core temperature is 823 K. The minimum and maximum sodium temperatures in the secondary circuit are 628 and 798 K, respectively. The steam temperature is 763 K at a pressure of 16.6 MPa. In the PFBR, 316L(N) austenitic stainless steel (SS) has been chosen as the structural material for components operating above 673 K. The liquid sodium coolant acts as a reducing agent and removes the protective oxide film present on the SS surface of the in-sodium components. Many of these components would be in contact with each other or would have relative motion during operation, and their exposure at high operating temperatures (typically 823 K) coupled with high contact stresses could result in self-welding of the clean metallic mating surfaces. In addition, the relative movement of mating surfaces could lead to galling, a form of high-temperature wear, in which material transfer occurs from one mating surface to another due to repeated self-welding and breaking at contact points of mating surfaces. Further, susceptibility to self-welding increases with temperature for 316 SS [1]. Hardfacing of the mating surfaces has been widely used in components of water-cooled and liquidsodium cooled FBRs to avoid self-welding and galling [2, 3]. Cobalt-base hardfacing alloys (e.g. Stellite©) have been traditionally used very extensively for high temperature application in many critical hardfacing applications due to their excellent wear-resistance properties [4]. However, when cobalt-base alloys were used in a nuclear reactor environment, the cobalt-60 isotope formed due to irradiation enhances the radiation dose rate to operating personnel during handling, maintenance or decommissioning of the hardfaced components. Hence, there is an emerging trend of avoiding the use of cobaltbase alloys for hardfacing of nuclear power plant components. Nickel-base hardfacing alloys (e.g. Colmonoy©) were developed mainly to replace the cobalt-base alloys for avoiding induced radioactivity problems in thermal and FBR applications. Accordingly, for the PFBR, selection of suitable hardfacing materials for various components was preceded by detailed induced radioactivity, dose rate and shielding computations to ensure that induced radioactivity from hardfaced components is kept to the minimum for maintenance and decommissioning purposes, and also to reduce the shielding thickness required for the component-handling flask, which in turn would reduce the flask weight, size of handling crane and loads on civil structures [5].

2. SELECTION OF HARDFACING MATERIAL Selection of hardfacing material was based on, for the first time, detailed calculations of induced radioactivity and radiation shielding during maintenance, handling and decommissioning for each of the PFBR components that are to be hardfaced [5]. For these computations, replacement of Stellite 6 and Stellite 12 by same amount of Ni-base E NiCr-B (Colmonoy 5) hardfacing alloy (nominal compositions given in Table 1) was considered. Based on these calculations, for the components of PFBR, E NiCr-B (Colmonoy) was chosen as the hardfacing material to replace the traditionally used Stellites. Colmonoys have already been used in FBRs with satisfactory results. Tests on six liquid sodium pumps, with 304 SS bearings hardfaced with Colmonoy 6 and the shafts/journals

Development of Hardfacing for Fast Breeder Reactors

151

hardfaced with Colmonoy 5, operating at 748–798 K, had accumulated 20000 h for each pump without failure of the bearing area [6]. Table 1. Nominal compositions (in wt. %) of the hardfacing alloys considered Alloy Stellite 6 Stellite 12 Colmonoy 5

B – – 2.5

C 1.0 1.8 0.65

Cr 27.0 30.0 11.5

Co 60.0 52.2 < 0.25

Fe < 2.5 < 2.5 4.25

Mn 1.0 1.0 –

Ni < 2.5 < 2.5 77.10

Si 1.0 1.0 3.75

W 5.0 9.0 –

The Rapsodie, Hallam, Fermi and EBR-II reactors used Colmonoy-faced sleeves and shafts in the hydrostatic sodium-lubricated pumps [7]. Bearing operation of the Hallam, Fermi and EBR-II pumps had been satisfactory, but all operations were below 813 K. However, a seizure occurred on the Rapsodie intermediate (secondary) pump before attaining an operating temperature of 823 K. The cause of the failure is not known. Another Rapsodie primary pump seizure occurred sometime later and its probable cause was lack of wear resistance in the bearing material. The temperature of the pumps was then limited to 723 K. All previous prototype bearings were made of Stellite but for the Rapsodie pumps, a change to Colmonoy was made. The use of the proven material, Stellite, might have eliminated the seizures [7]. To alleviate the main anxiety with NiCr hardface deposits, namely reduction in its hothardness, for the first-time, the hardness of long-term aged NiCr hardface deposits was studied using the Larsen-Miller parametric approach. For this purpose, a 316 SS plate was hardfaced with E NiCr-B alloy rods of 4 mm diameter by the gas tungsten arc welding (GTAW) process, with the hardface deposit thickness being about 2 mm. Samples with a hardface deposit thickness of 1.5 mm were cut from this hardfaced plate, and were subjected to ageing at three different temperatures (823, 873 and 923 K) for five different durations (200, 500, 1000, 2000 and 5000 h) at each temperature. The Vickers hardness (HV) of the asdeposited and all the aged hardface deposits were measured at room temperature (RT = 300K) using a load of 10 kg. These hardness values were then analysed to predict the hardness of the E NiCr-B deposit after long-term ageing at the service temperatures of 673 and 773 K. The hardness of as-deposited and all aged hardface deposits, measured at RT, are presented in Fig. 1. The time–temperature correlation for these hardness values were obtained using the Larson-Miller parametric approach, given by LMP = T(C + log t), where LMP is the Larson-Miller parameter, T is the temperature in Kelvin, t is the time in hours, and C is a constant. The constant C was determined as 14.4 for E NiCr-B deposit by least-square fitting with R2 of fit being about 0.97. Using C as 14.4, the RT hardness of E NiCr-B after ageing at 823 K for the service-life of the various PFBR components was estimated. Fig. 2 shows the estimated hardness after simulated service exposure of the E NiCr-B deposit for 2, 3, 5, 10, 15, 20, 25, 30, 35 and 40 years.

152

A. K. Bhaduri and S. K. Albert

Vickers hardness (HV, 10 kgf)

800 750 700 823 K 650 600

873 K

550 923 K 500 450 400 0

1000

2000

3000

4000

5000

Ageing time (h)

Figure 1. Variation of hardness at RT of E NiCr-B deposit with duration of ageing at 823-923 K.

HV (at RT) = -0.0872 * LMP + 1947.2697 R2 = 0.9745

750

450 400 13.0

823 K 873 K 923 K Estimated at 823 K Linear (LMP, C= 14.4) 13.5

14.0

14.5

15.0

15.5

35 yr

20 yr 25 yr

16.0

40 yr

500

15 yr

550

30 yr

600

10 yr

2 yr

650

5 yr

700

3 yr

Vickers hardness (HV, 10 kgf) at RT

800

16.5

17.0

Larson-Miller parameter = T(C + log t), T in K, t in h

Figure 2. Variation of hardness at RT of E NiCr-B deposit with Larson-Miller Parameter.

To estimate the hot-hardness of E NiCr-B deposit on prolonged exposure at the different operating temperature of the various PFBR components, namely 673 and 823 K, the average hot-hardness values of unaged E NiCr-B (Colmonoy 5) and Stellite 6, as shown in Fig. 3 [8], were used. The temperature dependence of the hardness of these hardface deposits was determined by an Arrhenius-type plot of ln(hardness at RT/hardness at temperature) vs. 1/T (K–1), as given in Fig. 4. Using the relationships for both hardfacing alloys over specific temperature ranges as in Fig. 4, the hardness of E NiCr-B deposit at 673 and 823 K was estimated for prolonged exposure at 823 K, as presented in Fig. 5. The hardness values of asdeposited Stellite 6 at 300, 673 and 823 K are also presented in Fig. 5 for comparison.

Vickers hardness (HV, 10 kgf)

Development of Hardfacing for Fast Breeder Reactors

153

600 500 400 300 200

Colmonoy 5 Stellite 6

100 273

373

473

573

673

773

873

973

1073

Temperature (K) of hardness measurement

Figure 3. Variation in average hot-hardness of unaged (as-deposited) of Stellite 6 and E NiCr-B (Colmonoy 5) deposits with temperature [8]. 1000 K

0.45

667 K

500 K

400 K

333 K

Colmonoy 5 (300-589K)

ln [HV at RT / HV at T(K)]

0.40

Colmonoy 5 (589-922K) Stellite 6 (300-700K)

0.35

Stellite 6 (700-922K) 0.30 0.25 0.20 0.15 0.10 0.05 0.00 0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

1000/T(K) Figure 4. Arrhenius-type plot showing temperature dependence of hot-hardness of Stellite 6 and E NiCr-B (Colmonoy 5) deposits.

Figs. 2 and 5 show that although there is expected to be about 43% reduction in the hardness of E NiCr-B deposit after 40 years of exposure at 823 K, the hardness of E NiCr-B deposit is expected to remain sufficiently higher than the hardness of as-deposited Stellite 6. Hence, E NiCr-B deposits are expected to retain adequate hardness of about 516 HV at RT and about 430 HV at 823 K after 40 years of exposure (ageing) at 823 K, i.e. up to the end of the components’ designed service-life.

A. K. Bhaduri and S. K. Albert Estimated Vickers hardness (HV, 10 kgf)

154 800

Colmonoy 5 at 300 K Colmonoy 5 at 673 K Colmonoy 5 at 823 K Stellite 6 at 300 K Stellite 6 at 673 K Stellite 6 at 823 K

700

600

500

400

300

200 0

5

10

15

20

25

30

35

40

Time of exposure (ageing) at 823 K (years)

Figure 5. Estimated hot-hardness of E NiCr-B (Colmonoy 5) deposit after ageing at 823 K.

3. SELECTION OF HARDFACING PROCESS AND HARDFACING ALLOY TYPE NiCr hardfacing alloys, which contain high chromium and boron, form very hard chromium borides and carbides that contribute to their high hardness in addition to the solid solution strengthening by the alloying elements [9]. The abrasive resistance of the NiCr alloys is a function of amount of hard borides present in the matrix. During deposition, dilution from the substrate material occurs and this could significantly alter the microstructure and mechanical properties of the hardface deposits near the deposit/substrate interface [10]. Further, the coating thickness is optimised from the consideration that, due to differential thermal expansion of the deposit and substrate, an increase thickness would cause an increase in the residual stress and the tendency of the deposit to crack and spall under thermal cycling conditions. Also, radiation-induced damage can aggravate the integrity of the hardface coatings. Finally, when designing coatings for wear resistance, corrosion resistance and other high temperature properties, the finished coating thickness is so chosen that it is greater than the permitted wear tolerance, especially for nuclear components in which refurbishing or repair is not envisaged. While the undiluted hardface deposit provides the required wear resistance, the dilution zone at the deposit/substrate interface partially accommodates the stresses that arise during deposition or due to differential thermal expansion of the deposit and substrate during high temperature service. It is for these considerations that the best deposition process has to be adopted so that the width of the dilution zone is optimum and sufficient undiluted zone is available within the desired deposit thickness. Hardfacing with NiCr alloys by weld deposition is usually carried out using GTAW process, for which no major technological development is involved. A major problem with weld deposition by GTAW is the high dilution and tendency for cracking of the weld deposit, necessitating stress relieving at high temperature. One possible way to alleviate these problems, at least partially, is to deposit

Development of Hardfacing for Fast Breeder Reactors

155

thinner coatings using the plasma transferred-arc welding (PTAW) process. However, the problems associated with the weld deposition of the nickel-base NiCr hardfacing alloys include low fluidity, generation of residual stress in the weld deposits that can lead to cracking, hard microstructure and significant dilution of deposit by the substrate material due to the large difference in their respective melting points. Since, the cracking resistance of hardfacing alloys is very poor, preheating and controlled slow cooling often needs to be adopted to avoid cracking. Selection of hardfacing process also depends on the form of filler material available. However, non-conventional weld deposition techniques like laser welding and PTAW are found advantageous over the other processes that generally used for hardfacing. Hence, the effect of GTAW and PTAW processes on the dilution, and the effect of stress relieving (SR) heat treatment on the properties of NiCr hardfacing alloys deposited on 316L SS were studied. For this purpose, E NiCr-A (Colmonoy 6, C-6) and E NiCr-B (Colmonoy 5, C-5) rods were deposited by the GTAW process and E NiCr-A (WT-60), E NiCr-B (WT-50) and E NiCr-C (WT-40) powders were deposited by the PTAW process. Specimens for metallography, hardness measurements and SR heat treatment (at 1123 K for 4 h) were extracted from the deposits. The effect of dilution on microstructure of hardface deposits was characterised by scanning electron microscopy (SEM), energy dispersive analysis of X-rays (EDAX) and electron probe micro-analysis (EPMA). The hardness profiles across the interface of GTA deposits (Fig. 6a) show that asdeposited hardness on the top surface of the C-5 deposit is 673 HV, while that of the C-6 deposit is 803 HV. However, the hardness of the C-5 deposit over a distance of about 1.5 mm from the substrate/deposit interface is only 350–400 HV, which increases to 550–650 HV over the next 1.5 mm of the deposit. For the as-deposited C-6 deposit, the hardness is about 575 HV over a distance of about 2.5 mm from the substrate/deposit interface, about 650 HV over the next 2.5 mm and about 800 HV over the remaining thickness of the deposit. In both the C-5 and C-6 GTA deposits, SR treatment does not seem to affect their hardness. The hardness profiles, across the interface of PTA deposits (Fig. 6b) show that in the asdeposited condition, the hardness of WT-40 deposit is 251 HV98N at the substrate/deposit interface and 350–360 HV over the rest of the deposit. Similarly, the hardness of WT-50 deposit is 317 HV at the substrate/deposit interface and 445-454 HV over the rest of the deposit. The corresponding values for WT-60 deposit are 437 and 612-663 HV, respectively. It is obvious that variation in hardness with increasing distance from the interface is much less in the PTA deposits than in the GTA deposits. A marginal decrease in hardness is observed after SR heat treatment of the WT-50 and WT-60 PTA deposits. SEM images for C-6 GTA deposit with increasing distance from the interface are shown in Fig. 7. The microstructure of the deposit at 1 mm from the interface is significantly different from that near the top (8 mm from the interface). The volume fraction of blocky (dark) precipitates is very low near the interface, while both the volume fraction and the size of these precipitates increase with increasing distance from the interface. Further, near the interface, a eutectic mixture with a lamellar-like structure is present that disappears as the distance from the interface increases.

156

A. K. Bhaduri and S. K. Albert As welded (C-5) As welded (C-6) 900 800

SR (C-5) SR (C-6)

Colmonoy-6

(a) 316L SS

Hardness, HV

700 600 500 400

Colmonoy-5

300 200 100 -4

-2

0

2

4

6

8

10

12

Distance Across the interface, mm As welded (WT-60) As welded (WT-50) As welded (WT-40)

(b)

700

Hardness, VH10

600

SR (WT-60) SR (WT-50) SR (WT-40)

WT-60

316L SS WT-50

500 400 300

WT-40

200 100 -2

-1

0

1

2

3

4

Distance Across the interface, mm

Figure 6. Variation in hardness across (a) GTA and (b) PTA deposits of NiCr hardface alloys.

Figure 7. SEM micrographs of E NiCr-A (Colmonoy 6) GTA deposit at different distances from deposit/substrate interface of: (a) 0 mm (at interface); (b) 1 mm; (c) 3.5 mm; (d) 8 mm.

X-ray intensity profiles for Fe and Ni of the as-deposited C-5 and C-6 GTA deposits across the 316L SS/hardface deposit interface were obtained by EPMA. In C-5 GTA deposit (Fig. 8), the average Fe count of 119 was higher over a distance of about 1.5 mm from the substrate/deposit interface than in the rest of the deposit (46 counts), while the average Ni count of 257 was lower over a distance of about 1.5 mm from the interface than in the rest of the deposit (308 counts). In the C-6 deposit (Fig. 9), the average Fe count of 75 was higher

Development of Hardfacing for Fast Breeder Reactors

157

(a)

500 400 300 200

Base metal

X-ray Intensity Fe100mSec (for 100 msec) Intensity–for

over a distance of about 2.5 mm from the interface than in the rest of the deposit (< 50 counts), while the average Ni count of 400 was lower over a distance of about 2.5 mm from the interface than in the rest of the deposit (500 counts). Thus, the X-ray intensity profiles for Fe and Ni across the substrate/GTA deposit interface confirmed dilution from the 316L SS substrate significantly affects the chemistry of these NiCr hardface deposits to the extent of about 1.5 mm into the E NiCr-B deposit and about 2.5 mm into the E NiCr-A deposit.

Diluted layer of deposit

Colmonoy-5 deposit

100 0 –0.5 -500

00

0.5 500

1.0 1500 1.5 2000 2.0 2500 2.5 3000 3.0 3500 3.5 4000 4.0 1000

X-ray Intensity Intensity for – Ni100mSec (for 100 msec)

Distancefrom across the fusion interface (mm) Distance the interface (microns) the line(mm) 400 350

Base metal

Diluted layer of deposit

Colmonoy-5 deposit (b)

300 250 200 150 100 50 0 –0.5 -500

0 0

0.5 500

1.0 1500 1.5 2000 2.0 2500 2.5 3000 3.0 3500 3.5 4000 4.0 1000

Distance across the interface (mm) Distance from the the interface (microns) Distance across fusion line(mm) Figure 8. X-ray intensity profiles for (a) iron and (b) nickel across E NiCr-B (Colmonoy 5) GTA deposit/316L SS substrate interface.

The microstructure of WT-60 PTA deposit at the interface is different from those at different distances away from the deposit/substrate interface (Fig. 10). However, there is no significant difference in the microstructure at about 2 mm from interface and at the top of the deposit (about 3.5 mm from interface), with the microstructure consisting of dendrites, carbides, borides and eutectic carbides. With increasing distance from the interface, the volume fraction of eutectic carbides decreases. The microstructures of WT-40 (Fig. 11a) and WT-50 (Fig. 11b) PTA deposits are considerably different from that of the WT-60 deposit.

158

A. K. Bhaduri and S. K. Albert

300

Colmonoy-6 deposit

Diluted layer of deposit

250

(a)

200

Base metal

X-ray Intensity Counts– Fe for(for Fe100 msec)

The microstructure of WT-40 deposit consists of a pro-eutectic dendritic matrix with interdendritic precipitates with rod-like precipitates being practically absent. In the case of the WT-50 deposit, the volume fraction of the eutectic phase is significantly larger than in the WT-40 deposit, with precipitates having fish-bone morphology being observed. As in the WT-60 PTA deposit, in these PTA deposits also no significant variation in the microstructure is observed with increasing distance from interface. The microstructure primarily consists of hypereutectic carbides, borides and a matrix with dendritic morphology. A comparison of the microstructure of the WT-50 PTA deposit after 1123 K/4 h SR heat treatment (Fig. 11c) with that of the as-deposited WT-50 (Fig. 9b), reveals that the SR heat treatment causes significant microstructural changes in the deposit, with the dendritic structure breaking down and the fish-bone type precipitates remaining unaltered.

150 100 50 0 -7

-6

-5

-4

-3

-2

-1

0

1

600

Colmonoy-6 deposit

Diluted layer of deposit

(b)

500

Base metal

X-ray Intensity – Ni Counts for(forNi100 msec)

Distance across the interface (mm)

400 300 200 100 0 -7

-6

-5

-4

-3

-2

-1

0

1

Distance across the interface interafce (mm) Figure 9. X-ray intensity profiles for (a) iron and (b) nickel across the E NiCr-A (Colmonoy 6) GTA deposit/316L SS substrate interface.

Development of Hardfacing for Fast Breeder Reactors

159

Figure 10. Microstructure of E NiCr-A (WT-60) PTA deposit at different distances from deposit/ substrate interface of: (a) 0 mm (interface); (b) 2 mm; (c) 3.5 mm.

Figure 11. Microstructure of PTA deposits at 3.5 mm from deposit/substrate interface for: (a) asdeposited E NiCr-C (WT-40); (b) as-deposited E NiCr-B (WT-50); (c) 850°C/4 h SR heat treated E NiCr-B (WT-50).

