. s. A 2.... a 3...: ¢‘t.ufiw.5 :29»: x. .5"; . '1! .L.I..... .. 2.1.x... . .. 3.... Fish. .. innit... v.5.{1 .2 8...“??? $3. . z 2 s. 3..» a: :2...» fl... . tlxxslr . n : I . 1 lo: .1... .I .41 l‘ri.‘ c357... 11...... 1'1. .3. . .1... it. axiiié . if: .a 399......511 F“ A ii's’x: . 5m“? y 3"! z w 3 .13.“ :r... e 5.1... v xéfl$i§fii§fi$ my..." :qu’EJ‘ouh- .u £33....iintix fikm. ! ‘1?! I3! éztlflr. , haulxiimei... .55 a i 1 5! . I... I.” an .I. 3‘0»! 4!] xv... “ 5 3.3.1.6.; .7. 21:!" 31- ab. .90. a. 9|?) . if .-Iv\.~i). .. ‘35.... ll . 15s . .2? fl. . Lit]: . . Lt. . 2 Eli‘s}! :13)?! til-BI L 1x10. 1231 {$1 nth... £1.16... . (-2.3: ilr.‘ in... (if... .27"? (K .. 3:1,... {3.1.x I... \‘lIVZtDIU \‘n . 4‘!) it}! 8“.le 5). .. i signal-5| .u‘ .lX-..\‘§;l.\ Eli... 1.1.}...65 I: »!;.3.!3?.: it“ ‘tur .t .I. L I? u... ‘ Vitillld.‘ .fiV . . i... a . «$15!. .- lxnl‘utlull, .I‘v .x:.~\ £61: 1...: Lciuz.ilsll On . 13.1.1.3. nival‘ :.¢ll IlJnotQ’H a... 1 i. ....E._2.§§. 56. ‘ . . .....;...J.....J...Wnu .....»...¢ .nar.5....i!n {Sell 1 3.3.3.5.11... air; .u . $1 I: .. 5.3,... 2... a: Z TH E‘. .51 S 1. 007/ This is to certify that the dissertation entitled SYNTHESIS AND CHARACTERIZATION OF POLYMANDELIDE AND LACTIDE/METHACRYLATE BLOCK COPOLYMERS presented by T1 anqi Li u has been accepted towards fulfillment of the requirements for Ph . D . degree in Chemj filmy Major professor ‘Date Z/Z'l /02— MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771 - "‘ ‘ — .——*«——k - LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 c:/CIRC/DateDue.p65-p.15 SYNTHESIS AND CHARACTERIZATION OF POLYMANDELIDE AND LACTIDE/METHACRYLATE BLOCK COPOLYMERS By Tianqi Liu A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 2002 ABSTRACT SYNTHESIS AND CHARACTERIZATION OF POLYMANDELIDE AND LACTIDE/METHACRYLATE BLOCK COPOLYMERS By Tianqi Liu Polymers derived from lactic acid are attractive alternatives to traditional petroleum-based polymers due to their degradability and biocompatibility. Once used exclusively in biomedical applications, polylactide is now being developed as a high volume commodity polymer for packaging and coatings. The task of replacing the non-degradable polymeric materials used in these fields with polylactides requires that a broad spectrum of physical properties be available from polylactides. This objective can be achieved, in part, by the synthesis of new derivatives and block copolymers of polylactide. Polymandelide is a derivative of polylactide where the methyl group has been replaced by a benzene ring. We synthesized high molecular weight polymandelide via ring opening polymerization of mandelide, the cyclic dimer of mandelic acid, and characterized its properties. Polymandelide is a clear amorphous material with physical properties that resemble polystyrene. In particular, polymandelide has a glass transition temperature of ~ 100 °C, higher than any known polylactide, and similar to that of polystyrene (109 °C). At pH 7.4 and 55 °C, polymandelide samples degrade at ~ 1/120 the rate of amorphous polylactide run under the same degradation conditions, and polymandelide’s degradation profile matches that for heterogeneous hydrolytic degradation. Copolymerizations of mandelide with racemic lactide resulted in homogeneous materials with glass transition temperatures that range from 60 - 95 °C. Copolymerizations using L-lactide and <12 mol% mandelide yielded semi- crystalline materials, but higher levels of mandelide inhibited crystallization of the L-lactide segments and gave amorphous materials. Block copolymers of lactide and two methacrylates, methyl methacrylate and methoxy-capped oligo(ethylene glycol) methacrylate (OEGMA), were synthesized via a combination of ring opening polymerization and atom transfer radical polymerization. Poly(lactide)-block-poly(methyl methacrylate) synthesized via an end group transformation approach exhibited much higher thermal stability than a comparable polymer prepared from a difunctional initiator. The block copolymers prepared from racemic lactide were homogeneous and exhibited a Single glass transition temperature that increased with the mole fraction of poly(methyl methacrylate) in the copolymer. The effect of poly(methyl methacrylate) on the crystallization of the poly(L-lactide) block was also studied. Longer poly(methyl methacrylate) blocks led to a decreased crystallization rate for the poly(L-lactide) block. Block copolymers of racemic and L-lactide with OEGMA were synthesized using a difunctional initiator. Miscibility of the