DEVELOPMENT AND CHARACTERIZATION OF L-TYROSINE BASED POLYURETHANES FOR TISSUE ENGINEERING APPLICATIONS. A Dissertation.

Transcription

1 DEVELOPMENT AND CHARACTERIZATION OF L-TYROSINE BASED POLYURETHANES FOR TISSUE ENGINEERING APPLICATIONS A Dissertation Presented to The Graduate Faculty of The University of Akron In Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy Debanjan Sarkar August, 2007

2 DEVELOPMENT AND CHARACTERIZATION OF L-TYROSINE BASED POLYURETHANES FOR TISSUE ENGINEERING APPLICATIONS Debanjan Sarkar Dissertation Approved: Accepted: Advisor Dr. Stephanie T. Lopina Department Chair Dr. Lu-Kwang Ju Committee Member Dr. H. Michael Cheung Dean of the College Dr. George K. Haritos Committee Member Dr. Bi-min Zhang Newby Dean of the Graduate School Dr. George R. Newkome Committee Member Dr. Stanley E. Rittgers Date Committee Member Dr. Jun Hu ii

3 ABSTRACT Natural amino acid based synthetic polymers have limited applicability as biomaterial due to several unfavorable material and engineering properties. This has led to the development of a new class of polymers known as pseudo poly(amino acid)s. Several L-tyrosine based pseudo poly(amino acid)s have been developed and characterized extensively for biomaterial applications. Desaminotyrosine tyrosyl hexyl ester (DTH), a diphenolic dipeptide molecule developed from L-tyrosine and its metabolite, is used to synthesize amino acid based pseudo polymers with improved physical and chemical properties. Polyurethanes are extensively used as biomaterials due to excellent biocompatibility and the ability to tune the structure for a wide range of properties. The uses of polyurethanes are mainly focused on biostable implants and biomedical devices. But polyurethanes have shown their susceptibility to degradation under the conditions of their performance. The use of polyurethanes for tissue engineering applications emerged mainly due to the degradability of the polyurethanes. Biodegradable polyurethanes with degradable linkages are developed by altering their structure and composition. The aim of the research presented in this dissertation is focused on developing L- tyrosine based polyurethanes for biomaterial applications including tissue engineering. L- tyrosine based polyurethanes can be developed by using DTH as the chain extender with iii

4 different polyols and diisocyantes. The use of amino acid based component will improve the biocompatibility and biodegradability of the polymers for tissue engineering application. In addition, by using the different components, the structure and composition of the polyurethanes can be altered to achieve a range of properties that are pertinent to biomaterial applications. This research describes the design, synthesis and characterization of L-tyrosine based polyurethanes with DTH as the chain extender. The polyurethanes are extensively characterized for different bioengineering properties, including surface characteristics, water absorption, degradation characteristics, and controlled release along with other important chemical, physical, thermal and mechanical characterizations. The structure-property relationships of the polyurethanes were investigated by developing a library of polyurethanes with different polyol and diisocyante. This library provides an important tool to design polyurethanes with relevant properties for biomaterial application. The effect of structure and composition of these polyurethanes in determining the material properties were studied in detail. In addition, blends of the polyurethanes were studied as an alternative to adjust different properties according to the requirements. The results show that L-tyrosine based polyurethanes are potential candidates for biomaterial applications including tissue engineering. The material characteristics are strongly dependent on the polyurethane structure and composition, and therefore a wide range of properties can be achieved by altering the structure and composition. iv

5 DEDICATION To all of my teachers, who made me what I am today, And Especially to my advisor, Dr. Stephanie T. Lopina A teacher affects eternity; he can never tell where his influence stops. -Henry Adams v

6 ACKNOWLEDGEMENTS The ups and downs that you endure in your graduate career is the part of the journey towards your destination. The successes and the failures that I have experienced as a graduate student will be the source of my motivation in the days to come. This period of my life and career at The University of Akron has made my dreams come true. This dissertation marks the end of a long and eventful period of my life for which there are many people that I would like to acknowledge for their support. I am fortunate to have Dr. Stephanie T. Lopina as my advisor. As an advisor, her determination and dedication against all the odds of her life is an inspiring example for me. It is her continuous guidance, support and encouragement that made this research and this dissertation a complete one. I am grateful to Dr. H. Michael Cheung, Dr. Bi-min Zhang Newby, Dr. Stanley E. Rittgers, and Dr. Jun Hu for serving on my committee and for their valuable suggestions and advices. My special thanks go to Dr. Hu with whom I spent the initial days of my graduate research in his lab. I am thankful to my Department of Chemical & Biomolecular Engineering for providing me the financial support and all other help to complete my graduate studies. I gratefully acknowledge the assistance provided by all the Faculty members and the staff of the Department. Thank you to Mr. Frank Pelc for providing the necessary help in the set-up of the lab. I also thank the Department of Chemistry for using the NMR and FTIR vi

7 facilities in this research. Special thanks to the Department of Polymer Engineering for the use of Instron and SEM facilities. Jon Page from the Department of Polymer Science is acknowledged for the help in GPC analysis. I also gratefully acknowledge the assistance provided by Michelle Miller of the Writing Lab to bear the pain of proof reading my dissertation. I express my thanks to Roulei and Fen in Dr. Chueng s lab, for the DSC and TGA analysis. I also thank Feng in Dr. Newby s lab for the assistance in contact angle measurements. I gratefully acknowledge the assistance and continuous friendship of all my research group members who made my life in the lab much more enjoyable. Special thanks to Peter, for his constant support and help. I gratefully acknowledge his hard work in helping me with the mechanical characterizations and the SEM analysis in this research. Words are not enough to describe the sacrifice of my parents who supported me in each and every aspect of my life. Their unflinching support and proper guidance have helped me to get to this point. I am thankful to my brother and all other relatives back at home for their help in this endeavor. Friends of old and friends recently acquired all need to be applauded. It is their persistent companionship that made my away-from-home life easier and memorable. Finally, it is my beloved wife Sukanya. I would especially thank her for the endurance and the patience in bearing the hardships of the graduate student life. Her love, care, support and everything she has done for me have made my life easier and enjoyable. vii

8 TABLE OF CONTENTS LIST OF TABLES.. LIST OF FIGURES. Page xiv xvi CHAPTER I INTRODUCTION Objective Layout of dissertation... 3 II BACKGROUND Tissue engineering and polymers Amino acid based polymer Polyurethanes as biomaterials Technical approach III SYNTHESIS AND CHARACTERIZATION OF L-TYROSINE BASED POLYURETHANES Experimental Synthesis of polymer DTH Synthesis Synthesis of Polyurethanes viii

9 3.1.2 Characterizations of polymer Structural Characterizations Thermal Characterizations Mechanical Characterization Results and Discussion Polymerization Reaction NMR Characterizations FT-IR Characterizations Molecular Weight of Polyurethanes Solubility of the Polyurethanes Thermal Characterizations Mechanical Characterizations IV 3.3 Conclusion CHARACTERIZATION OF L-TYROSINE BASED POLYURETHANES FOR BIOMATERIAL APPLICATIONS Experimental Preparation of Solvent Cast Films Water Contact Angle Water Vapor Permeation Release Study Water Absorption Hydrolytic Degradation Oxidative Degradation ix

10 4.1.8 Enzymatic Degradation Results and Discussion Water Contact Angle Water Vapor Permeation Release Characteristics Water Absorption Hydrolytic Degradation Oxidative Degradation Enzymatic Degradation Conclusion V STRUCTURE-PROPERTY RELATIONSHIP OF L-TYROSINE BASED POLYURETHANES Experimental Synthesis of polyurethane and casting of films Structural Characterizations Thermal Characterizations Mechanical Characterizations Water Contact Angle Water Vapor Permeability Water Absorption Hydrolytic Degradation Release Characteristics Statistical Analysis x

11 5.2 Results and Discussion Molecular Weight FTIR Analysis Thermal Characterizations Mechanical Properties of Polyurethanes Water Contact Angle Water Vapor Permeation Water Absorption Characteristics Hydrolytic Degradation Release Characteristics Conclusion VI CHARACTERIATION OF L-TYROSINE BASED POLYURETHANE BLENDS Experimental Fabrication of Blends Spectral Characterizations Microscopic Characterization Thermal Characterization Mechanical Characterizations Water Contact Angle Water Absorption Hydrolytic Degradation Statistical Analysis. 165 xi

12 6.2 Results and Discussion H NMR Characterization FTIR Characterization SEM Analysis Thermal Characterizations Mechanical Characterizations Water Contact Angle Water Absorption Characteristics Hydrolytic Degradation Conclusion VII CONCLUSIONS Summary Design, Synthesis and Characterization of L-tyrosine based Polyurethanes Characterization for Biomaterial Properties Structure-Property Relationship Blend Characterizations Principal Achievements Future Work REFERENCES APPENDICES xii

13 APPENDIX A. APPENDIX B. STATISTICAL ANALYSIS OF POLYURETHANE PROPERTIES BY ANOVA WITH MINITAB SOFTWARE STATISTICAL ANALYSIS OF POLYURETHANE BLEND PROPERTIES BY ANOVA WITH MINITAB SOFTWARE xiii

14 LIST OF TABLES Table Page 2.1 Polymers in tissue engineering Composition of the polyurethanes Molecular weight of the polyurethanes Solubility features of the polyurethanes Mechanical properties of the polyurethanes Water vapor permeation of the polyurethanes Value of fitted parameters k and n ATR-FTIR peaks of PEG-HDI-DTH ATR-FTIR peaks of PCL-HDI-DTH Polyurethane composition Weight fraction of different segments in the polyurethanes Representative molecular weight of polyurethanes Mechanical properties of polyurethanes (mean ± SD, n = 5) p-values for the mechanical properties of polyurethanes p-values for contact angle of the polyurethanes Water vapor permeability of polyurethanes (mean ± SD, n = 3) p-values for water vapor permeation of the polyurethanes p-values for water absorption of the polyurethanes xiv

15 5.10 p-values for mass loss (hydrolytic degradation) of the polyurethanes p-values for percent release of the polyurethanes Fitted values of k and n Formulation of blends Composition of polyurethane blends from 1 H-NMR Mechanical properties of the blends and polyurethanes p-values for the mechanical properties of the blends Contact angle of the blends and polyurethanes p-values for the contact angle of the blends p-values for the water absorption of the blends p-values for the hydrolytic degradation of the blends Comparison of thermal characteristics of tissue engineering polymers Comparison of mechanical properties of biological tissues with L-tyrosine based polyurethanes Comparison of contact angle and water absorption of L-tyrosine based polymers Comparison of hydrolytic degradation for tissue engineering polymers and L-tyrosine based polyurethane xv

16 LIST OF FIGURES Figure Page 2.1 Concept of tissue engineering Structure of L-tyrosine and its metabolites Structure of Desaminotyrosyl tyrosine hexyl ester (DTH) Structure of polyurethanes Scheme for synthesis of polyurethanes Components of L-tyrosine based polyurethanes Scheme for DTH Synthesis Structure of L-tyrosine based polyurethanes H NMR of PEG-HDI-DTH 28 1 H NMR of PCL-HDI-DTH C NMR of PEG-HDI-DTH C NMR of PCL-HDI-DTH FT-IR of L-tyrosine based polyurethanes FT-IR analysis of the components, prepolymer and polyurethane DSC heating curves of L-tyrosine based polyurethanes TGA analyses of L-tyrosine based polyurethanes Representative stress-strain curve of L-tyrosine based polyurethanes xvi

17 4.1 Water contact angle on PEG-HDI-DTH surface (A) Advancing mode (B) Receding mode Water contact angle on PCL-HDI-DTH surface (A) Advancing mode (B) Receding mode Water contact angle of L-tyrosine based polyurethanes Contact angle hysteresis of L-tyrosine based polyurethanes Plot of mass of water vapor permeated against time Structure of p-nitroaniline Plot of fractional release of p-nitroaniline versus square root of time Initial release characteristics of p-nitroaniline Fitted curves for (A) PEG-HDI-DTH and (B) PCL-HDI-DTH Water absorption of polyurethanes with respect to time Comparison of water absorption (17 hours) Effect of water absorption on dimension for PEG-HDI-DTH Effect of water absorption on dimension for PCL-HDI-DTH Mass loss of L-tyrosine based polyurethanes during hydrolytic degradation in PBS (ph 7.4) at 37 C Regression analyses for mass loss of L-tyrosine based polyurethanes Plot of mass loss rate with time of L-tyrosine based polyurethanes Different urethane linkages present in the polyurethane Effect of ph on hydrolytic degradation of PEG-HDI-DTH Effect of ph on hydrolytic degradation of PCL-HDI-DTH 76 xvii

18 4.20 FTIR spectra of PEG-HDI-DTH before and after 7 and 22 days of oxidative degradation Subtraction of spectra for PEG-HDI-DTH Change in CH 2 stretch intensity of PEG-HDI-DTH Degradation of PEG-HDI-DTH in CoCl 2 /H 2 O 2 at 37 C Effect of strength of H 2 O 2 in degradation of PEG-HDI-DTH (for 1617 and 1577 cm -1 normalized to 1658 cm -1 ) Mechanism of oxidative degradation of PEG-HDI-DTH FTIR spectra of PCL-HDI-DTH before and after 7 and 22 days of oxidative degradation Subtraction of spectra for PCL-HDI-DTH Degradation of PCL-HDI-DTH in CoCl 2 /H 2 O 2 at 37 C Effect of strength of H 2 O 2 in degradation of PCL-HDI-DTH (for 1640 and 1533 cm -1 normalized to 1167 cm -1 ) Mechanism of oxidative degradation of PCL-HDI-DTH FTIR analysis of residue of oxidative degradation (from solution) of L- tyrosine based polyurethanes SEM images of polyurethane surface A. Control PEG-HDI-DTH, B. PEG-HDI-DTH after 22 days C. Control PCL-HDI-DTH, B. PCL-HDI- DTH after 22 days for oxidative degradation in CoCl 2 /H 2 O 2 at 37 C Schematic representation of oxidative degradation of L-tyrosine based polyurethanes Enzyme activity measurements for free enzyme and in presence of 93 polymer at 37 C in PBS (ph 7.4) 4.35 Mass loss of L-tyrosine based polyurethanes with time due to enzymatic action FT-IR spectra of PEG-HDI-DTH and PCL-HDI-DTH before and after enzymatic degradation. 96 xviii

19 4.37 Comparison of mass loss between enzymatic and hydrolytic degradation Chemical structure of 1,4 cyclohexane dimethanol (CDM) Comparison of mass loss of polyurethanes from tyrosine based chain extender and non-amino acid based chain extender under enzymatic condition FTIR analysis of residue of enzymatic degradation (from solution) of L- tyrosine based polyurethanes SEM images of polyurethane surface after 6 days A. Buffer mediated PEG-HDI-DTH, B. Enzymatically degraded PEG-HDI-DTH C. Enzymatically degraded CDM based polyurethane with PEG soft segment D. Buffer mediated PCL-HDI-DTH E. Enzymatically degraded PCL-HDI-DTH F. Enzymatically degraded CDM based polyurethane with PCL soft segment [enzymatic degradation in α-chymotrypsin in PBS (ph 7.4) at 37 C and buffer mediated degradation in PBS (ph 7.4) at 37 C] Schematic representation of enzymatic degradation of polyurethanes Components used in L-tyrosine based polyurethanes FT-IR absorbance spectra of polyurethanes (A) Series based on different molecular weight of PEG (B) Enlarged in the region and 1200 cm FT-IR absorbance spectra of polyurethanes (A) Series based on different molecular weight of PCL (B) Enlarged in the region cm FT-IR absorbance spectra of polyurethanes of series based on different diisocyanates Hydrogen bonding interactions in the polyurethanes Phase morphology of polyurethanes DSC thermograms of polyurethanes (A) Series based on different molecular weight of PEG and PCL (B) Series based different diisocyanates TGA analyses of L-tyrosine based polyurethanes xix

20 5.9 Representative stress-strain curves of L-tyrosine based polyurethanes Advancing and receding water contact angle of polyurethanes (mean ± SD, n = 15) Contact angle hysteresis of L-tyrosine based polyurethanes Plot of water vapor transmitted against time of L-tyrosine based polyurethanes Water absorption of L-tyrosine based polyurethanes Hydrolytic degradation of L-tyrosine polyurethanes in PBS (ph 7.4, 37 C) (A) Series based on different molecular weight of PEG and PCL (B) Series based on different diisocyanates (mean ± SD, n = 4) Release of p-nitroaniline from L-tyrosine based polyurethane matrices in PBS (ph 7.4, 37 C) (A) Series based on different molecular weight of PEG and PCL (B) Series based on different diisocyanates (n = 4 error bars are omitted to make it clear) Curve fitting for release of p-nitroaniline from L-tyrosine based polyurethane matrices in PBS (ph 7.4, 37 C) Scheme for fabricating films of polyurethane blends Representative 1 H-NMR of L-tyrosine based polyurethane blend H-NMR spectra of the blends for integration (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) FT-IR spectra of the of pure polyurethanes and blends (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) Quantitative estimation of absorbance ratio (1730 cm -1 /1100 cm -1 ) of the of pure polyurethanes and blends (error bars represent standard deviation of measurement from 3 samples) (ratio indicates ratio of PEG-HDI- DTHG to PCL-HDI-DTH) FTIR analyses of the blends in the region of cm -1 (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) SEM images of the polyurethane blends (ratio indicates ratio of PEG- HDI-DTHG to PCL-HDI-DTH) xx

21 6.8 DSC thermograms of pure polyurethanes and blends (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) Representative stress-strain curve of pure polyurethane and blends (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) Deviation of mechanical properties of blends from calculated values (from additive rule) Histogram of distribution of contact angle on blend surface (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) Deviation of contact angle values from the calculated values Water absorption characteristics of polyurethane blends (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) Water absorption characteristic as a function of blend composition from both experimental and calculated (additive rule) values Hydrolytic degradation of blends (n=3) Mass loss by hydrolytic degradation as a function of blend composition for both experimental and calculated (additive rule) values SEM images of degraded samples after 30 days of hydrolytic degradation (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) 190 xxi

22 CHAPTER I INTRODUCTION Polymers are extensively used for different biomaterial applications. The demands to develop new polymers and improve the performance of existing polymers as biomaterials are increasing. Tissue engineering is one of the well known fields for the application of polymers. Polymers are used for fabricating scaffolds for the regeneration of tissues from the cellular level. Several important physicomechanical properties are required for the polymers for use as tissue engineering scaffolds. These are biodegradability, biocompatibility, structural strength, easy processability etc 1. Polyurethanes are widely used as biomaterials for different applications due to excellent properties and good biocompatibility 2. Segmented polyurethanes synthesized from various polyols, diisocyanates and chain extenders can be structurally manipulated to achieve a wide range of properties suitable for various biomaterial applications. The use of polyurethanes as biomaterials has been exploited for various implants, including pacemakers, vascular graft etc. where biostability of the polyurethane is of prime concern. The degradability of polyurethanes in biological environments have led to the development of degradable polyurethanes for applications where biodegradation is a requirement e.g. tissue engineering. Several biodegradable polyurethanes has been 1

23 synthesized and used for different purposes including tissue engineering 3. Biodegradable polyurethanes have been synthesized from different types of polyols, diisocyanates and chain extenders. The main criteria for the selection of ingredients depend on the biocompatibility and non-toxicity of the component materials and the degradation products as well. Amino acid based synthetic polymers are developed and studied as biomaterials due to biocompatibility and/or biodegradability. Several amino acid based polymers are used for biomaterial applications including tissue engineering 4. The use of amino acids as one component will enhance the compatibility of the polymers for biomedical applications. Polyurethanes based on amino acids offer several advantages including biocompatibility, biodegradability and a range of material properties which can be tuned by changing the structure of the polymer. 1.1 Objective This dissertation describes the study of polyurethanes on based natural amino acid L- tyrosine for biomaterial applications with the main focus on tissue engineering. The use of L-tyrosine based polyurethane will impart the required biocompatibility along with the chance to tune the properties of the polymers. The understanding of the structure property relationship with respect to tissue engineering application is the main focus of this work. This study seeks to explore the synthesis and characterization of the polyurethanes and to investigate the materials at the structural level. Based on different combinations, a library of polyurethanes will be constituted having a wide range of properties. The goal in this project is to use a derivative based on L-tyrosine as a chain extender in the synthesis, 2

24 with different polyols and diisocyanates. The structural contribution of the different components in the polyurethanes will be examined in terms of properties. The characterization of the polymers for different physical, chemical, and engineering properties and relating those properties to the structure and vice versa will ensure the development of polymers with controlled structure and defined properties. The specific aims of this dissertation are specified as: Synthesis and characterization of polyurethanes based on L-tyrosine based chain extender with different polyols and diisocyantes Characterization of the polyurethanes for physicomechanical and engineering properties including the surface characteristics, degradation performances, controlled release etc. relevant for tissue engineering applications Study of structure-property relationship of L-tyrosine based polyurethanes with different polyol and diisocyanate components for the range of properties to develop L-tyrosine based polyurethanes with a range of properties Characterization of blends based on L-tyrosine based polyurethanes 1.2 Layout of the dissertation The rest of the dissertation is divided into six chapters. Chapter II describes the current state of art in the field of polymers in tissue engineering. It also analyzes the pros and cons of the existing situation particularly in the context of this research. Chapter III describes the synthesis and characterization of L-tyrosine based polyurethanes. It includes detailed structural, thermal and mechanical characterization of the polymers. Chapter IV deals with the characterization of the polyurethanes for tissue engineering 3

25 applications. It includes analysis of different types of degradation characteristics, water absorption, surface properties, release characteristics, etc. Chapter V includes detailed analyses of structure-property relationship of the polyurethanes. A library of polyurethanes is developed from different combinations of the structural components. The effect of the structural variation is examined in terms of different physicomechanical properties to understand the underlying principle of structure-property correlation. Chapter VI includes a brief analysis of blends developed from L-tyrosine based polyurethanes and Chapter VII finally summarizes the conclusions and mentions the direction of future work. 4

26 CHAPTER II BACKGROUND The focus of this chapter is to describe and analyze the background knowledge in the context of the research presented in the dissertation. The following sections briefly explain the importance of polymers in tissue engineering, development of different polymers including polyurethanes and amino acid based polymers in biomaterial applications and the importance of structure-property correlation of polymers for biomaterial development. The final section presents the approach used in this dissertation in the background of the existing state of the art. 2.1 Tissue engineering and polymers Tissue engineering is an interdisciplinary field that applies the principles of engineering and life sciences towards the development of biological substitutes that restore, maintain, or improve tissue functions 5. Tissue engineering has emerged as a complementary and an alternative solution to regenerate tissues which are damaged either through injury or disease 5,6. The acute shortage of organ/tissue donors and the risk of rejection by the host due to incompatibility have led to the development of tissue 5

27 engineering 5,7. The concept of tissue engineering is to provide the cells an appropriate biological environment either in-vivo or ex-vivo to regenerate into fully functional tissue. The function of polymeric material in tissue engineering is to provide a 3-dimensional (3- D) architecture for the cells to grow into tissues. Cells are allowed to grow on bioactive degradable polymeric scaffolds that provide physical and chemical cues to guide the differentiation and transformation of the cells into 3-D tissues 8. These polymeric scaffolds act as an artificial extra cellular matrix (ECM) for the cells and are gradually replaced by original ECM as the cells are grown into tissues without any immunogenic or toxic effects 1,5,6. In addition to cells and scaffolds, several active ingredients e.g. genes, growth factors, drugs, etc. are required for promoting proper tissue regeneration. These materials are delivered by a proper mechanism through the polymer scaffolds during the process of tissue regeneration 9. The total concept of tissue engineering is represented in Figure 2.1. Polymers Designing Scaffold + Cells Drug/Active Agents Impaired or defective tissue/organ Tissue Engineering Reactor Regenerated tissue/organ Figure 2.1 Concept of tissue engineering 6

28 Polymer scaffolds used for restoration of tissue/organ function utilizing the tissue engineering concept must possess certain physicochemical characteristics, e.g. easy processability, thermal properties, structural strength, and material properties 1. Definitely, biocompatible polymers are used for such applications. However, the biodegradable polymers are preferred, as the scaffold is only a temporary one. Tissue engineering has benefited from the discovery of a wide range of biodegradable polymers 10. The essential criteria for the polymers for tissue engineering application are: The polymer should be tissue compatible and non-toxic The polymer should have appropriate cellular and tissue response; and inhibit any adverse effects The polymer must degrade into fully biocompatible degradation products, both in terms of local tissue response and systematic response. The polymer must possess physicomechanical and engineering properties suitable for the intended application. The polymer must degrade within a clinically useful range of several weeks to several months. The polymer must have drug delivery compatibility in applications that call for release or attachment of bioactive molecules. Several natural, synthetic and semi synthetic polymers have been used for tissue engineering applications. Different polysaccharides (e.g. cellulose 11, chitin 12 etc.) and proteins (e.g. collagen 13, elastin 14, fibrin 15 etc.) have been used as natural material for tissue engineering. Natural polymeric materials are bioactive (through cell specific interactions), non-toxic and non-antigenic and therefore biocompatible. The main 7

29 disadvantage of natural materials is related to reproducibility in terms of degradation, structural strength due to variation of sources and purification. The use of synthetic biocompatible polymers for tissue engineering applications is well known 1. Tissue engineering has benefited from the discovery of a wide range of biodegradable polymers. However; the choices for the polymers are limited due to the demand for materials having a combination of properties that are pertinent to tissue engineering applications. The advantages of synthetic biodegradable materials are: large scale production with reproducibility, easily variable micro- and macro-structure, and easily controlled physicochemical properties. Several synthetic polymers have found wide-ranging applications in tissue engineering products. The promising candidates are: polylactide (PLA) 16, polyglycolide (PGA) 17, poly lactide-co-glycolide (PLGA) 18, poly(εcaprolactone) 19, polydioxanone 20, polyanhydrides 21, poly(propylene fumarate) 22, poly(ortho esters) 23, polyurethanes 24, poly(amino acids) 25 etc. However, several drawbacks e.g. inflammatory response, aberrant cellular response, adverse effect of polymer erosion, absence of chances for structural modifications etc. are associated with these polymers. Among the different synthetic polymers, PLA, PGA and PLGA have been extensively used for fabrication of scaffolds for different types of tissue engineering. These polymers are biocompatible, easily synthesized and processible and have non-toxic degradation products. Moreover, the polymers are approved by the Food and Drug Administration (FDA) for clinical applications 26. Several devices and systems are made from these polymers for tissue engineering application and are tested. But several studies have also indicated some potential drawbacks e.g. inflammatory responses, lack of bioactivity and flexibility for chemical modifications etc. 8

30 The area of semi-synthetic materials involves a hybrid of natural and synthetic polymers. Mainly two approaches are used to make semi-synthetic polymers 27. Chemical modification through grafting and copolymerization is used to develop such materials e.g. peptide grafted PEG. The other technique involves composite structure obtained by blending and other scaffold fabrication techniques. Table 2.1 summarizes a list of different polymers used in tissue engineering. Table 2.1 Polymers in tissue engineering Natural Polymers Collagen Chitosan Alginate Starch Hyaluronic acid Gelatin Cellulose Semi-synthetic/ Hybrid Polymers PEG-Alginate PEG-Chitosan PLA-Starch CSA-hyaluronate Synthetic Polymers Poly(lactic acid) Poly(glycolic acid) Poly(lactide-glycolide) Poly(capro lactone) Polyanhydrides Polydioxane Polyesters Polyamides Poly(amino acid) Poly(hydroxylbenzene) 2.2 Amino acid based polymer The problems associated with natural and synthetic polymers have led to the development of polymers from naturally occurring nutrients and metabolites as monomers. These polymers are therefore expected to be biocompatible and biodegradable. Amino acids are the monomeric units of the proteins and a major part of natural metabolites. The polymer of amino acid linked by peptide linkages are called poly (amino acids) or peptides and these polymers have several advantages as biomaterials 25. The main advantages are: biocompatibility, non-immunogenecity, and non-toxic degradation products, in addition to enzyme specificity, protein folding, and tertiary 9

31 structure. Moreover, the side chain modification offers the chance to attach drug molecules, small peptides or pendant groups for changing the physicomechanical properties of the polymers. Poly(amino acids) have found applications in different biomaterial applications including suture materials, artificial skin substitutes, and as drug delivery systems 28. However, unfavorable synthetic and engineering properties of poly(amino acids) have restricted the use of these materials. The main disadvantages are the difficulties associated with synthesis and processing. Synthesis of these polymers involves the use of N-carboxy anhydride, a highly expensive, very reactive and moisture sensitive material. This problem leads to the difficulties in the formation, isolation and purification of the polymers 1. These polymers are highly crystalline in nature due to the inter-molecular H-bonding of the amide linkages of the polymer chains. Due to this fact, most of the poly(amino acids) are insoluble in common organic solvents and have high glass-transition and melting temperatures. Most of the polymers degrade before reaching the melting temperature. These features make the polymers highly insoluble and nonprocessible materials. Apart from these, several other problems are also related to poly(amino acids). Unpredictable swelling characteristics, change in conformation, uncontrolled and varied enzymatic degradation of the polymer in-vivo has limited the use of poly(amino acids) as a biomaterial 1. The practical difficulties associated with poly(amino acids) are due to the structure of the polymer itself. Structural modification by introducing non-amide linkages to replace the peptide linkages in the polymer backbone by pseudo-peptide chemistry is a tool to generate pseudo-poly(amino acids) or pseudo-peptides 29. The non-amide linkages refer to ester, imminocarbonates 30, carbonate 31, urethane 32, or phosphate bonds 33. Thus, the term 10