Results from the GTA deposits clearly indicate that both the microstructure and hardness variation observed with increasing distance from the interface can be attributed to dilution of the deposit by the substrate. The step-wise increase in hardness can be attributed to the deposition of multiple layers. Dilution from the substrate is the maximum in the first layer and hence its hardness is the lowest. During the deposition of the second layer, the molten metal mixes with the re-melted diluted first layer of the deposit and hence the effect of dilution is reduced. The results of EPMA studies (Figs. 8 and 9) are in agreement with the results from microstructural examination and hardness measurements. There is almost one to one correspondence between the hardness profile and the EPMA profiles for elements Fe and Ni for both C-5 and C-6 GTA deposits. In C-6 GTA deposits; the high-Fe and low-Ni region extends over a distance of about 2.5 mm from deposit/substrate interface, indicating the extent of dilution of the C-6 GTA deposit by the substrate material. This is approximately the same distance over which the hardness was low in this deposit. Results for C-5 GTA deposits are also similar except that the distances over which these changes are observed are lower at about 1.5 mm. The reason for the differences in the width of the diluted zones for the two deposits was not clear. Being multipass welds, it can be that these distances correspond to thickness of the first layer of the deposit at the location where both hardness measurements and EPMA analysis were carried out. As already stated, it is the first layer of the deposit that is most affected by dilution from the substrate. In contrast to the results obtained for GTA deposits, the hardness and microstructural changes in the PTA deposits are confined predominantly over a short distance of about 0.5 mm near the interface. Fairly uniform microstructure and hardness beyond this distance suggests that dilution from the substrate material is significantly low in these PTA deposits.

160

A. K. Bhaduri and S. K. Albert

Considering the various design requirements such as (i) minimising residual stresses (due to differential thermal expansion) both during deposition and service, (ii) avoiding cracking, (iii) ease of deposition and (iv) post-deposition machining etc., the thickness of the deposit recommended for finished components of the PFBR is 1.5 mm. From the results discussed above, it is clear that the hardness of GTA deposits of 1.5 mm thickness would be much lower than the minimum hardness achievable in the undiluted hardface alloy deposits. Hence, the PTAW process has been selected for hardfacing of the components. Among the hardfacing alloys considered for deposition by the PTAW process, the hardness of E NiCr-C (WT-40) alloy is quite low while that of E NiCr-A (WT-60) alloy is too high. Also, the poor weldability of Ni base alloys makes it very difficult to achieve crack-free deposits using the E NiCr-A (WT-60) alloy. Hence, hardfacing alloys conforming to AWS specification E NiCr-B has been chosen for hardfacing of the PFBR components. The hardness of the E NiCr-B hardfacing alloy also meets the minimum hardness requirement (40 RC ≡ 392 HV) specified for the hardface deposits of the PFBR components. The SR heat treatment at 1123 K for 4 h is specified for many of the hardfaced components of the PFBR to ensure dimensional stability of these components during final machining and high temperature exposure during service. Since it was reported that hightemperature hardness of NiCr hardfacing alloys reduces significantly with increase in temperature above 723 K, it was required to ensure that SR heat treatment at 1123 K does not adversely affect the hardness of the hardface deposit. The hardness of GTA deposits subjected to SR heat treatment indicates that this heat treatment does not have any adverse effect on the properties of the GTA deposits (Fig. 6a). The small differences in hardness observed between the as-deposited and SR heat treated GTA deposits, is attributed to non-uniform distribution of precipitates. However, SR heat treatment of the PTA deposits seems to have some effect on its hardness and microstructure. As seen in Fig. 11(c), the dendritic microstructure of the matrix breaks down resulting in a slight reduction in hardness (Fig. 6b). However, as the hardness reduction after SR heat treatment is only marginal, it is unlikely that the performance of hardfaced components would be adversely affected.

4. HARDFACING OF TAPER ROLLER BEARINGS OF THE TRANSFER ARM The Transfer Arm of the PFBR, which is removable for maintenance, is designed for normal operation at 523 K and for exposure at 823 K during reactor operation. As part of the development of high-temperature liquid-sodium bearing for the Transfer Arm, the surfaces of the roller bearings were to be hardfaced for imparting adequate wear resistance to the contacting surfaces. Using a suitably optimised deposition procedure, 4 sets of the taper roller bearings (4 cups and 4 cones) were the first to be hardfaced with E NiCr-B alloy by the selected PTAW process by an indigenous manufacturer. The cups and cones of the taper roller bearing, made of 316LN SS, were received in the pre-machined condition, and Fig. 12 shows the drawings of these components along with the locations on the outer diameter (OD) of the cones and the inner diameter (ID) of the cups that were to be hardfaced. The E NiCr-B (WT-50) hardfacing alloy, with hardness of 52 RC, was used in the form of powders of size –150/+53. As the dimensions of cups and cones were

Development of Hardfacing for Fast Breeder Reactors

161

small, no preheating of the components was carried out prior to hardface deposition by automatic PTAW process. However, the interpass temperature was meticulously maintained during the deposition and also after the completion of deposition. The deposition in each of the components was completed in about 340 seconds using two passes with 50% of the second pass deposit overlapping the first pass deposit. As the final thickness of the hardface coating is specified as 2–3 mm, a hardface deposit of 3.0–3.5 mm was made to provide allowances for rough and final machining in each component. +0

φ111 −2

+1

65 −0

+2

+0 −2

15

φ85 −0

+0 +2

φ78 −0

+1

φ130

35 −0

φ100 −2

(a) (b)

Figure 12. Drawings of the pre-machined (a) cups and (b) cones of the taper roller bearing, with the locations to be hardfaced indicated as JJJ

After completion of deposition the components were cooled very slowly in vermiculitepowder. Then the hardface deposits on all the cups and cones were inspected by liquid penetrant test (LPT) and were found to be free of cracks. Subsequently, all the hardface deposits were rough machined and subjected to the SR heat treatment at 1123 K for 4 h. This was followed by a final round of LPT. All the hardface deposits were also inspected by ultrasonic testing, and found to be free of defects. Dimensional measurements carried out on all the hardfaced cups and cones were found to be acceptable. Fig. 13 shows the hardfaced cups and cones in the as-deposited condition.

Figure 13. Outer surface of cones (left) and inner surface of cups (right) of a taper roller bearing hardfaced with E NiCr-B alloy by the PTAW process.

5. HARDFACING OF INNER SURFACE OF GRID PLATE SLEEVES One of the critical components of the PFBR, the Grid Plate (GP) sleeves made of 316L(N) SS that holds the core subassemblies, are to be hardfaced to prevent galling, minimize wear caused by subassembly insertion and removal and erosion due to high velocity

162

A. K. Bhaduri and S. K. Albert

of liquid sodium at 673 K. The hardface deposit on the sleeves must have good thermal shock resistance for reliable operation during the 40-years design life of the reactor, during which they would be subjected to a large number of thermal cycles due to shut downs and reactor scrams. The sleeve, an internally bored tube of about 1000 mm length, are to be hardfaced at two locations where it comes in contact with the core subassembly – one on the top chamfered portion, and the other on the inner diameter at a depth of about 500 mm from either ends with the internal bore diameter at the location of hardfacing being less than 80 mm (Fig. 14).

Figure 14. Drawing of Grid Plate sleeve, showing the two hardfacing locations

When technology development of hardfacing of grid plate was taken up there was no process or equipment commercially available to carry out the job. Even attempts to fabricate the sleeve with internally hardfaced ring electron beam welded on either side could not achieve the required dimensional tolerance during welding. It was at this stage that an indigenous manufacturer designed and developed a suitable miniature PTAW torch for hardfacing of the internal surface of the sleeve. Hardfacing on the ID of the sleeve was simulated by hardfacing a 316 SS mock-up sleeve on its inside surface with E NiCr-B hardfacing alloy powder using the PTAW process (Fig. 15a). After hardfacing, the sleeve was cooled slowly in vermiculite powder, machined to the required thickness of 1.5 mm, and examined by LPT. This hardfaced mock-up hardfaced sleeve was used to study the effect of thermal cycling during service on the residual stress distribution. For this purpose, the hardfaced sleeve was cut into two halves along AB (Fig. 15a). One half (with location C at the middle) was given a SR heat treatment at 1123 K for 35 min, using a heating rate of 150 K/h and holding time of 2.5 min/mm of thickness. The other half of the sleeve (with location D at the middle) was retained in the asdeposited condition, for comparison. Both the sleeve-halves were then subjected to thermal cycling between 473 and 823 K for 20 cycles, with holding duration of 1 h at both the temperatures and immediate transfer of samples between furnaces maintained at the two temperatures. The thermal cycling temperatures of 473 and 823 K that were used correspond to the minimum and maximum temperature of liquid sodium that would be encountered. Residual stress measurements were carried out by X-ray diffraction technique on the substrate

Development of Hardfacing for Fast Breeder Reactors

163

and across the coating. For both the half-sleeves, in-plane residual stress measurements were carried out in axial direction across the coating at three locations (Fig. 15b) in the asdeposited and SR conditions, and after 5 and 20 thermal cycles.

Figure 15. (a) E NiCr-B (WT-50) hardfaced mock-up sleeve of 316 SS and (b) one half of the hardfaced sleeve, with arrow showing direction of residual stress measurements

Axial residual stresses at all three locations across the hardface deposit on both halfsleeves (Fig. 16) showed very high compressive residual stress in as-deposited condition, due to difference in the coefficient of thermal expansion (CTE) between the E NiCr-B deposit (14-15 μm/m/K) and the 316 SS substrate (17-18 μm/m/K). During post-deposition cooling, the austenitic SS substrate shrinks more due to its higher CTE resulting in tensile residual stresses in the substrate and balancing compressive residual stresses in the deposit. The residual stress at the centre of the deposit (location 2 in Fig. 15b) is higher than those at the periphery of the deposit (locations 1 and 3 in Fig. 15b), because the total restraint of the substrate and deposit is higher at the centre than at the periphery. The SR heat treatment at 1123 K significantly reduces compressive residual stresses across the hardface deposit at all locations (Fig. 16a) as tensile thermal stresses generated during SR heat treatment offsets compressive stresses present in the as-deposited condition. Thermal cycling reduces peak compressive residual stress and residual stress gradient across the deposit (Fig. 16). Local yielding due to repeated expansion and contraction during thermal cycling relaxes prior residual stresses resulting in smoothening of residual stress distribution. After thermal cycling, compressive residual stresses increase at peripheral locations in the SR deposit. Differential shrinkage between coating and substrate, which depends on the cooling rate and difference in CTE, increases the compressive residual stress. On the other hand, local yielding decreases compressive residual stresses. The combined effect of these two factors results in the observed changes in residual stress in the peripheral locations. However, these changes in residual stress distribution on thermal cycling did not have any adverse effect on the integrity of the deposit, as LPT, UT and radiography of the hardfaced sleeves showed no evidence of cracking either in the deposit or at the deposit/substrate interface. The microhardness profile across the E NiCr-B deposit/316 SS substrate interface (Fig. 17) shows an appreciable rise in hardness over a distance of 0.4 mm across the interface – from about 175 VHN in the substrate to about 475 VHN in the deposit. The short distance over which the hardness rises across the interface is indicative of the narrow dilution zone obtained in this PTA deposit. It is also observed that the hardness of the E NiCr-B PTA deposit varies from about 475 VHN near the interface to 500–530 VHN in undiluted deposit.

164

A. K. Bhaduri and S. K. Albert

Figure 16. Residual stresses across E NiCr-B deposit in (a) half-sleeve C and (b) half-sleeve D in asdeposited, SR (only in half-sleeve C) and thermal cycled conditions.

The indigenously designed and developed miniature PTAW torch was successfully demonstrated for hardfacing deep inside the inner surface of the GP sleeves (Fig. 18). To eliminate the risk of micro-cracking and delamination of the deposit, and to minimise the magnitude of residual stress, an optimised PTAW deposition procedure was qualified. By controlling deposition parameters, groove design, preheat temperature etc. it was possible to avoid any for cracking, debonding and other form of defects on the hardface deposit. Subsequently, as a part of technology development in collaboration with fabricators, a large number of these sleeves were successfully hardfaced with E NiCr-B alloy by this procedure.

Development of Hardfacing for Fast Breeder Reactors

165

550 500

400

316 SS

Delero-50

350 300 250

Fusion line

Hardness (VHN)200 gm

450

200 150 100 -4

-3

-2

-1

0

1

Distance across the fusion line (mm) Figure 17. icrohardness profile across the 316 SS substrate/E NiCr-B (Deloro 50) deposit interface in the as-deposited condition in the grid plate sleeve.

Figure 18. Hardfacing of the grid plate sleeves using specially designed miniature PTAW torch.

6. HARDFACING OF THE BOTTOM PLATE OF THE TECHNOLOGY-DEVELOPMENT GRID PLATE The Grid Plate was one of the components selected for technology development prior to construction of the PFBR. One of the important manufacturing activities was the hardfacing with E NiCr-B alloy by the PTAW process. Hardfacing on the inner surface of sleeves (discussed above) and the bottom plate were among the most difficult challenges that had to overcome during technology development of this component. For the bottom plate, a welded circular plate of diameter 6830 mm and thickness 65 mm, an annular outer ring of about 21 m circumferential length and about 40 mm width had to be hardfaced. The sheer area and quantum of deposition were challenging. The PTAW process was used for deposition of E NiCr-B (Deloro 50) powders. Although the hardfacing procedure for the bottom plate was qualified, when in collaboration with fabricators, hardfacing of the bottom plate of actual dimensions was taken up (Fig. 19) difficulties had to be overcome at various stages during the hardfacing of the bottom plate due to large volume of the hardface deposit.

166

A. K. Bhaduri and S. K. Albert

6.1. Initial Hardfacing on the Bottom Plate For the initial hardfacing of the bottom plate, a 45 mm wide annular ring of OD 6420 mm and ID 6330 mm was hardfaced. For deposition of the hardfacing alloy, a groove of depth 6 mm with groove angle of 30° from the normal was machined on the bottom plate. The entire bottom plate was preheated and maintained at 723 K prior to hardfacing in a special electrical furnace. During hardfacing, three PTA machines were used simultaneously to deposit in three equally divided sectors, and the deposition was completed as a single layer using four passes. After completion of hardfacing the entire bottom plate was cooled slowly. However, LPT of the hardface deposit revealed transverse cracks at many locations during deposition, repair and SR heat treatment.

Figure 19. Bottom plate of technology-development grid plate after hardfacing of annular ring.

6.2. Modifications in Hardfacing Procedure and Groove Design To reduce cracking susceptibility of the hardface deposit during deposition, repair and SR heat treatment, modifications were carried out in the groove design and hardfacing procedure. To confirm the adequacy of these modifications, a mock-up circular plate of diameter 980 mm and thickness 50 mm was hardfaced. As this mock-up piece was considerably smaller than the actual bottom plate, the hardfacing was carried out in 360-mm long sectors on diametrically opposite sides leaving a gap of 100 mm, which were filled after all the 360mm sectors were deposited. LPT of the hardface deposit before SR heat treatment did not reveal any cracks; however, some porosity clusters were found at locations where the deposit sectors overlapped. The mock-up piece was then subjected to SR heat treatment at 1123 K for 2 h, and subsequent LPT revealed only one crack close to a deposit-overlap location. This crack was repaired using a pre-qualified repair welding procedure using the GTAW process. After the repair, the mock-up piece was subjected to another SR heat treatment at 1123 K for 2 h. Inspection using both LPT and UT, after rough machining, did not reveal any unacceptable indications. Based on the feedback from successful hardfacing of the mock-up

Development of Hardfacing for Fast Breeder Reactors

167

piece, including demonstration of GTAW-based repair procedure, and discussions with the fabricators, additional modifications were made to the PTA machine for hardfacing of deposit-overlap regions, and SR heat treatment and preheat temperatures.

6.3. Second Hardfacing on Bottom Plate A second hardfacing on the bottom plate was carried out in another area – an annular ring just inside the initial hardfaced ring using all the modifications to the hardfacing procedure as also the experience of the successful hardfacing of the mock-up piece. After completion of hardfacing, SR heat treatment at 1023 K was immediately carried out without allowing the job to cool down to room temperature. LPT of the deposit after SR heat treatment revealed only a few cracks. These cracks were repaired using the GTAW-based repair procedure already qualified during hardfacing of the mock-up piece. After all the cracks were repaired, the bottom plate was directly heated to 1123 K for carrying out the SR treatment. LPT after the SR heat treatment showed that no cracks were present.

Figure 20. Fabrication sequence for E NiCr-B (Colmonoy 5) alloy bushes

7. FABRICATION OF HARDFACING ALLOY BUSHES Wear-resistant bushes for high-temperature application, made of hardfacing alloys are required in various components for in-sodium service in the PFBR. To substitute for very expensive import of precision castings of E NiCr-B hardfacing alloy bushes, they were fabricated using a novel procedure involving weld deposition of the hardfacing alloy on austenitic SS rods by GTAW process followed by precision machining of the hardface

168

A. K. Bhaduri and S. K. Albert

deposits (Fig. 20) [11]. Ultrasonic examination, hardness measurements, dimensional stability on high-temperature ageing, as also achieving the dimensional tolerance and surface finish on the bushes as per specification, confirmed the success of this fabrication procedure. This procedure has been successfully implemented for fabricating wear-resistant bushes for the Transfer Arm gripper assembly, and is now being transferred to industry.

8. CONCLUDING REMARKS The developments in hardfacing technology have gained from developments worldwide, and in turn have contributed significantly to these technologies. Many challenges were faced while evolving a robust hardfacing strategy for the components of PFBR. At first, based on radiation dose rate and shielding considerations during maintenance, handling and decommissioning, nickel-base E NiCr-B hardfacing alloy was chosen to replace the traditionally used cobalt-base Stellite alloys. Also, it was demonstrated that the hot-hardness of E NiCr-B deposits after exposure at service temperatures would retain adequate hardness at end of the components’ designed service-life of up to 40 years. Further, based on detailed metallurgical studies, including residual stress measurements after thermal cycling, the more versatile PTAW process was chosen for hardfacing, so that the width of the dilution zone could be minimised. Hardfacing with E NiCr-B alloy by the selected PTAW process was first successfully implemented on the taper roller bearings of the Transfer Arm. Hardfacing deep inside the inner surface of the sleeves and on the bottom plate of the Grid Plate were among the most difficult challenges that were overcome during technology development, involving hardfacing inside the sleeves using an indigenous miniature PTAW torch and hardfacing of an annular ring of about 21 m circumferential length on the bottom plate. A novel procedure, involving hardfacing alloy deposition followed by precision machining, was also developed for fabrication of high-temperature wear-resistant hardfacing alloy bushes. Thus, adaptation of the hardfacing technology for PFBR, through collaborative effort with industries, has to use of semi-automatic PTAW process that has now been qualified and demonstrated for hardfacing of various technology-development components of the PFBR.

REFERENCES [1]

[2] [3] [4] [5] [6]

E.Yoshida, Y.Hirakawa, S.Kano and I.Nihei, Proceedings of International Conference on Liquid Metal Technology, Societé Francaise d’ Energie Atomique, Paris (1988) 5021. R.A.Douty and H.Schwartzbart, Welding Journal 51 (1972) 406s. E.Lemaire and M.Le Calvar, Wear 249 (2001) 338. S.K.Albert, I.Gowrisankar, V.Seetharaman and S.Venkadesan, Proceeding of National Welding Seminar, Indian Institute of Welding, Bangalore (1987) A1. A.K.Bhaduri, R.Indira, S.K.Albert, B.P.S.Rao, S.C.Jain and S.Asokkumar, Journal of Nuclear Materials 334 (2004) 109. N.J.Allnatt and G.R.Bell, Proceedings of International Colloquium on Hardfacing Materials in Nuclear Power Plants, Avignon (1980).

Development of Hardfacing for Fast Breeder Reactors [7]

169

“Large Sodium Pump Design Study”, Report no. WARD-3762-1, Westinghouse Advanced Research Division, USA. [8] Deloro Stellite Limited, “Tribaloy” Product Catalogue, Deloro Stellite Limited, Swindon, UK. [9] ASM Metals Handbook, Volume 6, 9th edition, ASM International, Materials Park, Ohio, USA (1993) 794. [10] C.R.Das, S.K.Albert, A.K.Bhaduri, C.Sudha and A.L.E.Terrance, Surface Engineering 21 (2005) 290. [11] C.R.Das, S.K.Albert, A.K.Bhaduri and G.Kempulraj, Journal of Materials Processing Technology 141 (2003) 60.

In: Materials Science Research Horizons Editor: Hans P. Glick pp. 171-192

ISBN 978-1-60021-481-3 © 2007 Nova Science Publishers, Inc.