two blocks increased with decreases in the length of the OEGMA. Block copolymers of L-lactide and OEGMA were used to prepare nanoparticles via dialysis. To my family iv ACKNOWLEDGMENTS First, I would like to thank my advisor, Professor Gregory L. Baker, for his guidance all through the years. During my stay at “Baker Lab”, l was given the freedom not only to make mistakes (thanks for the tears we Shed together!) but also Ieam from them. His encouragement and patience have led me through this Ieaming process and will also affect my future life. I would like to thank my committee members: Professors Ned Jackson, Milton Smith and David Weliky for their suggestions and comments. I would also like to thank past and current members of the Baker group: Yiyan, Gao, Mao, Tara, Chun, Cory, Wenxi, Micah, JB, Kirk, Erin, Fadi, Zhiyi, quei, Leslie, Ping, Feng and Ying, for their assistance and support. Thanks also go to Dr. Andre Lee and Dr. Haiping Geng for their assistance with polymer characterization. I also must acknowledge the NMR and X-ray staff: Long Le, Kermit Johnson, Art Bates and Rui H. Huang, for their help on the instruments. Thanks also go to Lisa Dillingham in the Graduate Office, for going out of her way to make my life in this department so much easier. Special thanks also go to Kunxiu for his constant support and friendship. Finally, I would like to thank my Mom, Dad, sister and brother-in-Iaw for their love and encouragement. TABLE OF CONTENTS List of Figures ..................................................................................... x List of Schemes ................................................................................ xiv List of Tables .................................................................................... xvi List of Abbreviations ......... '. ................................................................ xvii Chapter 1 Introduction ..................................................................................... 1 1.1. History of polylactide ................................................................................. 2 1.2. Applications of polylactide and other biodegradable polymers ................. 2 1.2.1. Polylactide as a commodity polymer ................................................. 2 1.2.2. Medical applications of polylactide and other biodegradable polymers ..................................................................................................................... 6 1.3. Synthesis of polylactide ............................................................................ 9 1.3.1. Synthesis of lactic acid .................................................................... 9 1.3.2. Manufacture of polylactide ................................................................. 9 1.4. Catalysts and mechanism of ring opening polymerization ...................... 14 1.4.1. Metal catalysts ................................................................................. 14 1.4.2. Transesterification ........................................................................... 18 1.5. Thermal degradation and stability ........................................................... 19 1.5.1. Thermal degradation pathways ....................................................... 21 1.5.2. Factors affecting thermal degradation of polylactide ....................... 22 1.6. Hydrolytic degradation ............................................................................ 25 vi 1.6.1. Degradation mechanisms ................................................................ 26 1.6.2. Factors affecting the hydrolytic degradation of polylactide .............. 27 1.6.3. Degradation models ........................................................................ 30 1.6.4. Enzymatic degradation .................................................................... 32 1.6.5. Biodegradable polymers as drug carriers ........................................ 32 1.7. Random copolymers of polylactide ......................................................... 35 1.7.1 Copolymerization with carbonates .................................................... 36 1.7.2. Copolymerization with caprolactone and its derivatives .................. 37 1.7.3. Copolymerization with morpholine-2,5-dione derivatives ................. 38 1.7.4. Copolymerization with glycolide and other substituted glycolides 39 1.8. Block copolymers .................................................................................... 