32 pseudo-poly(amino acids) refers to the family of polymers in which naturally occurring polymers are linked together by nonamide bonds. Kohn and Langer showed the way of inducing direct polymerization reactions between the suitably protected amino acids or dipeptides, involving the functional groups in the amino acids 34. The first investigated were a polyester from N-protected trans 4-hydroxy-L-proline and a poly(imminocarbonate) from tyrosine dipeptide 34,35. Backbone modification of conventional poly(amino acids) by nonamide linkages in general improves the physicomechanical properties e.g. solubility, thermal property, moldability etc. along with the desired properties of biocompatibility, non-toxicity, and non-immunogenecity. Several amino acids are used for this technique: serine, hydroxyproline, tyrosine, cysteine, glutamic acid etc. OH OH OH CH 2 CH 2 CH 2 CH CH CH H COOH H 2 N COOH H 2 N H Desaminotyrosine (Dat) Tyrosine (Tyr) Tyramine (Tym) Figure 2.2 Structure of L-tyrosine and its metabolites L-tyrosine (Figure 2.2) is an amino acid having a phenolic hydroxyl group. This feature makes it possible to use derivatives of tyrosine dipeptides as a motif to generate monomers, which are the important building blocks of the polymers. L-tyrosine (Tyr) and its metabolites desaminotyrosine (Dat) and tyramine (Tym) (Figure 2.2) can be used to 11

33 form four structural dipeptides 31. Synthesis and characterization of polycarbonates, polyimminocarbonates, polyphosphates and polyarylates based on L-tyrosine have been studied. Using specific groups to protect the amino group and/or acid group in L-tyrosine and its metabolites, different combination of dipeptides can be obtained. The most common dipeptides used are the desaminotyrosyl tyrosine alkyl esters. It is synthesized by coupling of L-tyrosine alkyl ester with desaminotyrosine. Several different alkyl groups have been used to form the dipeptide but desaminotyrosyl tyrosine hexyl ester (DTH) is mostly exploited for different polymers due to better physical properties 31. The structure of DTH is shown in Figure 2.3. HO O CH 2 CH NH C CH 2 CH 2 O C O (CH 2) 5 CH 3 OH Figure 2.3 Structure of Desaminotyrosyl tyrosine hexyl ester (DTH) Kohn et al. and Sen Gupta et al. have studied the synthesis and characterization as well as the structure property relationship of modified L-tyrosine polymers as poly carbonates 31 and poly imminocarbonates 30,36. These materials show significant improvement in the physical and chemical properties suitable for biomaterial applications over poly(l-tyrosine). L-tyrosine based polyarylates and copolymers with PEG have also been investigated as an alternative material for tissue engineering application 37. Biocompatibility studies using the tyrosine derived degradable polymer, poly(desaminotyrosyl tyrosine hexyl carbonate) (poly (DTH carbonate)) have been favorable, suggesting the material is suitable for tissue engineering application 38. The hydrolytic stability of poly(dth carbonate) shows that the polymer is relatively stable 12

34 and not degrading until 800 days or more 31,39. Integra Lifesciences, NJ is trying to commercialize the tyrosine based polycarbonate and polyimminocarbonate material as Tyrosorb. This is an indication that tyrosine based pseudo poly(amino acid) possess the potential for being used as biomaterial for different applications. The structure-property relationship of polymers based on different derivatives of tyrosine including DTH (desaminotyrosinetyrosylhexyl ester) has been investigated for different polymers. A series of polyiminocarbonates based on different tyrosine based derivatives show that the material properties are extensively varied by the structural variation 39. A combinatorial approach of developing a library of degradable polyarylates based on different desaminotyrosyl alkyl esters and diacids demonstrates the concept of structure property correlation 40. Several properties (e.g. thermal behavior, cellular response, surface characteristics etc.) of polycarbonates, polyarylates and copolymers with polyethylene glycol based on tyrosine have been studied for a large series of polymers to illustrate versatility of tyrosine derived esters as a building block for biodegradable polymer. All these studies show that structural variation in the polymer composition leads to the change in property and such systematic study permits to constitute a library of polymers with variable material properties. The synthesis of L-tyrosine based polyphosphate as a pseudo-poly(amino acid) has been reported, where the non amide linkage is the phosphate bonds 33. The presence of the phosphate bond in the polymer backbone enhances the physicomechanical properties and the degradation rate. The synthesis and characterization of L-tyrosine based polycarbonate, polyimminocarbonate, polyphosphate and polyarylates have shown significant potential of amino acid based polymers for biomaterial application. However, certain limitations of these polymers in controlling 13

35 hydrolytic degradation rates and other physicomechanical properties have led to the further development of L-tyrosine based polymers. 2.3 Polyurethanes as biomaterials Polyurethanes are widely used as biomaterials, mainly with the development of biomedical polyurethanes for long term applications e.g. vascular graft, pace maker applications etc., where biostability of the polyurethane is of prime concern 41. The polyurethanes have structures consisting of polyol, which constitutes the soft segment, and the polyfunctional isocyanate (mainly diisocyanate) and the chain extender (or crosslinker), which constitutes the hard segment 42. The general structure of polyurethanes is shown in Figure 2.4. A wide range of properties can be obtained by tuning the structure of the polyurethane. Hard Segment Soft Segment Polyol Diisocyanate Chain Extender Y HO M OH OCN NCO Y,Y = NH 2 / OH Urethane Linkages NH O C NH NH O C O Figure 2.4 Structure of polyurethanes 14

36 The biphasic nature of the polymer is due to the presence of hard and soft segment in the polymer structure. The ratio of hard segment to soft segment, co-existence and/or microphase separation of the two segments could adjust the different properties over a wide range. However, the polyurethanes have shown their susceptibility to degradation under the conditions of their performance. Poly(ester) urethanes and poly(ether) urethanes, widely used for long term applications, have been shown to degrade under hydrolytic 43 conditions and in oxidative 44 environment respectively. In addition, environmental stress cracking (ESC) of polyurethanes is also another important way of polyurethane degradation 45. All these have led to an extensive research of polyurethane degradation. The use of polyurethanes for tissue engineering applications emerged mainly due to the degradability of the polyurethanes. Since polyurethane structures can be tailored to have degradable linkages and a range of chemical, physical and mechanical properties, polyurethanes have been studied as an alternative material for tissue engineering application 46.The versatility of polyurethanes lies in the phasic behavior, elastomeric as well as thermoplastic nature and the easy structural tunability. Biodegradable polyurethanes have been synthesized from different types of polyols, polyisocyanates and chain extenders. The main criteria for the selection of ingredients depend on the biocompatibility and non-toxicity of the component materials and the degradation products as well. The biodegradation of the polyurethanes are largely controlled by the soft segment polyols. The variety of the soft segment includes polylactide or polyglycolic acid 47, polyethers and polyesters 48. The diisocyanates are mainly aliphatic (e.g. butane diisocyanate 46, hexamethylene diisocyantes 48 ) and amino acid based (e.g. lysine based diisocyanate 47 ).Chain extenders for the polyurethanes are 15

37 usually diol or diamine compounds. Amino acid based chain extender based on phenylalanine has also been used for synthesis of degradable polyurethanes 48,59. Biocompatibility studies of the polyurethanes show that these are potential materials for tissue engineering applications. The structure-property relationship of biodegradable polyurethanes has been studied but any systematic approach in developing a correlation between the two is missing. The effect of different polyols as soft segments has been studied on polyurethanes based on phenyl alanine based chain extender 49. Study on using triblock copolymer with different block lengths as soft segments and its effect on the polyurethane properties shows that the properties can be varied by changing the composition 46, Technical approach The approach followed in this dissertation is to develop polyurethanes based on L- tyrosine. The general scheme of two-step polyurethane synthesis is shown in Figure 2.5. L-tyrosine can be introduced in the polyurethane structure using DTH. Two phenolic hydroxyl group of DTH can be used for chain extension as diol chain extender. Different polyols and aliphatic diisocyanates will be used to synthesize and characterize a group of L-tyrosine based polyurethanes. An extensive characterization of the material properties of the polyurethanes will be done to study the effect of structure on the polyurethane properties. Structure-property relationship of these polyurethanes will be investigated in terms of biomaterial applications with particular reference to tissue engineering by altering the structure of the polyols and diisocyanate. In addition, blends of different 16

38 polyurethanes will be examined for tissue engineering to tune the properties of individual polymers. O C N R N C O + HO OH diisocyanate macrodiol H O O C N R N C O O C O prepolymer HO R 1 OH N R N C H diol chain extender O H H H H C O N R N C O O O C polyurethane O N R N C O O R 1 O n prepolymer H 2 N R 2 NH 2 diamine chain extender H H H H H H C O N R N C O O O C O N R N C O N R 2 N n polyurethane urea Figure 2.5 Scheme for synthesis of polyurethanes 17

39 CHAPTER III SYNTHESIS AND CHARACTERIZATION OF L-TYROSINE BASED POLYURETHANES Polyurethanes are one broad class of polymer with one common aspect: the presence of urethane linkages. Segmented polyurethanes are synthesized by two-step condensation polymerization reactions. The components for the synthesis of polyurethanes are: polyol, isocyanate and chain-extender 2,42. Depending on the functionalities different types of polyurethanes are synthesized. The polyols are typically high molecular weight (with average molecular weight up to 8000) diols. Among the different polyols mainly ether, ester, and carbonate type diols are used for biomaterial polyurethanes. The properties of the polyols actually contribute significantly to the property of the final polyurethanes. Both aromatic and aliphatic diisocyanates are used as diisocyanates with the aromatic one predominantly used in the design of biostable polyurethanes. Chain extenders are usually diol or diamine based compounds. The two phase structure of polyurethanes arises from the biphasic nature due to the differences in the physical and chemical nature of the components. The soft segment arises from the polyol fraction while the hard segment is from the chain extender and diisocyanate. The stoichiometry of the components and number of steps in the polymerization reaction determines the relative distribution of the segments in the final polyurethane composition. Both the physical and chemical nature of 18

40 the components along with the composition determine the properties of the polyurethane. Polyurethanes for biomaterial applications have been investigated for variety of applications. The essential criteria for such polyurethanes depend on the biocompatibility of the components. The polyols that are typically used for biomaterial polyurethanes are polyethylene glycol (PEG), polytetramethylene glycol (PTMG), polycaprolactone diol (PCL) etc. Several aromatic and aliphatic diisocyanates are used e.g. 4,4 diphenylmethane diisocyanate (MDI), toluene diisocyanate (TDI), hexamethylene diisocyanate (HDI) etc. Examples of chain extenders are 1,4-butanediol (BD), ethylenediamine (EA) etc. Amino acid based polyurethanes have been developed in the context of biodegradable polyurethanes for tissue engineering applications. Amino acid can be incorporated into the polyurethane structure as diisocyanate or the chain extender. Lysine based diisocyanate (LDI) has been used for the development of several biocompatible polyurethanes. Chain extenders based on several amino acids e.g. phenyl alanine, tyrosine, lysine, ornithine has been used to design amino acid polyurethanes 46,48,51,59. The prime objectives of developing such polyurethanes are biocompatibility and biodegradability for in vivo applications. The synthesis of segmented polyurethanes involves two steps: (i) first the reaction of polydiol with diisocyanate in a stoichiometric ratio such that isocyanate terminated prepolymer is formed and (ii) finally the reaction of the isocyanate terminated prepolymer with a low molecular weight diol or diamine compound to extend the chain. The research described in this chapter of the dissertation is focused on synthesis and characterization of the polyurethanes based on L-tyrosine. The components used for the development of the polyurethanes are shown in Figure 3.1. Two different polyols are 19

41 used: polyethylene glycol (M w : 1000) (PEG) and polycaprolactone diol (M w : 1250) (PCL) because of the biocompatible characteristics of segments. The diisocyanate used is aliphatic hexamethylene diisocyanate (HDI) due to its potential biocompatibility 91. The chain extender is desaminotyrosyl tyrosine hexyl ester (DTH) is a diphenolic, dipeptide molecule based on L-tyrosine and its metabolite, desaminotyrosine (DAT). HO CH 2 CH 2 O H n Poly ethylene glycol (PEG) HO CH 2 C O CH 2 CH 2 O C CH 2 O 5 5 m O O H n Poly caprolactone diol (PCL) OCN NCO Hexamethylene Diisocyanate (HDI) HO O Desaminotyrosyl tyrosine hexyl CH 2 CH NH C CH 2 CH 2 OH O C O (CH ester (DTH) 2) 5 CH 3 Figure 3.1 Components of L-tyrosine based polyurethanes 3.1 Experimental The following sections describe the details of the experimental procedure used in the synthesis and characterizations of the polyurethanes Synthesis of polymer The synthesis of polymer involves primarily two steps: (i) Synthesis of the chain extender DTH and (ii) Synthesis of the polyurethane. All the chemicals and solvents were 20

42 used as received, unless otherwise stated and were purchased from Sigma Aldrich. Distilled water was used for all purposes DTH Synthesis The synthesis of DTH is described in details in the literature 32. Briefly, DTH is synthesized from hexyl ester of L-tyrosine (TH) and desaminotyrosine through carbodiimide coupling reaction. The reaction steps are summarized below and scheme of the reaction is shown in Figure 3.2. i) The carboxylic acid group of the L-tyrosine (0.05 mole) is esterified by 1- hexanol (50 ml) in presence of thionyl chloride (0.05 mole) at 0 C initially, followed by reaction at 80 C for 12 hours. The reaction product obtained after cooling down the reaction to room temperature was completely precipitated in cold ethyl ether. The product was then filtered and washed with cold ether to obtain white solid, which is the chloride salt of hexyl ester of L- tyrosine. ii) The white solid was re-dissolved in distilled water and subsequently neutralized by 0.5 M sodium bicarbonate solution till the ph of the solution is slightly basic (ph~7.5). At this point solution turns turbid due to formation of TH. TH was extracted in ether, and the ether was evaporated to complete dryness to obtain tyrosine hexyl ester (TH) as an off-white solid. iii) Coupling of TH with DAT was mediated through hydrochloride salt of N- Ethyl-N -dimethylaminopropyl carbodiimide (EDC.HCl). Typically TH, DAT and EDC.HCl were added in equimolar proportion in 99% pure 21

43 tetrahydrofuran (THF) as solvent at 0 C. After that, the reaction was allowed to continue at room temperature for 12 hours. At the end of 12 hours, the reaction mixture was poured into four times its volume of distilled water and was extracted in the organic phase by dichloromethane (DCM). iv) The organic DCM phase was washed with 0.1 N HCl solution, 0.1 N sodium carbonate solution and concentrated sodium chloride solution to remove the by products. The organic DCM phase was dried, and the solvent was evaporated under vacuum to obtain desaminotyrosyl tyrosine hexyl ester (DTH) as yellow, viscous oil. HO SOCl 2 1-hexanol 80 C COOH NH 2 L-Tyrosine HO COO(CH 2 ) 5 CH 3 NH 2 TH EDC.HCl HO DAT COOH HO O CH 2 CH NH C CH 2 CH 2 O C O (CH 2) 5 CH 3 OH DTH Figure 3.2 Scheme for DTH Synthesis 22

44 Synthesis of Polyurethanes The synthesis of polyurethane is a condensation type polymerization typically involving the reaction of isocyanate (-NCO) and hydroxyl (-OH) to form the carbamate (- NHCO) linkages. The polymerization is usually a two-step process leading to the formation of segmented polyurethane: (i) Reaction of polyol with diisocyanate to form isocyanate terminated prepolymer and (ii) Chain extension through the reaction of prepolymer and chain extender. Two different polyurethanes were synthesized using PEG and PCL as the polyol with HDI (diisocyanate) and DTH (chain extender). The reactions were carried out in a completely dry and moisture-free environment under inert (completely dry nitrogen, N 2 ) atmosphere. Both PEG and PCL were dried under vacuum for 48 hours at 40 C to remove entrapped water. N,N -Dimethyl formamide (DMF) used as solvent, was dried over calcium hydride (CaH 2 ) followed by molecular sieve. Diisocyanate of high (>99%) purity grade was used. The detailed protocol for the synthesis of polyurethane is summarized below: i) The polyol (PEG or PCL) was reacted with HDI at a 1:2 molar ratio in DMF as solvent and 0.1% stannous octoate catalyst to form the prepolymer. Typically, 5 mmol of polyol was added into 40 ml of DMF and 10 mmol of HDI and 2~3 drops of stannous octoate was added to the reaction mixture under dry and inert atmosphere with continuous stirring. ii) The temperature was increased to 110 C and the reaction was allowed for 3 hours at this temperature. After 3 hours, the reaction cooled down to room temperature (~25 C) with continuous stirring. The temperature of reaction was carefully maintained within the range of ±3 C. 23

45 iii) DTH was added in the second step at a 1:1 molar ratio with the prepolymer. Typically, 5 mmol of DTH in 10 ml of DMF was added. iv) The temperature of reaction was then gradually increased to 80 C and the reaction was allowed to continue for 12 hours. The temperature of reaction was controlled within the range of ±3 C. After 12 hours the reaction was quenched by pouring the reaction into cold concentrated aqueous solution of sodium chloride. At this point, solid polyurethane polymer precipitates out from the reaction mixture. v) For PEG based polyurethanes, the polymer is suspended in the form of gel in the water. The final polymer is centrifuged out and re-suspended in water and then centrifuged. This process is repeated for at least three times to remove the impurities and unreacted materials. The final polymer is then dried in vacuum at 40 C for 48 hours. The polymer is yellowish white sticky solid. The nomenclature used for the PEG based polyurethane is PEG-HDI-DTH. vi) For PCL based polyurethanes, the polymer is suspended as solid polymer. The final polymer is filtered out and washed with water. This washing is repeated for at least three times to remove the impurities and unreacted materials. The final polymer is then dried in vacuum at 40 C for 48 hours. The polymer is yellowish white solid. The nomenclature used for the PCL based polyurethane is PCL-HDI-DTH. The polyurethanes synthesized were stored is desiccators for the purpose of characterization and future experiments. The structure of the two polyurethanes is shown in the Figure

46 3.1.2 Characterizations of polymer The polymerization and the polyurethanes were characterized extensively by various techniques to determine the structure and understand the basic properties of the polymers. The preliminary characterization studies include structural, thermal and mechanical characterization. N H (CH 2 ) 6 N H O O O n O H N H N (CH 2 ) 6 m O O O H N O O O p PEG-HDI-DTH O (CH 2 ) 5 CH 3 N H (CH 2 ) 6 N H O O (CH 2 ) 5 O O n O H N H N (CH 2 ) 6 m O O O H N O O O p PCL-HDI-DTH O (CH 2 ) 5 CH 3 Figure 3.3 Structure of L-tyrosine based polyurethanes Structural Characterizations The structural characterizations were done by 1 H-NMR, 13 C-NMR and FT-IR study. NMR was carried out in 300 MHz Varian Gemini instrument with d-dimethyl sulfoxide (δ = 2.50 ppm for 1 H NMR and 39.7 ppm for 13 C NMR as internal reference) solvent for PEG-HDI-DTH and d-chloroform (δ = 7.27 ppm for 1 H NMR and 77.0 ppm for 13 C NMR as internal reference) for PCL-HDI-DTH. FT-IR analysis was performed with a Nicolet NEXUS 870 FT spectrometer for neat samples with 16 scans. FT-IR analysis was 25

47 also used to study the progress of polymerization reaction. The molecular weights of polymers were determined by gel permeation chromatography (GPC) using tetrahydrofuran (THF) as solvent and polystyrene as internal standard. The solubility of the polymers was checked in a variety of solvents by dissolving ~ 10 mg of solid polymer in 10 ml of the solvent at room temperature Thermal Characterizations The thermal behaviors of the polyurethanes were characterized by differential scanning calorimetry (DSC) and thermo gravimetric analysis (TGA). DSC was performed with a DSC Q100V7.0 Build 244 (Universal V3. 7A TA) instrument at a scanning rate of 10 C/min from -80 to 250 C. TGA was performed with a TGA Q50V5.0 Build 164 (Universal V3. 7A TA) instrument from 0 to 600 C under nitrogen atmosphere at a rate of 20 C/min. An average of 10 mg of solid sample was used for both the experiments Mechanical Characterizations The tensile properties of the polyurethanes films were measured by Instron Tensile Testing Machine with a load cell of 100 N and cross-head speed of 100 mm/min at room temperature. The films were cast from 10% wt solution of polymers (DMF for PEG-HDI- DTH and chloroform for PCL-HDI-DTH) and solvent was allowed to evaporate at room temperature and then subsequently dried in vacuum oven at 50 C for 48 hours to remove the residual solvent. The sample dimension was 20 mm 6 mm ~ 0.3 mm with a free length of 10 mm. The average of five measured values was taken for each sample. 26

48 3.2 Results and Discussion The following sections includes the results of the experiments and its explanation for the synthesis and characterization of the polyurethanes Polymerization Reaction Table 3.1 summarizes the composition of the two polymers with the relative contribution of hard and soft segment. The yield for the synthesis of DTH was about 85% and for the polyurethanes was about 70-80%. The results were reproducible within a range of ±5% with reasonable purity of the polyurethanes. Table 3.1 Composition of the polyurethanes Polymer PEG-HDI- DTH PCL-HDI- DTH Soft Segment Molecular Weight (M w ) Soft Segment Content (wt. %) Hard Segment Content (wt. %) Diisocyanate Chain (HDI) extender (DTH) NMR Characterizations The 1 H NMR (along with the peak assignments) of PEG-HDI-DTH and PCL-HDI- DTH is shown in Figure 3.4 and 3.5 respectively. PEG-HDI-DTH: δ 0.8 (CH 3 - in hexyl group, DTH), 1.2 (-CH 2 - in hexyl chain in DTH), 1.3 (-CH 2 - in hexyl chain in HDI), 1.4 (-NH-CH 2 -CH 2 - in HDI), 2.7 (-CH 2 -CH 2 -CO- in DTH), 2.9 (-NH-CH 2 - in HDI and -C 6 H 4 -CH 2 -CH 2 - in DTH), 3.0 (-C 6 H 4 -CH 2 -CH in 27

49 DTH), 3.5 (-O-CH -CH -O- in PEG), 3.6 (-CH -CH -O-CO- in PEG), 4.0 (-CO-O-CH CH2- in DTH), 4.4 (-NH-CH-(CO)-CH2- in DTH), 6.9 and 7.1 (two -C 6 H4- in DTH) ppm Figure H NMR of PEG-HDI-DTH ppm Figure H NMR of PCL-HDI-DTH 28

50 PCL-HDI-DTH: δ 0.8 (CH3- in hexyl group, DTH), (CH2 in DTH, HDI and PCL), 2.3 (-CO-CH2- in PCL), 2.8 (-CH2-CH2-CO- in DTH), 2.9 (-NH-CH2- in HDI and -C6H4-CH2-CH2- in DTH), 3.1 (-C6H4-CH2-CH in DTH), 4.0 (-CO-O-CH2-CH2- in DTH and PCL), 4.8 (-NH-CH-(CO)-CH - in DTH), 6.7 and 6.9 (two -C H - in DTH) The 13 C NMR (along with the peak assignments) of PEG-HDI-DTH and PCL-HDI- DTH is shown in Figure 3.6 and 3.7 respectively ppm Figure C NMR of PEG-HDI-DTH PEG-HDI-DTH: δ 13.9 (CH3- in hexyl group, DTH), (CH2 in hexyl chain in DTH, HDI), (CH2 in hexyl chain in DTH, HDI), 36.0 (-CH2-CH2-CO in DTH), 37.5 (-C H -CH -CH in DTH), 54.6 (-NH-CH-(CO)-CH - in DTH), 62.9(-CH

51 CH2-O-CO-NH- in PEG), 64.8 (-CH2-CH2-CO- in DTH), 68.9 (-CH2-CH2-O-CO-NH- in PEG), 69.8 (-O-CH2-CH2-O- in PEG), and (two -C6H 4 - in DTH), (-NH-CO-O- in urethane carbonyl), (ester and amide carbonyls in DTH) ppm Figure C NMR of PCL-HDI-DTH PCL-HDI-DTH: δ 14.2 (CH 3 - in hexyl group, DTH), (CH 2 in hexyl chain in DTH, HDI, PCL), (CH2 in hexyl chain in DTH, HDI), 34.1 (-CH2-CH2-CO in DTH), 34.3 (-CH 2 -CH2-CO-O in PCL), 40.5 (-C6H 4 -CH 2 -CH in DTH), 53.5 (-NH-CH- (CO)-CH2- in DTH), 64.3 (-CO-O-CH2-CH2- in PCL), and (two -C6H4- in DTH), (-NH-CO-O- in urethane carbonyl) and 172.0(ester and amide carbonyls in DTH), (ester carbonyls in PCL) 30

52 The peak assignment from 1 H and 13 C NMR show that all the three components are present in the polymer chains. However due to the presence of similar chemical environments for certain protons and carbons, there is considerable overlap of the peaks which makes the assignment a difficult task. In general, for both the PEG and PCL based polyurethanes the presence of the characteristic peaks indicate that the polymers are composed of the corresponding soft segments along with HDI and DTH. Most important is the presence of urethane link indicated by the 2.9 ppm in 1 H NMR and 156 ppm in 13 C NMR for both in PEG and PCL based polyurethanes. This clearly shows that urethane linkages are formed by the condensation polymerization. However some unassigned peaks in the spectra corresponds to materials formed by possible side reactions and from of unreacted materials/solvent. But the intensity of such peaks are considerably lower than the assigned peaks which indicates that polymers are of reasonable purity FT-IR Characterizations The FT-IR spectra of the polyurethanes are shown in Figure 3.8. The spectra of both the polymers show the characteristic peaks for the polyurethane. For PEG-HDI-DTH, the characteristic 1100 cm -1 represents aliphatic ether linkage of the PEG segment and the peak around 1540 cm -1 represent N-H bending/c-n stretching of urethane linkages and the amide linkage of DTH segment. Moreover, 1620 cm -1 represents the aromatic stretch of DTH segment. The characteristic peaks in the region of cm -1 represents the carbonyl of the urethane linkages. The distribution of the carbonyl peak indicates a degree of hydrogen-bonding of urethane carbonyl group indicating interactions between different segments. The broad shoulder around 3330 cm -1 is indicative of hydrogen 31

53 bonded N-H stretching. For PCL-HDI-DTH, similar peaks are observed but the peaks around the region of ~1730 cm -1 is masked due to strong carbonyl absorption of carprolactone unit of PCL. The FT-IR analysis supports the structure of the polyurethanes. PCL-HDI-DTH PEG-HDI-DTH Wave numbers (cm -1 ) Figure 3.8 FT-IR of L-tyrosine based polyurethanes The FT-IR of the starting materials, intermediate prepolymer and the final polymer is shown together in Figure 3.9. The immergence of peaks around cm -1 represents formation of urethane bonds in the prepolymers compared to PEG and PCL. Peak at ~1630cm -1 represents the stretching of C=O (amide I) and 1540 cm -1 represents N-H bending vibrations (amide II) indicating the formation of urethane linkages. The peak at 2280 cm -1 comparable to the isocyanate peak of HDI indicates that both the prepolymers are isocyanate terminated. The addition of DTH results in the complete disappearance of the isocyanate peaks at 2280 cm -1 in the final polymer which indicates completion of reaction to the formation of final polyurethane. Moreover, the peak around 32

54 ~1620 cm -1 in the final polymer indicates C=C of aromatic ring structures of DTH. The peak at ~1715 cm -1 represents combined free non hydrogen bonded C=O in amide I of urethane and amide(in DTH)and shoulder at 1740 cm -1 represents ester C=O of DTH in PEG-HDI-DTH. Similarly, ~1715 cm -1 represents combined free non hydrogen bonded C=O of amide I of urethane and amide (in DTH) and at 1730 cm -1 represents ester C=O of caprolactone unit and DTH in PCL-HDI-DTH. PCL-HDI-DTH PEG-HDI-DTH DTH PCL-prepolymer PEG-prepolymer HDI PCL PEG Wave numbers (cm -1 ) Figure 3.9 FT-IR analysis of the components, prepolymer and polyurethane Molecular Weight of Polyurethanes Table 3.2 summarizes the molecular weight of the polymers which shows that both the polyurethanes have significantly high molecular weight. Compared to the molecular 33

55 weight of PEG and PCL as starting material, the molecular weight of the final polymers indicates the formation of polyurethanes. The low poly-dispersity indices of the polyurethanes indicate that the distribution of molecular weight is not broad and the polymerization is controlled. However, PEG based polyurethanes are lower in molecular weight compared to PCL based polyurethanes. This is probably due to presence of residual water in precursor PEG which inhibits high molecular weight of polymer by reacting away the diisocyanate 48. Considering different factors that contribute to the molecular weight of polymers in solution polymerization, these results were reproducible within range of ±10%. Table 3.2 Molecular weight of the polyurethanes Polymer M n (10 3 ) M w (10 3 ) Poly Dispersity Index PEG-HDI-DTH PCL-HDI-DTH Solubility of the Polyurethanes Table 3.3 shows the solubility features of the polyurethanes in the common solvents. Table 3.3 Solubility features of the polyurethanes Solvent\Polymer PEG-HDI-DTH PCL-HDI-DTH Methylene chloride Almost Soluble Almost Soluble Chloroform Almost Soluble Soluble DMF (Dimethyl formamide) Soluble Almost Soluble THF (Tetrahydrofuran) Soluble Soluble Methanol Insoluble Insoluble Ethanol Insoluble Insoluble Ethyl Acetate Insoluble Insoluble Acetone Insoluble Insoluble 34