Chapter 6

TISSUE ENGINEERING OF CARTILAGE IN BIOREACTORS Nastaran Mahmoudifar∗ and Pauline M. Doran School of Biotechnology and Biomolecular Sciences University of New South Wales Sydney NSW 2052, Australia

ABSTRACT The main goal of cartilage tissue engineering is to generate three-dimensional cartilage and osteochondral tissues for use in repair of large cartilage injuries. Cartilage constructs are generated by seeding and culturing viable cells in biodegradable polymer scaffolds under conditions suitable for tissue formation. In this chapter, current developments in cartilage tissue engineering are reviewed, focusing on the source of cells, the polymer scaffolds, seeding systems, bioreactors and application of mechanical stimulation for cell differentiation and tissue production. The generation of cartilage tissue constructs in the laboratory using a bioreactor system is also described. Chondrocytes were isolated from human foetal epiphyseal cartilage, expanded in monolayer, dynamically seeded into poly(glycolic acid) (PGA) polymer scaffolds and cultured in recirculation bioreactors. Composite scaffolds were used to improve the initial distribution of cells within the scaffolds and to develop cartilage constructs that were homogeneously cartilaginous throughout their thickness. The quality of the engineered cartilage was assessed after 5 weeks of bioreactor culture in terms of tissue wet weight, cell, glycosaminoglycan (GAG), total collagen and collagen type II contents, histological analysis of cell, GAG and collagen distributions, immunohistochemical analysis of collagen types I and II, and ultrastructural analysis using transmission electron microscopy.

∗

Correspondence to: Nastaran Mahmoudifar, telephone: +61-2-9385-2086; fax: +61-2-9313-6710, e-mail: [email protected]

172

Nastaran Mahmoudifar and Pauline M. Doran

INTRODUCTION Functional cartilage and osteochondral tissues are needed for implantation to repair large or full-thickness cartilage injuries. Cartilage in adults has a very limited capacity for self repair once it is damaged due to injury or disease; conventionally, autografts or allografts are implanted to repair the damage. However, the limited availability of autografts and the problems of immunorejection and transmission of infectious disease in the case of allografts have made tissue engineering of cartilage a promising alternative. The success of repairing small cartilage injuries by injecting autologous cartilage cells (chondrocytes) into the damaged site has been encouraging for tissue engineering, which aims to generate threedimensional cartilage and osteochondral biomaterials by seeding and culturing viable cells in biodegradable polymer scaffolds. The role of the polymer is to provide an initial scaffold for cell attachment and production of cartilage extracellular matrix (ECM). The polymer gradually dissolves and disappears as the tissue is formed. The main goal for engineered cartilage tissues is the repair of articular cartilage; however, other applications include plastic and reconstructive surgery of ears and noses.

CELL SOURCE Differentiated chondrocytes or undifferentiated stem cells may be used to generate tissueengineered cartilage. Consistent with the clinical practice of injecting autologous chondrocytes into damaged joints to treat small articular cartilage injuries (Brittberg et al., 1994), chondrocytes isolated from native cartilage have been applied in most tissue engineering studies. As indicated in Table 1, chondrocytes from a variety of animal sources have been tested experimentally; human cells have also been used for cartilage generation. Most researchers isolate chondrocytes from foetal or juvenile individuals for cartilage production in vitro. Although better results in terms of cartilage ECM development have been reported using immature rather than adult chondrocytes (Carver and Heath, 1999b), the presence of undesirably high levels of collagen type I in engineered cartilage has been attributed to the use of foetal cells and the developmental plasticity of foetal chondrocytes in the production of both bone and cartilage tissues (Mahmoudifar and Doran, 2005a). The multipotency of adult stem cells is being exploited increasingly to produce tissueengineered cartilage. An important advantage of using stem cells rather than autologous chondrocytes for cartilage engineering is that removal of healthy cartilage from the patient is not required, thus eliminating the risk of morbidity at the donor site. Mesenchymal stem cells are present in many tissues including synovium, muscle, adipose, bone marrow and bone (Jorgensen et al., 2004; Tuli et al., 2003) and have the capacity to differentiate along multiple lineages to form chondrocytes, osteoblasts or adipocytes under the direction of appropriate differentiation factors (Awad et al., 2003; Johnstone et al., 1998; Pittenger et al., 1999; Winter et al., 2003; Zuk et al., 2001). Typically, chondrogenesis is induced using high-density cell culture in a three-dimensional environment and supplementation of the medium with growth factors from the transforming growth factor beta (TGF-β) family; insulin and dexamethasone may also be added. The application of bone-marrow-derived, trabecular-bone-derived and adipose-derived stem cells for tissue engineering of cartilage is described in the recent

Tissue Engineering of Cartilage in Bioreactors

173

literature (Awad et al., 2004; Caterson et al., 2001; Li et al., 2005; Martin et al., 1998, 2001; Meinel et al., 2004; Tuli et al., 2004). Chondrocyte-specific gene expression and the synthesis of cartilage ECM components such as GAG and collagen type II are used to monitor the differentiation of stem cells into chondrocytes. Although mesenchymal stem cells develop chondrogenic properties when cultured as cell aggregates or pellets to promote cell–cell interactions, better results in terms of tissue weight and the production of GAG and collagen have been obtained after seeding the cells into biodegradable polymer scaffolds (Li et al., 2005; Martin et al., 1998). Table 1. Examples of sources of chondrocytes in studies of three-dimensional cartilage tissue engineering Animal source Cow

Dog Horse

Cartilage location

Age

Reference

Femoropatellar grooves (articular) and anterior ribs (costal) Femoropatellar grooves (articular)

1–2 weeks

Freed et al., 1993

2–3 weeks

6 months Not reported Calf 6–9 months Adult 1 week 1 month or less 24 months or less Adult 11 years Not reported 27 years (mean age) Foetal (17–20 weeks gestation) 2–8 months

Freed and Vunjak-Novakovic, 1995; Vunjak-Novakovic et al., 1999; Seidel et al., 2004 Wendt et al., 2003 Grande et al., 1997 Pazzano et al., 2000 Waldman et al., 2004 Nehrer et al., 1997 Carver and Heath, 1999a Carver and Heath, 1999c Carver and Heath, 1999b Carver and Heath, 1999b Freed et al., 1993 Aigner et al., 1998 Grigolo et al., 2002 Mahmoudifar and Doran, 2005a, 2005b, 2006 Freed et al., 1994a

4–8 months 7 days Up to 3 weeks

Dunkelman et al., 1995 Chen et al., 2004 Davisson et al., 2002

Ankle (articular) Glenohumeral joints (articular) Metacarpal–carpal (articular) Knee (articular) Stifle joints (articular)

Human

Ribs (costal) Nose (nasoseptal) Knee (articular) Knee and hip (articular)

Rabbit

Femur, tibia, patella, glenoid, humeral head (articular) Not reported (articular) Not reported (articular) Patellofemoral grooves (articular)

Rat Sheep

POLYMER SCAFFOLDS The function of the polymer scaffold in cartilage tissue engineering is to provide the initial support for cell growth and formation of tissue. Polymer scaffolds should be biodegradable with a degradation rate matching the rate of cartilage tissue development, and should also be highly porous for the cells to have room to grow and for permeation of nutrients and metabolites. Polymer scaffolds must have appropriate surface properties for chondrocyte attachment and growth, must be biocompatible with the tissue, and must not form toxic degradation products (Ma and Langer, 1999).

174

Nastaran Mahmoudifar and Pauline M. Doran

Polymer scaffolds used for tissue engineering are divided into two groups: synthetically made and naturally derived. Synthetic polymer scaffolds made of PGA, poly(L-lactic acid) (PLLA) and their copolymer poly(lactic-co-glycolic acid) (PLGA) are used extensively for tissue engineering. These polymers are bioresorbable materials and have been approved for clinical and surgical use. They are degraded by hydrolysis to glycolic acid and lactic acid, which are converted to CO2 and eliminated from the body via the respiratory route (Gilding, 1981; Hatton et al., 1994; Kim and Mooney, 1998). PGA degrades faster than PLLA and has been reported to support higher cell growth rates, cell densities and GAG formation than PLLA in 6–8-week in vitro cartilage engineering studies (Freed et al., 1993). A comparison of non-woven PLLA and PLGA (PGA:PLLA ratio of 90:10) polymers seeded with human septal chondrocytes in an in vivo study of 24 weeks showed that the PLGA constructs had collagen type II contents comparable to native cartilage whereas the PLLA constructs contained mainly collagen type I (Rotter et al., 1998). In other work, PGA films were shown to have better surface adhesion properties for chondrocytes isolated from human articular cartilage than PLLA films (Ishaug-Riley et al., 1999). Non-woven PGA and woven PLGA meshes supported higher proteoglycan synthesis than type I collagen and nylon scaffolds, while collagen synthesis was higher in collagen scaffolds although collagen typing was not performed (Grande et al., 1997). Overall, PGA has been the most extensively used polymer scaffold for in vitro production of cartilage. Naturally-derived polymer scaffolds include collagen and hyaluronan. Type I and II collagen sponges cross-linked with and without chondroitin sulphate have been used for tissue engineering of cartilage. Type II collagen scaffolds encouraged cell differentiation and the formation of hyaline-appearing cartilage containing a higher GAG content per cell than in tissues produced using collagen type I scaffolds, which promoted cell proliferation and higher DNA content (Nehrer et al., 1997). However, in contrast, the results of Pieper et al. (2002) indicated that collagen type I and type II matrices performed similarly in terms of GAG and DNA contents and collagen type II expression. In addition, cross-linking of chondroitin sulphate with type II collagen did not have a major effect on the construct biochemical composition. The opposite was reported for collagen type I matrix, where attachment of chondroitin sulphate resulted in higher GAG and DNA contents in cartilage constructs (van Susante et al., 2001). Non-woven mesh of hyaluronan benzyl-ester (Hyaff®-11) has been used as a scaffold for tissue engineering of cartilage and was reported to encourage the differentiation of chondrocytes (Aigner et al., 1998; Grigolo et al., 2002). Disadvantages associated with the use of natural polymers include the potential risk of pathogen transmission from the animal source (Cancedda et al., 2003), and concerns about the availability and quality of the materials (Kim and Han, 2000).

SEEDING POLYMER SCAFFOLDS WITH CELLS Polymer scaffolds are seeded with cells using static or dynamic methods. In the static method, typically, a small volume of highly concentrated cell suspension containing the desired number of cells is loaded into the polymer scaffold using a pipette or syringe (Freed et al., 1993; Schreiber et al., 1999). Static seeding is normally performed in tissue-culturetreated dishes. Dynamic seeding has been carried out in tissue culture dishes (Freed et al.,

Tissue Engineering of Cartilage in Bioreactors

175

1994b; Vunjak-Novakovic et al., 1996) or tubes (Kim et al., 1998) mixed on orbital shakers, in spinner flasks mixed using magnetic stirrers (Vunjak-Novakovic et al., 1996), in rotating bioreactors (Freed and Vunjak-Novakovic, 1995), and in perfusion bioreactor systems (Wendt et al., 2003). For dynamic seeding, the polymer scaffold may be stationary, as in spinner flasks and culture dishes, while the cells are suspended in the culture medium by mixing. In rotating bioreactors, both the cells and scaffolds are suspended in the culture medium by rotating the bioreactor vessel around its central axis; in tubes, the polymer scaffolds and cells are kept in suspension by shaking. In perfusion chambers or bioreactors, the polymer scaffolds are stationary while the culture medium containing the cells is forced to flow through the scaffold pores. In general, compared with static seeding, seeding under mixed conditions results in a more uniform cell distribution throughout the polymer scaffold and the generation of tissueengineered cartilage of enhanced quality (Freed et al., 1994b).

Seeding in Mixed Spinner Flasks Spinner flasks have been used extensively for seeding polymer scaffolds. Typically, the scaffolds are threaded onto syringe needles hung from the mouth of spinner flasks containing medium and cells, and the cell suspension is mixed using a magnetic stirrer. The efficiency and quality of seeding have been shown to be superior in spinner flasks compared with mixed culture dishes (Vunjak-Novakovic et al., 1996). After 3 days of seeding, polymer scaffolds seeded with chondrocytes in spinner flasks had significantly higher cell contents compared with those seeded in mixed culture dishes. In addition, the distribution of the attached cells was more uniform in spinner flasks; this was attributed to turbulent mixing in the flasks compared with orbital fluid motion in the mixed culture dishes. Mixing in spinner flasks was found to promote the formation of cell aggregates, which varied in size from 20 μm (4 chondrocytes) to 32 μm (16 chondrocytes) at, respectively, initial cell concentrations of 0.75 × 105 and 7 × 105 cells per mL (Vunjak-Novakovic et al., 1998). As cell aggregates attached to the PGA fibres in the scaffold faster than single cells, mixing effectively enhanced the kinetics of cell attachment in spinner flasks. The distribution of cells was uniform throughout the scaffold cross-section except for a 50-μm-thick surface layer in which the cell density was 60–70% higher than in the bulk of the scaffold. The rate of seeding increased with increasing initial concentration of cells in the suspension: this was attributed to cell aggregation. In other work, the distribution of chondrocytes seeded into 4.75-mm-thick PGA scaffolds in mixed flasks was heterogeneous, with significantly higher cell density in the upper half of the scaffold than in the lower half (Mahmoudifar and Doran, 2005a). These results highlight the difficulties that can be encountered in achieving a completely uniform cell distribution in thick polymer scaffolds.

Seeding in Rotating Bioreactors Rotating bioreactors, including a high-aspect-ratio vessel (HARV) and a slow-turning lateral vessel (STLV), were developed by the National Aeronautics and Space Administration

176

Nastaran Mahmoudifar and Pauline M. Doran

(NASA) Johnson Space Centre. Rotating bioreactors have been used to seed 2-mm-thick PGA polymer pieces with chondrocytes (Freed and Vunjak-Novakovic, 1995, 1997). A spatially uniform cell distribution was produced in scaffolds seeded in rotating bioreactors (Freed and Vunjak-Novakovic, 1995). However, the use of rotating bioreactors for seeding polymer discs larger than 5-mm-diameter × 2-mm-thick has not been reported extensively and there is a possibility that uniform seeding throughout the depth of the scaffold may not be achievable at higher scaffold thicknesses due to diffusion limitations.

Seeding in Perfusion Bioreactors Perfusion bioreactors of various design have been used to seed cells into polymer scaffolds. An oscillating perfusion bioreactor has been described for seeding of chondrocytes into Polyactive foams and Hyaff®-11 non-woven meshes (Wendt et al., 2003). Two pieces of polymer 4.0–4.3 mm thick were fixed in place while a concentrated cell suspension was made to oscillate in a U-shaped glass tube by the action of a vacuum pump, thus forcing the suspension alternately into and out of the scaffolds. The seeding efficiency, defined as the percentage of the cells added initially to the bioreactor that attach to the scaffolds, was significantly higher than that obtained in spinner flasks. The most important advantage of seeding in perfusion bioreactors was the enhanced uniformity of cell distribution throughout the depth and along the radius of the scaffolds compared with the results obtained using spinner flasks and static seeding (Wendt et al., 2003).

BIOREACTORS Bioreactors provide a dynamic culture environment for generation of tissue-engineered cartilage. Bioreactor culture conditions such as mixing and fluid flow affect the transfer of nutrients, removal of wastes and gas exchange to cells within the developing tissue, thus potentially enhancing the quality of the construct compared with tissues generated in static systems. Bioreactor culture conditions can also influence cell function and regulate chondrogenesis (Vunjak-Novakovic et al., 2002). Both the biochemical composition as well as the mechanical properties of engineered cartilage have been shown to be modulated by the conditions and duration of bioreactor cultures (Mahmoudifar and Doran, 2006; Martin et al., 2000; Vunjak-Novakovic et al., 1999, 2002). The most widely used bioreactors for tissue engineering of cartilage are spinner flasks, rotating vessels and perfusion systems.

Spinner Flasks Spinner flasks used for dynamic seeding of cells into polymer scaffolds have also been employed for cultivation of cartilage cell–polymer constructs. Tissue constructs were exposed to turbulent flow in spinner flasks stirred at 50 rpm (Vunjak-Novakovic et al., 1996), which improved mass transfer. As a result, tissues cultivated for 8 weeks maintained their original scaffold thickness of 5 mm and contained up to 70% more cells, 60% more GAG and 125%

Tissue Engineering of Cartilage in Bioreactors

177

more collagen compared with those cultivated under static conditions. However, the turbulent flow conditions in the flasks induced the formation of an outer fibrous capsule of thickness approximately 300 μm around the tissue constructs. This capsule contained multiple layers of elongated cells and collagen but little GAG (Vunjak-Novakovic et al., 1996); the collagen in the outer capsule was mainly collagen type I (Freed et al., 1994b).

Rotating Bioreactors Rotating bioreactors operate under conditions which simulate microgravity. The STLV configuration, which consists of two horizontal concentric cylinders rotated around their central axis, has been applied for tissue engineering of cartilage (Freed et al., 1998; Freed and Vunjak-Novakovic, 1995, 1997; Vunjak-Novakovic et al., 1999). Tissue constructs cultivated in the annular space between the cylinders were maintained in a state of continuous free-fall by adjusting the rotational speed to between 15 and 28 rpm. Gas exchange was provided by pumping filter-sterilised incubator gas (5% CO2 in air) through the inner concentric cylinder, which consisted of a silicone rubber membrane. Rotating bioreactors provide a low-shear (0.15 Pa) laminar flow field in which fluid mixing is generated by the settling tissue constructs (Freed and Vunjak-Novakovic, 1995). Cartilage constructs cultivated for 1–5 weeks in rotating bioreactors had higher GAG content and thinner outer fibrous capsules compared with those cultured in spinner flasks (Freed and Vunjak-Novakovic, 1995, 1997). The mechanical properties of the constructs cultivated in rotating bioreactors were also better than those from spinner flasks (Vunjak-Novakovic et al., 1999). However, cultivation of cellseeded polymer discs larger than 5-mm-diameter × 2-mm-thick has not been reported extensively in rotating bioreactors.

Perfusion (Flow-Through) Bioreactors Perfusion bioreactors operate on a similar basis as immobilised-cell or packed-bed bioreactors. Polymer scaffolds seeded with cells are placed in the bioreactors (typically one scaffold per bioreactor) and culture medium is forced to flow through the scaffold. A pump is usually used for continuous medium perfusion. The culture medium transports nutrients and oxygen to cells attached to the polymer and removes waste. An essential requirement of a perfusion bioreactor is a tight fit between the scaffold and the walls of the vessel to ensure that medium flows through the scaffold rather than bypassing around the edges. The flow of culture medium within the scaffold creates shear forces which provide mechanical stimulation to the cells and may influence the quality of the developing ECM (Darling and Athanasiou, 2003). The level of shear stress exerted on the cells can be changed by varying the medium flow rate (Bancroft et al., 2003). Perfusion bioreactors of various design operated at a range of medium flow rates have been used by different research groups (Davisson et al., 2002; Dunkelman et al., 1995; Mahmoudifar and Doran 2006; Pazzano et al., 2000). The flow rate is an important operating parameter and should be selected carefully to provide adequate oxygen to the cells, considering the oxygen uptake rate of the specific cell type seeded into the scaffold. On the other hand, high medium flow rates can dislodge cells from the scaffold and should be avoided, especially during the

178

Nastaran Mahmoudifar and Pauline M. Doran

early stages of culture when the ECM is not very well formed. Alternatively, rather than increasing the flow rate, the concentration of oxygen in the culture medium may be increased to provide the cells with adequate oxygen. Perfusion bioreactors made of polycarbonate have been used for three-dimensional cultivation of chondrocytes (Davisson et al., 2002; Dunkelman et al., 1995). Each bioreactor system consisted of five polycarbonate chambers which were connected to a medium bag and operated in parallel. A multi-channel peristaltic pump was used to recycle culture medium between the bag and the bioreactors. No medium change was performed during the bioreactor cultivation period in this work. Dunkelman et al. (1995) reported the formation of cartilage constructs containing 25% GAG and 15% collagen on a dry weight basis after cultivation of rabbit chondrocytes for 4 weeks at a flow rate of 50 μL min-1 (superficial velocity of 10.6 μm s-1). However, the ECM was not homogenous with more residual PGA fibres, fewer cells and less ECM in the middle of the constructs than at the periphery, suggesting that medium flow was not uniform throughout the construct. Davisson et al. (2002) reported that perfusion significantly increased the DNA content of constructs, while GAG synthesis and deposition depended on the velocity of medium flow and the duration of perfusion. Nine days of continuous perfusion increased GAG synthesis and deposition compared with static controls when constructs were cultured at a medium superficial velocity of 11 μm s-1 for the first 7 days and then at a superficial velocity of 170 μm s-1 for the following 2 days. The use of perfusion during the initial 3-day culture period suppressed the synthesis and retention of GAG compared with static control cultures. Pazzano et al. (2000) cultivated bovine chondrocytes seeded into PGA scaffolds coated with PLLA in a perfusion bioreactor for 4 weeks. Ten seeded polymer discs were placed in the bioreactor and culture medium from a reservoir was pumped continuously through the discs at a superficial velocity of 1 μm s-1. The constructs from the perfusion bioreactor had significantly higher DNA, GAG and hydroxyproline contents compared with staticallycultured controls. In addition, chondrocytes in the constructs from the bioreactor were aligned in columns in the direction of medium flow. We have previously generated cartilage and composite osteochondral tissues in a recirculation bioreactor (Mahmoudifar and Doran, 2005a, 2005b). Each bioreactor contained one scaffold and medium was pumped through the constructs at a superficial velocity of 19 μm s-1 (0.2 mL min-1). The average hydrodynamic shear stress experienced by the cells in freshly seeded scaffolds was estimated as 1.7 × 10-3 Pa.