41 1.8.1. General ............................................................................................ 41 1.8.2. Phase separation and morphology .................................................. 42 1.8.3. Block copolymers with poly(e-carprolactone) ................................... 44 1.8.4. Block copolymers with poly(ortho ester)s ........................................ 46 1.8.5. Copolymers with rubbery blocks ...................................................... 47 1.8.6. Block copolymers with poly(ethylene oxide) .................................... 50 1.9. Atom transfer radical polymerization (ATFiP) .......................................... 55 1.10. Combination of ring opening polymerization and (controlled) radical polymerization ................................................................................................. 61 References ..................................................................................................... 67 Chapter2 Polymandelide .............................................................................. 75 2.1. General ................................................................................................... 75 vii 2.2. Monomer ................................................................................................. 82 2.2.1. Synthesis of mandelide ................................................................... 82 2.2.2. Purification of mandelide ................................................................. 84 2.2.3. Physical properties of mandelide isomers ....................................... 84 2.3. Polymerization of mandelide ................................................................... 85 2.3.1. Melt polymerization .......................................................................... 85 2.3.2. Solution polymerization ................................................................... 90 2.3.3. Purification of polymandelide ........................................................... 94 2.3.4. Characterization of polymandelide .................................................. 96 2.4. Copolymerization of mandelide with lactide .......................................... 106 2.5. Hydrolytic degradation of polymandelide .............................................. 115 2.6. Experimental section ............................................................................. 123 References ................................................................................................... 128 Chapter 3 Block copolymers of lactide and methyl methacrylate (MMA) ..... 130 3.1. General ................................................................................................. 130 3.2. Results and discussion ......................................................................... 133 3.2.1 Synthesis of polylactide macroinitaitors .......................................... 133 3.2.2. Synthesis of polylactide-b-PMMA .................................................. 137 3.2.3. Kinetics of the ATRP of MMA using polylactide macroinitiators 142 3.2.4. Thermal properties of polylactide macroinitiators and block copolymers ............................................................................................... 147 3.2.5. Miscibility and crystallization of poly(L-lactide) blocks in the copolymers. .............................................................................................. 151 viii References ................................................................................................... 1 63 Chapter 4 Block copolymers of lactide and OEGMA ................................... 166 4.1. General ................................................................................................. 166 4.2. Results and discussions ........................................................................ 167 4.2.1. Polymerization of lactide and OEGMA ........................................... 167 4.2.2. Purification of block copolymers .................................................... 173 4.2.3. Thermal properties of lactide and OEGMA copolymers ................. 174 4.3. Preparation of polylactide-b-POEGMA nanoparticles ........................... 178 4.4. Experimental Section ............................................................................. 184 References ................................................................................................... 1 87 Figure 1.1. Figure 1.3. Figure 1.4. Figure 1.5. Figure 1.6. Figure 1.7. Figure 1.8. Figure 1.9. List of Figures Structure of D,L-lactic acid ................................................................ 