56 The solubility of the polymers shows that the polyurethanes are soluble in polar aprotic solvents and insoluble in water and protic solvents. The polyurethanes are also insoluble in acetone, ethyl acetate which is polar and aprotic, indicating that the different phases of the polyurethanes contribute differently towards solubility. But in general, the solubility features indicate that the polyurethanes are soluble for practical purposes Thermal Characterizations The DSC thermograms of the polyurethanes are shown in Figure Endotherm PCL-HDI-DTH PEG-HDI-DTH Temperature ( C) Figure 3.10 DSC heating curves of L-tyrosine based polyurethanes The differential scanning calorimetry (DSC) analysis of both the polymers indicates very important information regarding the morphology of the polyurethane structure. The biphasic morphology of the polyurethane is due to the presence of soft and hard segment. Considerable phase mixing or segregation occurs due to the difference in the 35

57 compatibility of the segments. The compatibility of the segments arises from different interactions including hydrogen bonding, dipolar interactions, van der Waals interaction etc. The DSC thermograms of the polyurethanes show distinct glass transition (T g ) at -40 C for PEG-HDI-DTH and at -35 C for PCL-HDI-DTH which correspond to the soft segment glass transition temperature. The shift from the T g s of the pure homopolymer T g s (-67 C for PEG and -62 C for PCL) indicates some degree of phase mixing between the soft and hard segment of the polyurethanes 48. For PEG-HDI-DTH, three additional endotherms were observed: at 0, 50 and 162 C. Similar endotherms are also observed for PCL-HDI-DTH at 5, 52 and 173 C with an additional one at 31 C. The absence of hard segment T g indicates that hard segments are relatively crystalline domains due to presence of aromatic ring structure in the back bone of polymer 52. Woodhouse et. al. observed hard segment T g probably due to amorphous hard segment with aromatic group as pendant groups from the backbone of the polymer 48. Moreover, absence of melting endotherms for the phenyl alanine based polyurethanes indicates that the hard segment is largely amorphous. The endotherms at 162 C represent the melting of the microcrystalline hard segment domain while the other transitions at 0 and 50 C represents the dissociation of short range and long range order of the hard segment domain 53. Short range order of polyurethane actually represents the interaction between the soft segment and hard segment that actually contributes to the phase mixing behavior of the polyurethane. Long range order represents unspecified interactions within the hard segment domain. Absence of soft segment melting endotherm for PEG-HDI-DTH indicates the amorphousness of the soft segment. The crystallinity of PEG is reduced due 36

58 to the presence of hard segment at the PEG chain ends and due to partial dispersion of the hard segment within the soft segment of the polyurethane. The low molecular weight of PEG and high hard segment content in PEG-HDI-DTH favors this feature. Similar observations for PTMO 53 based polyurethanes and phenyl alanine 48 based polyurethanes are made. The similar endotherms for PCL-HDI-DTH at 173 C represent the melting of the microcrystalline hard segment domain while the other transitions at 0 and 52 C represents the dissociation of short range and long range order of the hard segment domain respectively. The additional endotherm at 31 C is probably due to the melting of soft segment. PCL being relatively more crystalline shows melting due to chain mobility at this temperature. The crystallinity of PCL soft segment is less affected in spite of phase mixing due to the dipolar interaction of ester bonds and relatively lower hard segment content. The phase mixing phenomenon is present in both the polyurethanes but PCL based polyurethane exhibits comparatively lesser degree of mixing than PEG based polyurethane. The crystalline PCL soft segment is more cohesive in nature which prevents the mixing of hard and soft segment at the molecular level whereas relatively amorphous and non-polar PEG soft segment provides more integration in between the different segments. These characteristic features of the polyurethanes indicate that two phase morphology of the polyurethanes are present with variable degree of phase mixing/segregation behavior. The relative crystallinity of the polymers is mainly contributed by the H-bonded hard segment. The DSC analysis of the polyurethanes provides significant information about phase morphology of the polyurethanes. The thermogravimetric analysis (TGA) analysis of the polyurethanes is shown in Figure The TGA analysis shows that these polymers are thermally stable as the 37

59 onset of degradation for PEG-HDI-DTH is around 250 C and that for PCL-HDI-DTH is around 300 C. The earlier onset for PEG based polyurethane is probably due to associated water molecules of the PEG soft segment. Both the polyurethanes exhibit two stage degradation which is qualitatively in agreement with the two phase structure of the polyurethanes Weight (%) PEG-HDI-DTH PCL-HDI-DTH Temperature ( C) Figure 3.11 TGA analyses of L-tyrosine based polyurethanes The melting of the polymers is at relatively lower temperature compared to pure polytyrosine indicates its applicability in the processing of the material for practical purposes of scaffolding in tissue engineering applications. The high degradation temperature indicates that the range of temperature within which the polymers are processible is sufficiently large. 38

60 3.2.7 Mechanical Characterizations The typical stress-strain curve of the polyurethanes is shown in Figure The tensile properties of the polyurethanes are summarized in Table 3.4. Table 3.4 Mechanical properties of the polyurethanes Polymer Ultimate Tensile Strength (MPa) Modulus of elasticity Elongation at break (%) (MPa) PEG-HDI-DTH 2.81 ± ± ± 9 PCL-HDI-DTH 7.05 ± ± ± Stress (MPa) PEG-HDI-DTH PCL-HDI-DTH Strain (%) Figure 3.12 Representative stress-strain curve of L-tyrosine based polyurethanes The mechanical properties of polyurethanes show that PEG based polyurethane is lower in mechanical strength compared to PCL based polyurethane. The mechanical properties of the polyurethanes are mainly controlled by the dominant soft segments. The lower tensile strength, modulus of elasticity and elongation (at break) of PEG-HDI-DTH 39

61 is largely due to amorphous and flexible PEG soft segment compared to relatively more crystalline PCL. The contribution of hard segment is relatively less due to phase mixing of the hard segment with the soft segment. Thus, the mechanical properties of the polyurethanes are more controlled by the soft segment morphology. The difference in the mechanical properties of the polyurethanes can be directly correlated to structure and morphology of the polyurethanes. Polyurethanes with higher degree of phase separation exhibits better tensile properties than the phase mixed polyurethanes. This is probably due to disordering of hard segment domains. As indicated by DSC analysis, crystalline PCL soft segment inhibits phase mixing and therefore leads to more phase segregated morphology leading to higher tensile properties. In addition to this, the effect of molecular weight is directly related to the tensile property. PCL based polyurethane have significantly higher molecular weight which improves the tensile properties compared to the PEG based polyurethane. Moreover, the high hydrophilicity of PEG often leads to lower mechanical property of the polymer. 3.3 Conclusion The synthesis and characterizations of polyurethane based on L-tyrosine based chain extender provides an alternative to develop polymers for biomaterial application. The use L-tyrosine based diphenolic dipeptide, DTH, as a chain extender shows that amino acid based polyurethanes can be developed from biocompatible components e.g. PEG, PCL and aliphatic diisocyanate through easy and simple chemical syntheses. The biphasic morphology of the polyurethane due to the presence of hard and soft segment domains and its effect on the property of the material is demonstrated by the characterization 40

62 studies. The chemical and structural characterization by 1 H and 13 C NMR and FTIR confirms the structure and composition of the polyurethanes. The other physical properties, e.g. molecular weight, solubility etc. shows that L-tyrosine based polyurethanes are potentially useful for fabricating and designing biomedical systems. The thermal characteristics and the mechanical properties of the polyurethanes show that these polymers possess useful material properties for biomaterial application. Moreover, the results show that the composition of polyurethane plays a crucial role in determining the property of the material. In general, these results indicate that L-tyrosine based polyurethanes are useful as biomaterial for tissue engineering applications. 41

63 CHAPTER IV CHARACTERIZATION OF L-TYROSINE BASED POLYURETHANES FOR BIOMATERIAL APPLICATIONS The functional characteristics of biomaterial are required so that the material has specific property to perform the required task 2. The essential characteristics of polymeric biomaterials for tissue engineering application have been discussed in details in Chapter II. The surface characteristics of polymeric material are crucial for cell attachment, proliferation and differentiation. The interaction between the scaffold surface and the cell is largely dependent on the surface characteristics of the materials. Water absorption and permeation characteristics are also important for tissue engineering. These characteristics directly relates to the appropriate cellular environment and for transport of materials to and from the cells. Moreover, the amount of water absorbed is directly related to the degradation of the polymers. Delivery of drug molecules and other active ingredients is crucial in tissue engineering application. The ability of the polymer to release active molecules in response to tissue regeneration is an important characteristic feature. The degradation characteristics of a tissue engineering polymer are the most important feature. Different modes and mechanism of degradation and proper understanding of these mechanisms by mimicking the in vivo environment are important for the tissue regeneration. An extensive characterization these features are crucial for appropriate app- 42

64 -lication of the polymer in tissue engineering application. The use of polyurethanes in tissue engineering application is relatively new area of research. Extensive characterizations of the polyurethanes are required to establish the applicability of these materials for tissue engineering application. Surface characteristics of polyurethanes are interesting due to the two phase morphology and its distribution. The variations in the morphology along with other physical features (e.g. porosity, roughness etc.) of polyurethane surfaces are determining factors for appropriate response of the material to a particular environment. Water absorption and permeation characteristics of polyurethanes are largely guided by their chemical and physical structure. The relative hydrophilicity/hydrophobicity and the interactions between the different segments and their morphological distribution are the guiding parameters for these characteristics. The release characteristics of the drug and other active ingredients have not been investigated in details. The structure and morphology of the segmented polyurethanes plays an important in determining the release pattern and mechanism. In addition to the polyurethane characteristic, the drug-polymer interactions and the distribution of the drug within the polyurethane matrix are also controlling factors for the release of the drugs. The effect of polyurethane on the release characteristics depends on the physical and chemical nature of both the polyurethane and the drug. The degradation of polyurethanes has been researched widely for various biomaterial applications. The stability of polyurethanes is greatly affected due to several factors in a biological environment. The most common pathways for polyurethane degradation are hydrolytic, enzymatic, oxidative and environmental stress cracking 45. The susceptibility of polyurethane degradation is due to the segmented structure of the polymer which is comprised of soft 43

65 and hard segment. To understand the performance of the polyurethanes, it is necessary to investigate the different degradation effects on the material. Polyurethanes are considered to hydrolytically stable under physiological conditions. Hydrolysis of urethane linkages is unlikely unless augmented by catalytic conditions e.g. elevated temperature, presence of ions etc 45. Implantable polyurethanes consists of hydrophobic soft and hard segments and therefore are least affected by hydrolysis. The physical and chemical characteristics of the soft segment primarily control the hydrolysis of the polyurethanes. Polyurethanes with polyether soft segment are hydrolytically stable compared to polyester type due to the presence of hydrolysable ester linkages e.g. polyurethanes based on lactide and/or glycolide soft segment degrades hydrolytically 45,54. The susceptibility of the polyester type soft segment to hydrolysis also depends on several factors. The relative hydrophilicity/hydrophobicity of the soft segment controls the water absorption and therefore is crucial is hydrolytic degradation. In addition, the permeation of water within the polymer matrix to access the hydrolysable bonds of the polyurethane structure also controls the hydrolytic degradation. The contribution of the hard segment in the degradation is also important. The morphological distribution of the phases and the interactions between the soft and hard segment has significant effect on the hydrolysis of polyurethanes. Enzymatic degradations of polyurethanes have received a great attention in the recent years 55,56. Due to the presence of enzymes in physiological environment, it is important to understand the enzymatic degradation characteristics of polyurethanes for biomaterial applications. Polyurethanes with polyether, polyester and polycarbonate soft segments have shown degradation in presence of several hydrolytic (e.g. papain) and inflammatory cell derived (e.g. esterase, elastase) enzymes 55,57. These studies show that 44

66 the soft segments of the polyurethanes are degraded specifically by the enzymes. Moreover, the indirect effect of hard segment domain in the enzymatic degradation is explained in terms domain size and structure 58. The effect of proteolytic enzymes on the degradation has been studied for amino acid based polyurethanes. Phenyl alanine based polyurethanes have shown α-chymotrypsin mediated degradation in vitro conditions 59,60. These studies show that the presence of specific sites enhances the tendency toward enzymatic degradation of the polyurethanes. Polyurethane degradation by oxidative method has been studied widely for different type of polyurethanes. Several studies have been done to investigate the mechanism of oxidation by reproducing the physical and chemical environment in vitro. The presence of oxidative environment in vivo conditions leads to biodegradation of polyurethanes due to the attack from the immune system via macrophages, phagocytes, foreign body giant cells etc 61. In addition, the oxidation of polyurethanes is catalyzed by the presence of metal ions within the polymers or in the environment. Different conditions have been exploited to mimic the in vivo oxidative environment in vitro. Hypohalous and nitric oxide based oxidants have been used to study the degradation of the polyurethanes. The use of hydrogen peroxide (H 2 O 2 )/ cobalt chloride (CoCl 2 ) solution at 37 C reproduces the in vivo oxidative environment 62. Polyurethanes with different soft segment chemistry, e.g. polyether, polyester, polycarbonate has shown variable degree of degradation in oxidative environment. Polyether type urethanes are more susceptible to oxidative attack compared to polyester and polycarbonate based polyurethanes 62,63. Different mechanisms have been proposed for such degradations which show that structural variation of the polymers lead to different pathways of degradation 62. In general, these studies show that the soft segments 45

67 of the polyurethanes are controlling factor in oxidative degradation. The effect of hard segment on the oxidation of the polyurethane is not clear. The focus of this chapter is to characterize the L-tyrosine based polyurethanes for tissue engineering applications. The properties of the polyurethanes those are characterized are contact angle measurement, water vapor permeation, release characteristics, water absorption and different types of degradations. The release characteristics were examined by studying the release of model hydrophobic drug under physiological condition. Among the different types of degradation, hydrolytic degradation was examined by mimicking physiological condition of ph 7.4 and temperature 37 C. In addition to that, the effect of ph on the degradation was tested by using two different ph s in comparison to the neutral ph 7. For this purpose acidic ph of 4 and basic ph of 10 was selected. The degradation was measured by the weight loss of the polymers. The effects of oxidation on the polyurethanes were evaluated by using 0.1 M cobalt chloride solution in hydrogen peroxide of specific strength, which closely mimics the oxidative environment in contact cell-biomaterial interface. The degradation was examined spectroscopically by FTIR analysis and by scanning electron microscope (SEM) images. The enzymatic degradation of polyurethanes was tested by using a proteolytic enzyme α-chymotrypsin at 37 C. The effect of enzymatic degradation was compared to two controls: (i) buffer mediated hydrolytic degradation of L-tyrosine based polyurethanes (ii) enzymatic degradation of non-amino acid based polyurethanes to understand the effect of the enzyme and the amino acid based component in the degradation characteristics. The degradation was measured gravimetrically and its effect 46

68 in the chemical structure and the morphology of the polyurethanes were examined by FTIR and SEM images respectively. 4.1 Experimental The following sections describe the details of the experimental procedures related to the characterization of the polyurethane properties Preparation of Solvent Cast Films The polymer films were cast from 5 wt % solutions of PEG-HDI-DTH in DMF and PCL-HDI-DTH in chloroform. Accurately weighed polymers (~500 mg) was dissolved in 10 ml of solvent and allowed to form a homogeneous solution through constant stirring at room temperature for 24 hours. The polymer solutions were filtered through Teflon syringe filter to remove any undissolved residue and were cast onto poly(tetrafluoroethylene) (PTFE) pertidishes. The solvents were initially allowed to evaporate at room temperature followed vacuum drying at 50 C for another 48 hours to remove the residual solvents. Films of about thickness 0.15 mm were obtained by this method. These films were used for water vapor permeation, water absorption and different degradations experiments Water Contact Angle Thin films of polyurethanes were prepared on thoroughly cleaned and dried glass slides by dip coating the slides into the 5 wt % solution of PEG-HDI-DTH in DMF and PCL-HDI-DTH in chloroform for 12 hours. The films were initially dried at room 47

69 temperature for 24 hours followed by vacuum drying at 50 C for another 48 hours to remove the residual solvents. Water contact angle was measured by sessile method using a Ramé-Hart goniometer at room temperature in an air atmosphere both in advancing and receding modes. The averages of five readings from three different parts of the films were taken for each sample Water Vapor Permeation To measure the water vapor permeability, discs of polymer films were cut and placed on open vials containing 5 gm of silica gel (mesh size 6-16) and held in place with a screw lid having a diameter of 2 cm (test area: 1.33 cm 2 ). The vials were then placed in desiccator containing saturated aqueous solution of sodium chloride to maintain constant relative humidity (R.H. ~75 %, 21 C). The moisture transmitted through the polymeric films was determined gravimetrically over 48 hour period. The rate of water vapor transmitted was calculated from slope of the linear curve of water vapor transmitted versus time plot. The water vapor permeability (WVP) and water vapor permeability coefficient (WVPc) was calculated from the following equation: WVP = W / A. ΔP WVPc = W. t / A. ΔP where, W is the rate of water vapor transmitted, A is the cross sectional area of the film, ΔP is the vapor pressure difference, and t is the thickness of the film. The results reported are average of three values for each polymer film. 48

70 4.1.4 Release Study Release of model hydrophobic drug p-nitroaniline from the polymer films was studied. Accurately weighed amounts of p-nitroaniline and the polymer was dissolved in 10 ml of solvent (DMF for PEG-HDI-DTH and chloroform for PCL-HDI-DTH) such that a 20:1 weight ratio of polymer to p-nitroaniline was obtained. These polymer- p- nitroaniline solutions were used for solvent casting to obtain polymer films. Circular disk sample (diameter: 10 mm and weight: mg) were cut from the films and immersed in 15 ml of phosphate buffer saline (PBS; 0.1 M; ph 7.4) and was incubated at 37 C. The release of p-nitroaniline was measured spectrophotometrically at 410 nm with 1 ml aliquot and the volume was maintained constant at 15 ml by adding PBS. The cumulative release of the p-nitroaniline was measured over 8 week period using the following equation: M i 1 = CiV + Ci Vs where C i is the concentration of p-nitroaniline in the release solution at time i, and V is the total volume of the release solution, V s is the sample volume. The diffusion coefficients of p-nitroaniline in the polymers were calculated using the following equation 67 : M t M 1 2 Dt 2 2 πδ where, M t /M is the fractional mass of p-nitroaniline released at time t, D is the diffusion coefficient, and 2δ is the thickness of the polymer film. 49

71 4.1.5 Water Absorption The circular sample were cut from dried films (diameter: 1.5 cm and thickness: 0.15 mm) and immersed in 20 ml of deionized water at room temperature. At predetermined time intervals the hydrated samples were taken out and weighed after the surface water was blotted with Kimwipes. The water absorption was calculated on the basis of the weight difference of the film before and after swelling. The percentage of water absorption was calculated using the following equation: Water Absorption (%) = w w ) / w 100 ( where, w 2 and w 1 are the weight of sample films after and before being immersed in water, respectively. The averages of three values are reported for each polymer. The effect on water absorption on the dimensional stability of the polymer was assessed by measuring the change in the size and shape of the polymer Hydrolytic Degradation The circular samples (diameter: 1.0 cm and thickness: 0.15 mm) were cut from dried films. The samples were incubated at 37±1 C in 10 ml of phosphate-buffered saline (PBS; 0.1 M, ph 7.4) containing 200 mgl -1 of sodium azide to inhibit bacterial growth in a sealed vial placed within constant temperature water bath. Samples were taken at intervals, weighed for mass loss after drying under vacuum at 40 C for 2 days. The hydrolytic degradation was calculated from the weight loss (%) using the following equation: Weight Loss (%) = w w ) / w 100 (

72 where, w 2 and w 1 are the weight of sample films after and before degradation, respectively. The averages of three values are reported for each polymer. To examine the effect of ph on the degradation of the polyurethanes under hydrolytic conditions, two different buffer solutions were used at ph 4 (acidic) and ph 10 (basic) in comparison to neutral ph 7 solutions. The samples were incubated under similar conditions, and the degradation of the samples was measured by weight loss as described before Oxidative Degradation The polyurethanes films were cut approximately into 1cm 1cm squares with thickness of approximately 1 mm. 0.1 M cobalt chloride solution in 20% H 2 O 2 were prepared from 30% H 2 O 2 solution by proper dilution with distilled water. Different strength of H 2 O 2 solutions (5 % and 10%) were also used to understand the effect of peroxide concentration on the degradation of the polymers. The polymer films were added to these solutions at 37±1 C temperature (physiological body temperature). Samples from each of these solutions were taken out at 3, 7, 15, and 22 days interval and dried in vacuum oven at 40 C for two days prior to any characterization. The films were then characterized by ATR-FTIR and SEM. The test solutions were changed every 7 days to maintain the ionic concentration relatively constant. The degradation products were also analyzed by FTIR analysis of the residue after evaporating the degradation medium. FT-IR characterizations were done in Nexus 870-FTIR fitted with attenuated total reflection (ATR) attachment with germanium crystal. Spectra were collected at a resolution of 2 cm -1 with a sampling area of 3mm 2. The results presented here are the 51

73 average of the three spectrums recorded for each sample, i.e. a total of six spectrums, each with 16 scans. The FTIR presented here represents one of the sample spectrums. The SEM images were recorded on silver sputtered samples in Hitachi S2150 (Operating Voltage: 20 kv) Enzymatic Degradation α-chymotrypsin is a proteolytic enzyme that preferentially cleaves the peptide linkages of amino acid containing hydrophobic group e.g. phenyl alanine, tyrosine etc. It also catalyzes the cleavage of ester bonds. The activity of the enzyme is 47 units/g (one unit of enzyme hydrolyzes 1 micromole of the substrate at specified temperature and ph). The molecular weight of α-chymotrypsin used is and the source is bovine pancreas. The activity of α-chymotrypsin enzyme was measured over the period of 7 days under similar experimental conditions prior to the degradation study of the polyurethanes. The activity of the enzyme was measured estimating the release of p-nitroaniline by the reaction of α-chymotrypsin with the substrate N-succinyl-Ala-Ala- Pro-Phe-p-nitroanilide (Suc-AAPF-pNA). α-chymotrypsin reacts with the substrate to cleave the peptide linkage to release p-nitroaniline. p-nitroaniline was estimated by measuring absorbance at 410 nm in UV-vis spectrophotometer. The decrease in enzyme activity corresponds to decreased release of p-nitroaniline and the percentage decrease in activity is based on the activity at beginning of the experiment. The activity of α-chymotrypsin was measured both as free enzyme and also in presence of the polyurethane to check the effect of polymer and the degraded products on the activity of the enzyme. Enzyme solution of 52

74 concentration 1 mg/ml was prepared in PBS (ph 7.4) and 5.5 mm solution of Suc-AAPFpNA in 5 % (v/v) DMF in water was prepared. 1 ml of enzyme solutions (maintained at 37 C both as free enzyme and in presence of polymers) were taken out definite time intervals and 1ml of the stock substrate solution was added to it. The released p- nitroaniline was measured by UV-vis. The polyurethane films were cut into samples of 1 cm diameter and thickness approximately 0.03 mm and placed in vials containing 10 ml of α-chymotrypsin solution (concentration: 1mg/ml) in PBS (ph 7.4). The vials were placed in constant temperature water bath maintained at 37±1 C temperature. Samples were taken out from these solutions at 0.5, 1, 2, 4, and 6 days interval and dried in vacuum oven at 40 C for two days prior to any characterization. The mass loss of the polymers was measured gravimetrically to examine the effect of enzymatic degradation. Moreover, FT-IR and SEM characterization of the degraded polymer films was studied to analyze the degradation characteristics. The degradation products were also analyzed by FTIR analysis of the residue after evaporating the degradation medium. UV-visible spectra were collected on a Beckman DU640 spectrophotometer. 4.2 Results and Discussion The following sections describe the results of the experiments and its explanation related to the characterization of the polyurethane properties. 53

75 4.2.1 Water Contact Angle Figure 4.1 and 4.2 shows representative images of the water contact angle on the PEG-HDI-DTH and PCL-HDI-DTH surface both in advancing and receding modes. Figure 4.3 shows that the water contact angle values for PEG-HDI-DTH is 33 for the advancing mode and 21.4 for the receding mode while those for PCL-HDI-DTH are 75 and 50.5 respectively. The contact angle values (both advancing and receding) of PEG- HDI-DTH are lower compared to PCL-HDI-DTH indicating that the surfaces of the PEG based polyurethanes are more hydrophilic than PCL based polyurethanes due to hydrophilic nature of PEG. A B Figure 4.1 Water contact angle on PEG-HDI-DTH surface (A) Advancing mode (B) Receding mode A B Figure 4.2 Water contact angle on PCL-HDI-DTH surface (A) Advancing mode (B) Receding mode 54

76 Since the composition of the hard segment is the same for both the polyurethanes, the relative contribution of hard segment on the surface by both the polyurethanes is relatively similar. Moreover, the relatively crystalline soft segments of PCL based polyurethanes lead to decreasing value of contact angles. These results follow the similar trend as observed by Woodhouse et. al. and others 46,48. Contact angle hysteresis is the difference between the advancing contact angle and receding contact angle. Hysteresis of contact angle occurs due to surface heterogeneity which leads to the difference in the surface energy at the microscopic level 2. Surface roughness also leads to the hysteresis of contact angle. Researchers have attributed to the contact angle hysteresis due to the rapid orientation of the surface in order to reduce interfacial tension. The hysteresis values of the polyurethanes indicate the change at the surface of the polyurethanes due to rapid reorientation Adv. Rec. Contact Angle ( ) PEG-HDI-DTH PCL-HDI-DTH Figure 4.3 Water contact angle of L-tyrosine based polyurethanes 55

77 The higher hysteresis value for PCL based polyurethanes (24.5 compared to 11.4 for PEG based polyurethanes shown in figure 4.4) indicates that the surface of the PCL based polyurethane is more heterogeneous compared to the PEG based polyurethane. Increased phase separation of PCL based polyurethanes, as indicated by DSC, also supports that more polar urethane linkages present within the hard segment domain are preferentially oriented towards the surface in response to receding polar water droplet. Thus, the driving force for surface reorientation is higher in PCL based polyurethanes than PEG based polyurethanes where the surface is already hydrophilic in nature. These features indicate that the distribution of the domains at the surfaces is controlled by soft segment and the polar urethane linkages present at the interphasic region. 30 Contact Angle Hysteresis ( ) PEG-HDI-DTH PCL-HDI-DTH Figure 4.4 Contact angle hysteresis of L-tyrosine based polyurethanes Water Vapor Permeation The importance of permeation in tissue engineering application is immense. Figure 4.5 shows the plot of the amount of water vapor transmitted with respect to time for both PEG-HDI-DTH and PCL-HDI-DTH. 56

78 The amount of water vapor transmitted through PEG based polyurethane is higher than PCL based polyurethane due to the hydrophilic nature of PEG soft segment compared to PCL soft segment. The water vapor permeance of polyurethanes (Table 4.1) shows that PEG based polyurethane allows more water to permeate through the polymer films. The hydrophilic and amorphous PEG soft segment enables more permeation than the hydrophobic and relatively crystalline PCL soft segment. However, PEG being a highly water absorbing polymer, absorbs significant amount of water vapor during permeation. For such materials, water vapor permeance is not appropriate to describe the permeation effect. The effect of thickness on water vapor transmission is described by water vapor permeability coefficient, which is also higher for PEG based polyurethane Mass of Water Vapor (mg) PEG-HDI-DTH PCL-HDI-DTH Time (hour) Figure 4.5 Plot of mass of water vapor permeated against time Both of these values indicate that soft segment of the polyurethane plays a significant role in the pe rmeation of water. The hard segment of both polymers being similar, the 57

79 effect of hard segment in the permeation is not clear. For hydrophilic PEG based polyurethane, the mechanism of transmission is divided primarily through adsorption of water, dissolution and diffusion and then desorption. The hydrophobic hard segment, which is partially phase mixed with soft segment, forms a barrier to the permeation of water. But for PCL based polyurethane, the water vapor permeates through non adsorbing pores of the polymer as both the soft segment and hard segment of the polymer is hydrophobic in nature 64. The effect of polymer film thickness therefore decreases water permeation for PEG based polyurethane but has an opposite effect for PCL based polyurethane. Table 4.1 Water vapor permeation of the polyurethanes Polymer Water Vapor Permeance (10 6 mg/hr.mm 2. mm of Hg) Water Vapor Permeability Coefficient (10 6 mg/hr.mm. mm of Hg) PEG-HDI-DTH ± ± 1.14 PCL-HDI-DTH 9.11 ± ± Release Characteristics The structure of p-nitroanil ine is shown in Figure 4.6. p-nitro aniline is moderately soluble in water with a solubility of 0.79 mg/ml at ph 7.4 and at 37 C in phosphate buffer solution 89. The solubility of p-nitro aniline in water is attributed to the contribution of the charge separated structure due to resonance 90. H 2 N NO 2 Figure 4.6 Structure of p-nitroaniline 58

80 The appearance of polymer films with dispersed p-nitroaniline indicated uniform dispersion of the model hydrophobic drug. The cumulative release was calculated for an 8 week period of time bas ed on the total drug released at the end of the time period of the experiment. The fractional mass release was plotted against the square root of time for both the polyurethanes in Figure 4.7. The release patterns were similar for the both PEG and PCL based polyurethane. The majority of p-nitroaniline (about 80%) was released rapidly within the first five hours followed by release of the remaining 20% throughout the 8 week period. This apparently similar release pattern can be explained by the nature of the drug and its probable interaction with biphasic polyurethane. 1.2 PEG-HDI-DTH PCL-HDI-DTH Fractional Release (Mt/M ) Square root of time ( s) Figure 4.7 Plot of fractional release of p-nitroaniline versus square root of time 59