Use of Mechanical Stimulation in Bioreactors The tissue-engineered cartilage constructs generated to date contain lower concentrations of collagen and are inferior in functional properties compared with native articular cartilage. As the mechanical properties of engineered cartilage correlate with the concentrations of collagen and GAG present in the tissues (Vunjak-Novakovic et al., 2002), improving the biochemical composition results in better mechanical properties. One approach to improve the quality of engineered cartilage is to employ physical factors/mechanical stimulation to enhance the development of cartilage ECM (Chen et al., 2004; Seidel et al., 2004; Waldman et al., 2004; Wimmer et al., 2004). It is well known that chondrocytes perceive and respond to

Tissue Engineering of Cartilage in Bioreactors

179

physical treatments such as joint loading during normal daily activities, e.g. walking and running (Guilak et al., 1997). In vitro studies have shown that mechanical stimuli including cyclic compression (Hunter et al., 2002; Lee and Bader, 1997; Lee et al., 1998; Sah et al., 1989; Stoltz et al., 2000), cyclic hydrostatic pressure (Parkkinen et al., 1993) and cyclic tensile load (stretch) (De Witt et al., 1984; Honda et al., 2000) regulate gene expression and biosynthetic activity in chondrocytes in monolayer cultures, in agarose and collagen gels, and in cartilage explants. Chondrogenesis of mesenchymal progenitor cells has also been shown to respond to mechanical factors (Angele et al., 2004). A number of bioreactor/culture systems has been developed to incorporate mechanical stimulation such as semi-continuous hydrostatic pressure (Carver and Heath, 1999a, 1999b, 1999c) and dynamic compression (Seidel et al., 2004; Waldman et al., 2004; Wimmer et al., 2004). Carver and Heath (1999a, 1999b, 1999c) designed a perfusion chamber system that was intermittently pressurised at 500 and 1000 psi to enhance the formation of cartilage ECM. Culture medium was pumped through the chambers either continuously at a flow rate of about 3 mL min-1 when the chambers were not pressurised, or at a flow rate of 6.7 mL min-1 during a period of 3 min prior to each pressurisation. Culture of chondrocytes with perfusion and intermittent pressure accelerated cartilage matrix formation; enhancement of collagen synthesis required treatment at higher pressures than those found to increase GAG production. Waldman et al. (2004) subjected chondrocytes on the surface of porous calcium polyphosphate to intermittent compression at a compressive amplitude of 5% and frequency of 1 Hz for 400 cycles every second day for a period of 4 weeks. Compared with unstimulated chondrocyte cultures, this treatment resulted in a significant increase in tissue dry weight, 30 and 40% increases in proteoglycan and collagen contents, respectively, and a 2–3-fold increase in compressive mechanical properties. In contrast, a shorter culture duration of 1 week under the same compression conditions resulted in a significant increase only in collagen synthesis. Seidel et al. (2004) reported that dynamic compression of cartilaginous tissue developed by culturing chondrocytes in PGA scaffolds at a dynamic strain amplitude of 5% superimposed onto a static offset of 2% and frequency of 0.3 Hz for 1 h per day for a period of 37 days had little effect on the composition, morphology and mechanical properties of the construct interior, while the outer rings obtained after coring the interior from the tissue discs retained larger amounts of GAG. These studies indicate the importance of the compression conditions applied in determining tissue quality. Lower compressive amplitudes (5%) seem to favour collagen synthesis, whereas higher amplitudes (10–20%) appear to favour proteoglycan synthesis (Waldman et al., 2004). The frequency of compression also plays an important role: a frequency of 1 Hz has been shown to stimulate GAG synthesis in chondrocytes embedded in agarose gels (Lee and Bader, 1997).

BIOREACTOR CULTURE OF CARTILAGE: EXPERIMENTAL STUDY Fibrous PGA mesh was purchased from Albany International Research (Mansfield, MA, USA) as sheets of thickness 4.75 mm (thick) and 2.15 mm (thin). The mesh was cut into 15mm-diameter discs, dynamically seeded in mixed flasks using 2.2 × 107 chondrocytes per scaffold, and cultured in recirculation bioreactors as described previously (Mahmoudifar and

180

Nastaran Mahmoudifar and Pauline M. Doran

Doran, 2005a). For seeding, each PGA disc was threaded onto a 10–11-cm-long 23-gauge stainless steel syringe needle in a 250-mL side-arm flask and exposed to medium containing cells for 3 days, with mixing provided by a magnetic stirrer. Scanning electron micrographs of the PGA fibres in the scaffolds before and after seeding are shown in Figure 1. Seeding produced a non-homogeneous distribution of cells in the scaffolds: the cell density was substantially greater near the top surface of the PGA discs than throughout the remainder of the scaffold (Mahmoudifar and Doran, 2005a). To study the influence of initial cell distribution on the quality of the cartilage tissues generated, single thick scaffolds, composite thin scaffolds consisting of two thin discs sutured together after seeding, and composite thin

Figure 1. Scanning electron micrographs of 2.15-mm-thick PGA scaffolds: (a) before seeding, showing PGA fibres of diameter 12–14 µm; (b) after seeding with chondrocytes for 3 days, showing cells attached to the PGA fibres and almost completely covering the polymer surface.

Figure 2. (a) Orientation of seeded discs in single and composite scaffolds. After seeding, each disc contained a non-homogeneous distribution of cells with higher cell density near the top surface of the scaffold, as represented by the shaded regions. To form composite scaffolds, two seeded discs were sutured together with their top surfaces from seeding facing each other internally, as illustrated. (b) The culture chamber of a recirculation bioreactor used for cartilage production. The experiments were carried out using triplicate bioreactors, each connected to a separate medium reservoir. Medium was recirculated through the scaffolds using a peristaltic pump. In some experiments, the direction of medium flow through the constructs was periodically (every three days) reversed.

Tissue Engineering of Cartilage in Bioreactors

181

and thick scaffolds consisting of a thin and a thick disc sutured together after seeding, were used in this work. As illustrated in Figure 2a, for the composite scaffolds, the two discs were sutured together with their top surfaces from seeding facing each other internally. All scaffolds were cultured in recirculation bioreactors for 5 weeks; a bioreactor culture chamber is shown schematically in Figure 2b indicating the location of the seeded scaffold. Single thick scaffolds were oriented in the bioreactors so that medium entered the top surface from seeding; composite thin and thick scaffolds were oriented so that medium entered the thin disc. To improve the quality of the generated cartilage tissues, in some experiments, the direction of medium flow through the constructs was periodically (every three days) reversed during bioreactor culture. The biochemical composition, histology and immunohistochemistry of the constructs were examined after harvest of tissues from the bioreactors; the ultrastructure of the generated cartilage was also examined and compared with native cartilage tissues.

Results Biochemical Composition of Tissue-Engineered Cartilage There was no significant difference in the weight of the tissues or the concentration of cells found in the constructs at harvest between the single thick and composite PGA scaffolds (Fig. 3a, 3b). The constructs generated from composite thin scaffolds contained a 2.8-fold higher (p < 0.05) concentration of GAG than those produced from composite thin and thick scaffolds; there was no significant difference (p < 0.05) in GAG concentration using the single thick and composite thin scaffolds (Fig. 3c). The concentration of total collagen was 1.6- and 2.3-fold higher (p < 0.01) in the constructs generated from single thick scaffolds than in those produced from composite thin and composite thin and thick scaffolds, respectively (Fig. 3d). Whereas the concentration of collagen type II in the constructs generated from single thick scaffolds was similar to that obtained using composite thin scaffolds, the result for single thick scaffolds was 5.7-fold higher (p < 0.01) than in tissues generated using composite thin and thick scaffolds (Fig. 3e). There was no significant difference (p < 0.05) in collagen type II as a percentage of total collagen between the constructs generated using single thick and composite thin scaffolds; however, the constructs produced using composite thin scaffolds contained 4.1-fold higher (p < 0.01) levels of collagen type II as a percentage of total collagen than those from composite thin and thick scaffolds (Fig. 3f). Operating the bioreactors with periodic medium flow reversal was beneficial for constructs generated from composite thin scaffolds compared with those produced using single thick PGA discs (Fig. 4). The weight of the tissue constructs and the concentrations of GAG, total collagen and collagen type II were 1.5-, 2.8-, 1.5- and 2-fold higher (p < 0.05 or p < 0.01), respectively, in the constructs generated from composite thin scaffolds relative to those produced from single thick scaffolds (Fig. 4a, 4c, 4d, 4e).

182

Nastaran Mahmoudifar and Pauline M. Doran

Figure 3. Effect of using single thick and composite PGA scaffolds on the properties of engineered cartilage. (a) Tissue wet weight, (b) cell concentration, (c) GAG concentration, (d) total collagen concentration, (e) collagen type II concentration, (f) collagen type II as a percentage of total collagen. The constructs were cultured in bioreactors for 5 weeks. Significant differences are indicated * (p < 0.05) and ** (p < 0.01). The error bars represent standard errors from triplicate bioreactor cultures.

Histology The histological appearance of cross-sections of the tissue constructs is shown in Figure 5. Collagen (blue–green) staining is visible except in areas where it is masked by safranin-O (orange–red) staining of GAG. Undegraded PGA fibres present in the samples appear as red flecks. Tissue constructs produced using composite scaffolds were thicker than those generated using single thick scaffolds. All tissues had an outer capsule which contained elongated cells and stained for collagen but not for GAG. The distribution of GAG in the cross-sections of tissue constructs prepared using single thick (Fig. 5a–5c) and composite thin and thick (Fig. 5d–5f) scaffolds with uni-directional medium flow was non-homogeneous; more intense staining was observed within the top half of the cross-sections than in the

Tissue Engineering of Cartilage in Bioreactors

183

bottom half. Periodically reversing the direction of medium flow resulted in a more homogeneous distribution of GAG throughout the cross-sections of both the tissue constructs produced from single thick scaffolds (Fig. 5g) and those produced using composite thin scaffolds (Fig. 5h).

1000

Total collagen concentration (μg mg-1 dry weight)

800

Tissue wet weight (mg)

120

(a)

600 400 200

(d)

100 80 60 40 20 0 60

(b)

8

Collagen type II concentration (μg mg-1 dry weight)

Cell number concentration (mg-1 dry weight x 10-5)

10

6 4 2 0

30 20 10

80

Collagen type II as a percentage of total collagen

(c)

200

GAG concentration (μg mg-1 dry weight)

40

0

250

150 100 50 0

(e)

50

(f)

60

e)

40

20

0 Single thick

Composte thin

Scaffold type

Single thick

Composte thin

Scaffold type

Figure 4. Effect of using periodic medium flow reversal on the properties of engineered cartilage generated using single thick and composite thin PGA scaffolds. (a) Tissue wet weight, (b) cell concentration, (c) GAG concentration, (d) total collagen concentration, (e) collagen type II concentration, (f) collagen type II as a percentage of total collagen. The constructs were cultured in bioreactors for 5 weeks. There was a significant difference (*) (p < 0.05) or (**) (p < 0.01) in tissue wet weight and in the concentrations of GAG, total collagen and collagen type II between the constructs obtained using single thick and composite thin scaffolds. The error bars represent standard errors from triplicate bioreactor cultures.

184

Nastaran Mahmoudifar and Pauline M. Doran

Figure 5. Histological appearance of cross-sections of cartilage constructs cultured in bioreactors for 5 weeks. (a)–(c) Tissue constructs generated using single thick scaffolds cultured with uni-directional medium flow. (d)–(f) Tissue constructs generated using composite thin and thick scaffolds with unidirectional medium flow. (g) Tissue construct generated using a single thick scaffold cultured with periodic medium flow reversal. (h) Tissue construct generated using a composite thin scaffold with periodic medium flow reversal. Cell nuclei are stained black; GAG is shown orange–red; collagen is shown blue–green; residual PGA fibres appear as red flecks.

Tissue Engineering of Cartilage in Bioreactors

185

Immunohistochemistry Cross-sections of tissue constructs were stained with antibodies against collagen type I and collagen type II as described previously (Mahmoudifar and Doran, 2005a). All constructs stained positively for both collagen types I and II. Representative samples of tissue generated using single thick and composite thin scaffolds cultured with periodic medium flow reversal are shown, respectively, in Figure 6a, 6b and Figure 6c, 6d, illustrating positive reaction with both antibodies. The distributions of collagen types I and II were relatively uniform throughout the cross-sections of all composite constructs cultured using uni-directional medium flow. For the constructs prepared using single thick scaffolds and periodic medium flow reversal, the distributions of collagen types I and II were similar. More homogeneous staining was observed along the top and bottom of the cross-sections than in the middle, where holes or empty spaces were apparent. For the constructs prepared using composite thin scaffolds, the distribution of collagen type II was relatively uniform throughout the crosssections, whereas collagen type I, although present throughout, was more abundant along the top and bottom surfaces.

Figure 6. Immunohistochemical sections of cartilage constructs showing positive staining with antibodies against collagen type I (a) and (c), and collagen type II (b) and (d). (a) and (b): tissues generated using single thick PGA scaffolds cultured in bioreactors with periodic medium flow reversal. (c) and (d): tissues generated using composite thin PGA scaffolds cultured in bioreactors with periodic medium flow reversal.

Transmission Electron Microscopy of Tissue-Engineered and Native Cartilage The ultrastructure of tissue-engineered cartilage generated using composite thin scaffolds cultured with periodic medium flow reversal was examined and compared with human foetal and adult cartilage tissues using transmission electron microscopy (Fig. 7).

186

Nastaran Mahmoudifar and Pauline M. Doran

Regions of cartilage ECM adjacent to individual chondrocytes in the tissues are shown in Figure 7a, 7c, 7e. The presence of assembled extracellular collagen fibrils in the tissueengineered cartilage is indicated by the fibrous material outside of the cell in Figure 7a. The ECM of the native tissues also contains collagen fibrils (Fig. 7c, 7e). The higher density of collagen in the adult cartilage is evident in the micrographs; the tissue-engineered cartilage contains relatively small amounts of collagen. Images at higher magnification (Fig. 7b, 7d, 7f) show the presence of much thicker and distinctly banded collagen fibrils in the adult cartilage (Fig. 7f) compared with thinner, more sparse fibrils in the tissue-engineered and foetal samples (Fig. 7b, 7d).

Discussion When cultured in bioreactors with uni-directional medium flow, cartilage constructs generated using composite thin scaffolds were in tissue weight and biochemical composition similar to those produced from single thick scaffolds, except for total collagen concentration (Fig. 3). However, because the concentration of collagen type II and the result for collagen type II as a percentage of total collagen were not significantly different between these constructs (Fig. 3e, 3f), the higher concentration of total collagen in the tissues developed from single thick scaffolds could be due to the production of collagen types other than collagen type II. Among the composite scaffolds, tissues developed from composite thin scaffolds were more cartilaginous, i.e. they contained higher concentrations of GAG (Fig. 3c) and collagen type II (Fig. 3e), and levels of collagen type II as a percentage of total collagen (Fig. 3f), than the constructs generated from composite thin and thick scaffolds. This result may reflect the higher initial cell density of 5.8 × 107 cells cm-3 used in the thin sections of the composite scaffolds compared with 2.6 × 107 cells cm-3 in the thick sections, as well as the more uniform cell distribution in the thin disc of the composite scaffolds relative to that in the thick disc (Mahmoudifar and Doran, 2005a). Tissue constructs developed from composite thin scaffolds cultured in bioreactors with periodic medium flow reversal were larger (Fig. 4a), more cartilaginous (Fig. 4c, 4e) and had a more homogenous distribution of GAG in their cross-sections (Fig. 5g, 5h) than those generated from single thick scaffolds cultured under the same conditions. It is possible that this result was due to better retention of cells and ECM components within the composite constructs than within the single thick scaffolds. Periodically reversing the direction of medium flow could have caused cells and ECM to be washed out from the top surfaces of the single scaffolds, which contained much higher cell densities after seeding than the bottom surfaces (Mahmoudifar and Doran, 2005a). The presence of collagen type I in cartilage tissues is undesirable as it indicates dedifferentiation of chondrocytes and formation of fibrocartilage rather than hyaline cartilage. In this work, the presence of collagen type I in the same regions of the tissue as collagen type II can be attributed to the plasticity in terms of differentiation potential of the foetal cartilage cells used in this study and the use of cells passaged in monolayer prior to three-dimensional cultivation in bioreactors (Mahmoudifar and Doran, 2005a).

Tissue Engineering of Cartilage in Bioreactors

187

Figure 7. Transmission electron micrographs of: (a) and (b) tissue-engineered cartilage produced using composite thin PGA scaffolds and periodic medium flow reversal, (c) and (d) human foetal epiphyseal cartilage, and (e) and (f) human adult articular cartilage. (a), (c) and (e) show the ECM surrounding individual chondrocyte cells; all images show the distribution and thickness of collagen fibrils in the ECM.

CONCLUSION This chapter reviews the literature on cartilage tissue engineering with an emphasis on cell source, polymer scaffold, seeding method, bioreactor design and the role of mechanical stimulation on tissue development. The choice of seeding method influences the initial spatial distribution of cells within the scaffold. A uniform initial cell distribution together with an appropriate choice of bioreactor configuration, culture conditions and culture duration, have a

188

Nastaran Mahmoudifar and Pauline M. Doran

significant influence on the biochemical composition and homogeneity of the ECM at harvest. Mechanical stimulation, for example in the form of hydrostatic pressure or dynamic compression, can be applied to improve the biochemical composition and mechanical properties of engineered cartilage. The possibility of using adult stem cells for tissue engineering of cartilage will allow more flexibility in the choice of cell source and presents an alternative to using autologous chondrocytes, which requires removal of a piece of healthy cartilage from the patient. The generation of human tissue-engineered cartilage is described using single and composite scaffolds in recirculation bioreactors, thus demonstrating the influence of seeding and bioreactor culture conditions on the quality of the engineered tissue. The use of composite scaffolds and periodic medium flow reversal in this study resulted in cartilage constructs that were larger and more cartilaginous than those developed using single scaffolds and uni-directional medium flow.

ACKNOWLEDGEMENTS This work was funded by the Australian Research Council (ARC). We thank Gavin McKenzie for assistance with the histology and immunohistochemistry, Sigfrid Fraser for assistance with the electron microscopy, staff of the Sterilization Department, Prince of Wales Hospital, Sydney, for sterilizing the PGA scaffolds, Zbigniew Suminski for assistance with the vacuum packaging, Malcolm Noble for assistance with the photomicrography, and Russell Cail and staff at the UNSW Science Faculty workshop for assistance with equipment construction.