9 The Cargill/Dow continuous process for the synthesis of polylactide.3 Aluminum Schiff-base initiator systems .......................................... 15 Bulk erosion and surface erosion in biodegradable polymers ......... 25 Various morpholine-2,5-dione derivatives copolymerized with lacticclae9 Morphologies of AB block copolymers. White portions represent block A, while dark portions represent block B of the AB block copolymer ........................................................................................ 43 Sol-gel transition in lactide and ethylene oxide ............................... 54 Examples of nitrogen-containing ligands used in ATRP ................. 58 Figure 1.10. Difunctional monomers used for ring opening polymerization ..... 65 Figure 2.1 . Figure 2.2. Figure 2.3. Figure 2.4. Figure 2.5. Figure 2.6. Kinetics of solution polymerization of mandelide in CHacN at 70 °C under argon. [mandelide]:[Sn(Oct)2]:[BBA] = 100:1 :1; [mandelide]: 0.93 mol/L (75% R,S mandelide and 25% R,R/S,S mandelide) ...... 92 Molecular weight versus conversion during solution polymerization of mandelide in CHaCN at 70°C under argon. [mandelide]:[Sn(Oct)2]:[BBA] =100:1:1; [mandelide]: 0.93 moVL(75°/o R,S mandelide and 25% R,R/S,S mandelide) ................................. 93 FTIR of a polymandelide film spin-coated on a gold-coated silicon wafer. .............................................................................................. 97 ”C NMR of polymandelide in ds-DMSO ......................................... 98 080 of polymandelide samples, showing the molecular weight dependence of T9. A: Mn = 60,000 g/mol; B: 16,000 g/mol. Samples were heated at 10 °/min under helium ........................................... 101 Thermal Gravimetric Analysis of polylactide (A) and polymandelide (B) before washing with dilute HCl. Samples were heated at 40°C/min under N2 ........................................................................ 103 X Figure 2.7. Thermal Gravimetric Analysis of polymandelide (A) and polylactide (B) after washing with dilute HCI. Heating rate: 40°C/min. under N2 ...................................................................................................... 104 Figure 2.8. Thermal Gravimetric Analysis of polymandelide before and after washing with dilute HCI. Samples were heated at 40°C/min under N2 ...................................................................................................... 105 Figure 2.9. Glass transition temperatures of poly(mandelide-co-rad-lactide) copolymers. Samples were heated at 10 °C/min under helium ..... 110 Figure 2.10. Glass transition temperatures of poly(mandelide-co-rac-lactide) copolymers fitted to the Fox Equation ........................................... 111 Figure 2.1 1. Thermal properties of poly(mandelide-co-L-lactide) copolymers. Samples were heated at 10 °C/min under helium ......................... 113 Figure 2.12. Thermal Gravimetric Analysis of poly(mandelide-co-rao-lactide) copolymers. A: polymandelide; B: mandelidezlactide = 3:1; C: mandelidezlactide = 1:3; D: polylactide. Samples were washed with dilute HCI after precipitation and heated at 40 °C/min under N2.... 114 Figure 2.13. GPC traces of polymandelide samples during hydrolytic degradation at pH=7.4 and 55 °C .................................................. 120 Figure 2.14. Molecular weight change of polymandelide (A) and polylactide (A) during hydrolytic degradation in phosphate buffer at 55 °C. The lines are fit to a random chain scission model. The inset shows the molecular weight data before 97 days ........................................... 121 Figure 2.15. Weight loss during hydrolytic degradation of polymandelide in phosphate buffer (pH=7.4) at 55 °C .............................................. 122 Figure 3.1. 1H NMR of polylactide end capped with a-bromoisobutyryl bromide ...................................................................................................... 135 Figure 3.2. GPC traces of the polylactide macroinitiator and the resulting block copolymer. A: RO—PLA-Br (Mn: 7,200, PDI=1.24); B: poly(lactide-b- MMA) (Mn: 20,500, PDI=1.31). ..................................................... 140 Figure 3.3. ‘H NMR of poly(lactide-b-MMA) .................................................... 140 Figure 3.4. FT-IR spectra of polylactide, PMMA, and poly(lactide-b-MMA) films spin-coated on gold-coated silicon wafers ..................................... 