81 The hydrophobic model drug preferentially interacts with the hydrophobic part of the biphasic polymer and is mainly localized in hydrophobic pockets of the polymer matrix 65. The hydrophobic drug p-nitroaniline is likely to form H-bond with the urethane linkages within the hard segment. The hydrophobic part of the PEG based polyurethane is the hard segment whereas that for PCL based polyurethane is both the hard segment and the soft segment. But similar release pattern from both the polyurethanes indicates that the drug is mainly located in the hard segment domain and is released from the hard segment of the polyurethanes. However, for PCL based polyurethane, a fraction of the model drug is distributed in the hydrophobic PCL soft segment domain due to dipolar interactions and H-bonding. Figure 4.8 shows the initial time period of release for the first 6 hours. The initial release (for the first 6 hours) shows a brief lag period for 0.5 hours for both the polymers, indicating a period of hydration followed by relatively slower release rate for PCL based polyurethanes compared to PEG based polyurethane. The lag period for both the polymers are similar which supports that drug-polymer interaction is localized in the hydrophobic part and the release is initiated after hydration of the polyurethane matrix. The slightly slower rate of drug release after the lag period is due to the fact that in PEG based polyurethane additional effect of swelling plays an important role. The diffusion coefficient for p-nitroaniline was determined from the slope of the linear fit of the curve in this region. Interestingly, for both the polyurethanes the diffusion coefficient of p nitroaniline is cm /s. This indicates that the distribution of the drug in both the polyurethanes is similar and is mainly localized in the hard segment domain. Relatively constant and sustained release is achieved for the remaining period of the release. The similar release pattern observed for PEG-HDI-DTH and PCL-HDI-DTH at 60 66

82 the later period indicates that hydrolytic degradation has no practical effect on the release of p-nitroaniline. This suggests the release of p-nitroaniline during this period is controlled by diffusion and change in the domain morphology of the polyurethanes. Moreover, the hydrophobic p-nitroaniline largely interacts with the hydrophobic region of the polymers therefore the mechanism of release of p-nitroaniline is different for the PEG and the PCL based polyurethane. Fractional Releas e (M t/m ) PEG-HDI-DTH R 2 = PCL-HDI-DTH R 2 = Square root of time ( s) Figure 4.8 Initial release characteristics of p-nitroaniline Mechanistically, for PEG based polyurethanes, the drug is mainly released by diffusion controlled mechanism where water molecule penetrates into the polymer matrix leading to the release of the drug. But for PCL based polyurethanes, the swelling of the polymer controls the release pattern. The imbibition of water molecule in predominantly hydrophobic polyurethane facilitates the release of the drug. Moreover, hydrophobic p- 61

83 nitroaniline, which crystalline in nature disrupts the crystallinity of the polyurethane through formation of intermolecular H-bonds. This also allows the drug to be released from the polymer matrix. To better understand the release mechanism, a more general power law equation is used investigate the release of the drug from the polyurethanes 67. M The following equation is used: = M t kt n α where, Mt/M α is the fractional cumulative release at time t, k is the release constant and n is the release exponent signifying the release mechanism. The validity and applicability of the equation is within the range M t /M α <0.6. The experimental data for both PEG-HDIconsidering the lag time. The fitted curve for both PEG-HDI-DTH and PCL-HDI-DTH is DTH and PCL-DTH was fitted into this equation using MS Excel solver. During the fitting, the values of k and n were calculated both by considering the lag time and without shown in Figure (A) PEG-HDI-DTH Fractional R elease (M t /M ) Experimental 0.2 Fitting without lag time 0.1 Fitting with lag time Time(s) Figure 4.9 Fitted curves for (A) PEG-HDI-DTH and (B) PCL-HDI-DTH 62

84 F ractional Release (Mt/M ) (B) PCL-HDI-DTH Experimental Fitting without lag time Fitting with lag time Time(s) Figure 4.9 Fitted curves for (A) PEG-HDI-DTH and (B) PCL-HDI-DTH. Continued The values of the fitted parameters for both the polyurethanes are given in Table 4.2. Table 4.2 Value of fitted parameters k and n PEG-HDI-DTH without lag time with lag time PCL-HDI-DTH without lag time with lag time k (10 3 ) k (10 3 ) n n For slab geometry, n value equal to 0.5 indicates diffusion controlled release mechanism and 1.0 indicates swelling controlled release mechanism, provided the assumptions behind this power law analysis are satisfied 67. PEG based polyurethanes shows n value closer to 0.5 indicates that the drug is released predominantly by a diffusion mechanism. Highly hydrophilic PEG soft segment absorbs a large amount of water and therefore the mobility of the solvent molecules are greater compared to the relaxation of the polymer structure to accommodate the solvent. This implies that solvent is easily imbibed within the polymer matrix an d subsequently the dru g is released by diffusion. The slightly 63

85 higher n value (0. 67) is obtained when the lag time is not considered. Lag time indicates the period of hydration required for the water molecules to penetrate the matrix and initiate the drug release. Thus, higher n value without lag time consideration supports the hypothesis that during this initial period, hydration of the polymer takes place and no drug is released. The same explanation is valid for lower release rate when lag time (k value is 1.35 compared to 2.71) is not considered. The value of n within the range of 0.5 to 1.0 signifies an anomalous release mechanism, which is a combination of both mechanisms. For PCL based polyurethanes, the n value is within this range which indicates that the drug is released by a combination of diffusion and swelling mechanisms. However, the values closer to 1.0 indicate that the release is predominantly controlled by swelling mechanism. The hydrophobic PCL soft segment does not allows water molecules to penetrate within the bulk; therefore, the relaxation of the polymer structure is less compared to the solvent mobility. The relaxation of the polymer structure in PCL based polyurethane refers to two types of relaxation: (i) the crystalline structure of the PCL soft segment and (ii) the interaction of drug with hard segment (and soft segment) of the polyurethane. Thus the swelling of water molecules within the bulk of the polymer controls the release of the drug in PCL based polyurethane through the relaxation of the polymer. No significant difference is observed in the n value depending on the lag time consideration. This indicates that the hydration of the polyurethane is prevalent after the initial lag time. This supports the hypothesis that swelling mechanism (which is the same as hydration) controls the release of the drug from PCL-HDI-DTH matrix. The lower release rate of the PCL based polyurethane indicates that the drug is released slowly compared to the release from PEG based 64

86 polyurethanes. This is directly related to the mechanism which actually controls the release of the drug from the polyurethane matrix. Highly hydrophilic PEG soft segment absorbs more water to facilitate the release compared to PCL soft segment. The difference in the polyurethane structure and the interactions between the drug and the polymer is very important in determining the release characteristics. The similar diffusion coefficient value (of p-nitroaniline) but different n value for the polyurethanes indicates several important features of the release characteristics. The diffusion coefficient of p-nitroaniline probably signifies the release of the drug from the hard segment of the polyurethanes. Since for both the polyurethanes, the hard segment is identical, similar diffusion coefficient indicates that the drug is primarily distributed within the hard segments of the polyurethanes. Therefore, the diffusion coefficient values correspond to the release of the drug from the hard segment of both the PEG and PCL based polyurethane. However, in this case, the n value is an indication of the release mechanisms that characterizes the release, after the drug is diffused out from the hard segment domain of the polyurethane into the soft segment. The soft segment of the PEG based polyurethanes is highly hydrophilic and therefore, the release of the drug from the PEG soft segment is controlled by diffusion (n value closer to 0.5). But for hydrophobic PCL based polyurethane, the release from the soft segment is dominated by swelling mechanism (n value closer to 1.00). Thus, different n values of the polyurethanes indicate the release of the drug from the soft segment to the release medium. This suggests that a two stage release mechanism is operative for the release of hydrophobic drug from the polyurethanes. The first stage corresponds to the diffusion of the drug from the hard segment (which is similar in both PEG and PCL based 65

87 polyurethane) into the soft segment. The second stage corresponds to the release of the drug from the soft segment into the release medium. For the L-tyrosine based polyurethanes, it is the second stage that differs mechanistically due to the different characteristics of the soft segment Water Absorption The water absorption property of the polyurethanes is important for tissue engineering application. Figure 4.10 depicts the amount of water absorbed by the polymer with respect to the time. Amount of water abso rbed (%) PEG-HDI-DTH PCL-HDI-DTH Time (Hours) Figure 4.10 Water absorption of polyurethanes with respect to time Water uptake of the polymers is controlled by the bulk hydrophilicity of the polyurethanes. PEG based polyurethane shows significantly higher water absorption 66

88 values compared to PCL based polyurethane due to the difference in hydrophilicity of PEG. The PCL based polymer absorbs practically no water. The water absorption for PEG based polyurethane is very rapid and reaches a constant value within a short period of time (3 hour). However the decrease in water uptake for PEG-HDI-DTH after a 17 hour period indicates that degradation of polymers dominate over the water absorption. The lower water absorption value of PCL-HDI-DTH can also be correlated to the crystallinity of PCL soft segment compared to relatively amorphous PEG soft segment. The amorphous PEG allows more water to penetrate within the bulk as ordered crystalline soft segment of PCL based polyurethane inhibits the water absorption. Similar results were observed by others for the water absorption results 46,48. Figure 4.11 compares the water a bsorption of the polyurethanes for 17 hour at room temperature of 25 C. 100 Amount of water absorbed (%) PEG-HDI-DTH PCL-HDI-DTH Figure 4.11 Comparison of water absorption (17 hours) The effect of water absorption on the dimensional stability of PEG-HDI-DTH is shown in Figure The images show that after 17 hours in spite of absorbing ~70% of water, the size and shape of the polyurethane discs remains unchanged. 67

89 0 Hour 17 Hour Figure 4.12 Effect of water absorption on dimension for PEG-HDI-DTH This indicates that the polymers polyurethane has porous structures within the bulk which allows accommodating the water molecule in the bulk of the polymer. Moreover, the hydrophobic hard segment of the polyurethanes acts as a crosslink holding the soft segments together. This allows the water molecules to penetrate and confine within the porous voids of the polymer without significant change in the dimension. This fact is of immense significance, particularly for fabrication of scaffolds in tissue engineering applications, as maintaining the dimensional stability of the scaffold during tissue regeneration is very important. 0 Hour 17 Hour Figure 4.13 Effect of water absorption on dimension for PCL-HDI-DTH 68

90 The effect of water on the dimensional stability is on the expected line for PCL-HDI- DTH which retains its size and shape during the water uptake process as shown in Figure Since hydrophobic PCL based polyurethanes do not absorb any significant amount of water, the dimension of the polymer remains unchanged Hydrolytic Degradation Figure 4.14 shows the loss of mass due to hydrolytic degradation over the 8 week period for both PEG-HDI-DTH and PCL-HDI-DTH. The mass loss profile of the polyurethanes due to hydrolytic degradation shows that PEG based polymers degrades at faster rate due to the hydrophilicity of PEG soft segment. About 45% of the mass is lost for PEG based polyurethane compared to only 13% for PCL based polyurethanes within the 8 week period. 50 PEG-HDI-DTH PCL-HDI-DTH 40 Mass Loss (%) Time (Day) Figure 4.14 Mass loss of L-tyrosine based polyurethanes during hydrolytic degradation in PBS (ph 7.4) at 37 C 69

91 The hydrophilicity of PEG soft segment facilitates more water to penetrate the bulk of the polymer and hydrolytically degrade the polymer. In addition to the soft segment chemistry, the morphology of the soft segment plays an important role in the degradation. The amorphous soft segment domain of PEG based polyurethane allows more water to penetrate into bulk enhancing the degradation. The combined effect of soft segment chemistry and morphology controls the degradation pattern of the polyurethanes 48. Moreover, the diffusion and solubility of the degradation products have an obvious effect on mass loss. For PEG based polyurethanes, the polymer matrix is highly swollen and the degraded PEG readily dissolves in PBS whereas for PCL based polymers, the degradation products are largely insoluble, which is reflected on the mass loss profile y = x x x R 2 = Mass Loss (%) y = x x x R 2 = Time (Day) Figure 4.15 Regression analyses for mass loss of L-tyrosine based polyurethanes A simple nonlinear-polynomial regression was performed to fit the data of percent mass loss of polyurethanes with respect to time for the both PEG-HDI-DTH and PCL- HDI-DTH. Figure 4.15 shows the non-linear fit for the polyurethanes. 70

92 The corresponding equation and R 2 value for PEG-HDI-DTH is: y = x x x. R 2 = The corresponding equation and R 2 value for PCL-HDI-DTH is: y = x x x. R 2 = where, y is the numeric value of percent mass loss and x represents time (in days). The regression analysis indicates that a third order regression curve fits the data for the mass loss of the polyurethanes with an acceptable value of R 2. This means that there are multiple sites of degradation for the polyurethanes that undergo hydrolytic degradation. To gain further insight into the degradation characteristics, the rate of degradation i.e. the mass loss with respect to time (dw/dt) was calculated and is plotted against the time. Figure 4.16 shows the plot of dw/dt against time for both the polyurethanes PEG-HDI-DTH PCL-HDI-DTH dw dt Time (Days) Figure 4.16 Plot of mass loss rate with time of L-tyrosine based polyurethanes The rate analysis of the polyurethane degradation shows that for any given time the rate of mass loss is less for PCL-HDI-DTH than PEG-HDI-DTH. This is obvious since PCL 71

93 based polyurethane is hydrophobic and absorbs less water than PEG based polyurethane. The mass loss rate is significantly high for both the polyurethanes at the beginning and thereafter the rate starts decreasing. The initial burst is mainly due to loss of unreacted monomers, and oligomers of the polyurethane system. The trend of a decreasing rate continues for 15 days for both the polymers. After the initial 15 days, the decreasing trend disappears. For PEG-HDI-DTH, the rate starts to increase steadily after initial 15 days and continues for the rest of the period. However, the increase in the rate for the PCL based polyurethanes is much slower with only slight increase in the rate is observed. The hydrolytic degradation of polyurethanes is mainly controlled by the structure and morphology of the polyurethanes. PEG based polyurethanes are hydrophilic and therefore absorbs higher amount of water to undergo rapid mass loss hydrolytic degradation compared to PCL based polyurethane. The initial rapid loss of mass particularly for PEG based polyurethane corresponds to the loss of small monomers, oligomers, salt and entrapped solvent. The effect of morphology of biphasic structure of the polyurethane on the hydrolytic degradation is obvious from the regression analysis and the rate of mass loss. The hydrolysis mainly occurs at the urethane linkages and in the amide and ester linkage present in the DTH chain extender of the hard segment. The ester units of caprolactone in PCL hydrolyze slowly due to hydrophobic methylene groups (-CH 2 -) present in each unit. For PCL based polyurethane the mass loss is slower due to the hydrophobicity and relatively crystalline nature of the polyurethane. Structurally, there are two types of urethane linkages present in the polyurethane chains (Figure 4.17): (i) urethane linkages that connect the polyol (PEG or PCL) and the diisocyanate (HDI) and are present at the interphase of soft and hard segment (ii) 72

94 urethane linkages that connect the diisocyanate (HDI) and the chain extender (DTH) and are present within the hard segment. The urethane linkages which are mainly present within the hard segment are inter-molecularly H-bonded. The interphasic urethane linkages also form H-bonding with the soft segment to form a phase mixed morphology as indicated by the DSC analysis. O O O O O NH (CH 2 ) 5 O NH Urethane linkage between PEG and HDI Urethane linkage between PCL and HDI NH O O Urethane linkage between HDI and DTH Figure 4.17 Different urethane linkages present in the polyurethane The third order regression analysis clearly supports this fact that there are multiple sites of degradation. The initiation of hydrolysis takes place at the interphasic urethane linkages. The interphasic urethane links which are intermixed with the soft segment are mainly susceptible to hydrolysis. For PEG-HDI-DTH, the PEG soft segment absorbs water and the water molecules approach the urethane linkages those are phase mixed to cleave the links through hydrolysis. The initial period of degradation is characterized by the imbibitions of the water molecule to cleave the phase mixed interphasic urethane 73

95 links. As the water uptake for the PEG segment is very rapid (reaches to saturation level within three hours), the hydrolysis of these urethane linkages initiates the hydrolytic degradation and corresponding mass loss at the beginning. However, the mass loss rate decreases with time up to 15 days. This is due to the fact that only a small portion of the interphasic urethane linkages forms a phase mixed morphology with the soft segment. Most of these linkages are cleaved at the beginning and there are no additional linkages available for further cleavage. The decrease in the mass loss rate supports this hypothesis. After the initial 15 days when most of the phase-mixed interphasic urethane linkages are cleaved, the mass loss rate starts increasing. This occurs probably due to the combination of two effects: (i) the urethane linkages present within the hard segment become vulnerable to hydrolytic cleavage and (ii) the free water soluble polyethylene glycol (PEG) soft segment starts to diffuse out from the polymer bulk into the degradation medium. As time progresses, more urethane linkages undergo hydrolysis and the rate of mass loss increases. In addition to this, the amide and ester linkage present within DTH of the hard segment also becomes accessible for the hydrolysis. The period of increasing rate of mass loss is mainly characterized by this feature. Thus, the interphasic (both phase mixed and phase segregated) urethane linkages and the H-bonded urethane links (present within the hard segment) along the hydrolysable amide and ester links within DTH corresponds to the multiple sites for hydrolysis. This explanation corroborates the regression analysis of the mass loss data. For PCL based polyurethanes, a similar phenomenon is observed. First period characterizes decreasing mass loss rate. However, the rate does not increase significantly after 15 days. Moreover, compared to PEG-HDI- DTH the mass loss rate is much lower for PCL-HDI-DTH. This is mainly due to 74

96 hydrophobic nature of the polyurethane. Due to its hydrophobicity, the water uptake of PCL-HDI-DTH is very low. Thus, there is not enough water molecules present in the bulk to hydrolyze the urethane linkages. This is reflected by very a low rate of mass loss and virtually constant rate during the period of 15 to 40 days. The effect of ph of the degradation medium on the degradation characteristics of the polyurethanes are shown in Figure 4.18 for PEG-HDI-DTH and in Figure 4.19 for PCL- HDI-DTH. The practical significance of this analysis indicates that polyurethane structure has significant impact in controlling the degradation rate. The initial slow period followed a relatively faster period of mass loss (particularly for PEG-HDI-DTH) may be significantly important for tissue engineering application to sustain the cell growth and proliferation during tissue regeneration. 35 Mass Loss (%) ph4 ph7 ph10 ph Time (Day) Figure 4.18 Effect of ph on hydrolytic degradation of PEG-HDI-DTH 75

97 The general trend of mass loss in different phs are similar for PEG and PCL based polyurethanes. For both PEG-HDI-DTH and PCL-HDI-DTH, the polyurethanes exhibits relatively less degradation in acidic (ph 4) and neutral (ph 7) conditions compared to basic (ph 10) conditions. More interestingly a slight deviation of ph from neutral condition to slightly basic physiological condition (ph 7.4) significantly increases the degradation of the polyurethane. Any further increase in ph (to ph 10) does not change the mass loss for the polyurethanes. This signifies that the polyurethanes are relatively stable under acidic and neutral conditions but has a higher tendency to hydrolyze under basic conditions, even at slightly basic condition. 15 Mass Loss (%) ph4 ph7 ph10 ph7.4 t Time (Day) Figure 4.19 Effect of ph on hydrolytic degradation of PCL-HDI-DTH The effect of ph is direct consequence of the reaction mechanism of acid catalyzed and based catalyzed hydrolysis. In acid catalyzed, the hydrolysis proceeds through 76

98 protonation and deprotonation steps which slowers the hydrolysis rate in comparison to the base catalyzed hydrolysis where such steps are not present. Moreover, the leaving group is in the form of carboxylate ion in base catalyzed hydrolysis instead of carboxylic acid in acid catalyzed reaction which makes reaction faster leading to more mass loss. This mechanistic difference in the hydrolysis makes the polyurethanes relatively stable in acidic and neutral condition compared to basic condition. This analysis indicates another practical significance of these polyurethanes as stable biomaterials for under the acidic condition e.g. stomach etc Oxidative Degradation A large number of techniques have been used to investigate the structure of both degraded and un-degraded polyurethanes. Oxidative degradation is a predominantly surface phenomenon. The changes on surfaces are different from the interior and even the degradation pattern is not uniform throughout the surface of the polymer films. The complexity in the assignment of peaks arises due to the similarity in the chemical structure of the different (i.e. hard and soft) segments and results in substantial overlap of the peaks in the same region. Therefore, the changes in the structure of the polyurethane can be investigated only through proper selection of a representative peak so that minimum interference of the other peaks is observed. The complete FTIR spectrum reveals the overall structural changes of the polymer and by measuring the changes in the relative intensities of the representative peak(s) with respect to a standard peak offers a way to quantify the change occurring in the polymer structure. 77

99 22 day 7 day 0 day Wavenumbers (cm -1 ) Figure 4.20 FTIR spectra of PEG-HDI-DTH before and after 7 and 22 days of oxidative degradation Table 4.3 ATR-FTIR peaks of PEG-HDI-DTH -1 Wavenumber (cm ) Assignments 1040(sh) C-O stretch in C-O-C=O of urethane + C-O symmetric stretch 1100 Asymmetric C-O-C stretch in aliphatic ether C-N stretch + CH 2 twisting + C-O-C in vinylic ether a C-N stretch + C-O asymmetric stretch 1349 Aliphatic CH 2 wagging 1456 Aliphatic CH 2 bending 1516 Aliphatic CH 2 wagging + bending and urethane/amide N-H bending + C-N stretch 1533 Urethane/amide N-H bending + C-N stretch 1577 a N-H in primary amine + urethane/amide N-H bending + C-N stretch 1618 C=C aromatic stretch + C=C in vinylic ether a 1658 C=O in amide I bonds 1718 Hydrogen bonded C=O in urethane 1730(sh) Non- hydrogen bonded C=O in urethane Aliphatic CH 2 stretch 3332 (b) N-H stretch (Hydrogen bonded) sh = shoulder, a = degradation product, 0 = overlap, b = broad 78

100 The degradation of the polyurethanes due to oxidation is assessed by the change in the structure of the polymer and is examined by FT-IR spectra. The FT-IR spectra of the undegraded polyurethane and the degraded polymers are compared for PEG-HDI-DTH in Figure The spectral assignment 68 for the peaks of the control and degraded PEG- HDI-DTH is shown in Table 4.3. The change in peak position and intensity was analyzed by subtracting the control spectra from the spectra of the degraded sample (22 days) as shown in Figure Subtracted Spectra Absorbance day day Wavenumbers 1400 (cm-1) Wave numbers (cm ) Figure 4.21 Subtraction of spectra for PEG-HDI-DTH Figure 4.21 shows a decrease in the peak height at 1100 cm -1 corresponding to ether of PEG soft se gment and decrease in the peak height at 1533 and 1718 cm -1 corresponding to the urethane linkages. Moreover, a substantial increase in peak heights was observed at 1214, 1577 and 1617 cm -1. These features indicate that the polyurethanes soft segments and the urethane linkages present at the interphase of the soft and hard segment domains 79

101 are affected by oxidation. The increase in 1214 cm -1 peak is attributed to the formation of vinylic ether (C=C-O) and 1617 cm -1 is attributed to the formation of a vinylic double bond (C=C). It is well known that oxidative degradation of polyurethane proceeds via abstraction of α-methylene hydrogen 62. The PEG soft segment contains two methylene groups in between the ether linkages. This structural feature allows abstraction of α- methylene hydrogen from two adjacent methylene groups leading to the formation of vinylic double bond and thus the vinylic ether links. The formation of vinylic ether corresponds to the loss of the aliphatic C-O-C ester stretch (at 1100 cm -1 ). Moreover, the loss of aliphatic α-ch 2 stretch at 2865 cm -1 ( compared to α + β+ γ CH 2 stretch at 2900 cm -1 ) also supports the fact that vinylic double bonds are formed during degradation as shown in Figure α + β+ γ CH 2 α-ch Wave numbers (cm -1 ) Figure 4.22 Change in CH stretch intensity of PEG-HDI-DTH 2 80

102 Cross linking and/or chain scission, as reported for polytetramethylene glycol(ptmo) soft segment 62, is less likely for PEG soft segment as the adjacent α-methylene groups readily form the double bonded structure. However, a small shoulder at 1170 cm -1 indicates the possibility of such cross linking leading to the formation of branched ether. The decreased intensity of the peaks at 1533 and 1718 cm -1 indicates that urethane linkages are degraded by oxidation which also leads to the formation of an amine group which corresponds to the increase in peak at 1577 cm -1. This feature can be indicative of hard segment degradation. It is generally accepted that the urethane linkage present at the interphase of the hard and soft segment are either non-hydrogen bonded or dispersed within the soft segment domain and therefore is more prone to degradation 58. The urethane linkage formed due to the chain extension by DTH is not likely to be affected by o xidation as this urethane link is pres ent in the more crystalline and ordered hard segment domain. The absence of new peak(s) and/or in crease in intensity of the existing peaks in the region of 3500 cm -1 (corresponding to OH) and 1730 cm -1 (corresponding to C=O) indicates that no or minimal generation hydroxyl and carbonyl group. This indicates that chain scission is less likely mode of degradation for PEG based polyurethane. The degradation of the polyurethane due to oxidation is reported as the change in the polymer structure expressed as the percentage change in the peak intensity normalized to a peak which is assumed to be unaffected in the degradation. The peak at 1658 cm -1, assigned to the amide I linkage present in the DTH segment, is assumed to be non degraded as it is present within the ordered hard segment domain that is less likely to be affected by the oxidation. The degradation of the soft segment of the polyurethane is thus expressed by the change in peak intensity at 1617 cm -1 (corresponding to formation of vinylic C=C 81

103 bond) normalized to peak 1658 cm -1 as shown in Figure 4.23 which shows 64 % of the soft segment of the polyurethane is affected in 22 days by oxidation. 80 % Degradation (1617 cm -1 /1658 cm -1 ) (1577 cm -1 /1658 cm -1 ) cm cm Time (Days) Figure 4.23 Degradation of PEG-HDI-DTH in CoCl 2 /H 2 O 2 at 37 C Similarly, the hard segment degradation is represented by the change in peak intensity at 1577 cm -1 (corresponding to formation of amine group) normalized to the peak at 1658 cm -1 as shown in Figure 4.23 which shows 60 % of the urethane linkages (at the interphase of hard and soft segment) are affected in 22 days by oxidation. To verify the effect of hydrogen peroxide in oxidative degradation, 5 and 10% hydrogen peroxide solution were used under similar conditions of 0.1 M CoCl 2 and 37 C. The results indicate degradation represented by a change at 1617 cm -1 is 27 % and 31 % in 5 and 10 % hydrogen peroxide solutions respectively (Figure 4.24). 82

104 % Degradation cm cm % 10% 5% Strength of H 2 O 2 Figure 4.24 Effect of strength of H 2 O 2 in degradation of PEG-HDI-DTH (for 1617 and 1577 cm -1 normalized to 1658 cm -1 ) a. Formation of double bond O CH 2 CH 2 O CH 2 + HO O CH CH O CH 2 O CH CH O CH 2 O CH CH O CH 2 b. Chain Scission O CH 2 CH O CH 2 + HO O CH 2 CH O CH 2 OH HO CH 2 C OH+ CH 2 C H + CH 2 OH c. Cross linking O O O CH 2 CH O CH 2 O CH 2 CH O CH 2 O CH 2 CH O CH 2 d. Degradation of urethane link O CH2 NH2 CH 2 C CH2 NH O + HO NH 2 O C O CH 2 + HO O C Figure 4.25 Mechanism of oxidative degradation of PEG-HDI-DTH 83

105 Similarly, degradation represented by a change in 1577 cm -1 is 23 % and 30 % in 5 and 10 % hydrogen peroxide solution respectively. The dependence of degree of degradation on the peroxide concentration verifies that the degradation is caused by oxidation. Based on FT -IR evidence, a plausible mechanism for the degradation of PEG-HDI- DTH is shown in Figure The soft segment is mainly degraded due to the formation of a vinylic double bond and the hard segment is affected by the degradation of the urethane linkages. Although less probable, a possible mechanistic pathway for cross linking and/or chain scission in the soft segment is also shown. The FT-IR spectra of the un-degraded polyurethane and the degraded polymers are compared for PCL-HDI-DTH in Figure day 7 day 0 day Wavenumbers (cm -1 ) Figure 4.26 FTIR spectra of PCL-HDI-DTH before and after 7 and 22 days of oxidative degradation 84

106 The spectral assignment 68 for the peaks of the control and degraded PCL-HDI-DTH is shown in Table 4.4 Table 4.4 ATR-FTIR peaks of PCL-HDI-DTH Wavenumber (cm -1 ) Assignments 1045 C-O stretch in C-O-C=O of urethane 1100 C-O-C stretch 1167 (1187*) C-(C=O)-O in ester 1213 C-N stretch Aliphatic CH2 wagging 1460 Aliphatic CH 2 bending 1505 Aliphatic CH 2 wagging + bending and urethane/amide N-H bending + C-N stretch 1533 Urethane/amide N-H bending + C-N stretch 1577(sh) Urethane/amide N-H bending + C-N stretch 1620 C=C aromatic stretch 1640 C=O in amide I bonds 1721(sh) Hydrogen bonded C=O in urethane and/or ester 1733 Non- hydrogen bonded C=O in urethane and/or ester Aliphatic CH 2 stretch 3332 (b) N-H stretch (Hydrogen bonded) * = in pure PCL, sh = shoulder, b = broad The change in peak position and intensity was analyzed by subtracting the control spectra from the spectra of the degraded sample (22 days) as shown in Figure The spectrum in Figure 4.26 and 4.27 shows that the peak heights at 1533 and 1640 cm -1 are decreased in the degraded sample compared to the control. No significant increase in peak heights was observed. However, the spectra of the degraded sample at 22 days shows a shift in the peak positions in the region of 1000 to 1500 cm -1. The shifted peak positions are similar to polycaprolactone pure polymer (not shown) and indicate that the degraded sample closely resembles pure PCL which is the soft segment of the PCL-HDI- DTH. The decreased peak height at 1533 cm -1 is indicative of degradation of urethanes linkages and that at 1640 cm -1 indicates that the amide bond present in the DTH of the 85