REFERENCES Aigner, J., Tegeler, J., Hutzler, P., Campoccia, D., Pavesio, A., Hammer, C., Kastenbauer, E., Naumann, A. (1998) Cartilage tissue engineering with novel nonwoven structured biomaterial based on hyaluronic acid benzyl ester. J Biomed Mater Res 42, 172–181. Angele, P., Schumann, D., Angele, M., Kinner, B., Englert, C., Hente, R., Füchtmeier, B., Nerlich, M., Neumann, C., Kujat, R. (2004) Cyclic, mechanical compression enhances chondrogenesis of mesenchymal progenitor cells in tissue engineering scaffolds. Biorheology 41, 335–346. Awad, H.A., Halvorsen, Y.-D., Gimble, J.M., Guilak, F. (2003) Effects of transforming growth factor β1 and dexamethasone on the growth and chondrogenic differentiation of adipose-derived stromal cells. Tissue Eng 9, 1301–1312. Awad, H.A., Wickham, M.Q., Leddy, H.A., Gimble, J.M., Guilak, F. (2004) Chondrogenic differentiation of adipose-derived adult stem cells in agarose, alginate, and gelatin scaffolds. Biomaterials 25, 3211–3222. Bancroft, G.N., Sikavitsas, V.I., Mikos, A.G. (2003) Design of a flow perfusion bioreactor system for bone tissue-engineered applications. Tissue Eng 9, 549–554.

Tissue Engineering of Cartilage in Bioreactors

189

Brittberg, M., Lindahl, A., Nilsson, A., Ohlsson, C., Isaksson, O., Peterson, L. (1994) Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. New Eng J Med 331, 889–895. Cancedda, R., Dozin, B., Giannoni, P., Quarto, R. (2003) Tissue engineering and cell therapy of cartilage and bone. Matrix Biol 22, 81–91. Carver, S.E., Heath, C.A. (1999a) Semi-continuous perfusion system for delivering intermittent physiological pressure to regenerating cartilage. Tissue Eng 5, 1–11. Carver, S.E., Heath, C.A. (1999b) Increasing extracellular matrix production in regenerating cartilage with intermittent physiological pressure. Biotechnol Bioeng 62, 166–174. Carver, S.E., Heath, C.A. (1999c) Influence of intermittent pressure, fluid flow, and mixing on the regenerative properties of articular chondrocytes. Biotechnol Bioeng 65, 274–281. Caterson, E.J., Nesti, L.J., Li, W.-J., Danielson, K.G., Albert, T.J., Vaccaro, A.R., Tuan, R.S. (2001) Three-dimensional cartilage formation by bone marrow-derived cells seeded in polylactide/alginate amalgam. J Biomed Mater Res 57, 394–403. Chen, H.-C., Lee, H.-P., Sung, M.-L., Liao, C.-J., Hu, Y.-C. (2004) A novel rotating-shaft bioreactor for two-phase cultivation of tissue-engineered cartilage. Biotechnol Prog 20, 1802–1809. Darling, E.M., Athanasiou, K.A. (2003) Articular cartilage bioreactors and bioprocesses. Tissue Eng 9, 9–26. Davisson, T., Sah, R.L., Ratcliffe, A. (2002) Perfusion increases cell content and matrix synthesis in chondrocyte three-dimensional cultures. Tissue Eng 8, 807–816. De Witt, M.T., Handley, C.J., Oakes, B.W., Lowther, D.A. (1984) In vitro response of chondrocytes to mechanical loading. The effect of short term mechanical tension. Connect Tiss Res 12, 97–109. Dunkelman, N.S., Zimber, M.P., LeBaron, R.G., Pavelec, R., Kwan, M., Purchio, A.F. (1995) Cartilage production by rabbit articular chondrocytes on polyglycolic acid scaffolds in a closed bioreactor system. Biotechnol Bioeng 46, 299–305. Freed, L.E., Vunjak-Novakovic, G. (1995) Cultivation of cell–polymer tissue constructs in simulated microgravity. Biotechnol Bioeng 46, 306–313. Freed, L.E., Vunjak-Novakovic, G. (1997) Microgravity tissue engineering. In Vitro Cell Dev Biol Animal 33, 381–385. Freed, L.E., Marquis, J.C., Nohria, A., Emmanual, J., Mikos, A.G., Langer, R. (1993) Neocartilage formation in vitro and in vivo using cells cultured on synthetic biodegradable polymers. J Biomed Mater Res 27, 11–23. Freed, L.E., Grande, D.A., Lingbin, Z., Emmanual, J, Marquis, J.C., Langer, R. (1994a) Joint resurfacing using allograft chondrocytes and synthetic biodegradable polymer scaffolds. J Biomed Mater Res 28, 891–899. Freed, L.E., Marquis, J.C., Vunjak-Novakovic, G., Emmanual, J., Langer, R. (1994b) Composition of cell–polymer cartilage implants. Biotechnol Bioeng 43, 605–614. Freed, L.E., Hollander, A.P., Martin, I., Barry, J.R., Langer, R., Vunjak-Novakovic, G. (1998) Chondrogenesis in a cell-polymer-bioreactor system. Exp Cell Res 240, 58–65. Gilding, D.K. (1981) Biodegradable polymers. In: D.F. Williams (ed), Biocompatibility of Clinical Implant Materials, vol II, CRC Series in Biocompatibility (pp 209–232), Boca Raton, Florida: CRC Press.

190

Nastaran Mahmoudifar and Pauline M. Doran

Grande, D.A., Halberstadt, C., Naughton, G., Schwartz, R., Manji, R. (1997) Evaluation of matrix scaffolds for tissue engineering of articular cartilage grafts. J Biomed Mater Res 34, 211–220. Grigolo, B., Lisignoli, G., Piacentini, A., Fiorini, M., Gobbi, P., Mazzotti, G., Duca, M., Pavesio, A., Facchini, A. (2002) Evidence for redifferentiation of human chondrocytes molecular, grown on a hyaluronan-based biomaterial (HYAFF®11): immunohistochemical and ultrastructural analysis. Biomaterials 23, 1187–1195. Guilak, F., Sah, R., Setton, L.A. (1997) Physical regulation of cartilage metabolism. In: V.C. Mow, W.C. Hayes (eds), Basic Orthopaedic Biomechanics, 2nd ed (pp 179–207), Philadelphia: Lippincott–Raven. Hatton, P.V., Walsh, J., Brook, I.M. (1994) The response of cultured bone cells to resorbable polyglycolic acid and silicone membranes for use in orbital floor fracture repair. Clin Mater 17, 71–80. Honda, K., Ohno, S., Tanimoto, K., Ijuin, C., Tanaka, N., Doi, T., Kato, Y., Tanne, K. (2000) The effects of high magnitude cyclic tensile load on cartilage matrix metabolism in cultured chondrocytes. Eur J Cell Biol 79, 601–609. Hunter, C.J., Imler, S.M., Malaviya, P., Nerem, R.M., Levenston, M.E. (2002) Mechanical compression alters gene expression and extracellular matrix synthesis by chondrocytes cultured in collagen I gels. Biomaterials 23, 1249–1259. Ishaug-Riley, S.L., Okun, L.E., Prado, G., Applegate, M.A., Ratcliffe, A. (1999) Human articular chondrocyte adhesion and proliferation on synthetic biodegradable polymer films. Biomaterials 20, 2245–2256. Johnstone, B., Hering, T.M., Caplan, A.I., Goldberg, V.M., Yoo, J.U. (1998) In vitro chondrogenesis of bone marrow-derived mesenchymal progenitor cells. Exp Cell Res 238, 265–272. Jorgensen, C., Gordeladze, J., Noel, D. (2004) Tissue engineering through autologous mesenchymal stem cells. Curr Opin Biotechnol 15, 406–410. Kim, B.-S., Mooney, D.J. (1998) Development of biocompatible synthetic extracellular matrices for tissue engineering. Trends Biotechnol 16, 224–230. Kim, H.W., Han, C.D. (2000) An overview of cartilage tissue engineering. Yonsei Med J 41, 766–773. Kim, B.-S., Putnam, A.J., Kulik, T.J., Mooney, D.J. (1998) Optimizing seeding and culture methods to engineer smooth muscle tissue on biodegradable polymer matrices. Biotechnol Bioeng 57, 46–54. Lee, D.A., Bader, D.L. (1997) Compressive strains at physiological frequencies influence the metabolism of chondrocytes seeded in agarose. J Orthop Res 15, 181–188. Lee, D.A., Noguchi, T., Knight, M.M., O’Donnell, L., Bentley, G., Bader D.L. (1998) Response of chondrocyte subpopulations cultured within unloaded and loaded agarose. J Orthop Res 16, 726–733. Li, W.-J., Tuli, R., Okafor, C., Derfoul, A., Danielson, K.G., Hall, D.J., Tuan, R.S. (2005) A three-dimensional nanofibrous scaffold for cartilage tissue engineering using human mesenchymal stem cells. Biomaterials 26, 599–609. Ma, P.X., Langer, R. (1999) Fabrication of biodegradable polymer foams for cell transplantation and tissue engineering. In: J.R. Morgan, M.L. Yarmush (eds), Tissue Engineering Methods and Protocols (pp 47–56). Totowa, New Jersey: Humana Press.

Tissue Engineering of Cartilage in Bioreactors

191

Mahmoudifar, N., Doran, P.M. (2005a) Tissue engineering of human cartilage in bioreactors using single and composite cell-seeded scaffolds. Biotechnol Bioeng 91, 338–355. Mahmoudifar, N., Doran, P.M. (2005b) Tissue engineering of human cartilage and osteochondral composites using recirculation bioreactors. Biomaterials 26, 7012–7024. Mahmoudifar, N., Doran, P.M. (2006) Effect of seeding and bioreactor culture conditions on the development of human tissue-engineered cartilage. Tissue Eng 12, 1675-1685. Martin, I., Padera, R.F., Vunjak-Novakovic, G., Freed, L.E. (1998) In vitro differentiation of chick embryo bone marrow stromal cells into cartilaginous and bone-like tissues. J Orthop Res 16, 181–189. Martin, I., Obradovic, B., Treppo, S., Grodzinsky, A.J., Langer, R., Freed, L.E., VunjakNovakovic, G. (2000) Modulation of the mechanical properties of tissue engineered cartilage. Biorheology 37, 141–147. Martin, I., Shastri, V.P., Padera, R.F., Yang, J., Mackay, A.J., Langer, R., Vunjak-Novakovic, G., Freed, L.E. (2001) Selective differentiation of mammalian bone marrow stromal cells cultured on three-dimensional polymer foams. J Biomed Mater Res 55, 229–235. Meinel, L., Hofmann, S., Karageorgiou, V., Zichner, L., Langer, R., Kaplan, D., VunjakNovakovic, G. (2004) Engineering cartilage-like tissue using human mesenchymal stem cells and silk protein scaffolds. Biotechnol Bioeng 88, 379–391. Nehrer, S., Breinan, H.A., Ramappa, A., Young, G., Shortkroff, S., Louie, L.K., Sledge, C.B., Yannas, I.V., Spector, M. (1997) Matrix collagen type and pore size influence behaviour of seeded canine chondrocytes. Biomaterials 18, 769–776. Parkkinen, J.J., Ikonen, J., Lammi, M.J., Laakkonen, J., Tammi, M., Helminen, H.J. (1993) Effects of cyclic hydrostatic pressure on proteoglycan synthesis in cultured chondrocytes and articular cartilage explants. Arch Biochem Biophys 300, 458–465. Pazzano, D., Mercier, K.A., Moran, J.M., Fong, S.S., DiBiasio, D.D., Rulfs, J.X., Kohles, S.S., Bonassar, L.J. (2000) Comparison of chondrogenesis in static and perfused bioreactor culture. Biotechnol Prog 16, 893–896. Pieper, J.S., van der Kraan, P.M., Hafmans, T., Kamp, J., Buma, P., van Susante, J.L.C., van den Berg, W.B., Veerkamp, J.H., van Kuppevelt, T.H. (2002) Crosslinked type II collagen matrices: preparation, characterization, and potential for cartilage engineering. Biomaterials 23, 3183–3192. Pittenger, M.F., Mackay, A.M., Beck, S.C., Jaiswal, R.K., Douglas, R., Mosca, J.D., Moorman, M.A., Simonetti, D.W., Craig, S., Marshak, D.R. (1999) Multilineage potential of adult human mesenchymal stem cells. Science 284, 143–147. Rotter, N., Aigner, J., Naumann, A., Planck, H., Hammer, C., Burmester, G., Sittinger, M. (1998) Cartilage reconstruction in head and neck surgery: comparison of resorbable polymer scaffolds for tissue engineering of human septal cartilage. J Biomed Mater Res 42, 347–356. Sah, R.L.-Y., Kim, Y.-J., Doong, J.-Y.H., Grodzinsky, A.J., Plaas, A.H.K., Sandy, J.D. (1989) Biosynthetic response of cartilage explants to dynamic compression. J Orthop Res 7, 619–636. Schreiber, R.E., Dunkelman, N.S., Naughton, G., Ratcliffe, A. (1999) A method for tissue engineering of cartilage by cell seeding on bioresorbable scaffolds. Ann New York Acad Sci 875, 398–404.

192

Nastaran Mahmoudifar and Pauline M. Doran

Seidel, J.O., Pei, M., Gray, M.L., Langer, R., Freed, L.E., Vunjak-Novakovic, G. (2004) Long-term culture of tissue engineered cartilage in a perfused chamber with mechanical stimulation. Biorheology 41, 445–458. Stoltz, J.F., Dumas, D., Wang, X., Payan, E., Mainard, D., Paulus, F., Maurice, G., Netter, P., Muller, S. (2000) Influence of mechanical forces on cells and tissues. Biorheology 37, 3– 14. Tuli, R., Seghatoleslami, M.R., Tuli, S., Wang, M.L., Hozack, W.J., Manner, P.A., Danielson, K.G., Tuan, R.S. (2003) A simple, high-yield method for obtaining multipotential mesenchymal progenitor cells from trabecular bone. Molec Biotechnol 23, 37–49. Tuli, R., Nandi, S., Li, W.-J., Tuli, S., Huang, X., Manner, P.A., Laquerriere, P., Nöth, U., Hall, D.J., Tuan, R.S. (2004) Human mesenchymal progenitor cell-based tissue engineering of a single-unit osteochondral construct. Tissue Eng 10, 1169–1179. van Susante, J.L.C., Pieper, J., Buma, P., van Kuppevelt, T.H., van Beuningen, H., van der Kraan, P.M., Veerkamp, J.H., van den Berg, W.B., Veth, R.P.H. (2001) Linkage of chondroitin-sulfate to type I collagen scaffolds stimulates the bioactivity of seeded chondrocytes in vitro. Biomaterials 22, 2359–2369. Vunjak-Novakovic, G., Freed, L.E., Biron, R.J., Langer, R. (1996) Effects of mixing on the composition and morphology of tissue-engineered cartilage. AIChE J 42, 850–860. Vunjak-Novakovic, G., Obradovic, B., Martin, I., Bursac, P.M., Langer, R., Freed, L.E. (1998) Dynamic cell seeding of polymer scaffolds for cartilage tissue engineering. Biotechnol Prog 14, 193–202. Vunjak-Novakovic, G., Martin, I., Obradovic, B., Treppo, S., Grodzinsky, A.J., Langer, R., Freed, L.E. (1999) Bioreactor cultivation conditions modulate the composition and mechanical properties of tissue-engineered cartilage. J Orthop Res 17, 130–138. Vunjak-Novakovic, G., Obradovic, B., Martin, I., Freed, L.E. (2002) Bioreactor studies of native and tissue engineered cartilage. Biorheology 39, 259–268. Waldman, S.D., Spiteri, C.G., Grynpas, M.D., Pilliar, R.M., Kandel, R.A. (2004) Long-term intermittent compressive stimulation improves the composition and mechanical properties of tissue-engineered cartilage. Tissue Eng 10, 1323–1331. Wendt, D., Marsano, A., Jakob, M., Heberer, M., Martin, I. (2003) Oscillating perfusion of cell suspensions through three-dimensional scaffolds enhances cell seeding efficiency and uniformity. Biotechnol Bioeng 84, 205–214. Wimmer, M.A., Grad, S., Kaup, T., Hänni, M., Schneider, E., Gogolewski, S., Alini, M. (2004) Tribology approach to the engineering and study of articular cartilage. Tissue Eng 10, 1436–1445. Winter, A., Breit, S., Parsch, D., Benz, K., Steck, E., Hauner, H., Weber, R.M., Ewerbeck, V., Richter, W. (2003) Cartilage-like gene expression in differentiated human stem cell spheroids. Arthritis Rheum 48, 418–429. Zuk, P.A., Zhu, M., Mizuno, H., Huang, J., Futrell, J.W., Katz, A.J., Benhaim, P., Lorenz, H.P., Hedrick, M.H. (2001) Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng 7, 211–228.

In: Materials Science Research Horizons Editor: Hans P. Glick pp. 193-216

ISBN 978-1-60021-481-3 © 2007 Nova Science Publishers, Inc.

Chapter 7

HETEROGENEOUS COMBUSTION SYNTHESIS Hung-Pin Li1 Jin-Wen University of Science and Technology Hsintien, Taipei County, Taiwan

ABSTRACT Many exothermic non-catalytic solid-solid or solid-gas reactions, after being ignited locally, can release enough heat to sustain the self-propagating combustion front throughout the specimen without additional energy. Since the 1970’s, this kind of exothermic reaction has been used in the process of synthesizing refractory compounds in the former Soviet Union. This novel technique, so-called Combustion / Micropyretic synthesis or Self-propagating High-temperature Synthesis(SHS), has been intensively studied for process implication. This technique employs exothermic reaction processing, which circumvents difficulties associated with conventional methods of time and energyintensive sintering processing. The advantages of combustion synthesis also include the rapid net shape processing and clean products. In addition, the combustion-synthesized products have been reported to possess better mechanical and physical properties. Heterogeneous distributions of reactants, diluents, and pores are common during combustion synthesis when powders are mixed, and this directly leads to the variations of the thermophysical / chemical parameters of the unreacted compacts. Since combustion synthesis is sustained by the sequences of the local chemical reactions, the propagation manner is strongly dependent on the parameters of each portion of the reactants. Thus, the variation of thermophysical / chemical parameters of reactants caused by heterogeneities in composition and porosity is thought to significantly change the processing parameters, such as combustion temperature and propagation velocity; and further affect the product properties. This chapter systematically introduces the impact of heterogeneities during combustion synthesis with Ni + Al. Correlations of heterogeneities in the reactants and a diluent with the propagation velocity and combustion temperature are discussed. In addition, a map, considering concurrent heterogeneities in the composition and porosity, has been generated to provide a better understanding of the change in propagation velocity on account of the heterogeneous combustion synthesis. 1

Correspondence: Hung-Pin Li, Ph.D., Professor, Dean of R&D Office, Jin-Wen University of Science and Technology, Taiwan, e-mail : [email protected],, TEL: +886-932383482 FAX:+886-282122209

194

Hung-Pin Li

Keywords: composition heterogeneity, porosity heterogeneity, self-propagating hightemperature synthesis (SHS), heterogeneous micropyretic synthesis, heterogeneous combustion synthesis

INTRODUCTION Combustion synthesis, also referred to as micropyretic synthesis or self-propagating hightemperature synthesis (SHS), is a novel processing method for the production of intermetallics, ceramics, composites, and other materials. The technique employs exothermic reaction processing which circumvents difficulties associated with conventional methods of time and energy-intensive sinter processing. Two basic combustion synthesis modes are commonly employed, namely the wave propagation mode and the thermal explosion mode. In the wave propagation mode, the compacted powders are ignited at a point by a heat source. After ignition, the heat to propagate the combustion wave is obtained from the heat released by the formation of the synthesized product, as shown in figures 1 [1] and 2. The unreacted portion in front of the combustion wave is heated by this exothermic heat, undergoes synthesis, the wave propagates, thus causing further reaction and synthesis. In the thermal explosion mode, the specimen is heated in a furnace. The furnace may be kept at the ignition temperature or the specimen may be heated in the furnace at a predetermined heating rate to the ignition temperature. The combustion reaction in this mode may occur more or less simultaneously at all points in the specimen. Although the synthesized product phases obtained by both techniques are similar [2], there may be differences in the amount of residual porosity, final dimensions, and the thermal gradient during the processing. In both the modes, solid-solid reactions are most commonly encountered, sometimes solid-gas reactions are also noted as in the case of synthesis of refractory nitrides like TiN where nitrogen gas is used [3].

Figure 1. The combustion front propagates from right to left in the combustion synthesis of 95 wt.%(Ti+2B) + 5 wt.% Cu. [1].

Heterogeneous Combustion Synthesis

195

Figure 2. Schematic representation of the wave propagation mode in combustion synthesis.

The advantages of combustion synthesis techniques include rapid net shape processing and clean products. When compared with conventional powder metallurgy operations, combustion synthesis not only offers shorter processing time but also excludes the requirement for high-temperature sintering. Volatile contaminants or impurities may be eliminated as the high temperature combustion wave propagates through the sample, and thus the synthesized products have the higher purity [4,5]. The steep temperature gradient also gives rise to the occurrence of metastable or non-equilibrium phases, which are not available in the conventional processing [4,5]. Combustion synthesized products have also been reported to have better mechanical and physical properties [5,6]. An example is the formation of shape-memory alloys of nickel and titanium [6]. It has been reported that those prepared by combustion synthesis, possess greater shape-recovery force than corresponding alloys produced by conventional methods [6]. On account of the high thermal gradients encountered in combustion synthesis, it has been speculated that the products of such a process may contain a high defect concentration. The presence of high levels of defects has led to expectation of higher reactivity, namely higher sinterability [7]. Combustion synthesis continues to generate interest because of current and potential applications. Such applications include the use of combustion synthesized materials for [4]: (1) electrodes for electrolysis of corrosive media - TiN and TiB2; (2) abrasive, cutting tools, and polishing powders - TiC, carbonitrides, and cemented carbides; (3) high temperature structural intermetallics - NiAl; (4) resistive heating elements - MoSi2; (5) steel processing additives - nitrided ferroalloys; (6) shape-memory alloys - TiNi; (7) composites - TiC + Al2O3. Several numerical and analytical models of combustion synthesis in a composite system have been well developed [8-19]. Lakshmikantha and Sekhar firstly explored the numerical model that includes the effects of dilution and porosity, and melting of each constituent of the reactants and products [13,14]. The analytical modeling of the propagation of the combustion

196

Hung-Pin Li

front in solid-solid reaction systems is also reported [15]. The analytical model gives good results when compared with the experimentally determined numbers and the numerically calculated values. In addition, a dynamic modeling of the gas and solid reaction has also been carried out to illustrate the effects of various parameters on the combustion synthesis [16]. These numerical and analytical analyses provide the better understanding of the reaction sequence during combustion synthesis reactions. However, heterogeneities in initial composition and porosity are common during combustion synthesis when powders are pressed or mixed and the conventional modeling treatments [13-19] thus far have only considered uniform systems. Since combustion synthesis is sustained by the sequences of the local chemical reactions, the propagation manner is strongly depending on the parameters in the previous portion (node). Thus, the variation of thermophysical / chemical parameters of reactants caused by composition and porosity heterogeneity is thought to significantly change the processing parameters, such as combustion temperature and propagation velocity; and further affect the product properties. In this chapter, a numerical simulation is used to characterize the effect of heterogeneities in composition and porosity on combustion synthesis with Ni + Al. Firstly, the effects of heterogeneities in reactants, diluent, and porosity on the propagation velocity and combustion temperature are investigated. The influence of the variation of each individual reactant parameter caused by heterogeneities in composition on the propagation velocity is also carried out. The heterogeneity maps, considering the heterogeneities in initial composition and porosity concurrently, are also generated. From the knowledge of heterogeneity maps, the effects of heterogeneities in initial composition and porosity on combustion reaction can be acquired.

NUMERICAL CALCULATION PROCEDURE During the passage of a combustion front in the reaction, the energy equation for transient heat conduction, including the source term, containing heat release due to the exothermic combustion reaction is given as [13,15,20]:

ρC p (

∂T ∂ ⎛ ∂T ⎞ 4h(T − To ) ) = ⎜κ ( ) ⎟ − + ρQΦ (T ,η ) ∂t d ∂z ⎝ ∂z ⎠

(1)

Each symbol in the equation is explained in the nomenclature section. The reaction rate, Φ (T ,η ) , in Eq.(1) is given as :

Φ (T ,η ) =

∂η E = K o (1 − η ) exp( − ) RT ∂t

(2)

In this study, a numerical calculation for Eq.(1) is carried out with the assumption of the first order kinetics. In the Eq. (1), the energy required for heating the synthesized product from the initial temperature to the adiabatic combustion temperature is shown on the left-hand side. The terms on the right-hand side are the conduction heat transfer term, the surface heat

Heterogeneous Combustion Synthesis

197

loss parameter, and the heat release due to the exothermic combustion reaction, respectively. The surface heat loss is assumed radically Newtonian in this study. The previous studies [13,21] have shown that the surface heat loss is much less than the exothermic heat of the reaction, thus, the surface heat loss is taken to be zero in the numerical calculation. The middle-difference approximation and an enthalpy-temperature method coupled with Guass-Seidel iteration procedure are used to solve the equations of the combustion synthesis problems. In the computational simulation, a one-dimensional sample of 1 cm long is divided into 1201 nodes (regions) to calculate the local temperature using an enthalpy-temperature method. The choice of 1 cm sample length is only for computational purpose, and the simulation results are applicable to practical experimental conditions. Firstly, the proper initial and boundary conditions are used to initialize the temperatures and enthalpies at all nodes. The initial conditions in the simulation are taken as follows: (1) At the ignition node, at time t ≥ 0, the temperature is taken to be the adiabatic combustion temperature, (T = Tc and η = 1). (2) At the other nodes, at time t = 0, the temperatures are taken to be the same as the substrate temperature, (T = To and η = 0). Depending on the values of the temperature and enthalpy occurred in the reaction, the proper thermophysical / chemical parameters are considered and the limits of the reaction zone are determined for each node in the numerical calculation. At any given time, the fraction reacted and enthalpy of the current iteration are calculated from the previous fraction reacted and enthalpy of the earlier iteration. The range of the enthalpy as well as the molar ratio among each material for each node is thus determined, and the values of temperature, density, and thermal conductivity at each node can be further calculated in appropriate zone. Since concurrent heterogeneous distributions of reactants and diluent are common when powders are mixed, the effects of heterogeneities in the reactant and diluent are also both considered in this study. Composition at each node is calculated from the random number ( f R ( j ) at node j) and the assigned heterogeneity ( Hetero react and Hetero diluent ) that determines the magnitude of the variation. The sequence of the random numbers (-0.5 ~ +0.5) generated from the computation is repeatedly used in the specimens with different heterogeneities to compare the magnitude of heterogeneity effect. In this chapter, the heterogeneities in the reactant (Ni/Al) and diluent (NiAl) are respectively considered in the numerical calculation. The heterogeneities in compositions are calculated to be as: NiAl diluent molar fraction at node j:

X diluent , j = X

o diluent

(1 + Heterodiluent ⋅ f R ( j ) )

Ni molar fraction at node j:

(

)

X Ni , j = 1 − X diluent , j ⋅ X

o Ni

⋅ (1 + Hetero react ⋅ f R ( j ) )

Al molar fraction at node j: X Al , j = 1 − X diluent , j − X Ni , j

(3a)

(3b)

(3c)

where j = 1,2,.....,1201 . In order to assure the sum of the compositions for all 1201 nodes equal to the stoichiometric values, the calculated Ni and Al composition of each node is adjusted so that the average value of each composition is equal to the original homogeneous

198

value, i.e.,

Hung-Pin Li

1 n =1201 1 n =1201 o X = X = 50 at .% ; ∑ Ni , j Ni ∑ X Al , j = X Alo = 50at.% . For example, as n j =1 n j =1

30% heterogeneity in initial Ni composition is occurred, Ni composition is correspondingly varied within 15 at.% (=50.0 at.% × 30%). Thus, Ni composition is noted to vary from 42.5 at.% to 57.5 at.% and Al composition is correspondingly determined. After the molar fractions of reactants are determined, the reaction yield at node j can be further determined from the molar fractions of reactants and diluent:

X ⎫ ⎧X Ryield , j = min ⎨ Nio, j , Alo, j ⎬ ⎩ X Ni X Al ⎭

(4)

In addition, the porosity effects of the reactants and product that influence the density (ρ) and thermal conductivity (κ) profiles are also considered in this chapter. During the numerical calculation, the average global porosities (Po) of the reactants and product are both taken to be 25%. The initial porosity of the reactants at each node is calculated from the random number ( f R ( j ) at node j) and the assigned heterogeneity (Heteroporosity) that determines the magnitude of the porosity variation:

(

porosity at node j: Pj = P 1 + Hetero porosity ⋅ f R ( j ) o

)

(5)

− 0.5 ≤ f R ( j ) ≤ +0.5 , P o = 25 %, 0% < Heteroporosity < 100%, and j = 1,2,.....,1201 . The studied compositions with the different initial heterogeneities in

where

compositions and porosity are also shown in Table I. Once the initial composition and porosity at each node are set to given heterogeneities, the thermophysical/chemical parameters at node j can be thus calculated as: density at node j :

ρ j = ∑ [ ρ s ⋅Vs , j ⋅ (1 − Pj )]

(6)

s

thermal conductivity at node j :

κ j = ∑ [κ s ⋅Vs , j ⋅ (1 − Pj ) (1 + Pj 2)]

(7)

⋅X s , j )

(8)

s

heat capacity : Cp j =

∑ (Cp

s

s

where s denotes the component involved in the reaction, including Ni and Al in this study. The effect of melting of reactants and product is also included in the calculation. In this chapter, the surface heat loss is taken as zero in the numerical calculation. Using Eqs. (1)-(8), the energy equation on nth time step at node j can be written as:

Heterogeneous Combustion Synthesis

⎧⎪⎛ K j +1 + K j ⎨⎜⎜ 2 Tm − Tm −1 ⎪⎝ )=⎩ ρ jC pj ( Δt

⎞⎛ T j +1 − T j ⎟⎟⎜ ⎜ ⎠⎝ z j +1 − z j

o + ρ j Q j (1 − X diluent ) K o (1 − η m −1 ) exp(− E

199

⎞ ⎛ K j + K j −1 ⎞⎛ T j − T j −1 ⎞⎫⎪ ⎟−⎜ ⎟ ⎟⎟⎜ ⎟ ⎜ ⎜ z − z ⎟⎬⎪ 2 j −1 ⎠ ⎭ ⎠⎝ j ⎠ ⎝

(z

j +1

− z j −1

) / ⎛1 + K o exp(− E )Δt ⎞ RTm −1 ⎜⎝ RTm −1 ⎟⎠ (9)

Table 1. The examples of the studied Ni + 50 at.% Al compositions with different heterogeneities in initial composition and porosity Heterogeneity in composition, %

Heterogeneity in Ni composition, porosity, % at.%

Al composition, at.%

Porosity

0

0

50.0

50.0

25.0

0

60

50.0

50.0

17.5 – 32.5

30

0

42.5 – 57.5

57.5 – 42.5

25.0

30

60

42.5 – 57.5

57.5 – 42.5

17.5 – 32.5

Using Eq. (9), the temperature, fraction reacted, and enthalpy on nth time step at node j can be thus determined by the Guass-Seidel iteration procedure. The various thermophysical / chemical parameters, such as thermal conductivity, density and heat capacity of the reactants and product, are assumed to be independent of temperature, but they are different in each state. The average values of these parameters vary as the reaction proceeds, depending upon the degree of reaction. In addition, the higher preexponential factor (Ko) value, 8 x 108 1/s, is used to be capably of illustrating the variation of the propagation velocity for the NiAl combustion reaction. The parameter values used in the computational calculation are shown in Table II [22-24] and Table III [22,25]. In this chapter, the combustion temperature is defined as the highest reaction temperature during combustion synthesis, and the propagation velocity is the velocity of the combustion front propagation. Table 2. The thermophysical/chemical parameters for the reactants (Ni and Al) and product (NiAl) at 300 K and liquid state [22-24] Thermophysical/chemical parameters Heat capacity (at 300 K) (J/(kgK)) Heat capacity (liquid state) (J/(kgK)) Thermal conductivity (at 300 K) (J/(msK)) Thermal conductivity (liquid state) (J/(msK)) Density (at 300 K) (kg/m3) Density (liquid state) (kg/m3)

Al 902 [22] 1178 [22] 238 [24] 100 [24] 2700 [24] 2385 [24]

Ni 445 [22] 735 [22] 88.5 [24] 53 [23] 8900 [24] 7905 [24]

NiAl 537 [22] 831 [22] 75 [23] 55 [23] 6050 [23] 5950 [24]

200

Hung-Pin Li

Table 3. The values of various parameters used in the numerical calculation [22,25] Parameters Combustion Temperature (K) Activation Energy (kJ/mole) Exothermic Heat (kJ/mole) Pre-Exponential Factor (1/second) Time Step (second)

NiAl 1912 139 [22] 118.5 [25] 8 x 108 0.00025

The criterion used to ascertain whether the fraction reacted and enthalpies at each time level converge or not, is determined from the relative error criterion. Once the convergence criterion for every node is met, the enthalpy and fraction reacted of the last iteration in a time step are considered to be the corresponding final values. The calculations are normally performed from 500 to 2000 times, depending upon the calculated thermal parameters; to make all 1201 sets (nodes) meet the criterion for each time step. At least 600 time steps are calculated to allow the combustion front propagate the 1-cm-long specimen completely.

RESULTS AND DISCUSSION 1. Effect of Heterogeneity in Reactants Figure 3 shows the variations of the Ni composition and the correspondingly reaction yield along the Ni + 50 at. % Al composition with 20 % maximum heterogeneity in reactants. For this composition, Ni and Al compositions are respectively set to vary within 10 at. % (= 50 at.% x 20 %). On account of the variations of compositions with the distance, the reaction yields are correspondingly altered at all nodes. It is noted that as the Ni composition is deviated far from the stoichiometric value (50 at. %), the difference between Ni and Al compositions increases and the reaction yield for NiAl combustion reaction decreases, as shown in figure. 3. A decrease in the reaction yield is expected to decrease the exothermic heat of the reaction, further reducing the reactivity of the combustion reaction.

Figure 3. The variations of the Ni composition and the correspondingly reaction yield with the distance for the NiAl compound with 20 % maximum heterogeneity in reactants. No diluent is added in this composition.

Heterogeneous Combustion Synthesis

201

In addition, the occurrence of heterogeneities in reactants also influences the thermophysical/chemical parameters of the reactants at each node, as expected. Since the thermophysical/chemical parameters (including thermal conductivity, heat capacity, and density) are respectively calculated from the ratio of the reactants at each node (Eqs. (6)-(8)), the variations of these parameters are strongly correlated with the change of the compositions. figure 4 shows the variations of density and thermal conductivity along the specimen. In the ideal homogeneous specimen (0 % heterogeneity in reactants), thermal conductivity and density remain as constants at all nodes. As heterogeneities in reactants increase, density and thermal conductivity are noted to vary with the distance. Figure 4 also shows that the magnitudes of variations of density and thermal conductivity are increased as the heterogeneities in reactants are increased.

Figure 4. The variations of density (  and thermal conductivity (κalong the NiAl reactants. The horizontal line denotes the values for the ideal homogeneous specimens. The black and gray curves denote the specimens with 10 % and 20 % heterogeneities in reactants, respectively.

The variations of reaction yield, density, and thermal conductivity may further influence the reactivity of synthesis reaction at each node and thus change the propagation pattern and combustion corresponding parameters (i.e., combustion temperature and propagation velocity). Figure 5 shows the temperature profiles of combustion fronts at various times along the Ni + Al specimen for 0 %, 10 % and 20 % maximum heterogeneities in reactants, respectively. The combustion reaction is ignited at the position 0 cm and the combustion front starts to propagate from left to right. It is noted from figure 5(a) that the combustion front propagates at steady state for the ideal homogeneous specimen (0% heterogeneity in reactants). The highest reaction temperature (i.e., combustion temperature) and the instantaneous propagation velocity of the homogeneous specimens are at a steady value during front propagation. When the non-homogeneous specimens are ignited, it is found that the temperature and the instantaneous propagation velocity are altered with the distance (figures 5(b) and (c)). The magnitude of temperature variation is also increased with the increase in the heterogeneity in reactants. The average propagation velocity is calculated to decrease from 73.4 cm/s (for the ideal homogeneous specimen) to 72.9 cm/s (for the specimen with 10 % maximum heterogeneity in reactants). When the heterogeneity in

202

Hung-Pin Li

reactants is further increased to 20 %, the average propagation velocity is noted to dramatically decrease in 9.13 % (from 73.4 cm/s to 66.7 cm/s). The propagation front even stops half way for the composition with a further increase in the heterogeneity in reactants.

Figure 5. Time variations of the combustion front temperature along the Ni + Al specimen. The interval time between two consecutive time steps (profiles) is 0.00025 s. The number 20 denotes the twentieth time step (0.005 s) after ignition. The heterogeneities in reactants for figures (a), (b), and (c) are 0 %, 10 %, and 20 %, respectively.

To carefully investigate the variation of the combustion temperature with the distance, the combustion temperatures of specimens with different reactant heterogeneities are calculated at each node and plotted in figure 6. As expected, the combustion temperature is changed periodically with the distance for the heterogeneous specimen. In addition, an increase in the heterogeneity in reactants increases the magnitude of the combustion temperature variation. Also shown in the figure is the variation of Ni content. It is noted that the combustion temperature does not strongly correlate with the variation of Ni content in time. However, as the variation of Ni content is accumulated to a certain level, the combustion temperature is found to alter periodically.

2. Effect of Heterogeneity in Diluent To illustrate the effect of the heterogeneities in NiAl diluent on combustion reaction, the heterogeneities in reactants are temporarily neglected in this section, i.e., the heterogeneities in reactants are taken to be zero. With the heterogeneity in reactants is set to zero and the diluent amount is set to 20% of composition, the diluent composition will therefore vary within 4 % as the 20% heterogeneity in NiAl diluent is taken. Such a small composition variation correspondingly results in a tiny variation of the thermophysical / chemical parameters. It is thus found from figure 7 that the magnitudes of the variations for thermal conductivity and density are dramatically decreased as compared with those in figure 3. A

Heterogeneous Combustion Synthesis

203

decrease in the variations of the reaction yield and the thermophysical/chemical parameters is correspondingly to reduce the effects of heterogeneity in diluent on lowering propagation velocity and combustion temperature.

Figure 6. The plots of the combustion temperatures at different positions for the specimens with 0 %, 10 %, and 20 % heterogeneities in reactants, respectively. Also shown in the upper part of the figure is the variation of Ni composition.

Figure 7. The plots of density (  and thermal conductivity (κalong the NiAl reactants with 20 at. % diluent. The horizontal line denotes the values for the homogeneous specimens. The black and gray curves denote the specimens with 10 % and 20 % heterogeneities in diluent, respectively.

Figure 8 shows the temperature profiles of combustion fronts at various times along the NiAl specimen with 20 at. % diluent for three different diluent heterogeneities, 0 %, 10 %, and 20 % respectively. Consistent with the fact that the heterogeneities in diluent only slightly affect density and thermal conductivity (figure 7), the combustion temperature is also only slightly altered with the distance (figure 8). Figure 8 also shows that the propagation velocity

204

Hung-Pin Li

is slightly dependent on the heterogeneity in diluent. The combustion fronts for the specimens with different diluent heterogeneities are noted to propagate at the same propagation velocity, even the maximum heterogeneity in diluent has been enhanced to 20 %.

Figure 8. Time variations of the combustion front temperature at various times along the NiAl specimen with 20 at.% diluent. The interval time between two consecutive time steps (profiles) is 0.00025 s. The number 20 denotes the twentieth time step (0.005 s) after ignition. The heterogeneities in diluent for figures (a), (b), and (c) are 0 %, 10 %, and 20 %, respectively.

The combustion temperatures for specimens with different diluent heterogeneities are also calculated at each node and are plotted with the distance in figure 9. It is noted from Fig. 9 that the combustion temperature is also changed periodically with the distance for the heterogeneous specimen, but the range of variation is significantly decreased as compared with those plots for the specimens with different reactant heterogeneities in figure 6. As expected, the magnitude of the combustion temperature variation is also increased with the increasing heterogeneity in diluent. It is also noted that the oscillatory (variation) frequency is decreased as the heterogeneity in diluent increases. In addition, Table IV shows that average values of the combustion temperature are only slightly decreased as the maximum heterogeneity in diluent is increased. These observations suggest that the heterogeneity in diluent has a smaller effect than the heterogeneity in reactants, on changing the propagation pattern. The reductions in the propagation velocity caused by the composition heterogeneity and the addition of diluent are also shown in Table V. It is noted that the reduction in the propagation velocity caused by the 10 % Ni composition heterogeneity is equivalent to adding 5 % diluent for Ni + Al combustion reaction. It is found that the effect of the adding diluent on reducing propagation velocity is two to four times larger than the effect of composition heterogeneity.