107 hard segment is degraded. But the effect of degradation on urethane carbonyl around 1720 cm -1 cannot be observed due to the strong peak of the ester carbonyl of the PCL segment. Subtracted Spectra 22 day 0 day Wavenumbers (cm-1) Wave number (cm -1 ) 1000 Figure 4.27 Subtraction of spectra for PCL-HDI-DTH % Degradation (1640 cm -1 /1167 cm -1 ) (1533 cm -1 /1167 cm -1 ) cm cm Time (Days) Figure 4.28 Degradation of PCL-HDI-DTH in CoCl 2 /H 2 O 2 at 37 C 86

108 These features indicate that oxidation leaves the soft segment practically unaffected while the urethane linkages present in the hard segment and/or at the interphase (of hard and soft segment) are degraded. The degradation of the polymer is therefore represented as the change of intensity of the peaks 1533 and 1640 cm -1 compared to the control (Figure 4.28). The peak intensity is normalized to 1167 cm -1 corresponding to ester C- (CO)-O of the soft segment which is assumed to be not degraded by oxidation. It can be seen in Figure 4.28 that 38 % of the urethane linkages and 50 % of the amide linkages are degraded by oxidation. % Degradation cm cm % 10% 5% Strength of H 2 O 2 Figure 4.29 Effect of strength of H 2 O 2 in degradation of PCL-HDI-DTH (for 1640 and 1533 cm -1 normalized to 1167 cm -1 ) To verify the effect of hydrogen peroxide in oxidative degradation, 5 and 10% hydrogen peroxide solutions were used under similar conditions of 0.1 M CoCl 2 and 37 C. The results show degradation represented by a change at 1533 cm -1 is 20 % and 25 % in 5 and 10 % hydrogen peroxide solution respectively (Figure 4.29). Similarly, degradation represented by a change at 1640 cm -1 is 5 % and 6 % in 5 and 10 % hydrogen 87

109 peroxide solution respectively. The results verify that oxidation of PCL-HDI-DTH is primarily due to the effect of hydrogen peroxide. However, absence of any particular trend for the degradation as shown by changes in peak height (for 1533 and 1640 cm -1 peak) indicates the heterogeneous distribution of the domains on the polyurethane surface. a. Chain Scission O CH 2 O C b. Cross Linking CH + HO CH 2 C 2 O OH+ CH 2 C H + CH 2 OH O O CH 2 O C CH 2 + HO CH O C O CH 2 O CH O C CH 2 c. Degradationof urethane linkage CH O C CH 2 O O O CH 2 NH C O C CH 2 + HO O C NH 2 O CH 2 NH2 CH 2 + HO O C Figure 4.30 Mechanism of oxidative degradation of PCL-HDI-DTH Based on the FT-IR evidence, the mechanistic pathway of degradation of PCL-HDI- DTH is shown in Figure FT-IR data suggests the soft segment remains unaffected, but some degradation of the segment is possible by chain scission and/or cross linking. 88

110 FT-IR data suggests that the hard segment of the polyurethane is the primary site for degradation, which is significant compared to soft segment degradation. The residue from the oxidative degradation of both PEG-HDI-DTH and PCL-HDI- DTH are similar as seen from the FTIR analysis in Figure 4.31 which indicates the formation of amine group corresponding to 3340 cm -1 and 1570 cm -1. This suggests the formation of amine group from the degradation of the urethane linkages in the hard segment as proposed in the mechanistic pathway of the degradation of the polyurethanes. PCL-HDI-DTH PEG-HDI-DTH Wavenumber (cm -1 ) 1000 Figure 4.31 FTIR analysis of residue of oxidative degradation (from solution) of L- tyrosine based polyurethanes Figure 4.32 shows the SEM analysis for the change in the surface morphology of polyurethanes compared to the control. Both PEG-HDI-DTH and PCL-HDI-DTH show significant changes in the surfaces of the polyurethanes due to oxidation. Comparison between degraded PEG based polyurethane with that of PCL based polyurethane shows 89

111 that degradation of PEG-HDI-DTH generates a more uneven surface with larger cavities than PCL-HDI-DTH. The PCL based polyurethane shows a more uniform pitted surface. This difference in surface morphology pattern is mainly attributed to the mechanistic differences in the degradation pathway of the polyurethanes. Oxidative degradation of polyurethanes is usually initiated on the surface but can proceed within the bulk in due course of time. But bulk oxidation is highly improbable as free radicals are effective only within limited distances from the surface due to short half lives of the oxidative species 45. Thus, the surface characteristics of the polyurethanes play an important role in oxidation. The contact angle data suggests that the PEG based polyurethane surface is mainly dominated by the soft segment PEG in the oxidative environment due to the hydrophilic nature of PEG. Whereas, for PCL-HDI-DTH, the surface is comparatively more dominated by the hard segment consisting of HDI and DTH. The higher contact angle hysteresis (difference between advancing and receding contact angle) value for PCL-HDI-DTH confirms the same interpretation. The different degradation characteristics of the polyurethanes are supported by the structural features of the polyurethane. The impact of oxidation on the bulk of the polyurethanes cannot be A Figure 4.32 SEM images of polyurethane surface A. Control PEG-HDI-DTH, B. PEG- HDI-DTH after 22 days C. Control PCL-HDI-DTH, B. PCL-HDI-DTH after 22 days for oxidative degradation in CoCl 2 /H 2 O 2 at 37 C 90

112 B C D. 10μm Figure 4.32 SEM images of polyurethane surface A. Control PEG-HDI-DTH, B. PEG- HDI-DTH after 22 days C. Control PCL-HDI-DTH, B. PCL-HDI-DTH after 22 days for oxidative degradation in CoCl 2 /H 2 O 2 at 37 C. Continued 91

113 inferred from these results. However, large cavities on degraded PEG based polyurethanes indicate that PEG based polyurethanes are more susceptible to oxidative attack in the bulk than PCL based polyurethanes. This may be due to the hydrophilicity and less crystalline nature of PEG soft segment compared to PCL soft segment. Thus, PEG based polyurethane mechanistically degrades at the soft segment and at urethanes linkages present at the interphase of hard and soft segment, whereas for PCL based polyurethane the degradation is mainly localized at the interphasic urethane linkages. Figure 4.33 schematically represent the oxidation mechanism of the polyurethanes. Oxidative Solution Segmented Polyurethane PEG-HDI-DTH Oxidative Solution Water/ H 2 O 2 Hard Segment Soft Segment PCL-HDI-DTH Figure 4.33 Schematic representation of oxidative degradation of L-tyrosine based polyurethanes 92

114 4.2.7 Enzymatic Degradation The change in the activity of α-chymotrypsin with respect to time resulting from the incubation in PBS (ph 7.4) at 37 C is shown in Figure % Activ ity loss Free Chymotrypsin Chymotrypsin + PEG-HDI-DTH Chymotrypsin + PCL-HDI-DTH Time (Hours) Suc-AAPF-pNA + α-chymo trypsin PBS ph 7.4 p-nitroaniline 37 C p-nitroanilide (Monitored absorbance peptidic substrate at 410 nm) Figure 4.34 Enzyme activity measurements for free enzyme and in presence of polymer at 37 C in PBS (ph 7.4) The plot shows that the activity of the enzyme initially decreases only by 10 % and remains the same for a 6 day period without any significant change. Further decrease in the activity is observed only after 7 days when the activity of enzyme is reduced to 50 % of the initial activity. This indicates that the activity of the enzyme remains practically constant and unchanged during the period of the degradation study. Moreover, the change in enzyme activity follows a similar trend in presence of the polyurethanes which 93

115 indicates that the polymers and any degradation products do not have any significant effect on the activity of the enzyme PEG-HDI-DTH PCL-HDI-DTH % Mass loss Time (Days) Figure 4.35 Mass loss of L-tyrosine based polyurethanes with time due to enzymatic action The effects of enzymatic degradation on the polyurethanes are shown in Figure 4.35 which represents the mass loss of the polyurethanes with respect to time due to action of α -chymotrypsin. The mass loss for PEG based polyurethanes is significantly higher than that for PCL based polyurethanes. About 55 % of PEG-HDI-DTH mass is lost in 6 days compared to only 6 % for PCL-HDI-DTH. The difference in the degradation profile between the two polyurethanes can be related to the structure of the polymers. α- Chymotrypsin is a proteolytic enzyme and is known to degrade the peptide linkages at the carboxylic side of amino acids having aromatic side groups e.g. tyrosine, phenyl 69 alanine. The use of an L-tyrosine based dipeptide chain extender increases the tendency 94

116 for enzyme mediated degradation due to the presence of hydrolysable amide and urethane linkages in the hard segment domain of the polyurethanes. PEG-HDI-DTH 6 day 1 day 0 day Wave numbers (cm -1 ) PCL-HDI-DTH 6 day 1 day 0 day Wave numbers (cm -1 ) Figure 4.36 FT-IR spectra of PEG-HDI-DTH and PCL-HDI-DTH before and after enzymatic degradation

117 Characteristic FT-IR spectra for the enzymatically degraded polyurethanes, PEG-HDI- DTH and PCL-HDI-DTH, are compared with the un-degraded controls in Figure The decreased intensity of the bands in the region between 1500 and 1650 cm -1 (C-N stretching and N-H bending of urethane/amide linkages) for both PEG-HDI-DTH and PCL-HDI-DTH corresponds to degradation of urethane linkages. This indicates that primarily the hard segments of the polyurethanes are affected by enzymes. Specifically in PEG-HDI-DTH the band 1658 cm -1 assigned to amide linkage present in DTH segment is decreasing showing that the enzymes are capable degrading the amide bond in addition to the urethane linkages. Moreover, the disappearance of the shoulder at ~ 1730 cm -1 indicates that free urethane links are more prone to enzyme attack. Similarly, for PCL- HDI-DTH the decreased intensity of the peak at 1640 cm -1 represents the degradation of the amide linkage in the DTH segm ent. Analysis of the peaks for PCL-HDI-DTH around ~ 1730 cm -1 becomes difficult due to the strong peak of the ester carbonyl group of the caprolactone units. This qualitative analysis of FT-IR peaks shows that both the polyurethanes are degraded by a similar mechanism. But the mass loss profile shows PCL-HDI-DTH degrades at a much slower rate. The effect of enzymatic degradation of L-tyrosine based polyurethanes was compared to two different controls. First the effect of enzyme was examined by comparing the mass loss of L-tyrosine based polyurethanes in buffer (PBS ph 7.4) solution. The buffer mediated degradation of PEG-HDI-DTH shows that PEG based polyurethanes degrade at a significantly faster rate in the presence of α-chymotrypsin, whereas PCL-HDI-DTH shows a similar rate for enzymatic and buffer mediated degradation (Figure 4.37). 96

118 7 6 Buffer mediated Enzymatic Degradation 5 Mass loss (%) PEG-HDI-DTH PCL-HDI-DTH Figure 4.37 Comparison of mass loss between enzymatic and hydrolytic degradation Second control compares the mass loss of L-tyrosine based polyurethanes to the mass loss of polyurethanes based on non amino acid base chain extender 1,4 cyclohexane dimethanol (CDM as shown in Figure 4.38) under similar enzymatic condition. CDM based polyurethanes show significantly less mass loss compared to L-tyrosine based polyurethane indicating that the presence of amino acid based component enhances the enzymatic degradation (Figure 4.39). HO OH Figure 4.38 Chemical structure of 1,4 cyclohexane dimethanol (CDM) Both these controls indicates that effect of mass loss is both due to action of proteolytic enzyme α-chymotrypsin and also due to the presence of amino acid based component which provides enzyme specific site for the degradation. 97

119 70 60 Enzyme-Tyrosine based Enzyme-CDM based 50 Mass loss (%) PEG-HDI-X PCL-HDI-X Figure 4.39 Comparison of mass loss of polyurethanes from tyrosine based chain extender and non-amino acid based chain extender under enzymatic condition This discrepancy in the mass loss profile can be explained in terms of the difference in the soft segment chemistry. Enzyme mediated degradation is primarily located on the surface of the polymer matrix as the macromolecular enzymes are not able to penetrate the matrix. But high water absorption of hydrophilic PEG segment allows the enzyme to penetrate the polymer matrix and induce bulk erosion of PEG-HDI-DTH in addition to the surface degradation. Whereas in PCL-HDI-DTH, the enzymatic degradation is mainly confined to the surface of the polymer. Moreover, the degradation products of PEG based polyurethanes are soluble in water and therefore the polyurethane rapidly experiences mass loss as compared to the PCL based polyurethanes where the degraded polycaprolactone remains with the polymer even after degradation, showing lower mass loss. Moreover the ester group of the caprolactone unit seems to be resistant to hydrolysis 98

120 under action of enzyme, which may be due to the crystallinity of PCL. This fact is further corroborated by FTIR analysis of the residues from the enzymatic degradation as shown in Figure The residue from the enzymatic degradation of PEG-HDI-DTinantly contains PEG soft segment as seen from ~ 1100 cm -1 peak (corresponding predom to ether linka ge of PEG) whereas absence of strong peak ~1730 cm -1 (corresponding ester carbonyl of PCL) in PCL-HDI-DTH indicates that the PCL soft segment is absent in the residue. PCL-HDI-DTH PEG-HDI-DTH Wavenumber (cm -1 ) Figure 4.40 FTIR analysis of residue of enzymatic degradation (from solution) of L- tyrosine based polyurethanes Results reported for polyurethane synthesized from phenyl alanine based chain extenders show a similar trend 60. The degradation rates are comparatively higher for polyurethanes with a PEG soft segment, probably due to the pendant aromatic phenyl ring of the phenyl 99

121 alanine residue. The pendant aromatic ring may provide better enzyme-substrate interaction leading to more degradation. However, for L-tyrosine based polyurethanes the aromatic chain structure is present in the back bone of the polymer chain which might lead to restricted substrate enzyme interactions. However, the trends for degradation of the PCL based polyurethanes from phenyl alanine chain extenders were comparable. The surface morphology of the enzymatically degraded polyurethanes was investigated by SEM analysis and was compared with buffer mediated degradation and enzymatic degradation of CDM (non amino acid based) polyurethanes (Figure 4.41). The surface morphology of polyurethanes shows that α-chymotrypsin induced surface degradation of the polyurethanes, as evidenced by the large porous structures on the surface. Compared to buffer mediated controls, both PEG and PCL based polyurethanes exhibited significant erosion of the surface in the presence of enzymatic solution. The magnitude of surface morphology alteration for both PEG-HDI-DTH and PCL-HDI-DTH supports the hypothesis that these polyurethanes are degraded by a similar mechanism. SEM images show that both the polyurethanes have holes on the degraded surface which indicates a significant amount of bulk degradation, in addition to the surface erosion 70. Comparison of surface morphologies for enzymatic and buffer mediated degradation of PEG-HDI- DTH shows that in the presence of α-chymotrypsin, the polymer surface is eroded to a greater extent, which follows the pattern of mass loss for the polymer. Similar comparison for PCL-HDI-DTH also shows that in presence of enzyme the surface morphology changes to greater extent in the presence of α-chymotrypsin, but the mass loss under enzymatic treatment is similar to that in buffer mediated degradation. The relatively lower mass loss of PCL-HDI-DTH under enzymatic condition is mainly related 100

122 to the insolubility of degradation products in water and may not reflect the actual degradation of the PCL based polyurethane. In addition, comparison of surface morphologies of CDM based (non amino acid) polyurethanes and L-tyrosine based polyurethanes shows that relatively significant amount of degradation has occurred for the L-tyrosine based polyurethanes. This further indicates that presence of L-tyrosine moiety increases the enzymatic degradability of the polyurethanes. A B Figure 4.41 SEM images of polyurethane surface after 6 days A. Buffer mediated PEGbased polyurethane with PEG soft segment D. Buffer mediated PCL-HDI-DTH E. HDI-DTH, B. Enzymatically degraded PEG-HDI-DTH C. Enzymatically degraded CDM Enzymatically degraded PCL-HDI-DTH F. Enzymatically degraded CDM based polyurethane with PCL soft segment [enzymatic degradation in α-chymotrypsin in PBS (ph 7.4) at 37 C and buffer mediated degradation in PBS (ph 7.4) at 37 C] 101

123 C D E Figure 4.41 SEM images of polyurethane surface after 6 days A. Buffer mediated PEG- HDI-DTH, B. Enzymatically degraded PEG-HDI-DTH C. Enzymatically degraded CDM based polyurethane with PEG soft segment D. Buffer mediated PCL-HDI-DTH E. Enzymatically degraded PCL-HDI-DTH F. Enzymatically degraded CDM based polyurethane with PCL soft segment [enzymatic degradation in α-chymotrypsin in PBS (ph 7.4) at 37 C and buffer mediated degradation in PBS (ph 7.4) at 37 C] Continued 102

124 F 10μm Figure 4.41 SEM images of polyurethane surface after 6 days A. Buffer mediated PEG- HDI-DTH, B. Enzymatically degraded PEG-HDI-DTH C. Enzymatically degraded CDM based polyurethane with PEG soft segment D. Buffer mediated PCL-HDI-DTH E. Enzymatically degraded PCL-HDI-DTH F. Enzymatically degraded CDM based polyurethane with PCL soft segment [enzymatic degradation in α-chymotrypsin in PBS (ph 7.4) at 37 C and buffer mediated degradation in PBS (ph 7.4) at 37 C] Continued Segmented Polyurethane Morphology Under enzymatic condition PEG-HDI-DTH Enzyme Hard Segment Soft Segment PCL-HDI-DTH Figure 4.42 Schematic representation of enzymatic degradation of polyurethanes 103

125 Both PEG-HDI-DTH and PCL-HDI-DTH degrades enzymatically by similar mechanism but the mass loss profile is different due to difference in the solubility of soft segment. The mechanism of enzyme is specific and the presence of amino acid based component makes the polyurethane degradable under enzymatic condition. This phenomenon signifies the usefulness of this polymer for tissue engineering application since the degradation characteristic is important criteria for scaffold fabrications. The schematic representation of the enzymatic degradation is shown in Figure Conclusion The characterization of the polyurethanes for the material properties indicates the potential of L-tyrosine based polyurethanes for biomaterial applications including tissue engineering. The surface characteristics of the polyurethanes range from hydrophobic to hydrophilic surfaces depending on the soft segment. The water vapor permeation results indicate the ability of water vapor to permeate through the polymer matrix; and therefore, the use of these polyurethanes are useful for tissue engineering scaffold designing. In addition, the characterizations of release patterns of a model hydrophobic drug, provides important clues about the distribution and interaction of drug and/or other ingredients within the polymer matrix that can be useful is designing scaffolds which calls the release and delivery of drugs/active ingredients for the tissue regeneration. The water absorption and degradation characterization is the other important features of biomaterials for tissue engineering application. The detail analysis of different modes of degradations including the enzymatic and oxidative degradation along with the hydrolytic one provides 104

126 important insights about the material performance for tissue engineering application. All these results in combination provide a strong background for the use of the L-tyrosine based polyurethanes for tissue engineering applications. 105

127 CHAPTER V STRUCTURE-PROPERTY RELATIONSHIP OF L-TYROSINE BASED POLYURETHANES The properties of the segmented polyurethanes are very much dependent on the polyurethane structure and composition 42. Segmented polyurethanes are a unique class of block copolymers of alternating soft segment and hard segments. The soft segment of the polyurethanes consists of polydiol (moderately high molecular weight diol) which is relatively amorphous and rubbery in nature. The hard segment usually consists of the diisocyanate and a low molecular weight diol or diamine chain extender which is relatively crystalline and glassy. Depending on the physical and chemical nature of the segments, polyurethanes exhibits dual phase structure and therefore have an unmatched combination of different properties. The biphasic nature of the segmented polyurethanes arises from the difference in structure, morphology and distribution of the segments. A variety of polydiols, diisocyanates and chain extenders has been used in the synthesis of polyurethanes and their effects on the properties have also been investigated. Polyurethanes are becoming increasingly important biomaterial for tissue engineering applications. Degradable polymers are used for fabrication of 3-D scaffolds for tissue engineering. By altering the structure, polyurethanes with different properties are 106

128 developed for tissue engineering application. Degradable polyurethanes are developed by introducing hydrolysable linkages in the polyurethane structures. The use of hydrolysable soft segments e.g. polylactides, poly (ε-caprolactones) is most common way of developing degradable polyurethanes 71. Amino acid based chain extender has been used to incorporate degradable linkages in the polyurethane backbones 48. The diisocyantes used are mainly aliphatic or amino acid based to avoid the toxic effect of aromatic degradation products. Apart from degradability, these polyurethanes have shown to possess physicomechanical properties that are pertinent to tissue engineering application. Investigation of structure-property relationship of polymers for biomaterial applications is important for designing new polymers and also for the development of existing polymers 40. A methodical study of material dependent responses of biomaterial provides guidelines for selection and optimization of materials for particular use in tissue engineering application. The approach is to develop a library of material by systematic structural variation and to investigate the correlation between the change in polymer structure (and/or composition) with physicomechanical properties. Since polyurethanes are synthesized from three different components, the effect of structural variation and its correlation to the change in the material property will provide a tool to design new biomaterials for tissue engineering applications. The change in polyurethane morphology and phase characteristics is very useful in studying the structure property relationship for polyurethanes. Moreover, structure property relationships of the polyurethanes show that the properties of the material can be changed by altering the soft and hard segment of the polyurethanes. 107

129 HO CH 2 CH 2 O H n Poly ethylene glycol (PEG) HO CH 2 C O CH 2 CH 2 O C CH 2 O 5 5 m O O H n Poly caprolactone diol (PCL) OCN NCO Hexamethylene Diisocyanate (HDI) OCN CH 2 NCO Dicyclohexylmethane 4,4'- diisocyanate (HMDI) HO O CH 2 CH NH C CH 2 CH 2 O C O (CH 2) 5 CH 3 OH Desaminotyrosyl tyrosine hexyl ester (DTH) Figure 5.1 Components used in L-tyrosine based polyurethanes The development of L-tyrosine based polyurethanes with two different soft segments has shown the importance of structure property relationship in designing polyurethanes for biomaterial application. The detailed analysis of structure property relationship of a series of L-tyrosine based polyurethanes with different soft and hard segments with different structural variations will provide better understanding in the development of L- tyrosine based polyurethanes for tissue engineering application. The chain extenders for the polyurethanes are based on L-tyrosine based diphenolic dipeptide, desaminotyrosyl tyrosine hexyl ester (DTH). The effect of soft segment will be analyzed by using either poly ethylene glycol (PEG) or poly caprolactone diol (PCL) of different molecular weights. Two different aliphatic diisocyanates are used: hexamethylene diisocyanate (HDI), a linear diisocyanate and dicyclohexylmethane 4,4'-diisocyanate (HMDI), a cyclic 108

130 diisocyanate. DTH is used as a chain extender for all the polyurethanes. The structures of the components used in the polyurethanes are shown in Figure 5.1. The molecular weight of the PEG is varied by using three different molecular weights of PEG e.g. 400, 600, and Similarly two different molecular weights of PCL is used e.g. 530 and Using these combinations seven different polyurethanes are synthesized as shown in Table 5.1. Since polyurethanes exhibit complex phase behavior, it is reasonable to assume that the use of different soft and hard segments will impact the physicomechanical properties of the polyurethanes. This work aims to relate the effect of structural variation on the properties of L-tyrosine based polyurethane for selection of appropriate tissue engineering material. Table 5.1 Polyurethane composition Code Representative Codes Polyol (Molecular Weight) Diisocyanate PU1 PEG400-HDI-DTH PEG(400) HDI PU2 PEG600-HDI-DTH PEG(600) HDI PU3 PEG1000-HDI-DTH PEG(1000) HDI PU4 PCL530-HDI-DTH PCL(530) HDI PU5 PCL1250-HDI-DTH PCL(1250) HDI PU6 PEG1000-HMDI-DTH PEG(1000) HMDI PU7 PCL1250-HMDI-DTH PCL(1250) HMDI 5.1 Experimental The following sections describe the experimental procedures for the characterization of the polyurethanes related to structure-property relationships. 109

131 5.1.1 Synthesis of polyurethane and casting of films The polyurethanes were synthesized by the conventional two step method. The details of the synthetic process are described in Chapter I. Briefly, polydiol and diisocyanate was added to 50 ml dry DMF (solvent) in the molar ratio of 1:2 and was allowed to react for 3 hours at 110 C in presence of 0.1 % stannous octoate as catalyst and subsequently cooled down to room temperature. To it, DTH was added in the molar ratio of 1:1 to the polydiol and the reaction was allowed to continue at 80 C for another 12 hours. After 12 hours the reaction was quenched by precipitating the polyurethanes in concentrated aqueous solution of sodium chloride. Depending on the nature of the final polymer, the polyurethane was either filtered or centrifuged and washed for several times. The polyurethanes were dried in vacuum at 40 C for three days prior to any characterization. The detailed compositions (weight fractions) of the polyurethanes are shown in Table 5.2. Table 5.2 Weight fraction of different segments in the polyurethanes Hard Segment (wt %) Diisocyanate Code Representative Codes Soft Segment (wt %) PU1 PEG400-HDI-DTH PU2 PEG600-HDI-DTH PU3 PEG1000-HDI-DTH PU4 PCL530-HDI-DTH PU5 PCL1250-HDI-DTH PU6 PEG1000-HMDI-DTH PU7 PCL1250-HMDI-DTH Chain Extender (DTH) The polyurethane films were cast from 5 wt% solution of the polymers in chloroform as the solvent. The solutions were cast in polytetrafluroethylene (PTFE) petridishes and the solvent was allowed to evaporate at room temperature for 24 hours followed drying in 110

132 vacuum oven at 40 C to remove the residual solvent. The polymer films obtained by this process were used for all characterizations except mechanical testing. 10 wt% solutions were used to cast the polyurethane films for mechanical testing Structural Characterizations The molecular weights of the polyurethanes were determined by gel permeation chromatography (GPC) using tetrahydrofuran (THF) as solvent and polystyrene as internal standard. FT-IR analysis of the polyurethanes was performed with a Nicolet NEXUS 870 FT spectrometer for neat samples with 16 scans Thermal Characterizations The thermal characteristics of the polyurethanes were examined by differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA). DSC was performed with a DSC Q100V7.0 Build 244 (Universal V3. 7A TA) instrument at a scanning rate of 10 C/min from -80 to 250 C. TGA was performed with a TGA Q50V5.0 Build 164 (Universal V3. 7A TA) instrument from 0 to 600 C under nitrogen at a rate of 20 C/min. An average of 10 mg of solid sample was used for both the experiments Mechanical Characterizations The tensile properties of the films were measured by Instron Tensile Testing Machine with a load cell of 100 N and cross-head speed of 100 mm/min at room temperature. The sample dimension was 20 mm 6 mm ~ 0.3 mm with free length of 10 mm. 111

133 5.1.5 Water Contact Angle For contact angle measurement, thin films of polymers were prepared on thoroughly cleaned and dried glass slides by dip coating the slides into the 5 wt % solution of polyurethanes for 12 hours. The films were initially dried at room temperature for 24 hours followed by vacuum drying at 50 C for another 48 hours to remove the residual solvents. Water contact angle was measured by sessile method using a Ramé-Hart goniometer at room temperature in an air atmosphere both in advancing and receding modes Water Vapor Permeability The water vapor permeability of the polyurethanes was measured by calculating water vapor permeance (WVP) and water vapor permeability coefficient (WVPc). Discs of polymer films were cut and placed on open vials containing 5 gm of silica gel (mesh size 6-16) with a screw lid having a diameter of 2 cm (test area: 1.33 cm 2 ) and then placed in desiccator maintained at constant relative humidity (R.H. ~75 %, 21 C). The moisture transmitted through the polymeric films was determined gravimetrically over a 48 hour period. The rate of water vapor transmitted was calculated from slope of the linear curve of water vapor transmitted versus time plot. The water vapor permeability (WVP) and water vapor permeability coefficient (WVPc) was calculated from the following equation: WVP = W / A. ΔP WVPc = W. t / A. ΔP 112

134 where, W is the rate of water vapor transmitted, A is the cross sectional area of the film, ΔP is the vapor pressure difference, and t is the thickness of the film. The results reported are average of three values for each polymer film Water Absorption To measure water absorption, circular sample were cut from dried films (diameter: 1.5 cm and thickness: 0.15 mm) and immersed in 20 ml of deionized water. After 12 hours, the hydrated samples were taken out and weighed after the surface water was blotted with Kimwipes. The water absorption was calculated on the basis of the weight difference of the film before and after swelling. The percentage of water absorption was calculated using the following equation: Water Absorption (%) = w w ) / w 100 ( where, w 2 and w 1 are the weight of sample films after and before being immersed in water, respectively. The time period of 17 hour was chosen because the polyurethanes exhibit substantial hydrolytic degradation after 17 hours Hydrolytic Degradation For hydrolytic degradation, similar circular samples (diameter: 1.0 cm and thickness: 0.15 mm) were cut from dried films. The samples were incubated at 37±1 C in 10 ml of phosphate-buffered saline (PBS; 0.1 M, ph 7.4), containing 200 mgl -1 of sodium azide to inhibit bacterial growth, in a sealed vial placed within constant temperature water bath. Samples were taken at intervals, weighed for mass loss after drying under vacuum at