Heterogeneous Combustion Synthesis

205

Figure 9. The plots of the combustion temperatures at different positions for the specimens with 0 %, 10 %, and 20 % heterogeneities in diluent (20 at.%), respectively. Also shown in the upper part of the figure is the variation of Ni composition.

Table 4. The average combustion temperatures for the NiAl stoichiometric compositions with 20 at.% diluent for the different heterogeneities in diluent Heterogeneity in diluent, %

combustion temperature, K

0

1846.38 ± 0.09

5

1845.52 ± 0.10

10

1844.64 ± 0.11

15

1843.77 ± 0.12

20

1842.78 ± 0.11

Table 5. Propagation velocities (Vc) for the compositions with different amounts of diluent and composition heterogeneities NiAl diluent %

Vc, cm/s

0% 5% 10 % 15 %

72.5 70.0 66.0 64.5

Ni composition heterogeneity 0% 5% 10 % 15 %

Vc, cm/s 72.5 71.5 70.0 68.0

Figure 10 shows the plot of the average propagation velocity and combustion temperature for the NiAl stoichiometric combustion front with different Ni composition heterogeneities. This figure clearly illustrates that the average propagation velocity and combustion temperature are decreased as the Ni composition heterogeneity is increased. Generally a decrease in the propagation velocity and combustion temperature slows down the combustion

206

Hung-Pin Li

reaction. It is inferred that the occurrence of the composition heterogeneity also decreases the reactivity of the combustion reaction. Similar phenomena of the decreased propagation velocity and the reduced reactivity of the combustion reaction have also been reported when the diluent is present in combustion reactions [13,26]. Therefore, these observations suggest that the effects caused by the Ni composition heterogeneity have the similar effects as adding diluents on combustion synthesis.

Figure 10. Plots of the average combustion temperature and propagation velocity for the NiAl stoichiometric combustion front with the different Ni composition heterogeneities.

3. The Heterogeneous Effect for Each Parameter The discussion above shows that the heterogeneities in reactants and diluent first alter the compositions at each node, and then further change the reaction yield, density, thermal conductivity, and heat capacity. The temperature, propagation velocity, and propagation manner are thus correspondingly changed along the specimen. To illustrate the magnitude of the heterogeneous effects caused by each parameter on the propagation velocity, each parameter is also assumed as independent of composition heterogeneities to calculate propagation velocity. It is noted that as the exothermic heat of reaction is assumed as constant during the numerical calculation, the calculated propagation velocities are enhanced from 66.8 cm/s (normal 20% heterogeneity in reactants) to 70.9 cm/s (Table VI). The calculated results also show that the propagation velocities are respectively increased to 68.5 cm/s and 67.3 cm/s when the heat capacity and reactant density are taken as constant values. These calculated results show that the reduction in the exothermic heat and the variation of heat capacity caused by the heterogeneity in reactants are the major factors to reduce the propagation velocity.

Heterogeneous Combustion Synthesis

207

Table 6. A change in the propagation velocity caused by the variation of individual parameter for the specimens with 20 % maximum heterogeneity in reactants. In the numerical calculations of the heterogeneous specimens, the exothermic heat (Q), density (ρ), thermal conductivity (K), and heat capacity (Cp) are respectively taken as constant to calculate the propagation velocity conditions

propagation velocity

homogeneous reaction

73.4 ± 0.0 cm/s

reaction with 20 % heterogeneity in reactants

66.8 ± 3.4 cm/s

reaction with 20 % heterogeneity in reactants (except Q = constant) reaction with 20 % heterogeneity in reactants (except ρ = constant) reaction with 20 % heterogeneity in reactants (except K = constant) reaction with 20 % heterogeneity in reactants (except Cp = constant)

70.9 ± 3.7 cm/s 67.3 ± 2.9 cm/s 66.7 ± 4.7 cm/s 68.5 ± 2.3 cm/s

4. Effect of Heterogeneities in Porosity and Composition A study of the combustion front propagating across a non-uniform compact, where the porosity is monotonically decreased or increased from the surface on account of the higher die wall friction, is also carried out in this chapter. To investigate the influences caused by the heterogeneities in initial porosity and composition, the heterogeneity maps for the thermophysical/chemical reactant parameters (such as density, heat capacity, thermal conductivity, and reaction yield) and the corresponding combustion parameters (such as propagation velocity and thickness of pre-heat zone) are generated.figures 11 and 12 show the percentage variations in density and thermal conductivity for the specimens with different heterogeneities in Ni composition and porosity, respectively. Since the density and the thermal conductivity at each node are calculated from the composition and the porosity; the variations in these parameters are found to correlate strongly with the changes in composition and porosity. In the ideal homogeneous specimen (0 % heterogeneities in Ni composition and porosity), the thermal conductivity and density remain as constants at all nodes, and variations in these parameters are zero. Density and thermal conductivity start to vary with the distance for the heterogeneous specimens. figure 11 shows that the variation in thermal conductivity along the specimen is increased to 26.7 % with the increase in the heterogeneity in porosity to 60%. Under such a variation, the thermal conductivity is varied between 145.5 and 210.9 J/(msK) [= 178.2 (homogeneous value)·(1 ± 26.7%/2)].

208

Hung-Pin Li

Figure 11. A map for the variations in thermal conductivity with the different heterogeneities in initial composition and porosity. The thermal conductivity for the homogeneous specimen is 178.2 J/(msK).

Figure 12. A map for the variations in density with the different heterogeneities in initial composition and porosity. The density for the homogeneous specimen is 3.87 g/cm3.

On the other hand, the variation in thermal conductivity is only increased to 15.5 % upon increasing the heterogeneity in Ni composition to 30%. A 60 % heterogeneity in porosity (= 25% x 60% = 15%) and a 30% heterogeneity in Ni composition (= 50 at. % x 30% = 15%)

Heterogeneous Combustion Synthesis

209

both lead to 15% change in porosity or composition, as shown in Table I. The heterogeneity map in Fig. 11 reveals that the heterogeneity in porosity has stronger effects on changing thermal conductivity than the heterogeneity in Ni composition for a given change percentage. A similar phenomenon is also found in the variation in density. figure 12 shows that the variations in density are enhanced to 20.0 % and 17.2 % as the heterogeneities are increased to 60% for porosity and 30% for composition, respectively. In addition, Eqs. (4) and (8) also show that the reaction yield and the heat capacity at each node are only influenced by the heterogeneity in initial Ni composition. Thus, Figures 13 and 14 show that the variations in heat capacity and reaction yield are correlated with the changes in heterogeneity in Ni composition, but independent of the changes in porosity. Figure 13 shows that an increase in 30% heterogeneity in Ni composition enhances the variation in heat capacity to 20.4 % as compared with the homogeneous specimen. As also seen in Eq.(4) and figure. 14, the maximum decrease in the reaction yield is calculated to be 15 % for the specimen with 30 % heterogeneity in Ni composition. A decrease in the reaction yield correspondingly reduces the exothermic heat of the combustion synthesis and propagation velocity of the combustion front. The heterogeneity maps in figures 11 – 14 show that the variations in composition and porosity change density, heat capacity, thermal conductivity, and exothermic heat, further influencing the reactivity at each node. The corresponding combustion parameters (e.g. propagation velocity and thickness of pre-heat zone) and reaction temperature profiles are expected to change with the heterogeneities during combustion synthesis.

Figure 13. A map for the variations in heat capacity with the different heterogeneities in initial composition and porosity. The heat capacity for the homogeneous specimen is 673.5 J/(kgK).

210

Hung-Pin Li

Figure 14. A map for the maximum decrements in reaction yield with the different heterogeneities in initial composition and porosity. The reaction yield for the homogeneous specimen is 100 %.

Figures 15-17 show the temperature profiles of combustion fronts at different time points along the specimens for different heterogeneities in composition and porosity. For the homogeneous specimen, the combustion front propagates at the velocity of 45.4 cm/s (figure 15). The value is higher than the experimental result because a high pre-exponential constant is used in the numerical calculation to illustrate the heterogeneous effects. As the heterogeneity in porosity is increased to 60%, it is noted from figure 16 that the propagation velocity is slightly increased to 46.3 cm/s. However, as the heterogeneity in initial Ni composition is increased to 30% and the heterogeneity in porosity is kept at 0 %, the propagation velocity has been significantly reduced to 38.1 cm/s, as shown in figure 17. In addition, the combustion temperature and propagation velocity are noted to dramatically change with the distance. The heterogeneity maps for the pre-heat zone thickness and propagation velocity with the heterogeneities in composition and porosity are further generated in this study. figure 18 shows the heterogeneity maps for the pre-heat zone thickness. The average zone thickness is noted to increase with the heterogeneity in Ni composition, whereas the average zone thickness is only slightly changed as the heterogeneity in porosity is increased. A previous study [17] has indicated that a narrow pre-heat zone results is normally referred to a higher oscillatory frequency of the combustion front, further increasing the reaction temperature and propagation velocity. Thus, the stability and propagation velocity of the combustion front are expected to increase when the pre-heat zone becomes narrower.

Heterogeneous Combustion Synthesis

211

Figure 15. Time variations of the combustion front temperature along the Ni + Al specimens. The interval time between two consecutive time steps (profiles) is 0.0005 s. The number 20 denotes the 20th time step (0.010 s) after ignition. The heterogeneities in composition and porosity are respectively 0 % and 0 %.

Figure 16. Time variations of the combustion front temperature along the Ni + Al specimens. The interval time between two consecutive time steps (profiles) is 0.0005 s. The number 20 denotes the 20th time step (0.010 s) after ignition. The heterogeneities in composition and porosity are respectively 0 % and 60 %.

The heterogeneity map for the propagation velocity (Figure 19) shows a continuous decrease in the propagation velocity upon increasing the heterogeneity in Ni composition. As compared with the homogeneous specimen, the propagation velocity is decreased within 16.2% for the specimen with 30% heterogeneity in composition. However, the propagation velocity is only changed in < 2 % even though the heterogeneity in porosity has been enhanced to 60 %.

212

Hung-Pin Li

Figure 17. Time variations of the combustion front temperature along the Ni + Al specimens. The interval time between two consecutive time steps (profiles) is 0.0005 s. The number 20 denotes the 20th time step (0.010 s) after ignition. The heterogeneities in composition and porosity are respectively 30 % and 0 %.

Figure 18. A map for the changes in thickness of pre-heat zone with the different heterogeneities in initial composition and porosity. The average thickness of pre-heat zone for the homogeneous specimen is 0.714 mm.

Heterogeneous Combustion Synthesis

Figure 19. A map for the changes in propagation velocity of combustion front with the different heterogeneities in initial composition and porosity. The propagation velocity for the homogeneous specimen is 50.4 cm/s.

Figure 20. A map for the standard deviations of propagation velocity of combustion front with the different heterogeneities in composition and porosity.

213

214

Hung-Pin Li

It has been shown that the variations in thermal conductivity and density caused by the heterogeneity in composition are smaller as compared with those caused by the same heterogeneity in porosity. However, the variation in the propagation velocity is larger for the specimen with heterogeneous composition distribution. As already discussed, this is because a change in the composition at each node directly changes the ratio of reactants far from the stoichiometric ratio. The reaction yield is correspondingly decreased, further reducing the exothermic heat of reaction. The combustion temperature has thus been significantly reduced with the distance and the variations in the propagation velocity are correspondingly decreased, as shown in figure 19. On account of the variations in the reaction yield and exothermic heat of reaction along the specimen, the combustion front is found to propagate in a succession of rapid and slow movements. Therefore, the standard deviation of propagation velocity is also found to increase with the heterogeneity in composition and porosity, as shown in figure 20. Again, the heterogeneity in composition has stronger effects on increasing the standard deviation of the velocity than the effects caused by the heterogeneity in porosity. The maximum standard deviation of propagation velocity is found when the 30 % heterogeneity in composition and 40% heterogeneity in porosity occur.

CONCLUSION The effect of heterogeneities in composition on combustion synthesis is investigated by using numerical simulation. It is found that the heterogeneities in reactants and diluent directly change the reaction yield and the thermophysical / chemical parameters of reactants, such as thermal conductivity and density, at each node. The combustion temperature and the propagation velocity of the combustion front are thus altered, as a result. However, the combustion temperature does not directly correlate with the composition variation in time. As the composition variation is accumulated to a certain level, the combustion temperature is thus changed periodically. The propagation velocity of the Ni-Al reaction without diluent is decreased from 73.4 cm/s to 70.9 cm/s and 66.7 cm/s as the heterogeneity in reactants is increased from 0% to 10 % and 20 %, respectively. As the diluent becomes heterogeneous but the heterogeneity in reactants is neglected in the numerical calculation, the variations of density and thermal conductivity are less than those caused by heterogeneity in reactants. Thus, the magnitudes of the variations of propagation velocity and combustion temperature for the compositions with heterogeneity in diluent are correspondingly decreased. The generated results also show that the reduction of exothermic heat and the changes in the heat capacity caused by the heterogeneities in composition are key factors in reducing the propagation velocity of the combustion front during a Ni + Al heterogeneous combustion reaction. The heterogeneity maps for the thermophysical/chemical reactant parameters (such as density, heat capacity, thermal conductivity, and reaction yield) and the corresponding combustion parameters (such as pre-heat zone thickness and propagation velocity) with the heterogeneities in composition and porosity have been explored in this chapter. These heterogeneity maps for the thermophysical/chemical reactant parameters have shown that the heterogeneities in initial composition and porosity influence the thermal conductivity and density, whereas the heat capacity and reaction yield are only influenced by the heterogeneity

Heterogeneous Combustion Synthesis

215

in initial composition. Such variations change the nature of propagation of a combustion front. The calculations show that the heterogeneity in initial porosity has a stronger effect on thermal conductivity and density when compared with the heterogeneity in initial Ni composition. However, the heterogeneity maps for the corresponding combustion parameters suggest that an increment in the pre-heat zone thickness occurs only when the heterogeneity in initial Ni composition increases. An increment in the pre-heat zone thickness has been reported to decrease the oscillatory frequency of the unstable combustion front, further decreasing the reactivity of the combustion reaction. Therefore, a heterogeneity map also reveals that the propagation velocity is significantly decreased with the heterogeneities in initial composition. From the knowledge of heterogeneity maps, the effects of heterogeneities in initial composition and porosity on combustion reaction can be acquired.

NOMENCLATURE Cp heat capacity of product (general form), kJ/kg/K E activation energy, kJ/kg Heterocomp heterogeneity in composition, % Heteroporosity heterogeneity in porosity, % Ko pre-exponential constant, (s-1 for zero order reaction) Q heat of reaction, kJ/kg Po original porosity, % Pj porosity at node j, % R gas constant, kJ/kg/K Ryield,j reaction yield at node j, % T temperature, K Tc combustion temperature, K To initial temperature, K Vs volume fraction of component (species) s, % Vi,j volume fraction of component i at node j, % Vio original (homogeneous) volume fraction of component i, % Xi,j molar fraction of component i at node j, % Xo original molar fraction of component (species), % Xi,j molar fraction of component (species) i at node j, % z dimensional coordinate, m d diameter of the specimen, m fR(j)random number at node j h surface heat transfer coefficient, J/m2/K/s t time, s ρ density, kg/m3 κ thermal conductivity (general form), kJ/m/K/s η fraction reacted Φ(Τ, η) reaction rate, 1/s

216

Hung-Pin Li

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26]

Li, H. P., Bhaduri, S.B., and Sekhar, J.A. Metall. Mater. Trans. A 1993, 24A, 251-261. Naiborodenko, Y. S., Itin, V. I., and Savitskii, K. V. Powder. Metall. Met. Ceram. 1970, 7(91), 562. Munir, Z. A. and Holt, J. B., J. Mater. Sci. 1987, 22, 710-714. Munir, Z.A., Am. Ceram. Bull. 1988, 67(2), 342-349. Munir, Z.A. and Anselmi-Tamburini, U. Mater. Sci. Reports 1989, 3, 277-365. Booth, F. Trans. Farad. Soc., 1953, 49, 272-281. Walton, J. D. Jr. and Poulos N. E. J. Am. Ceram. Soc. 1959, 42(1), 40-49. Li, H. P. Modelling Sim. in Mater. Sci. Eng. 2005, 13, 1331-1339. Li, H. P. Acta Mater. 2005, 53(8), 2405-2412. Li, H. P. J. Mater. Res. 2002 , 17(12), 3213-3221. Li, H. P. Mater. Sci. Eng. A 2003, 345(1-2), 336-344. Li, H. P. Mater. Sci. Eng. A 2005, 404(1-2), 146-152. Lakshmikantha, M. G.., Bhattacharys, A., and Sekhar, J. A. Metall. Mater. Trans. A 1992, 23A, 23. Lakshmikantha, M. G. and Sekhar, J. A., J. Mater. Sci. 1993, 28, 6403-6408. Lakshmikantha, M. G., and Sekhar, J. A. J. Am. Ceram. Soc. 1994, 77(1), 202. Subramanian V, Lakshmikantha M. G. , and Sekhar J. A. J. Mater. Res. 1995, 10(5), 1235-1246. Li H. P. Scripta Mater. 2003, 50(7), 999-1002. Li, H. P. Metall. Mater. Trans. A 2003, 34A(9), 1969-1978. Dey, G.. K., Arya A., and Sekhar J. A. J. Mater. Res. 2000, 15(1), 63-75. Merzhanov, A. G.. and Khaikin, B. I. Prog. Energy Combust. Sci. 1988, 14, 1-98. Li, H. P. Mater. Chem. Phys. 2003, 80(3), 758-767. Brain, I., Knacke, O., and Kubaschewski, O. Thermochemical Properties of Inorganic Substances; Springer-Verlag : New York, NY, 1973. Lide, D. R. CRC Handbook of Chemistry and Physics CRC : Boca Raton, FL, 1990. Brandes E. A., Brook G. B. Smithells Metals Reference Book; Butterworth-Heinemann Ltd.: Washington, DC, 1992. Naiborodenko, Y. S. and Itin, V. I. Combust. Explos. Shock Waves, 1975, 11(3), 293300. Li, H. P., and Sekhar, J.A. J. Mater. Sci. 1995, 30(18), 4628-4636.

In: Materials Science Research Horizons Editor: Hans P. Glick pp. 217-234

ISBN 978-1-60021-481-3 © 2007 Nova Science Publishers, Inc.

Chapter 8

RECYCLING OF ECOCOMPATIBLE TREATED RED MUD AND COMPOST FROM SS-MSW: EXAMPLES OF USE ON SEDIMENT AND MINE SOIL SAMPLES P. Massanisso1, E. Nardi, R. Pacifico, L. D’Annibale, C. Cremisini, and C. Alisi ENEA – C.R. Casaccia Via Anguillarese 301, 00060 Roma- Italy

ABSTRACT Ecological restoration of polluted areas is an increasing necessity for many countries around the world. Current technologies used to recover polluted soil and sediment are in general too costly. Recently, on-site approaches such as metal trapping and phytoremediation have attracted attention for their ability to meet criteria of economicity. Metal trapping is based on the diminution of metal mobility and availability as a result of applying soil amendments, for example particular industrial residues. Phytoremediation is an appealing environmental cleanup technology but a deeper understanding of the complex interactions in the soil-plant system is still needed. In this study, the effect of adding treated red mud (BauxsolTM - material with the potential to immobilise metal) on mine soil and on sediment (from a volcanic coastal lagoon in Southern Italy) and of adding both red mud and compost (produced from Source-Separated Municipal Solid Waste) on trace elements fractionation and mobility, have been investigated. Barley (Hordeum vulgare) was used as a plant model to follow any change in matrices phytotoxicity: seedlings were transplanted in pots containing the contaminated mine soil or sediment and a mixture of the investigated matrices with different percentages of treated red mud and compost. Plant growth was studied also by controlling the total protein content, biomass and enzyme activity. The knowledge of trace elements mobility and “speciation” in contaminated soils and sediments is an important requisite for any further environmental evaluation and these features can be evaluated through leaching tests or by "sequential extraction procedure". In this work, total concentration of selected trace elements, their fractionation

1

Corresponding author: [email protected]; tel +390630484935; fax +390630484525

218

P. Massanisso, E. Nardi, R. Pacifico et al. by sequential extraction procedure (BCR standardised) and leaching batch tests using a kinetic approach, were studied. The most evident result in the soil trials was that the utilization of amendments, used both separately and in a mixture, always improved the growth of barley plants. In particular, barley seedlings were practically not able to grow on the polluted mine soil and the simple adding of red mud resulted in a significant improvement in plant development. An even more drastic improvement was obtained with the addition of compost and compost plus treated red mud. In the sediment trials, the best yield in plant growth was obtained in the pot with the addition of treated red mud alone. The necessity of a delicate compromise between the maintaining of an acceptable plant viability and the control of metal mobility seems to be achievable through a careful balancing of the percentages of compost and red mud utilized as amendments.