135 C for 2 days. The hydrolytic degradation was calculated from the weight loss (%) using the following equation: Weight Loss (%) = w w ) / w 100 ( where, w 2 and w 1 are the weight of sample films after and before degradation, respectively Release Characteristics Release of model hydrophobic drug p-nitroaniline from the polymer films was studied. Accurately weighed amount of p-nitroaniline and the polymer was dissolved in 10 ml of solvent (chloroform) such that a 20:1 weight ratio of polymer to p-nitroaniline was obtained. These polymer- p-nitroaniline solutions were used for solvent casting to obtain polymer films. Circular disk sample (diameter: 10 mm and weight: mg) were cut from the films and immersed in 15 ml of phosphate buffer saline (PBS; 0.1 M; ph 7.4) and was incubated at 37 C. The release of p-nitroaniline was measured spectrophotometrically at 410 nm with 1 ml aliquot and the volume was maintained constant at 15 ml by adding PBS. The fractional cumulative release of the p-nitroaniline was measured over a 30 day period using the following equation: R i M = i L where, R i is the fraction of cumulative release on i th day, M i is the cumulative amount of p-nitroaniline released on i th day and L is the theoretical loading of p-nitroaniline. The fractional release of p-nitroaniline (R i ) is plotted against the square-root of time ( t). 114

136 Statistical Analysis The statistical analysis of the data was performed by using a generalized linear model of ANOVA with Minitab 15 software. Statistical analysis was performed specifically to examine the significant change in the polyurethane property by the effect of structural variation. The effect of structure was analyzed by four categories :(i) Effect of soft segment diol (ii) Effect of diisocyanate (iii) Effect soft segment PEG molecular weight and (iv) Effect soft segment PCL molecular weight. Results with p value less than 0.05 (p<0.05) was considered to be statistically significant. In general, the p values were used to interpret the effect of structural variations on the properties of the polyurethanes. 5.2 Results and Discussion The following sections describe the results of the characterizations of the polyurethanes and its explanation with respect to structure-property relationships Molecular Weight The molecular weights of the polyurethanes are shown in Table 5.3. Table 5.3 Representative molecular weight of polyurethanes Polyurethane Representative Codes M n M w Poly Dispersity Index PU1 PEG400-HDI-DTH 4,710 11, PU2 PEG600-HDI-DTH 7,520 12, PU3 PEG1000-HDI-DTH 78,980 98, PU4 PCL530-HDI-DTH 12,530 25, PU5 PCL1250-HDI-DTH 150, , PU6 PEG1000-HMDI-DTH 93, , PU7 PCL1250-HMDI-DTH 64,670 75,

137 The molecular weight of PU6 is low due to the difficulties encountered in filtering the polymer solution while determining molecular weight. The results show that PCL based polyurethanes are of comparatively higher molecular weight than the PEG based polyurethanes. This is mainly due to the presence of water with PEG that leads to low molecular weight 48. Lower soft segment molecular weight results in low molecular weight polyurethane in spite of having higher hard segment content 49. This indicates that the chain extension through DTH is random and the hard segment length is comparatively smaller in PU3, PU4, and PU5 compared to the rest of the polyurethanes. The higher polydispersity index of PU1, PU2 and PU4 is also an indication of uncontrolled polymerization reaction. The effect of structural variation of the diisocyanate is not evident from the molecular weight which indicates that the diisocyanate structures have practically no effect on the molecular weight. In general PEG based polyurethanes are tacky and soft compared to PCL based polyurethanes which are relatively stronger FTIR Analysis Figure 5.2 to 5.4 shows the FTIR analysis of polyurethanes. Figure 5.2 shows the effect of molecular weight of PEG soft segment. Figure 5.3 shows the effect of molecular weight of PCL. Figure 5.4 shows the effect of diisocyanate structure on the polyurethanes. The characteristic soft segment peak for PEG based polyurethanes is around 1100 cm - 1 representing aliphatic ether group (of PEG) and for PCL based polyurethane is around 1730~1725 cm -1 representing ester carbonyl group (of PCL). The effect of soft segment 116

138 A PU1 PU2 PU Wave numbers (cm -1 ) B PU1 PU1 PU2 PU2 PU3 PU1 (PEG400-HDI-DTH) PU2 (PEG600-HDI-DTH) PU3 (PEG1000-HDI-DTH) PU Wave numbers (cm -1 ) Figure 5.2 FT-IR absorbance spectra of polyurethanes (A) Series based on different molecular weight of PEG (B) Enlarged in the region and 1200 cm

139 A PU4 PU Wave numbers (cm -1 ) B PU4 PU4 (PCL530-HDI-DTH) PU25 (PCL1250-HDI-DTH) PU Wave numbers (cm -1 ) Figure 5.3 FT-IR absorbance spectra of polyurethanes (A) Series based on different molecular weight of PCL (B) Enlarged in the region cm

140 PU7 (PCL1250-HMDI- DTH) PU5 (PCL1250-HDI- DTH) PU6 (PEG1000-HMDI- DTH) PU6 (PEG1000-HDI- DTH) Wave numbers (cm -1 ) Figure 5.4 FT-IR absorbance spectra of polyurethanes of series based on different diisocyanates molecular weight for PEG based polyurethanes shows that increasing molecular weight leads to increasing H-bonding of the urethane carbonyl within the hard segment domain leading to a cohesive and ordered hard segment. The appearance of peak at 1702 cm -1 in PU1 and PU2 in addition to 1718 cm -1 compared to the single peak at 1715 cm -1 in PU3 shows that a fraction of urethane carbonyl is non H-bonded in PU1 and PU2. However no significant shift of N-H peaks around 3320 cm -1 indicates that the ether linkages are H- bonded and that leads to certain degree of phase mixing in PU1 and PU2 which is relatively less in PU3. For PEG based polyurethanes considerable phase-mixed morphology is obtained with low molecular weight soft segment which is comparable to observations made by other researchers 42,75.The phenomenon of phase mixing behavior at 119

141 low molecular weight of PEG is further supported by the appearance of asymmetric C-O stretch at ~1250 cm-1 in PU1 and PU2 compared to PU3. This indicates ether oxygen of PEG forms H-bonded structure with the urethane linkages to give rise to asymmetric C-O stretch which leads to a higher degree of phase mixed behavior. Similar effects due to the effect of molecular weight of PCL soft segment is less pronounced. The presence of strong ester carbonyl overlaps the peak due to the urethane carbonyl group. However, in PU4 a peak around 1690 cm -1 is observed which might represent H-bonded carbonyl group. This peak is either absent or merged with the strong ester carbonyl absorbance at 1726 cm -1 in PU5. This indicates that a fraction of ester carbonyl is H-bonded in PU4 due to phase mixing between the hard and soft segments. N C O H O O H O C N N C O H O O O C N C O H O O C urethane-urethane urethane-ester urethane-ether Figure 5.5 Hydrogen bonding interactions in the polyurethanes 2 This is indicative that with low molecular weight of PCL the degree of phase mixing between the hard segment and the soft segments increases. In general, the shift of ester carbonyl peak for PCL based polyurethane from 1730 cm -1 in PU5 to 1726 cm -1 in PU4 is attributed to the more crystalline nature of high molecular weight PCL and lattice effects 49,76. The effect of diisocyanate structure on the polyurethane characteristics is not 120

142 evident from the FTIR analysis. A comparison of the spectra in Figure 5.4 does not reveal any significant information about the polyurethane characteristics. The analysis of the FTIR spectra reveals important facts regarding the effect of the polyurethane structure on the morphological characteristics of the polymer. Figure 5.5 shows the different types of H-bonding interactions that are present within the polyurethanes. These interactions indicate that the hard segment of the polyurethanes is coherently associated to form ordered structures through urethane-urethane H-bonds. Moreover, the amide and ester linkage in DTH section of the hard segment also contributes to the formation of H-bonding leading to ordered structures. Moreover, the additional H-bonding interactions, shown in Figure 5.5, reveal the other possible interactions between the hard segment and the soft segment of the polyurethanes. For PEG based polyurethanes, the decreasing molecular weight of the segment leads to more phase mixed behavior. From Table 5.2, it is evident that that among the series of PU1, PU2, and PU3, PU1 has the highest hard segment content and PU3 has the lowest hard segment content. The increasing hard segment content leads to the formation of continuous hard segment domain in which the soft segment is discretely dispersed. This shows that with decreasing PEG molecular weight there is increasing hard segment content in the polyurethanes which leads to formation of more phase mixed morphology. The molecular weight of the soft segment is critical for the crystallization of the soft segments. Low molecular weight PEG is mainly amorphous in nature due to shorter chain length. The absence of enough soft segment crystallization acts as the driving force for the mixing of the hard and soft segments. The less cohesive amorphous soft segment tends to form a phase mixed morphology. It appears that PEG molecular weight of

143 is critical to induce soft segment crystallinity. Thus the soft segment crystallinity in PU3 prevents the formation of phase-mixed morphology. Thus higher the molecular weight of PEG, lower is the phase mixing characteristics in PEG based polyurethanes. The appearance on non-h bonded carbonyl and asymmetric C-O stretch peaks in FTIR analysis supports the formation of urethane-ether H-bonding interaction to form a phase mixed morphology. The increasing interaction between the hard and soft segment of the polyurethanes has a direct consequence on the hard segment characteristics. More phase mixed morphology of the polyurethanes means more interaction between the hard and soft segment through urethane-ether H-bonds. This indicates that the hard segment of low molecular weight PEG based polyurethanes are less coherent due to less urethaneurethane H-bonding. This in turn reduces the ordered structure of the hard segments and the hard segments are relatively less crystalline in PU1 and PU2 compared to PU3. The shorter soft segment probably leads to shorter hard segment as evident from the molecular weight of the polyurethanes as shown in Table 5.3. This characteristic leads to a random orientation of the polymer chain and therefore also contributes to less number of H-bond in the hard segment domain. Therefore, the hard segment crystallinity reduces with decreasing molecular weight of the PEG soft segment. This in general reduces the polyurethane crystallinity and polymers become more amorphous in nature. Figure 5.6 shows the phase characteristics of the polyurethanes with the variable composition of hard segment/soft segment of the polyurethane. From Figure 5.6 it is clear that at a low concentration of either component, the two phases form discontinuous immiscible phase morphology and with increasing concentration, at some optimal point, the two phases become continuous phase mixed morphology. The phase characteristics of PCL based 122

144 polyurethanes can be explained through similar explanation. With decreasing molecular weight of PCL soft segment, the crystallinity is reduced which favors phase mixing through more urethane-ester H-bonding. In addition to that, higher hard segment content in low molecular weight PCL based polyurethane, i.e. PU4, leads to the formation of continuous hard domain as shown in Figure 5.6. This phase mix morphology in turn reduces the hard segment crystallinity by reducing the number of urethane-urethane H- bonds. Hard Segment Soft Segment Hard Domain Soft Segment Hard Segment Phase separated domain morphology Phase mixed domain morphology Figure 5.6 Phase morphology of polyurethanes 42 Thus, the polymers are more amorphous with decreasing molecular weight of PCL soft segment. However, the phase mixing in PCL based polyurethanes are comparatively lesser than PEG based polyurethanes as the PCL soft segment is more ordered due to 123

145 dipolar interactions of the ester carbonyl of the caprolactone unit. But this generalization is not valid for low molecular weight soft segment. As the soft segments at low molecular weight becomes more amorphous, the extent of phase mixing is greater in PCL based polyurethanes compared to PEG based polyurethanes. This is probably due to stronger urethane-ester H-bonds (between hard and soft segment of PCL based polyurethane) than urethane-ether H-bonds (between hard and soft segment of PEG based polyurethane). The difference in the phase behavior of the polyurethanes with different soft segments and molecular weights reveals a complex structure-property relationship of the polyurethanes. The effect of cyclic diisocyanate on the polyurethane morphology is not evident from FTIR analysis but most likely cyclic structure of HMDI leads to more amorphous hard segment due to less packing and less number H-bonds Thermal Characterizations The differential scanning calorimetry (DSC) of the polyurethanes is shown in Figure 5.7. Figure 5.7(A) shows the effect of soft segment while figure 5.7(B) shows the effect of different diisocyanate. A comparison of thermograms for PU1, PU2 and PU3 shows that the molecular weight of PEG soft segment has a significant effect on the thermal properties and morphology of the polyurethanes. The soft segment T g (glass transition temperature) increases with decreasing molecular weight of PEG. The T g values for the soft segment are -15, -26 and -40 C in PU1, PU2 and PU3 respectively. The decrease in soft segment T g indicates that increasing molecular weight of PEG soft segment leads to lesser phase mixing between 124

146 A PU4 (PCL530-HDI-DTH) Endothermic PU2 (PEG600-HDI-DTH) PU1 (PEG400-HDI-DTH) PU5 (PCL1250-HDI-DTH) PU3 (PEG1000-HDI-DTH) Temperature ( C) B B PU7 (PCL1250-HMDI-DTH) Endothermic Endothermic PU6 (PEG1000-HMDI-DTH) PU5 (PCL1250-HDI-DTH PU3 (PEG1000-HDI-DTH) Temperature ( C) Figure 5.7 DSC thermograms of polyurethanes (A) Series based on different molecular weight of PEG and PCL (B) Series based different diisocyanates 125

147 the hard and soft segment of the polyurethanes. Similar results have been observed for other polyurethanes with different molecular weight soft segments 48,53,72. The presence of three endotherms in PU3 corresponds to disruption of short range and long range order of hard segments and melting of crystalline hard segments. These additional endotherms are different in PU1 and PU2. For PU1 only one broad endotherm was observed around 150 C representing the melting of the polyurethane whereas two endothermic transitions are observed for PU2 around 48 C and a broad one around 118 C. The first endothermic transition is probably due to a disruption of ordered hard segments and/or H-bonding interactions between the hard and soft segment. The second endotherm indicates melting of the polyurethane. This indicates that even with high hard segment content of PU1 and PU2, the hard segment is relatively less ordered and shows considerable phase mixing. The hard segments of PU1 and PU2 are relatively more amorphous. Thus the increasing molecular weight of PEG soft segment leads to more phase segregated morphology with a relatively ordered hard segment. The effect of molecular weight of PCL soft segment is less compared to PEG. The soft segment T g for PU4 is at -35 C compared to -37 C of PU5. This shows that there is significant phase mixing but it is practically unaffected by the change in molecular weight of PCL. This is in contrast to the observations made on polyurethanes based on PCL of molecular weight 530 by Skarja et. al 48. This is probably due to the asymmetrical lysine based diisocyante and the pendant group of the phenyl alanine based chain extender. But the effect of PCL molecular weight is consistent with other observations where PCL soft segment beyond molecular weight range is mostly phase separated 88. However, two endothermic transitions are observed for PU4 at 48 C and 66 C in comparison to four in PU5 at 0 C, 31 C, 52 C and 173 C. The 126

148 endotherms of PU4 correspond to a disruption of interactions between hard and soft segments and within the hard segment. The additional endotherms of PU5 represent soft segment and hard segment melting. Comparison of endotherms for PU4 and PU5 shows that PCL soft segment with higher molecular weight exhibits soft segment crystallinity and also leads to more crystalline and ordered hard segment. Although the change in soft segment T g is not appreciable but phase mixed morphology of PU4 is evident from the absence of melting endotherms. Comparison of PU3 and PU5 thermograms shows that at comparable soft segment molecular weight, PCL based polyurethanes exhibits soft segment crystallinity compared to PEG based polyurethane 48. But similar comparison of PU2 and PU4 shows that interaction between the soft and hard segments in much stronger in PCL based polyurethanes where H-bonding with ester carbonyl is stronger than H-bonding with ether oxygen of PEG soft segment 73. The effect of diisocyanate structure has significant impact on the polyurethane morphology 74. Comparison of thermograms of PU6 to PU3 and of PU7 to PU5 shows that changing from linear to cyclic structure changes the hard segment morphology. The soft segment T g of PU3 is - 40 C and that of PU6 is -28 C. This shows the extent of phase mixing is more in PU6. However the soft segment T g for PU7 is -39 C which very similar to PU5, indicating similar extent of phase mixing. This is probably due to the crystalline nature of PCL compared to PEG. But PU6 exhibits a small endotherm at 7 C compared to three endotherms of PU3. The absence of melting endotherm in addition to the other endotherm (around 50 C) in PU6 shows that with cyclic diisocyanate the hard segment is relatively less ordered and more amorphous in nature. The cyclic structure of the diisocyanate and the chain extender (DTH) prevents close packing of the hard segment 127

149 leading to a relatively amorphous nature. The small endotherms around 7 C is due to disruption of some order of the hard segment and the interactions between the soft and hard segment. The similar comparison of thermograms of PU5 and PU7 show that cyclic diisocyanate prevents hard segment crystallization leading to a nearly complete amorphous hard segment. The effect of structural variation in the soft segment and hard segment of the polyurethanes as analyzed by DSC analysis supports the observations from FTIR analysis. For both PEG and PCL based polyurethanes, the degree of phase mixing increases with decreasing molecular weight of the soft segment. As the molecular weight of the polyol decreases, the soft segment is largely amorphous which means more interactions between the soft and hard segment leading to phase mixed morphology. For PEG based polyurethanes there are more urethane-ether H-bonding and less urethaneurethane H-bonds which allows the formation of a phase mixed morphology. This in turn leads to less ordered and relatively less crystalline hard segment. Moreover, increasing hard segment contents forms a hard segment domain in which the soft segment is dispersed to form a continuous one phase structure. This fact is further supported by the absence of endotherms in the region 0-50 C in PU1 which indicates that short range and long range order of the hard segments are absent. Similar explanation is applicable for PU2 in which only one endotherm is observed in this region around 48 C. Whereas, for PU3 two distinct endotherms are observed in this region. Thus with increasing molecular weight of PEG the hard segment is ordered and more crystalline in a phase segregated morphology. For PCL based polyurethanes similar effects are observed. At low molecular weight of PCL, the polyurethane forms a phase mixed morphology due more 128

150 amorphous soft segment. This leads to formation of more urethane-ester H-bonds (the hard and soft segment) compared to urethane-urethane H-bonds (within the hard segment) which lower the crystallinity and ordering of the hard segment. The absence of melting endotherm of PU5 supports this fact. Comparison of thermograms of PCL and PEG based polyurethane at comparable soft segment molecular weight reveals two interesting phenomenon. At low molecular weight, the degree of phase mixing is higher in PCL based polyurethanes compared to PEG based polyurethanes. This is due to the amorphousness of the low molecular weight of the polyol. With more amorphousness, the less cohesive soft segment easily interacts with the hard segment through H-bonding. Comparison of PU1, PU2 and PU4 shows that phase mixing is more evident in the PU4 which is composed from low molecular weight PCL. As the molecular weight is low, the interaction between the soft and hard segment in PU4 is more effective due to stronger urethane-ester H bonds compared to urethaneether H-bonds. But a different observation is made when the molecular weight of the soft segment is high. At a higher molecular weight of the polyol, the degree of phase mixing is lower in PCL based polyurethanes compared to PEG based polyurethanes. At a higher molecular weight, both PEG and PCL are relatively crystalline due to more ordered structure. But a comparison of PU3 and PU5 shows that at high molecular weight of the polyol, the degree of crystallinity is relatively more in PCL based soft segment ( as evidenced by the presence of soft segment melting endotherm at 31 C in PU5). The increased crystallinity of PCL based segment is due to the dipolar interaction of the ester carbonyls in the caprolactone units. This increased crystallinity of PCL soft segment leads to more cohesive soft segment in PU5 which inhibits the soft and hard segment to 129

151 mix into one phase morphology. Similar observation is made from FTIR analysis. Thus at higher molecular weight, although PEG exhibits certain degree of crystallinity (at comparable molecular weight) compared to PCL but is not sufficient to inhibit the mixing of the two phases. The effect of diisocyante structure on the morphology of the polyurethane is also significant. Cyclic structure leads to the formation amorphous hard segment domain. This is evident from the absence of hard segment melting endotherm in PU6 and PU7. Moreover, only one endotherm is observed for PU6 and PU7 around 7 C compared to the two endotherms of PU3 and PU5 respectively (in the region of 0-50 C). This indicates that short and long range orders of the hard segment are absent with cyclic diisocyanate. However, this change in the hard segment morphology does not increase the phase mixing characteristics of the polyurethanes either for PEG based (comparison between PU3 and PU6) or for PCL based (comparison between PU5 and PU7) polyurethanes compared to linear diisocyante (HDI) based hard segment. This is probably due to irregular structure of the hard segment that inhibits any increased degree of phase mixing. The hard segment with HMDI and DTH is highly irregular due to two cyclohexyl rings of HMDI and two phenyl rings of DTH in the polyurethane structure. Thus, DSC analyses show that depending segment structures and compositions, the polyurethanes exhibit variable degree of phase behavior and morphology. The TGA analysis of all the polyurethanes shows very a similar pattern as shown in Figure 5.8. For PEG based polyurethanes the initial 30% of weight are lost slowly followed by relatively faster degradation and for PCL based polyurethanes the initial 70% of weight are lost slowly followed by relatively faster degradation. But, in general all the 130

152 polyurethanes show the onset of degradation around 300 C indicating the stability of the polyurethane up to a temperature of 300 C. The effect of structural variation has practical effect on the degradation temperature of the polyurethanes. All the polymers show reasonable thermal stability indicating the usefulness of the polyurethanes for thermal processing applications. The thermal characteristics of the polyurethanes indicate a wide range of temperature within which the polymer can be thermally processed for scaffold fabrication in tissue engineering application. 100 Weight (%) PU4 PU5 PU1 (PEG400-HDI-DTH) PU2 (PEG600-HDI-DTH) PU3 (PEG1000-HDI-DTH) PU4 (PCL530-HDI-DTH) PU5 (PCL1250-HDI-DTH) PU6 (PEG1000-HMDI-DTH) PU7 (PCL1250-HMDI-DTH) PU1 PU3 PU2 PU Temperature ( C) Figure 5.8 TGA analyses of L-tyrosine based polyurethanes PU Mechanical Properties of Polyurethanes The mechanical properties of the polyurethanes are summarized in Table 5.4. The representative stress-strain curves of the polyurethanes are shown in Figure

153 Table 5.4 Mechanical properties of polyurethanes (mean ± SD, n = 5) Polyurethane Ultimate Tensile Strength (MPa) Modulus of elasticity (MPa) Elongation at break (%) PU1(PEG400-HDI-DTH) 0.47 ± ± ± 7.6 PU2(PEG600-HDI-DTH) 0.93 ± ± ± 0.8 PU3(PEG1000-HDI-DTH) 2.81 ± ± ± 9 PU4(PCL530-HDI-DTH) 0.53 ± ± ± 11.3 PU5(PCL1250-HDI-DTH) 7.05 ± ± ± 87 PU6(PEG1000-HMDI-DTH) 3.73 ± ± ± 95 PU7(PCL1250-HMDI-DTH) ± ± ± 29 Stress (MPa) PU2 (PEG600-HDI-DTH) PU4 (PCL530-HDI-DTH) PU1 (PEG400-HDI-DTH) Strain (%) Figure 5.9 Representative stress-strain curves of L-tyrosine based polyurethanes 132

154 7 PU5 (PCL1250-HDI-DTH) 6 5 Stress (MPa) PU3 (PEG1000-HDI-DTH) Strain (%) 20 PU7 (PCL1250-HMDI-DTH) Stress (MPa) 8 4 PU6 (PEG1000-HMDI-DTH) Strain (%) Figure 5.9 Representative stress-strain curves of L-tyrosine based polyurethanes. Continued 133

155 The p-values from the statistical analysis of the mechanical properties are presented in Table 5.5. The detailed analysis is presented in appendix A. All the structural variations significantly change the modulus of elasticity and the ultimate tensile strength (p<0.05) of the polyurethanes. The elongation at break does change significantly with the change in the polyol but the p value is higher than 0.05 due to interaction between polyol and diisocyanate. But separate analysis of PEG and PCL based polyurethanes shows that for a given diisocyante, the change in the elongation at break is statistically significant with the p-value equal to On the other hand, for PCL based polyurethanes, (in PU5 and PU7) the change in elongation is insignificant (p=0.223) due to the change in the diisocyanate structure. Moreover, combined analysis to examine the effect of PEG molecular shows that changing the PEG molecular weight, significantly changes the elongation. But individual comparison between PEG (400) and PEG (600) shows that the difference in the elongation of PU1 and PU2 is statistically not very different with a p- value of Table 5.5 p-values for the mechanical properties of polyurethanes Structural Effects\Properties Ultimate Tensile Modulus of Elongation at Strength Elasticity Break Polyol (PEG/PCL) Diisocyanate (HDI/HMDI) PEG (Mol. Wt) PCL (Mol. Wt.) The results show that PEG based polyurethanes are relatively weaker in mechanical properties compared to PCL based polyurethanes due to relative crystallinity 48,49. The effect of soft segment molecular weight can be seen for both PEG and PCL based polyurethanes. For PU1, PU2 and PU3, increasing PEG molecular weight shows 134

156 increased mechanical properties although the hard segment content of the polyurethanes is decreasing. Phase mixed morphology and lack of ordered hard segment, as indicated by FT-IR and DSC results in poorer mechanical strength in low molecular weight PEG based polyurethanes. Moreover, the molecular weight of polyurethanes with lower molecular weight PEG is also low which is also possible for such pattern. Similar explanations are applicable for the mechanical properties of PU4 and PU5 with different molecular weight of PCL soft segment. In addition, increasing PCL molecular weight shows increasing crystalline nature of the soft segment which tends to improved mechanical property of PU5 compared to PU4. These results show the importance of phase segregated morphology (resulting from ordered and crystalline hard segment) in the mechanical property of the polyurethane. The effect of diisocyanate structure has significant impact on the mechanical properties of the polyurethanes. Both for PEG and PCL based polyurethanes, changing of diisocyanate from linear to cyclic structure improves the ultimate tensile strength and elongation but significantly reduces the modulus. Both PU6 and PU7 exhibit higher ultimate strength and very high elongation but reduced modulus of elasticity compared to PU3 and PU5 respectively. In general, cyclic structure improves the mechanical properties due to ordered and crystalline hard segment 54,74. But DSC and FT-IR indicate considerable phase mixing and disordered hard segment in PU6 and PU7. This explains the low modulus of elasticity of PU6 and PU7 compared to linear diisocyante based polyurethane. Therefore, increase in ultimate tensile strength and elongation is contrary to the general trend. This can be explained by the strain-induced crystallization and/or finite extensibility of the polyurethanes 77. Although cyclic structure of HMDI is symmetrical, it prevents close packing of the polymer chains. 135

157 Therefore, at higher strain the molecular chains of the polyurethanes are either able to reorient to form crystalline structures or change the conformation to absorb higher energy. But the higher elongation of PEG based PU6 compared to PCL based PU7 is most likely due to the combined effect of non-crystallizable hard segment and relatively amorphous PEG soft segment. The importance of phase morphology is crucial in determining the mechanical properties of the polyurethanes. The hard segment of the polyurethane is usually in a semicrystalline glassy state and the soft segment is in the viscoelastic state. Phase segregated morphology improves the mechanical properties of the polyurethanes. In phase segregated morphology, hard segments acts as physical crosslink for the soft segment domains and therefore act as reinforcing fillers. For a given soft segment concentration, polyurethanes usually exhibit improved mechanical properties with increased concentration and increased crystallinity of the hard segment domain. Similarly, for a given hard segment concentration, the polyurethane property increases with increasing crystallisability of the soft segment and the ability to dissipate the viscoelastic energy. For both PEG and PCL based polyurethanes, the soft segment of the polyurethanes are largely amorphous at low molecular weight. This is directly reflected from the mechanical property of PU1, PU2 and PU4. For higher molecular weight of PEG and PCL, the mechanical properties of PU5 is much more improved compared to PU3 due to increased crystallinity of the soft segment and more phase segregation. The effect of diisocyanate structure on the polyurethane properties is interesting. With cyclic diisocyanate, the hard segment is loosely packed leading to more free volume within the polymer structure. This affects the polyurethane property in two ways: (i) the crystallinity 136

158 and the ordering of the hard segment is destroyed resulting in lower modulus and (ii) the ability to deform and reorient at increasing strain leading toward high elongation and ultimate tensile stress Water Contact Angle The water contact angle values of the polyurethanes both in advancing and receding mode are shown in Figure Contact angle of PEG based polyurethane are lower due to hydrophilic PEG soft segment. As expected, with increasing molecular weight of PEG the contact angle values are lowered due to more hydrophilicity whereas for PCL based polyurethanes the contact angle increases with increasing molecular weight of PCL. The change of diisocyanate from linear to cyclic structure leads to higher hydrophobic surface as indicated by high contact angles and high hysteresis values. This is also indicative of a heterogeneous pattern of polyurethane surfaces with HMDI as diisocyanate. Moreover, the contact angle of lower molecular weight PEG based polyurethanes are similar to the contact angle of PCL based polyurethanes. These features indicate that the surface of the low molecular weight polyurethanes are relatively heterogeneous with mixed hard and soft segments. This is directly related to the phase mixed morphology of the polyurethanes with low molecular weight soft segment. The statistical analysis (Table 5.6) of the contact angle analysis shows that in advancing mode only the effect of PCL soft segment molecular weight is insignificant (p=0.181) in the contact angle value. This indicates that altering the PCL molecular weight does not change the surface characteristics of the polyurethane significantly. An individual comparison of PU2 and PU3 also shows that there is no significant (p=0.179) 137

159 change in the surface characteristics by changing the PEG molecular weight from 600 to For receding modes, all the effects of structural variations have significant impact on the contact angle values. But individual comparison of PU1 and PU3 shows no change with p-value of which indicates that with high hard segment content the polyurethanes behaves similarly to the polyurethanes with lower hard segment content. Moreover, the effect of diisocyanate structure is also not statistically significant (p=0.043 in advancing mode and p=0.023 in receding mode) for PCL based polyurethanes (PU5 and PU7). This indicates that change in diisocyante from HDI to HMDI does not change the surface characteristics. Table 5.6 p-values for contact angle of the polyurethanes Structural Effects\Properties Advancing Receding Polyol (PEG/PCL) Diisocyanate (HDI/HMDI) PEG (Mol. Wt) PCL (Mol. Wt.) The contact angle hysteresis (difference between advancing and receding contact angle) of the different polyurethanes are shown in Figure The hysteresis value is higher in low molecular weight soft segments. This indicates that the surfaces of these polyurethanes are more heterogeneous with more polar hard segment reoriented towards the surface. The similar hysteresis value of PU1 and PU4 indicates that for the polyurethanes with low molecular weight soft segments the extent of heterogeneity is similar and therefore the response of the surface toward receding water drop is similar. The effect of diisocyanate structure on the hysteresis value is in not evident from this analysis. 138

160 Contact Angle ( ) PU1 (PEG400-HDI-DTH) PU2 (PEG600-HDI-DTH) PU3 (PEG1000-HDI-DTH) PU4 (PCL530-HDI-DTH) PU5 (PCL1250-HDI-DTH) PU6 (PEG1000-HMDI-DTH) P U7 (PCL1250-HMDI-DTH) Advancing Receding 0 PU1 PU2 PU3 PU4 PU5 PU6 PU7 Figure 5.10 Advancing and receding water contact angle of polyurethanes (mean ± SD, n = 15) e Hysteresis ( ) Contact Angl PU1 (PEG400-HDI-DTH) PU2 (PEG600-HDI-DTH) PU3 (PEG1000-HDI-DTH) PU4 (PCL530-HDI-DTH) PU5 (PCL1250-HDI-DTH) PU6 (PEG1000-HMDI-DTH) P U7 (PCL1250-HMDI-DTH) PU1 PU2 PU3 PU4 PU5 PU6 PU7 Figure 5.11 Contact angle hysteresis of L-tyrosine based polyurethanes Water Vapor Permeation Figure 5.12 shows the plot of amount of water vapor transmitted with respect to time for the polyurethanes. Comparison of PU1, PU2 and PU3 shows that with increasing molecular weight of PEG, the amount of water permeated increases due higher 139

161 hydrophilic character of the soft segment. Moreover, the phase mixed morphology of low molecular weight PEG based polyurethanes leads to hydrophobic character leading to lesser permeation of water vapor. The effect of change in molecular weight of PCL soft segment practically has no effect in the permeation rate. This is simply because the hydrophobic/hydrophilic character of the polyurethane remains unchanged with change in molecular weight. At comparable soft segment molecular weight the permeation rate decreases by the change of diisocyanate from a linear to a cyclic structure. The cyclic HMDI increases the hydrophobic character due heterogeneous morphology of the polyurethanes and therefore the permeation rate decreases. The water vapor permeability of the polyurethanes is shown in Table 5.7. Water vapor permeance (WVP) and water vapor permeability coefficient (WVPc) of PEG based polyurethanes are higher for hydrophilic PEG soft segment. Both the values of WVP and WVPc decreases with increasing molecular weight of PEG due to lower hydrophilicity and heterogeneous phase mixed behavior whereas in the case of PCL based polyurethanes there is no effect. The presence of cyclic diisocyanate decreases the values of WVP and WVPc which is mainly due to heterogeneous phase mixed characteristic of the polyurethanes. Table 5.7 Water vapor permeability of polyurethanes (mean ± SD, n = 3) Water Vapor Permeance Water Vapor Permeability Polyurethane (10 6 mg/hr.mm 2. mm of Hg) Coefficient (10 6 mg/hr.mm. mm of Hg) PU1(PEG400-HDI-DTH) 8.50 ± ± 0.12 PU2( PEG600-HDI-DTH) ± ± 0.98 PU3(PEG1000-HDI-DTH) ± ± 1.14 PU4(PCL530-HDI-DTH) 8.74 ± ± 0.74 PU5(PCL1250-HDI-DTH) 9.11 ± ± 0.44 PU6(PEG1000-HMDI-DTH) ± ± 0.37 PU7(PCL1250-HMDI-DTH) 7.73 ± ±

162 Mass of Water Vapor (mg) PU3 (PEG1000-HDI- DTH) PU6 (PEG1000- HMDI-DTH) PU2 (PEG600-HDI- DTH) PU1 (PEG400-HDI- DTH) Time (hour) 0.1 Mass of Water Vapor (mg) PU5 (PCL1250-HDI-DTH) PU4 (PCL530-HDI-DTH) PU7 (PCL1250-HMDI- DTH) Time (hour) Figure 5.12 Plot of water vapor transmitted against time of L-tyrosine based polyurethanes Table 5.8 shows the p-vales from the statistical analysis of the WVP and WVPc values of the polyurethanes. For PCL based polyurethanes the changing of the molecular weight has no significant (p>>0.05) impact on the permeation characteristics. This 141

163 indicates that decreasing PCL soft segment molecular weight does increase the hydrophilicity and/or permeability of the polymer to improve the permeability of water vapor. The effect of diisocyante structure does not significantly (p=0.027) change the permeation characteristics. Both for PEG and PCL based polyurethanes, changing HDI to HMDI have no significant (p>>0.05) impact on the permeation of water vapor. Table 5.8 p-values for water vapor permeation of the polyurethanes Structural Effects\Properties WVP WVPc Polyol (PEG/PCL) Diisocyanate (HDI/HMDI) PEG (Mol. Wt) PCL (Mol. Wt.) Water Absorption Characteristics The water absorption of the polyurethanes is shown in Figure rbed (%) Amount of water abso PU1 PU2 PU3 PU4 PU5 PU6 PU7 PU1 (PEG400-HDI-DTH) PU2 (PEG600-HDI-DTH) PU3 (PEG1000-HDI-DTH) PU4 (PCL530-HDI-DTH) PU5 (PCL1250-HDI-DTH) PU6 (PEG1000-HMDI-DTH) PU7 (PCL1250-HMDI-DTH) Figure 5.13 Water absorption of L-tyrosine based polyurethanes 142

164 PEG based polyurethanes absorb more water than PCL based polyurethanes due to hydrophilicity of the soft segment. However, with decrease in molecular weight of PEG water absorption decreases due to relative decrease in hydrophilic nature of the polyurethane. Moreover, polyurethanes with low molecular weig ht PEG exhibit phase mixed morphology due to which hydrophilicity of the soft segment is reduced. For PCL based polyurethanes, the effect of molecular weight of PCL soft segment is not significant as observed by the water absorption of PU4 and PU5. A comparison of PU6 with PU3 shows that changing diisocyanate from linear to cyclic structure leads to more water absorption. This is attributed mainly to the phase mixed morphology and relative amorphous hard segment of the polyurethanes. Moreover, due to the cyclic structure of the diisocyanate, the polymer chains are less closely packed. This creates enough free space within the bulk of the polymer. The water molecules penetrate within the available free space of the polymer leading to higher water absorption. Similar feature, although in lesser extent, is observed for PCL based polyurethane as seen by the water absorption of PU5 and PU7. Table 5.9 p-values for water absorption of the polyurethanes Structural Effects Water Absorption Polyol (PEG/PCL) Diisocyanate (HDI/HMDI) PEG (Mol. Wt) PCL (Mol. Wt.) Table 5.9 shows the statistical significance of the change in water absorption characteristics. The p-values indicate that all the effect of structural variations has statistically significant impact on the water absorption characteristics. 143

165 5.2.8 Hydrolytic Degradation The hydrolytic degradation of the polyurethanes is shown in Figure The role of polymer morphology is important for the polymer degradation 78. Figure 5. 14(A) shows the effect of soft segment and its molecular weight on the hydrolytic degradation. PEG based polyurethanes degrades at a faster rate compared to PCL based polyurethanes 46,48. Since PEG is hydrophilic and absorbs more water, the degradation rate in PEG based polyurethanes is faster than PCL based polyurethane. Moreover, PCL is relatively crystalline compared to PEG. The extent of degradation decreases with decreasing molecular weight of PEG due to more hydrophobic nature of the polyurethane. Similar degradation characteristics were observed for PCL based polyurethanes where PU4 (lower molecular weight PCL based polyurethane) degraded at a slower rate compared to PU5 (high molecular weight PCL based polyurethane). But this is contrary to the expectation since PU4 is less hydrophobic and more amorphous compared to PU5. The phase mixed morphology of PU4, as seen from DSC and FT-IR, indicates that PU4 is more hydrophobic and thus shows relatively slower rate of degradation. The effect of diisocyanate structure on the degradation characteristics is shown in Figure 5.14(B). The change of diisocyanate slightly slow down the degradation rate for PEG based polyurethane whereas in PCL based polyurethanes it enhances the rate. This anomalous nature in degradation characteristics is due to the morphology of the polyurethanes. PU6 is relatively amorphous and absorbs more water than PU3 and therefore, is expected to degrade faster compared to PU3. In addition, PU6 exhibits considerable extent of phase mixing which indicates that urethane linkages are H-bonded with the soft segment. This lowers the number of urethane linkages available for hydrolytic degradation. 144

166 30 PU1 PU2 PU3 PU4 PU5 A Mass lost (%) PU1 (PEG400-HDI-DTH) PU2 (PEG600-HDI-DTH) PU3 (PEG1000-HDI-DTH) PU4 (PCL530-HDI-DTH) PU5 (PCL1250-HDI-DTH) Time (Day) PU6 PU7 PU3 PU5 Mass lost (%) 20 B Time (Day) PU3 (PEG1000-HDI-DTH) PU5 (PCL1250-HDI-DTH) PU6 (PEG1000-HMDI-DTH) PU7 (PCL1250-HMDI-DTH) Figure 5.14 Hydrolytic degradation of L-tyrosine polyurethanes in PBS (ph 7.4, 37 C) (A) Series based on different molecular w eight of PEG and PCL (B) Series based on different diisocyanates (mean ± SD, n = 4) The soft segment PEG in PU3 comparatively phase separated and therefore is readily dissolved after degradation. Since the hard segment in PU3 is ordered and relatively 145

167 crystalline, the urethane linkages present at the interphase is degraded initially and rapid mass loss is experienced by PU3 due to easy extraction of degraded PEG in water. However, opposite trend is observed in PU6. In PU6, the H-bonding interaction between the hard and soft segment prevents the dissolution of PEG after degradation which results into relatively smaller amount of mass loss. However, in case of PCL based polyurethanes, PU7 degrades at a faster rate compared to PU5. The cyclic structured diisocyanate leads to relatively less ordered hard segment. The interactions between the hard and soft segment leads to a phase mixed morphology due to which the crystallinity of PCL soft segment is lower substantially. This enables water to approach the more urethane linkages compared to PU3 and hydrolytically cleave the polymer chain. This s hows that hard segment structure and morphology controls the degradation of the polyurethane. Table 5.10 p-values for mass loss (hydrolytic degradation) of the polyurethanes Structural Effects Hydrolytic Degradation Polyol (PEG/PCL) Diisocyanate (HDI/HMDI) 0.33 PEG (Mol. Wt) 0.00 PCL (Mol. Wt.) The statistical significance of the effects of structural variation on the hydrolytic degradation is shown in Table The effect of PCL soft segment molecular weight is not exceedingly significant as indicated by the p-value of But the effect of diisocyanate structure has no statistically significant effect on the degradation properties. Particularly for PEG based polyurethane, changing of diisocyanate from HDI to HMDI 146

168 does not appreciably change the mass loss of the polyurethanes which indicates that the hydrolytic degradation effect is similar in PU3 and PU6. A 1 PU3 PU1 PU2 PU5 PU4 R i PU1 (PEG400-HDI-DTH) PU2 (PEG600-HDI-DTH) PU3 (PEG1000-HDI-DTH) PU4 (PCL530-HDI-DTH) PU5 (PCL1250-HDI-DTH) Time ( s) B 1.2 PU3 PU5 PU7 PU6 R i PU3 (PEG1000-HDI-DTH) PU5 (PCL1250-HDI-DTH) PU6 (PEG1000-HMDI-DTH) PU7 (PCL1250-HMDI-DTH) Time ( s) Figure 5.15 Release of p-nitroaniline from L-tyrosine based polyurethane matrices in PBS (ph 7.4, 37 C) (A) Series based on different molecular weight of PEG and PCL (B) Series based on different diisocyanates (n = 4 error bars are omitted to make it clear) 147

169 5.2.9 Release Characteristics The release of p-nitroaniline, a model hydrophobic drug, was studied in order to investigate effect of the polyurethane structure on the release pattern of the drug. The structure and morphology of the polymers are important controlling factors in the release of drugs 79. Figure 5.15 shows the release pattern of p-nitroaniline from polyurethane matrices where the fractional release is plotted against the square root of time. Figure 5.15(A) shows the effect of different soft segments with variable molecular weights. The series of polyurethanes based on PEG soft segment shows that more drugs are released for low molecular weight PEG soft segment. For PU1 and PU2 more than 80% of the drug is released compared to only 43% released from PU3. p-nitroaniline being hydrophobic drug, is mainly dispersed in hard segment of the polyurethanes rather than hydrophilic PEG soft segment. Since considerable phase mixing is observed in PU1 and PU2, the hydrophobic drug is uniformly distributed throughout the polymer matrix in PU1 and PU2. In PU3 the drug is mainly located in phase separated hard segment domains only preferably through the H-bonding interactions between the drug and the hard segment. The extent of degradation is highest for PU3 but the percentage release of the drug is lowest for the series of PEG based polyurethanes. This shows that release of p-nitroaniline is largely diffusion controlled and the hydrophobic drug is mainly localized within the phase separated hard segment domain in PU3 compared to PU1 and PU2. Similar observations are made for PU4 and PU5 for the effect of variable molecular weight of PCL soft segment. But in comparison to PU5 where release becomes constant after 2 days, PU4 continues to release drug till the end of 30 day period. Moreover, phase mixed morphology and low molecular weight 148

170 PCL soft segment has reduced the crystallinity of the soft segment in PU4 which improves the release of the drug. The release pattern of the drug is significantly changed by the structure of the diisocyanate as shown in Figure 5.15(B). For PEG based polyurethanes changing of diisocyante from linear to cyclic structure increased the release of p-nitroaniline from 43% (in PU3) to 100% (in PU6). This is due to the uniform dispersion of the drug in mixed phase structure of PU6 where the soft and the hard segments are inter-mixed. This allows the drug to be uniformly dispersed throughout the matrix resulting in substantial higher amount of release. However, for PCL based polyurethanes there was no change in the release pattern due to change in the diisocyanate structure. Both PU5 and PU7 shows about 39 % release of p-nitroaniline in 30 day period. PU7 shows phase mixed morphology and relatively amorphous hard segment which is expected to increase the release of p-nitroaniline. The lower release in PU7 indicates that amount of water absorbed is not sufficient in to swell the polyurethane matrix for releasing the drug. The high water absorption of PU6 compared to PU7 allows 100% release in PU6 compared to only 39% in PU7. This implies that amount of water absorption is important for the diffusion of the drug from the polyurethane matrices. Moreover, the hydrophobic drug mainly interacts with the relatively hydrophobic domains of the polyurethane. Table 5.11 p-values for percent release of the polyurethanes Structural Effects Percent Release Polyol (PEG/PCL) 0.02 Diisocyanate (HDI/HMDI) PEG (Mol. Wt) 0.00 PCL (Mol. Wt.)

171 Table 5.11 shows the p-values from the statistical analysis of the release characteristics. Overall analysis shows that all the effect of structural variation has statistically significant difference in the amount of drug released from the polyurethane matrix. Decreasing PEG molecular weight lower than 1000 results in significant increase in the release of the drug but comparison of PU1 and PU2 shows that the difference is not significant (p=0.664). Moreover, comparison of PU3 and PU5 shows that with HDI as the diisocyanate there is no significant difference (p=0.217) between the polyurethanes in terms of the release of p-nitroaniline. To investigate into the details of the release mechanism of the drug release characteristics, the following power law equation is used to fit the experimental data 67 : M = M α kt n where, M/Mα is the fractional cumulative release at time t( which is R i and M α is equal to L), k is the release constant and n is the release exponent signifying the release mechanism. The validity and applicability of the equation is within the range M/M α <0.6. This indicates that the following analysis of release characteristics pertains to the very initial period of release time (approximately from beginning to 4 hours). Figure 5.16 shows the fitting of experimental data to power law equation. The parameters k and n are estimated from the fitted equation by using MS Excel solver and is tabulated in Table The values of the param eters k and n can be correlated to the release mechanism of the drug from the polyurethane matrix. For slab geometry, the value of n generally ranges in between 0.5 to

172 0.5 PU1 (PEG400-HDI-DTH) Experimental 0.4 Fitting R i Time (s) R i Experimental Fitting PU2 (PEG600-HDI-DTH) Time ( s) Figure 5.16 Curve fitting for release of p-nitroaniline from L-tyrosine based polyurethane matrices in PBS (ph 7.4, 37 C) 151

173 Experimental Fitting PU3 (PEG1000-HDI-DTH) R i Time (s) Experimental Fitting PU4 (PCL530-HDI-DTH) R i Time (s) Figure 5.16 Curve fitting for release of p-nitroaniline from L-tyrosine based polyurethane matrices in PBS (ph 7.4, 37 C) Continued 152

174 0.5 PU5 (PCL1250-HDI-DTH) Experimental Fitting R i Time (s) 0.8 PU6 (PEG1000-HMDI-DTH) 0.6 Experimental Fitting R i Time (s) Figure 5.16 Curve fitting for release of p-nitroaniline from L-tyrosine based polyurethane matrices in PBS (ph 7.4, 37 C) Continued 153

175 Experimental Fitting PU7 (PCL1250-HMDI-DTH) R i Time ( s) Figure 5.16 Curve fitting for release of p-nitroaniline from L-tyrosine based polyurethane matrices in PBS (ph 7.4, 37 C) Continued When n is equal to 0.5 it signifies that the drug is released by diffusion controlled mechanism (known as Case I mechanism) whereas, when n is equal to 1.00 it signifies that the drug is released by swelling controlled mechanism (known as Case II mechanism). For Case I mechanism, the relaxation rate of the polymer structure is much higher compared to the solvent mobility and therefore the polymer bulk can easily accommodate the solvent molecules. In this case, the controlling mechanism is diffusion through which the drug is released after the polymer is imbibed by the solvent. For Case II mechanism, the relaxation rate of the polymer structure is much lower compared to the solvent mobility and therefore the polymer bulk cannot easily accommodate the solvent molecules. Therefore in this case, the controlling mechanism is swelling of the polymer bulk which allows the solvent molecule to penetrate the bulk of the polymer matrix to allow the drug to diffuse out of the matrix. Value of n in between 0.5 to 1.00 is indicative of an anomalous mechanism which is a combination of Case I and Case II mechanism. 154

176 Table 5.12 Fitted values of k and n k(10 3 ) n Polyurethane PU1(PEG400-HDI-DTH) PU2(PEG600-HDI-DTH) PU3(PEG1000-HDI-DTH) PU4(PCL530-HDI-DTH) PU5(PCL1250-HDI-DTH) PU6(PEG1000-HMDI-DTH) PU7(PCL1250-HMDI-DTH) For the polyurethane series with variable molecular weight of PEG soft segment, PU3 shows diffusion controlled release whereas PU2 represents a combined mechanism. Since PU3 is mostly phase separated and the hydrophobic drug is mainly localized in hard segment domain, the diffusion of the drug controls the release pattern. PU3 absorbs a significant amount of water to facilitate the diffusion of the drug from the polymer matrix. In comparison, PU2 shows anomalous release due to the combined mechanism. The soft segment of PU2 is less hydrophilic and therefore absorbs less water. Moreover, due to phase mixed morphology of PU2, the drug is distributed throughout the matrix which releases the drug through diffusion and swelling mechanism at the same time scale. Interestingly, PU1 which also absorbs less water compared to PU3 and PU2 and also exhibits a phase mixed morphology, shows predominantly diffusion controlled (n=0.46) mechanism. This is apparently in contradiction to the expected behavior. Most probably, high hard segment concentration (~65%) of PU1 distributes the drug within the polyurethane matrix through increased the drug-polymer interactions (H-bonding). This indicates that the release mechanism is compounded by some other mechanisms in addition to diffusion and swelling. The release mechanism for PCL based polyurethanes 155

177 (both PU4 and PU5) is a combination of diffusion and swelling mechanism (with 0.5<n<1.00). The hydrophobic PCL soft segment controls/restricts the imbibitions of water molecules within the bulk of the polymer and therefore, the release of the drug is controlled by the swelling mechanism. However, PU5 has n value of 0.87 (closer to 1.00) indicates that the release is dominated by th e swelling mechanism as the soft segment is more hydrophobic and the morphology is mainly phase s eparated compared to PU4. PU4 exhibits a phase mixed morphology with relatively less hydrophobic and less crystalline soft segment in comparison to PU5. The release exponent value of PU4 (n=0.69) indicates a combination of mechanism for the release characteristics. In addition, the interaction between PCL soft segment and the crystalline p-nitroaniline through H- bonding decreases the crystalline nature of the soft segment. In addition to swelling 67 controlled mechanism, the crystal dissolution of the PCL soft segment would be another probable mechanism which controls the release of hydrophobic drug in PU5. The use of cyclic diisocyanate changes the release mechanism for PEG based polyurethane. PU6 exhibits a phase mixed morphology and relatively disordered hard segment. This allows more uniform distribution of the drug within the polyurethane matrix. Moreover, with a very high water absorption characteristics of PU6, it is reasonable that both swelling and diffusion both plays a significant role at the same time scale to control the release pattern. Interesting comparison of the release exponents of PU2, PU4 and PU6 shows that all the polyurethanes exhibits similar release mechanism. However the effect of structural variation is different for PU6 compared to PU2 and PU4. The release rates are lower for PCL based polyurethanes compared to PEG based polyurethanes indicating that PEG based polyurethanes releases at higher rate at comparable structural composition due to 156

178 hydrophilicity of the soft segment. In case of phase mixed morphology, the other explanation is due to H-bonding interactions between the drug and the soft segment(s) of the polyurethane. A comparison between PU5 and PU7 shows that release mechanism and rate is same for both the polyurethanes. This indicates that in spite of difference in the diisocyanate structure, the drug is released by similar pattern. PU7 exhibits higher degree of phase mixing compared to PU5 which means that the drug is well distributed in PU7 and most likely interacts with PCL soft segment through H-bonding in phase mixed morphology. Thus, in spite improved distribution; the release pattern remains identical in PU5 and PU7. Among the series of polyurethanes, only PU3 and PU5 exhibit a lag period in the release profile which is significant both with respect the to polyurethane structure and the release mechanism. Both PU3 and PU5 exhibits relatively phase segregated morphology compared to other polyurethanes. But the hydrophobic drug mainly localized in the phase separated hard segment through H-bonding. Thus, hydration is required by both PCL and PEG to initiate the release through diffusion. The lag period signifies a period of hydration for PU3 and PU5 which is due to the phase segregation of the polyurethane and due to the drug polymer interactions. The subsequent release in PU3 in controlled by diffusion as the PEG soft segment absorbs enough water molecule to diffuse the drug. Whereas, in PU5 the release is subsequently controlled by the swelling of polyurethane by water molecule to relax the polymer structure and diffuse the drug from the polyurethane matrix. In general, the above analysis shows that the structure and composition of the polyurethanes plays an important role in controlling the release pattern. Moreover, the physical and chemical characteristics of the drug and the drugpolymer interactions are crucial in determining the release characteristics. 157

179 5.3 Conclusion The variations in the structure and composition of the polyurethane have significant effect on the properties of the polyurethane. Segmented polyurethanes exhibit biphasic characteristics due to the presence of soft and hard segments. The compatibility or incompatibility of the two phases leads to either phase mixed or phase segregated morphology. The series of polyurethanes synthesized by using L-tyrosine based chain extender shows variable degree of phase behavior depending on the structure and composition of the components. Detailed analyses of the characterization results show that the properties of these polyurethanes vary over a wide range. The structural characterizations along with the analysis of the biphasic characteristics directly indicate the structure-property correlation of the polyurethanes. This study of investigating the relation between the structure and its effect on the polyurethane property provides an important tool for designing the appropriate material for particular application. The library of polyurethanes set up by using different soft segments and hard segment with DTH as chain extender shows that the properties can be changed by manipulating the structure and therefore would be useful for different biomaterial application including tissue engineering. 158

180 CHAPTER VI CHARACTERIZATION OF L-TYROSINE BASED POLYURETHANE BLENDS Blending and copolymerization are the most commonly used techniques to combine the properties of individual polymers 80. However, blending is easier than preparation of copolymers to obtain the advantageous properties of the constituent polymers 81. Generally blends possess better physical and mechanical properties in comparison to the individual polymers and also suppress the disadvantageous properties. The concept of blends as biomaterials is increasing as an easy alternative to combine the properties of individual polymers e.g. blending of amorphous/crystalline or hydrophilic/hydrophobic polymers to control the degradation rate. Depending on the compatibility and miscibility of the constituent polymers, the blend exhibits a wide range of morphology, from phase mixed to phase segregated structure 82. Moreover, by changing the composition, the final properties of the material can be altered easily according to requirement for particular applications. Thus by changing the components and the composition of the individual polymers, the polymeric blends can be fabricated with a wide range of advantageous properties. Different types of polymeric blends have been fabricated and characterized for various applications. Polymeric blends are used for different biomaterial applications e.g. 159

181 drug delivery, scaffold for tissue engineering, implants etc. Various natural, synthetic as well as semi synthetic polymers are blended and characterized for several applications. The use of polyurethanes as tissue engineering material has received a great deal of attention due to its unique combination of physicomechanical properties and degradability. Polyurethanes with polyester soft segment e.g. polycaprolactone (PCL) and polyether soft segment e.g. polyethylene glycol (PEG) are mainly used for tissue engineering applications. PCL based polyurethanes are relatively hydrophobic and have slow degradation rate whereas PEG based polyurethanes are highly hydrophilic and therefore shows enhanced rate of degradation. But in terms of mechanical properties, PEG based polyurethanes possess poorer moduli and ultimate stress compared to PCL based polyurethanes. Thus PEG based polyurethanes lack the structural integrity required for tissue engineering scaffold formation 83. This shows that depending on the structure and composition, the properties of the polyurethanes vary over a wide range. Copolymerization and blending are the commonly used techniques to manipulate the properties the properties of the individual polyurethanes. Different techniques are used to characterize the blends and investigate the property of these materials for various applications. The previous chapters of this dissertation describe the design, synthesis and characterization of polyurethanes based on L-tyrosine based chain extender, DTH. The results show that the polyurethanes exhibit a wide range of properties. The concept of blending is used to combine some of these polyurethanes in different composition to achieve desirable properties for particular applications. To demonstrate the advantages of polymeric blends, this chapter focuses on development and characterization of three 160

182 different blends using two L-tyrosine based polyurethanes. The polyols used for these polyurethanes are either PEG or PCL and the diisocyanate is hexamethylene diisocyanate (HDI). The nomenclature for the polyurethanes used is: PEG-HDI-DTH and PCL-HDI- DTH, where PEG or PCL represents the soft segment and HDI and DTH represent the hard segment. The structure and composition of these polyurethanes are described in details in Chapter III. The detail characterizations of the polyurethanes are included in Chapter III and IV. In brief, the polyurethanes synthesized from PEG exhibits poor mechanical properties and high rate of degradation whereas PCL based polyurethanes have better mechanical property but very slow rate of degradation. Thus, blending of these materials would provide an easy and alternative technique to tune the properties. 6.1 Experimental The following sections describe the experimental procedures for the fabrication and characterizations of the polyurethane blends Fabrication of Blends Two different L-tyrosine based polyurethanes were synthesized from either polycaprolactone diol (PCL, M w =1250) or polyethylene glycol (PEG, M w =1000), as the polyol and hexamethylene diisocyanate (HDI) as the diisocyanate. The chain extender was desaminotyrosyl tyrosine hexyl ester (DTH), a diphenolic dipeptide derived from L- tyrosine and its metabolite desaminotyrosine. The polyurethane blends were fabricated by solvent casting technique. The polyurethanes were blended in three different weight ratio using chloroform as the 161

183 solvent and films were cast by solvent evaporation. Typically 5 % (w/v) solutions of the polyurethanes were prepared in 10 ml of chloroform. Accurately weighed polymers (a total of ~500 mg in definite weight ratio) was dissolved in 10 ml of solvent and allowed to form a homogeneous solution through constant stirring at room temperature for 48 hours. The polymer solutions were filtered through Teflon syringe filter to remove undissolved residue and were cast onto poly(tetrafluoroethylene) (PTFE) pertidishes. The solvents were initially allowed to evaporate at room temperature followed vacuum drying at 50 C for another 48 hours to remove the residual solvents. Films of about thickness 0.15 mm were obtained by this method. ~500 mg of PEG-HDI-DTH + PCL-HDI-DTH ~10 ml of Chloroform ~5% W/V solution of the components in chloroform Solution cast in teflon petridish Sonicated and filtered through ashless Whatman filter Films of the blends (stored in desiccator) Controlled evaporation of solvent at room temperature for 48 hours Dried in vacuum for 48 hours at 50 C Figure 6.1 Scheme for fabricating films of polyurethane blends 162