INTRODUCTION Mine tailings and soils are characterized by high levels of heavy metals concentration, low pH reaction grade and low organic carbon content; sediments can act as a potential sink for contaminants and then become a secondary source of contamination. Ecological restoration of such polluted areas and sediments is now required by law in many countries around the world and there is a clear indication to improve processes resulting in benefit to agriculture or ecological improvement (EC, 1975; EC, 2006; e-CFR, 2006; US EPA, 2006). The remediation of sites and the recovery of sediment contaminated with toxic metals is a complex problem to be solved, and has an increasing economic relevance considering the huge amount of such large polluted areas. At the same time this problem is extremely challenging due to the fact that, unlike most of the organic compounds, metals cannot be degraded and, consequently, the full cleanup requires their complete removal and treatment (Lasat, 2002). The technologies used to reach this target in mine soil sites are too costly. In addition, they may cause extra risks for the workers and produce secondary wastes. Moreover, engineering-based technologies are often environmentally invasive and do not permit natural reshaping of the environment (Lombi et al., 2002;). In-situ technologies, that necessitate low inputs and are low cost, are increasingly required by the environmental operators to meet the needs for soil remediation and community acceptance. When the contamination level in the sediment is considered too high, relocation at sea is not allowed and alternative management techniques need to be considered. The recovery of sediments by means of recycling, re-use or reclamation or any other process with a view to extracting secondary raw materials through processes which do not endanger human health or harm the environment, is a clever approach which has the great advantage of reducing the landfilling of contaminated sediments that, up till now, has been the most common method of disposal of contaminated dredged material. For contaminated dredged material, direct use or ways of processing and subsequent beneficial use are available; among the possible options it is important to remember the direct use of the sediments for soil improvement of agricultural land and the utilization of treated sediments in landfarming (ESPO, 2006). Recently, approaches that involve processes or methods which do not harm the environment, such as metal trapping and phytoremediation, have attracted attention for their

Recycling of Ecocompatible Treated Red Mud and Compost from SS-MSW...

219

ability to meet the criteria of clean and low-cost processes that can be utilized on-site, as in the case of mine soils remediation, and off-site, as in the case of dredged materials. Metal trapping is based on the diminution of metal mobility and availability as a result of applying amendments, for example particular industrial residues. During recent years, treated red mud residues, by-products of alumina production, have been widely used for their metal trapping capability in many environmental remediation activities (Qiao and Ho, 1997; Summers and Pech, 1997; Genç-Fuhrman et al., 2004; Brunori et al., 2005 AandB). Added to contaminated solid matrices, they can neutralise low pH and reduce metals mobility, through different chemical-physical mechanisms (including the increase of available adsorption sites). Phytoremediation (use of higher plants to remove -phytoextraction and phytoaccumulation- or to immobilize –phytostabilization- contaminants from polluted matrices) is an appealing environmental clean-up technology, but probably a deeper understanding of the complex interactions in the soil-plant system is needed to allow for "safe" metal translocation and accumulation in plants (Tyler, 2004; Ying, 2005). Due to the low organic carbon content in mine tailing soils and sediments, this action is necessary to restore their organic fraction and to reconstitute their structure in order to obtain a “healthy” soil or sediment. This step is also essential with the objective of a successive phytoremediation based approach. SourceSeparated Municipal Solid Waste (SS-MSW) compost, mainly produced as fertiliser for agriculture, could be used to this aim but total concentration, speciation and fractionation of heavy metals in compost-amended soils should be carefully evaluated for predicting elemental mobility and phytoavailability (Zheljazkov and Warman, 2004). The utilization of plants can also be seen as a test of toxicity for a contaminated soils or sediments (Massanisso et al., 2006). In the present work, different soils and sediments systems were studied: a contaminated soil or sediment (from an abandoned mine and from a lagoon), soil or sediment and treated red mud (BauxsolTM), soil and compost, soil or sediment and compost plus BauxsolTM. Barley seedlings were cultivated in these soils and differences in their development were studied, controlling several parameters after 30 and 60 days for soil and after 30 days for sediment: biomass, protein content and enzyme activity. Considering that trace elements are often important from a toxicological point of view, the determination of their total concentration may only grant information about the enrichment of soils and sediments and does not provide adequate information about their environmental impact (especially in terms of mobility, bioavailability and potential toxicity). The ecological relevance of trace elements in the water/solid matrix system is due to their mobility more than to their total content (Rauret, 1998; Guevara-Riba et al., 2004). Therefore, the knowledge of trace elements mobility and "fractionation" in contaminated solid matrices is an important requisite for any further environmental evaluation. In this work two approaches were used. The first one is the application, on both the solid matrices investigated, of a "sequential extraction procedure" based on the step by step fractionation of metals from samples, by using different reagents or extractants: the fraction of each metal determined depends on the extractants and the operating conditions under which the extraction is carried out (Mester et al., 1998). The well-standardised BCR 3-step sequential extraction procedure (Rauret et al., 2000) (a simplification of the first, and most applied, five-step sequential extraction procedure proposed and published by Tessier et al., 1979) was used: the fractions of metals established by the 3-step procedure are the exchangeable/carbonatic fraction, the easily reducible fraction, and the oxidable fraction.

220

P. Massanisso, E. Nardi, R. Pacifico et al.

The second approach is the application of the results of a leaching experiment to the six different systems of soil studied, using EDTA as extractant (Hage and Mulder, 2004; Song and Greenaway, 2004; Brunori et al. 2005 B). The use of a relatively non-specific extractant (EDTA) therefore suggests a kinetic approach because measurements of trace elements extracted at equilibrium cannot be related to their speciation. Using an excess of relatively non-specific extractant, the leaching reactions can be regarded as a non linear mathematic model of the type: y = a (1- e -k1t) + b (1- e -k2t) +…..+n (1-e-knt) The kinetic approach permits to subdivide into a labile metal fraction (quickly extracted) and a non-labile metal fraction (less quickly extracted) the trace metals extracted in the leaching test. This objective requires a non-linear equation with one or two components and an estimation by non-linear regression of the constant of the equation. The non linear equation can be written as: •

One Component: Y = a + b (1-e-k2t)

•

Two Components: Y = a (1-e-k1t) + b (1-e-k2t)

where y represents the amount of metal extracted at time t; a and b represent the labile and not-labile amounts, respectively; k1 and k2 are the kinetic constants associated with a and b for a given metal, respectively. The selection between the two models depends on the leaching rate: high leaching rate of labile fraction, associated with high k1, are better described by one component model. Results were analysed in the perspective of the use of both BauxsolTM and compost in the contaminated sites remediation, assessing the utilization of barley plants as indicator of contaminated soil or sediment toxicity, and studying if the utilization of a proper combination of the amendment(s) is able to recover the matrices for a following step of phytoremediation. Finally, the study of the elements mobility by sequential extraction procedure and kinetic approach will give information regarding the toxicity of these matrices in relation to the elements leachability.

MATERIAL AND METHODS Samples Soil samples (MT soil) were collected in the area around a mine dump at about 70 km (North-West) from Rome (Italy) and was chosen for its high levels of toxic metals concentration and the relatively low reaction grade (pH 4.5) and low organic carbon concentration. Sediment samples were collected, with a Peterson grab sampler, from Fusaro and Lucrino lagoons on the western coast of the South of Italy, within the polygenic volcanic complex of the Phlegrean Fields.

Recycling of Ecocompatible Treated Red Mud and Compost from SS-MSW...

221

Commercial Amended Soil for plant cultivation, "UNIVERSALE" from Agri Flor®, loc. Villa Pitignano, Ponte Felcino, Perugia, Italy. Main components: light-brown/dark-brown peat, heather earth, mixed vegetable substrates, mixed manure, volcanic sand, pumice, clays. Average composition: organic carbon 25%, humic and fulvic acids 7%, total N 2,5%, organic N 2%, C/N = 10, copper 5 mg/kg, zinc 27 mg/kg, conductivity 600 μS/cm (1:5 in water), pH 7, humidity 30%. Red mud sample was kindly supplied by Virotec Italia (Virotec International Ltd. Australia). Virotec optimised the process of red mud treatment with seawater and patented this technology and several products with the name BauxsolTM Compost sample was obtained from a production plant from the area of Maccarese (Italy).

Preparation of Soil Samples The following soil samples were used for the experiments. A B C D E F

100% abandoned mine toxic metal contaminated soil (MT soil) 80% MT soil + 20% BauxsolTM 80% MT soil + 10% compost + 10% BauxsolTM 80% MT soil + 20% compost 60% MT soil + 20% compost + 20% BauxsolTM 100% Commercial Amended Soil

The soil samples were obtained by thoroughly mixing, without modifying the physical characteristics of the single components, until an acceptable homogeneity was reached. Eighteen pots (φ = 17 cm, h = 14 cm) were filled with 1.6 kg of soil samples (A to F) in triplicate. Barley (Hordeum vulgare L. cv. Adonis) seeds were soaked in aerated tap water and grown on moistened filter paper for 3 days in the dark at 25°C. The seedlings were then transplanted in the pots (15 per pot) containing the different soil samples.

Preparation of Sediment Samples The following sediment samples were used for the experiments where differences in plants development were studied. S1 100% lagoon sediment S2 80% lagoon sediment + 20% BauxsolTM S3 70% lagoon sediment + 25% compost + 5% BauxsolTM. The following sediment samples were used for the experiments where the influence on the plants growth of a toxic solution added to the samples was studied. C1 100% lagoon sediment.

222

P. Massanisso, E. Nardi, R. Pacifico et al. C2 65% lagoon sediment + 15% compost + 20% Bauxsol.TM

The sediment samples were obtained by thoroughly mixing, without modifying the physical characteristics of the single components, until an acceptable homogeneity was reached. The same condition as for the soil samples were adopted for the study on barley plants. It is worth noting that to evaluate the variation in plant growth between the two tests (soil and sediment) the different time of the year in which the experiments have been carried out has to be taken in account: early spring for soil, late summer for sediment. Since barley has an optimum growth in a mild temperature, the plants grown in summer generally show a lower yield in biomass. Although this fact hinders the possibility of comparing the two tests, the comparison within the same set of experiments is still significant.

Determination of Elemental Total Content Aliquots of soils or sediments, BauxsolTM, compost and mixtures of 0.5 g, precisely weighted, were digested with a mix of conc. HNO3 (5 mL, 69%), conc. HF (2 mL, 30%), conc. HClO4 (1 mL, 70%) and with H2O2 (2 mL, 30% v/v) in the TFM vessels with a microwave system. The work program used in the microwave digestion was as follows: 5 min at 250 W of power, 10 min at 400 W, 10 min at 600 W and 5 min at 250 W. The microwave digestion was followed by an open vessel procedure: the samples were first slowly evaporated nearly to dryness in PFA vessels, and subsequently the residues were redissolved with 1 mL of HClO4 and the solutions again evaporated nearly to dryness. Hence the residues were submitted twice to an analogous treatment utilizing 1 mL of HNO3. Finally, 2 mL of HNO3 were added to each sample and the resulting solution was completely transferred into a 50 mL volumetric flask and made up to the final volume with ultrapure water.

pH of Soils pH was measured at the beginning of the experiment following the ISO 10390 procedure (ISO, 2005). The same checks were carried out after 30 days and at the end of the experiment (60 days) on representative samples from each pot.

Sequential Extraction The BCR three-step sequential procedure (table 1) was applied on soils and sediment, BauxsolTM, compost and mixtures. The residues from step 3 were treated with the same procedure used for the determination of elemental total content. The analytical performance of the laboratory in the sequential extraction procedure was evaluated by analysing the certified reference material BCR 483 and BCR 601: results were in good agreement with those reported in the certificate.

Recycling of Ecocompatible Treated Red Mud and Compost from SS-MSW...

223

Table 1. BCR 3-steps sequential extraction procedure STEP

EXTRACTANT

A: exchangeable and weakly adsorbed fraction

CH3COOH

B: reducibile fraction (bound to Fe and Mn oxides) C: oxidable fraction (organically- and sulphide-bound) Residual fraction

NH2OH*HCl H2O2, then CH3COONH4 HNO3, H2O2 and HF

Leaching Test For kinetic studies, 3 g of each soil were utilized with EDTA (0.05 mol L-1) and with a L/S ratio of 10. The tests were performed on soils, Bauxsol, compost and soil mixtures, collecting samples at the beginning of the experiment, after 30 days and at the end of the experiment (60 days). Polyethylene tubes containing samples and the extractant (EDTA) were stirred using an end-over-end shaker at a speed of 30 rpm for a given time, different for each tube: 15, 30, 60, 100, 150, 240, 360 and 960 minutes. At the end of the chosen mixing time, each tube was removed and the extract was separated from the solid residue by centrifugation at 400 rpm for 10 min and successive filtration with filter millex HA 0.45 μm (Millipore). For each sample, 15 mL of the filtrate (after the addition of 150 μL of nitric acid) were kept at 4°C until analysis. The non-linear regression study of the leaching results was carried out using SigmaPlot 8.0, a software package produced by SPSS Ltd.

Protein Extraction Leaf samples were collected at different stages (10, 30 and 60 days for soils samples; 30 days for sediments samples) of treatment and promptly ground in 10 volumes of cold (4°C) extraction buffer containing 100 mM Tris-HCl pH 8, 1 mM EDTA, 10 mM DTT and 5 mM phenyl-methylsulfonyl fluoride (PMSF, a protease inhibitor). The homogenates were centrifuged at 12,000 g for 20 min at 4°C, and the supernatants were assayed for total protein content using the Bradford method (Bradford, 1976) with bovine γ-globulin as standard. Since the imposition of biotic and abiotic stress condition can give rise to excess concentrations of active oxygen species, resulting in oxidative damage at the cellular level, antioxidant enzymes such as superoxide dismutase (SOD), peroxidase (POX) and catalase (CAT) play an important role in eliminating H2O2 and therefore they are good stress indicators. For this reason, we chose to investigate the levels of peroxidase activity in the barley plants, as a parameter of stress conditions. Total proteins from each extract were then separated by 10% non-denaturing PAGE on a Mini-protean apparatus (Bio-Rad Laboratories, CA, USA). Gels were loaded with equal amounts of total proteins (50 μg) per lane; after electrophoresis, peroxidases activity in the non-denaturing gels were visualized by soaking the gel in 50 ml of Na acetate 50 mM pH 5

224

P. Massanisso, E. Nardi, R. Pacifico et al.

containing CaCl2 20 mM for 20 min, then adding 2.5 ml of a staining solution (0.2 ml eugenol, 10 mg 3A9EC dissolved in 10 ml acetone) and 0.1 ml of H2O2 to catalyse the reaction. After the colour development was completed, band intensity was visually recorded.

TESTS ON THE UNTREATED AND AMENDED SOILS Chemical parameters of soils (A and F), Bauxsol, compost and soil mixtures (B, C, D, E) are reported in table 2. The soil from the abandoned mine (A) shows a relatively acidic pH, low organic carbon content and high concentration of toxic elements, particularly As and Pb. Bauxsol has already been characterised in several previous literature papers (McConchie et al, 2002; Brunori et al, 2005 A) showing a pH around 10 and a low content in toxic elements. The compost sample shows not negligible concentration of Cu and Zn, but significantly lower than the soil A. Data on the different soil mixtures are similar to the theoretical values derivable from data of the single components, and no significant variation was observed in the course of the experiment (60 days). However, it should be considered that, even if a careful mixing was operated, the necessity to preserve the structural characteristics of the single components did not permit the availability of an "analytical" grade of homogeneity of the mixtures. This point should be always considered in the following comments and evaluations of the results. In figure 1, plants after 30 days of growth at different soil conditions are shown. Plant growth was considerably reduced in the soil from the mining area (A) in comparison with the plants grown on the other soil mixtures. This evidence was confirmed by the observation of the root apparatus: roots were practically undeveloped in soil A and the comparison with those of the plants developed in the other soil mixtures is considerably demonstrative of the toxicity of mine soil. The simple adding of 20% of BauxsolTM (B) allowed a significant improvement in the root apparatus and in the plant development. The adding of 10% compost plus 10% BauxsolTM (C) produced a further improvement in the plant development, evidencing the fundamental role of the organic carbon, even more clear in soil from the mining area added with 20% compost (D, E). These data are in agreement with the total protein content and peroxidases activity shown in table 3: in the early stage (10 days after transplanting) the barley shoots grown on soils A and B show a significantly lower amount of proteins when compared with the other soil mixtures. After 30 days the total amount of protein in the plants grown in soil A shows a further reduction if compared with the control (F); on the other hand, plants grown in soils B, C, D and E show protein content similar to the control and the peroxidase activity is high in soil A and B, indicating the onset of serious stress conditions; after 60 days the soil B produced a further reduction in protein content and an increase in peroxidases activity indicating the stress persistence; for soils C, E and D no major effect is observed, while soil A confirms its heavy toxicity. The same behaviour can be evidenced on the biomass changes: the biomass measured in soil B is 5 times higher than the biomass in soil A, showing a appreciable reduction in soil toxicity only by the addition of treated red mud. A further reduction in soil toxicity is shown in soil C and E where the biomass increases 15 times when compared with the biomass in soil A, probably due to the synergic action of treated red mud and compost. The use of compost

Recycling of Ecocompatible Treated Red Mud and Compost from SS-MSW...

225

alone as amendment in soil D produces a biomass increase of 20 times with respect to the biomass in soil A. It is worth noting that when we compare the organic carbon data in table 2 with the biomass data in table 3, a correlation value of 0.94 is found, indicating that the amount of organic carbon is important in attenuating the soil toxicity.

Figure 1. Barley plants after 30 days of growth in different soil samples.

As, Pb and Zn, which were present in moderately high concentrations in the soil (about 600, 700 and 400 mg kg-1, respectively; see table 2) and Cd which was present in low but environmentally significant concentration (according to the Italian Decree on the recovery of contaminated sites: IMD, 1999), were selected as the elements to be studied both in the leaching test and sequential extraction experiments. The distribution in the different fractions obtained with the BCR sequential extraction procedure, for each of the investigated element, is very similar in the soil and in all the treated mixtures. Furthermore, there are not significant differences in the distribution of the samples analysed before the transplanting and at the end of the experiment. As an example, the results of the BCR procedure, expressed as percentage of element removed from the soil in steps A, B, C and residue are reported in figure 2 for the soil A and E, before the transplanting (t = 0, figure 2a, b) and at the end of the test (60 days, figure 2c, d).

226

P. Massanisso, E. Nardi, R. Pacifico et al.

Table 2. Organic carbon, pH and selected metals concentration in the different soil samples before the transplanting: A) 100% abandoned mine toxic metal contaminated soil (MT soil); B) 80% MT soil + 20% BauxsolTM ; C) 80% MT soil + 10% compost + 10% BauxsolTM; D) 80% MT soil + 20% compost; E) 60% MT soil + 20% compost + 20% BauxsolTM ; F) 100% commercial amended soil (CA soil)

Organic carbon pH As Cd Cu Mn Pb Zn

Unit

A

B

C

D

E

F

bauxsol

compost

(%)

0.61 4.5 569 1.0 283 130 654 383

0.47 7.9 470 0.8 230 114 579 295

3.22 7.4 477 0.8 225 159 607 301

6.01 6.3 465 0.9 240 200 599 313

5.54 8.0 379 0.7 190 197 474 263

17.26 6.6 14.1 0.7 155 706 120 333