184 The fabrication of the blends is schematically shown in Figure 6.1. The nomenclature and composition of the blends are summarized in Table 6.1. Table 6.1 Formulation of blends Code PU1 PU2 PU3 PU4 PU5 Composition (PEG-HDI-DTH/PCL-HDI- DTH) (wt %) 100/0 67/33 50/50 33/67 0/100 Weight (PEG-HDI-DTH + PCL-HDI-DTH) (mg) Volume of Solvent (Chloroform) (ml) Spectral Characterizations The blends were characterized by 1 H NMR and FTIR. NMR was carried out in 300 MHz Varian Gemini instrument with d-chloroform (δ = 7.27 ppm as internal reference) and FT-IR analysis was performed with a Nicolet NEXUS 870 FT spectrometer for neat samples with 16 scans Microscopic Characterization The domain morphology of the blends was characterized by scanning electron microscopic (SEM) images. The SEM images of the samples were recorded on silver sputtered samples in Hitachi S2150 (Operating Voltage: 20 kv) Thermal Characterization The blends were thermally characterized by differential scanning calorimetry (DSC) techniques. Differential scanning calorimetry (DSC) was performed with a DSC 163

185 Q100V7.0 Build 244 (Universal V3. 7A TA) instrument at a scanning rate of 10 C/min from -80 to 250 C Mechanical Characterizations The tensile properties of the films were measured by Instron Tensile Testing Machine with a load cell of 100 N and cross-head speed of 100 mm/min at room temperature. The sample dimension was 20 mm 6 mm ~ 0.3 mm with free length of 10 mm Water Contact Angle For contact angle measurement, thin films of the polymers were prepared on thoroughly cleaned and dried glass slides by dip coating the slides into the 5 wt % solution of polyurethanes for 12 hours. The films were initially dried at room temperature for 24 hours followed by vacuum drying at 50 C for another 48 hours to remove the residual solvents. Water contact angle was measured by sessile method using a Ramé- Hart goniometer at room temperature in an air atmosphere both in advancing and receding modes Water Absorption To measure the water absorption, circular sample were cut from dried films (diameter: 1.5 cm and thickness: 0.15 mm) and immersed in 20 ml of deionized water. At predetermined time intervals the hydrated samples were taken out and weighed after the surface water was blotted with Kimwipes. The water absorption was calculated on the 164

186 basis of the weight difference of the film before and after swelling. The percentage of water absorption was calculated using the following equation: Water Absorption (%) = w w ) / w 100 ( where, w 2 and w 1 are the weight of sample films after and before being immersed in water, respectively Hydrolytic Degradation For hydrolytic degradation, similar circular samples (diameter: 1.0 cm and thickness: 0.15 mm) were cut from dried films. The samples were incubated at 37±1 C in 10 ml of phosphate-buffered saline (PBS; 0.1 M, ph 7.4) containing 200 mgl -1 of sodium azide to inhibit bacterial growth in a sealed vial placed within constant temperature water bath. Samples were taken at intervals, weighed for mass loss after drying under vacuum at 40 C for 2 days. The hydrolytic degradation was calculated from the weight loss (%) using the following equation: Weight Loss (%) = w w ) / w 100 ( where, w 2 and w 1 are the weight of sample films after and before degradation, respectively. The SEM images of the degraded samples were recorded on silver sputtered samples in Hitachi S2150 (Operating Voltage: 20 kv) Statistical Analysis The statistical analysis of the data was performed by using a generalized linear model of ANOVA with Minitab 15 software. Statistical analysis was performed on four categories to examine the effect of blending with respect to pure polyurethanes: (i) All 165

187 the blends and the pure polyurethanes (ii) All the blends and pure PEG based polyurethane (iii) All the blends and (iv) All the blends and pure PCL based polyurethane. Results with p value less than 0.05 (p<0.05) was considered to be statistically significant. However, p values were used to interpret the significance of the blending in the material properties of the polyurethanes. 6.2 Results and Discussion The following sections describe the results of the experiments and its explanation related to the characterization of the blends H NMR Characterization The 1 H NMR spectra for all the blends are very similar due to the similarity in most of the proton environments. A representative spectrum for the blend (for PU3) is shown in Figure ppm Figure 6.2 Representative 1 H-NMR of L-tyrosine based polyurethane blend 166

188 Quantitative estimation by integrating a peak area is difficult due to similar chemical shift of the different protons present in soft and hard segment of the polyurethanes. Moreover, the results were not reproducible due to sample size and variation. Since the only difference between the constituent polyurethane is in the soft segment, two peaks were chosen that are characteristic of the PEG and PCL. The chemical shifts (δ) at 3.65 ppm corresponding to methylene protons (-CH 2 -CH 2 -) of the PEG soft segment and at 4.06 ppm corresponding to methylene protons (-CH 2 -O-CO-) of the PCL soft segment are integrated to estimate the relative contribution of the constituent polyurethanes. The interferences due to the peak present at 4.25 and 3.70 ppm from PCL based polyurethanes were taken into account for the calculation. Figure 6.3 shows the NMR spectra of all the blends in the region of interest. The results of the integration shown in Table 6.2 indicate that the composition of the blends follows the general trends i.e. PU1 fraction decreases and PU5 fraction increases from PU2 to PU4. However, a significant deviation from the theoretical composition was observed for all the blends. This can be attributed to the inhomogeneous mixing due to incompatibility, variation in sample size and similarity in the proton environment 87. All the blends shows considerable higher fraction of PEG based polyurethanes than the theoretical fraction which indicates some extent of immiscibility between the constituent polyurethanes. Table 6.2 Composition of polyurethane blends from 1 H-NMR Sample PU2 PU3 PU4 Theoretical Ratio (PEG-HDI-DTH/PCL- HDI-DTH) 2 : 1 1 : 1 1 : 2 Observed Ratio (PEG-HDI-DTH/PCL- HDI-DTH) 2.31 : : : 2 167

189 PU3 (1:1) ppm PU2 (2:1) ppm PU4 (1:2) ppm Figure H-NMR spectra of the blends for integration (ratio indicates ratio of PEG- HDI-DTHG to PCL-HDI-DTH) FTIR Characterization The FT-IR spectra of the blends are shown with pure polyurethanes in Figure 6.4. The characteristic peak for PU1 is the absorbance at 1100 cm -1 corresponding to the aliphatic ether linkages (C-O-C) present in the PEG soft segment and that for PU5 is the absorbance at 1730 cm -1 corresponding to the ester carbonyl linkages (C=O) present in the PCL soft segment. In addition, the spectra displays characteristics urethane absorbance at 1713 cm -1, aromatic C=C stretch at1620 cm -1 (in DTH) and C=O for amide 168

190 I bonds at 1640 cm -1. The intensity of the absorbance around 1100 cm -1 gradually decreases and that for 1730 cm -1 gradually increases from PU1 to PU5. This qualitatively suggests that the content of PEG based polyurethane is gradually decreasing from PU2 to PU4 while the content for PCL based polyurethane is increasing. PU5 (PCL-HDI- DTH) PU4 (1:2) PU3 (1:1) PU2 (2:1) PU1 (PEG-HDI- DTH Wave numbers (cm -1 ) Figure 6.4 FT-IR spectra of the of pure polyurethanes and blends (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) A quantitative characterization of the blends were attempted by calculating the ratio of the absorbance for 1730 cm -1 to1100 cm -1 and the variation of the absorbance ratio is plotted against the composition in Figure 6.5. This estimation shows that ratio is least in PU1 and increases for the blends from PU2 to PU4 and is highest in PU5. This estimation qualitatively agrees that PU2 has highest fraction of PEG-HDI-DTH and PU4 has highest fraction of PCL-HDI-DTH. 169

191 3 Absorbance Ratio 2 1 PU1 (PEG-HDI-DTH) PU2 (2:1) PU3 (1:1) PU4 (1:2) PU5 (PCL-HDI-DTH) 0 PU1 PU2 PU3 PU4 PU5 Figure 6.5 Quantitative estimation of absorbance ratio (1730 cm -1 /1100 cm -1 ) of the of pure polyurethanes and blends (error bars represent standard deviation of measurement from 3 samples) (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) PU5 PU1 (PEG-HDI-DTH) PU2 (2:1) PU3 (1:1) PU4 (1:2) PU5 (PCL-HDI-DTH) PU4 PU3 PU2 PU Wave numbers (cm -1 ) Figure 6.6 FTIR analyses of the blends in the region of cm -1 (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) A rule of thumb for blend miscibility involving intermolecular interactions can be explained as follows. To obtain strong molecular mixing between a strongly self associated polymer X, another polymer Y should be weakly self associated but has available site for relatively stronger association with X. 170

192 FTIR analysis can be useful for analyzing the stability of the blends. Figure 6.6 shows the FTIR analyses can be useful to interpret the mixing of the two L-tyrosine based polyurethanes with respect to the rule of thumb. In PU1, the carbonyl frequency is due to urethane carbonyl stretching only whereas in PU5, the carbonyl frequency is the combined frequency due to the urethane carbonyl and ester carbonyl. Thus, in PU1 most of the carbonyl groups are H-bonded with urethane N-H linkages to form an ordered crystalline hard segment. The carbonyl stretching frequency in PU1 represents mostly H- bonded carbonyl with only a small fraction of non H-bonded carbonyl (represented by the shoulder at 1735 cm -1 ). However, in PU5 the ester carbonyls are most likely non H- bonded whereas, the urethane carbonyls are H-bonded with the urethane N-H linkages. Due to very strong absorption of ester carbonyls, the FTIR spectra of PU5 exhibits strong absorption of free ester carbonyl groups. The blend characteristics show that as the fraction of PU5 increases within the blend, the carbonyl frequency shifts to a higher position. This represents a higher concentration of the free carbonyl due to the ester group of caprolactone unit of PCL soft segment. Thus, no additional interactions take place between the constituent polymers with the mixing of two polyurethanes PU1 and PU5. For PU5, the urethane carbonyls are H- bonded with the N-H of urethanes linkages and the ester carbonyls forms strong dipolar interactions. There are no are no additional sites in PU5 for improved interactions between PU1 and PU5 which can lead to increased blend miscibility. Thus combination of PU5 with PU1 leads poor mixing between the polyurethanes. Interaction within the soft segments is not evident from FTIR analysis. Relatively amorphous PEG soft segment either disrupts or improves the crystalline order of PCL soft through intermolecular 171

193 interactions. The hard segment of both the polyurethanes is similar. The effects of hard segment on the mixing of the polymers are not obvious from FTIR analysis SEM Analysis The SEM images of the blends are shown in Figure 6.7. All the images show the presence of two phases corresponding to the two polyurethanes present in the blend. For PU3, the two phases corresponding to the two polyurethanes are intermixed. But in PU2 and PU4 where one component is in higher concentration compared to the other, the existence of the separate phases can be seen distinctly. In PU4, which contains higher concentration of PCL based polyurethane; the presence of ordered structure probably corresponds to the relatively crystalline PCL-HDI-DTH fraction. The SEM images indicate the two phase morphology of the blends Thermal Characterizations The thermal characteristics of the blends were assessed by the differential scanning calorimetry (DSC). The DSC thermograms of the pure polyurethanes and the blends are presented in Figure 6.8. The glass transition temperature (T g ) of PU1 is -40 C and of PU5 is -35 C. Both PU1 and PU5 exhibit additional endotherms which are assigned as dissociation of ordered hard segment and melting of crystalline hard segment. No hard segment T g was observed for the pure polymers but PU5 exhibited a soft segment melting endotherm at -31 C as described in Chapter III. 172

194 PU2 (2:1) PU3 (1:1) PU4 (1:2) Figure 6.7 SEM im ages of the polyurethane blends (ratio indicates ratio of PEG-HDI- DTHG to PCL-HDI-DTH) 173

195 PU5 PU1 (PEG-HDI-DTH) PU2 (2:1) PU3 (1:1) PU4 (1:2) PU5 (PCL-HDI-DTH) PU4 PU3 PU2 PU Temperature ( C) Figure 6.8 DSC thermograms of pure polyurethanes and blends (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) T g for the blends is in between the range of -40 to -35 C. A shift in the glass transition temperature is indicative of polymer miscibility but any appreciable shift in the T g for the blends is hard to detect within this small range. However for PU2, which has higher fraction of PU1, the T g is close to -40 C and for PU4, which has higher fraction of PU5, the T g is close to -35 C. The T g for PU3 which has equal fraction of fraction of PU1 and PU5 is also close to -40 C, indicating the dominance of PU1 fraction over PU5. The decreases in T g values with increasing PEG containing polyurethanes in consistent with results observed for blends of phenyl alanine based polyurethanes. All the blends from PU2 to PU4 exhibit an endotherm around 10 C which is due to the dissociation of short

196 range order of the hard segment. But the endotherms observed due to dissociation of long range order in PU1 and PU5 around 50 C is only observed in PU2 and is present in PU3 as broad endothermic transition and is absent in PU4. This indicates that increasing amount of PU5 inhibits the formation of long range order of the hard segment. This feature is further substantiated by the absence of hard segment melting endotherm of PU3 and PU4 which indicates that with higher PU5 content the hard segment is more amorphous in nature. PU2 with lowest PU5 fraction only exhibits a melting endotherm at 152 C. The absence of soft segment melting in PU1 and PU2 indicates that blends with more PEG content are amorphous in nature. The soft segment melting endotherm starts to appear from PU3. This feature shows that relatively crystalline PCL components shows melting endotherm. The soft segment melting endotherm for PU5 appears at 31 C and for PU4 and PU3 at slightly lower temperatures around 25 C. This feature shows that at certain concentration, the presence of PEG helps to crystallize the PCL component. This feature is reported for similar systems 84. In general, the T g values of the sample show that the polyurethanes are not completely miscible and a possible phase separation occurs due to the incompatibility of hydrophilic PEG and hydrophobic PCL soft segment. However, due to the chemical similarity of the hard segments, the hard segments of the constituent polyurethanes are likely to form an amorphous one phase domain as evident from the disappearance of hard segment melting endotherm of PU3 and PU4. The analysis of DSC thermograms provides useful information about the blend characteristics. The interactions between the soft segments i.e. PEG and PCL soft depends on the relative crystallinity of the components. Although from hydrophobicity/hydrophilicity standpoint, the soft segments are incompatible but the 175

197 presence of one component induces change in the morphology of the other component. The blending of the polyurethanes creates one phase hard segment domain of the polyurethanes. Similar structure of the hard segments of the L-tyrosine based polyurethanes forms one phase structure. But the two different soft segments actually inhibit the ordering of the hard segment into ordered crystalline domain structure. This shows that the difference in the chemical structure of the soft segment has significant effect on the blending characteristics Mechanical Characterizations The typical stress-strain curve of the pure polyurethane and the blends are shown in Figure 6.9. The mechanical properties of the blends and the pure polyurethanes are summarized in Table 6.3. Table 6.3 Mechanical properties of the blends and polyurethanes Ultimate Tensile Modulus of Elongation at Sample* Strength Elasticity break (%) (MPa) (MPa) PU1(PEG-HDI-DTH) 2.81 ± ± ± 9 PU2 (2:1) 3.35± ± ± 57 PU3 (1:1) 3.90± ± ± 21 PU4 (1:2) 4.21± ± ± 64 PU5 (PCL-HDI-DTH) 7.05 ± ± ± 87 *Ratio indicates ratio of PEG-HDI-DTHG to PCL-H DI-DTH Detailed statistical analy sis of the data is presented in appendi x B. The samples were grouped into four categories and the p values from the analysis of each category are presented for all the mechanical properties in Table 6.4. The p values show that the mechanical properties of the blends are significantly different from the pure polyurethane with p<0.05. This indicates that blend mechanical properties significantly differs from the 176

198 polyurethanes. However, p values of blends show that the tensile strength (p= 0.07) and modulus (p=0.066) of the blends are not significantly different from each other. The change in composition of the blends does not alter the tensile strength and modulus considerably due to the physical characteristics of the blends. Table 6.4 p-values for the mechanical properties of the blends Category Ultimate Tensile Strength Modulus of Elasticity Elongation at Break All PEG-HDI-DTH + Blends Blends PCL-HDI-DTH+ Blends The result shows that the blend properties resemble the p roperty of PU1. Even for the blends with comparatively higher content PU5 content i.e. PU3 and PU4 the properties are closer to pure polyurethane PU1. The reason for inferior mechanical properties of the blend PU3 and PU4 is mainly attributed to the distribution of the hard segment fraction of the polyurethanes in the blend 77. In presence of PCL soft segment, it is likely that hard segments do not have any long range order and is more amorphous in nature. The absence of crystalline melting endotherm of PU3 and PU4 in DSC indicates that hard segments are not ordered and random in distribution. The random distribution of hard segment leads to inferior mechanical properties of the blends PU3 and PU4 in spite of higher content of PU5. Another possible explanation is the phase separation between the polyurethanes that is likely contributing to the lower mechanical properties 77. The phase separation can be due to several the factors with the incompatibility between the hydrophilic PEG and hydrophobic PCL soft segment is the major one. The phase separation between the two phases is likely to constitute a continuous matrix of PU1 in 177

199 which PU5 is dispersed as discrete domains. This probably explains that in spite of having higher PU5 fraction, PU3 and PU4 exhibits relatively poorer mechanical properties. 8 7 PU1 PU2 PU3 PU4 PU5 Stress (MPa) PU1 (PEG-HDI-DTH) PU2 (2:1) PU3 (1:1) PU4 (1:2) PU5 (PCL-HDI-DTH) Strain (%) Figure 6.9 Representative stress-strain curve of pure polyurethane and blends (ratio indicates ratio of PEG- HDI-DTHG to PCL-HDI-DTH) For a homogenized miscible blend, a given property of the blend follows the additive rule that corresponds to the property of the pure constituent polymer and the weight fraction 78 : M = x M + x X X Y M Y where, M is the property of the blend, M X is the property of the pure polymer X and x X is weight fraction of pure polymer X and M Y is the property of the pure polymer Y and x Y is weight fraction of pure polymer Y. 178

200 8 Ultimate Ten sil e Strength (MPa) 6 4 Calculated Experimental Weight Fraction of PCL-HDI-DTH 20 M odulus of Elasticity (MPa) Calculated Experimental Weight Fraction of PCL-HDI-DTH Elongation at Break ( %) Calculated Experimental Weight Fraction of PCL-HDI-DTH Figure 6.10 Deviation of mechanical properties of blends from calculated values (from additive rule) 179

201 Figure 6.10 shows the deviation in the mechanical properties of the blends from the calculated values (by additive rule) for different composition. Analysis of the deviation in the mechanical properties of the blend from the ideal values directly correlates to the miscibility and stability of the blends. The Figure 6.9 shows that all the mechanical properties are significantly deviated from the ideal values calculated from the additive rules. These deviations indicate that the blends are not homogeneous as the constituent polymers are immiscible. In polymer blends, the tensile characteristics significantly differ f rom the pure polymer due to micromechanical deformation process which dominates over molecular deformation of individual polymers 85. Due to the immiscibility of the phases, micromechanical deformation of individual polymer causes negative deviation in the mechanical properties. The prime reason is debonding or dewetting of the individual polymers under strain that causes to lower the mechanical properties of the blends. This phenomenon is more pronounced by significant reduction in the stiffness (modulus) of blends compared to the ultimate properties such as tensile strength and elongation at break. These results also support the observation made from the spectral and thermal analysis as discussed before Water Contact Angle The contact angle values both in advancing and receding modes are shown in Table 6.5. Detailed statistical analysis of the data is presented in appendix B. The samples were grouped into four categories and the p values from the analysis of each category are presented for the advancing and receding contact angle in Table

202 Table 6.5 Contact angle of the of the blends and polyurethanes Sample* Advancing Contact Receding Contact Hysteresis Angle [a] Angle [r] (a-r) PU1(PEG-HDI-DTH) PU2 (2:1) PU3 (1:1) PU4 (1:2) PU5 (PCL-HDI-DTH) *Ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH Table 6.6 p-values for the contact angle of the blends Category Advancing Receding All PEG-HDI-DTH + Blends Blends PCL-HDI-DTH+ Blends Both in advancing and receding modes, the contact angle values of the blends are significantly different from the pure PEG based polyurethane (p<0.05). There is no significant difference in the contact angle values of the blends both in advancing mode (p=0.265) and receding mode ( p= 0.217). This indicates that the blend surfaces are heterogeneous in nature with no practical differences. Moreover, in advancing mode the difference in the contact angle is statistically insignificant (p=0.133) between the pure PCL based polyurethane and blends but is statistically significant (p= 0.005) in the receding mode. This indicates that the blend s urfaces are initially dominated by PCL based polyurethanes but in response to the receding water drop the hydrophilic PEG soft segment migrates to the surface. The values are average of measurements from 6 readings taken from 3 replicates for each sample. The high contact angle values of the blend indicate that the surfaces are relatively hydrophobic. With increasing PCL-HDI-DTH concentration, the surfaces of the blends become increasingly hydrophobic which results 181

203 in higher water contact angle. The high hysteresis value indicates that the blend surfaces are heterogeneous. The hysteresis values of the blends are closer to pure PCL based polyurethane i.e. PU5, which shows that the blend surfaces are hydrophobic. The values show that the contact angle of the blend PU2 and PU3 is comparatively closer to PU5 in spite of having higher and equal fraction of PU1 respectively. This indicates a possible phase separation between the components of the polyurethanes with the PCL based polyurethanes migrating from the bulk toward the surface PU2 (2:1) PU3 (1:1) PU4 (1:2) Frequency PU2 PU3 PU4 Figure 6.11 Histogram of distribution of contact angle on blend surface (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) Figure 6.11 shows the distribution of the water contact angle on the surface of the polyurethane blends. PU2 which contains higher concentration of PEG based polyurethanes shows all contact angle within the range of which indicates that PEG-HDI-DTH is absent on the surface of PU2. This indicates that PU2 surface exclusively contains PCL based polyurethane. For PU3, which contains equal amount of both the polyurethanes, shows the majority of contact angles in the range of PU4 182

204 which contains highest fraction of PCL based polyurethane also shows the majority of contact angles in the range of This clearly indicates the presence of the hydrophobic PCL-HDI-DTH component at the surface of the blends. The deviation of the experimental values from ideal values is shown in Figure The ideal contact angle values are calculated using the additive rule: θ = x θ + x θ X X Y Y where θ is the contact angle (advancing or receding) of the blend, θ X is the contact angle (advancing or receding) of the pure polymer X and x X is weight fraction of pure polymer X and θ Y is the contact angle (advancing or receding) of the pure polymer Y and x Y is weight fraction of pure polymer Y Advancing Contact Angle ( ) Experimental Experimental Calculated Calculated Receding Weight Fraction of PCL-HDI-DTH Figure 6.12 Deviation of contact angle values from the calculated values The deviation of the contact angle values shows interesting trend in case of advancing and receding modes. For advancing mode, the positive deviation indicates that the surfaces of the blends are initially dominated by the presence of hydrophobic PCL based 183

205 polyurethane. But in case of receding contact angle the experimental values matches the real values without showing any deviation. This indicates that in response to the receding water drop the hydrophilic PEG segment migrates to the surface. Thus compared to pure polyurethanes, the blends react in a different mechanism in response to receding polar water drop. The distribution of the blend surfaces is primarily dominated by the hydrophobic PCL-HDI-DTH component (as indicated by positive deviation of advancing water contact angle) but the surface pattern is gradually dominated by the hydrophilic PEG-HDI-DTH with the migration of hydrophilic PEG based polyurethane. This characteristic behavior of the blend surfaces has significant implication in its use for biomaterial application Water Absorption Characteristics The water absorption characteristics of the blends are plotted against time is shown in Figure The water absorption values after 17 hours are not conclusive because of weight loss due to degradation of the polymers. The water absorption of the blends increases from PU2 to PU4 with increasing content of PEG based polyurethane i.e. PU1. As PEG is hydrophilic in nature, blends having higher PU1 content absorb more water. Blends PU4 and PU3 have water absorption characteristics similar to PU5 whereas for PU2 with highest fraction of PU1 is 25% compared to 73% of PU1. This is probably due to the surface characteristics of the PU2 which is predominantly hydrophobic with PCL based polyurethane phase separated and migrating to the surface. This feature is also evident from the contact angle measurements. The results presented follows the general 184

206 trend that water uptake increases with increasing PU1 i.e. PEG based polyurethane content. 40 PU3 PU2 PU4 %) Water Absorption ( PU2 (2:1) PU3 (1:1) PU4 (1:2) Time (hours). Figure 6.13 Water absorption characteristics of polyurethane blends (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) 80 Water Absorption (%) Experimental Calculated Weight Fraction of PCL-HDI-DTH Figure 6.14 Water absorption characteristic as a function of blend composition from both experimental and calculated (additive rule) values 185

207 The ideal values of water absorption are calculated using the additive rule: W = x W + X X where W is the amount of water absorbed by the blend, W X is th e amount of water absorbed by the pure polymer X and x X is weight fraction of pure polymer X and W Y is the amount of water absorbed by the pure polymer Y and x Y is weight fraction of pure polymer Y. x Y W Y Figure 6.14 shows the amount of water absorbed by the blends and the pure polymers after 17 hours at room temperature for both the experimental and calculated values. The negative deviation from the ideal values indicates that the presence of hydrophobic PCL based polyurethanes creates hindrance for the water molecules to penetrate the bulk of the polyurethane. As indicated in contact angle analysis, the surfaces of the all the blends are hydrophobic. This probably hinders the absorption of water during the initial period. Moreover, all the blends collectively contain higher concentration of hydrophobic component(s) which create a barrier for the water molecules to penetrate within the bulk. This indicates that by varying the composition of the polyurethane blends the amount of water abso rbed can be controlled. Detailed statistical analysis of the water absorption data is presented in appendix B. The samples were grouped into four categories and the p values from the analysis of each category are presented for the water absorption in Table 6.7. Table 6.7 p-values for the water absorption of the blends Category Water Absorption All 0.00 PEG-HDI-DTH + Blends 0.00 Blends PCL-HDI-DTH+ Blends

208 The analysis shows that difference in the amount of water absorbed is statistically significant for all the blends (p=0.001) and also in comparison to the pure polyurethanes (p=0.00) Hydrolytic Degradation The hydrolytic degradation of the samples was measured by the mass loss and is shown in Figure Mass loss (%) Time (Days) Weight Fraction of PCL-HDI-DTH Figure 6.15 Hydrolytic degradation of blends (n=3) The mass loss is higher with increasing PU1 content, which indicates that increased water absorption leads to more hydrolytic degradation. The degradation of the blends increases from PU2 to PU4 with increasing PU1 content. Blend PU4 with 33 weight percent of 187

209 PU1 has almost similar degradation characteristics compared to PU5. PU3 and PU2 have significantly higher degradation compared to PU4 in 30 days period. This indicates that the controlling factor in the degradation of the blends is the soft segment characteristics. The hydrophilic PEG component absorbs more water leading to the degradation of the hydrolytically labile urethane, amide and ester linkages present in the polymer leading to more mass loss. For all the samples the initial degradation was rapid followed by significantly slower degradation over the 30 day period. This is probably due to rapid and initial degradation of PEG based polyurethane, i.e. PU1 followed by the slow degradation of hydrophobic and comparatively crystalline PCL based polyurethane, i.e. PU5. Table 6.8 p-values for the hydrolytic degradation of the blends Category Hydrolytic degradation All 0.00 PEG-HDI-DTH + Blends 0.00 Blends PCL-HDI-DTH+ Blends 0.00 Detailed statistical analysis of the hydrolytic data is presented in appendix B. The samples were grou ped into four categories and the p values from the analysis of each category are presented for the hydrolytic degradation in Table 6.8.The analysis shows that difference in the amount of mass loss due to hydrolytic degradation is statistically significant for all the blends (p=0.003) and also in comparison to the pure polyurethanes (p=0.00). The ideal values of mass loss by hydrolytic degradation are calculated using the additive rule: D = x D + x D X X Y Y 188

210 where D is the amount of mass loss by the blend, D X is the amount of mass loss by the pure polymer X and x X is weight fraction of pure polymer X and D Y is the amount of mass loss by the pure polymer Y and x Y is weight fraction of pure polymer Y. 30 Mass loss (%) Calculated 15 Experimental Weight Fraction of PCL-HDI-DTH Figure 6.16 Mass loss by hydrolytic degradation as a function of blend composition for both experimental and calculated (additive rule) values The mass loss due to hydrolytic degradation of the blends closely follows the ideal values as calculated from the additive rules. No significant deviation was observed for any composition of the blends as shown in Figure This is probably indicates that amount of water absorption (which shows a negative deviation) is enough for the blends to degrade according to the composition of the blend. With more concentration of PEG- HDI-DTH in the blends, the amount of mass loss is higher compared to the mass loss of blends with higher PCL-HDI-DTH. This indicates that the PEG-HDI-DTH component of 189

211 blends absorbs water to degrade the urethane and other hydrolysable linkages. The mass loss characteristics show that water soluble free PEG segment is leaves the polymer film after degradation. This leads to higher mass loss in PU2 compared to PU3 and PU4. PU1 (PEG-HDI-DTH) PU5 (PCL-HDI-DTH) PU2 (2:1) PU3 (1:1) PU4 (1:2) Figure 6.17 SEM images of degraded samples after 30 days of hydrolytic degradation (ratio indicates ratio of PEG-HDI-DTHG to PCL-HDI-DTH) 190