141 Figure 3.5. Semi-logarithmic kinetic plot of the solution ATRP of MMA in anisole at 70 °C using a polylactide macroinitiator. [MMA]=0.474 mol/L, [MMA]: [polylactide]: [CuBr]: [Bipy] = 300 : 1 : 1 :2.5 .................... 144 xi Figure 3.6. Molecular weight of block copolymers vs. monomer conversion. . 145 Figure 3.7. ATRP kinetics of MMA in anisole initiated by ethyl-2- bromoisobutyrate with polylactide (Sn(Oct)2 and BBA) as the spectator. ...................................................................................... 146 Figure 3.8. Thermal Gravimetric Analysis of polylactide macroinitiators synthesized by the end capping (RO-PLA-Br) and difunctional initiator strategies (HO-PLA-Br). Heating rate: 40 °C/min. under N2. ...................................................................................................... 149 Figure 3.9. Thermal Gravimetric Analysis of poly(lactide-b-MMA) copolymers prepared from macroinitiators: a, HO-PLA-Br; b, RO-PLA-Br. Heating rate: 40 °C under N2 ...................................................................... 150 Figure 3.10. DSC second heating scans of poly(L-lactide-b-MMA) copolymers taken after cooling from 180 °C at 10 °C/min. Samples were heated at 10 °C/min. under helium. a: PLLA 100 (pure PLLA); b: PLLA 72 (PLLA:PMMA= 72:28); c: PLLA 54 (PLLA:PMMA= 54:46); d: PLLA 32 (PLLA:PMMA= 32:68) .............................................................. 154 Figure 3.11. Normalized DSC heating scans of poly(L-lactide-b-MMA) (54:46). Samples were heated at 10 °C/min. under helium. c: after cooling at 10 °C/min from 180 °C; 0’: taken after annealing overnight at 130 °C ...................................................................................................... 155 Figure 3.12. Optical micrograph of pure poly(L-lactide) (Mn: 21,800) annealed at 140 °C and observed through cross polarizers (black regions were due to air bubbles) ........................................................................ 156 Figure 3.13. Optical micrograph of PLLA 72 (PLLA: PMMA: 72:28) annealed at 140 °C and observed through cross polarizers(black regions were due to air bubbles) ........................................................................ 157 Figure 3.14. Wide-angle X-ray diffraction pattern of poly(L-Iactide) and poly(L- lactide-b-MMA) (72:28) after annealing at 130 °C overnight. (A) as precipitated from solution (B). — polylactide; — block copolymer ...................................................................................................... 158 Figure 4.1. GPC traces of polylactide macroinitiator (A: Mn: 26,150, PDI= 1.47) and polylactide-b-POEGMA (B: Mn: 45,790, PDI= 1.39) .............. 169 Figure 4.2. FTIR spectra of polylactide, POEGMA and polylactide-b-POEGMA) films spin coated on gold-coated silicon substrates. ..................... 171 Figure 4.3. NMR spectrum of polylactide-b-POEGMA .................................... 172 xii Figure 4.4. Differential Scanning Calorimetry of poly(raolactide), poly(OEGMA), and poly(raolactide)-b-poly(OEGMA) copolymers. A: polylactide; B: poly(rac-lactide)-b-poly(OEGMA) (9% OEGMA); C: poly(rac-lactide)- b-poly(OEGMA) (28.5% OEGMA); D: poly(OEGMA) .................... 175 Figure 4.5. Strain recovery of poly(rac-lactide)-b-poly(OEGMA) (9% OEGMA) at 37 °C. Stress was applied for the first 22 seconds. ...................... 176 Figure 4.6. Stress-strain curve of poly(rac-lactide)-b-poly(OEGMA) (9% OEGMA) measured at a frequency of 1Hz at 37 °C ...................... 177 Figure 4.7. Preparation of poly(L-lactide)-b-poly(OEGMA) nanoparticles via dialysis. ......................................................................................... 179 Figure 4.8. SEM photographs of poly(L-lactide)-b-poly(OEGMA) particles freeze-dried on a gold substrate .................................................... 180 Figure 4.9. The chemical structure of lidocaine .............................................. 181 Figure 4.10. UV absorbence vs. concentration of lidocaine in methanol (9.: 262 nm) ................................................................................................ 183 xiii Scheme 1.1. List of Schemes Three routes for polylactide syntheses: A, azeotropic condensation; B, condensation followed by chain coupling; C, ring opening polymerization ................................................................... 12 Scheme 1.2. Mechanism of ring opening polymerization of lactide catalyzed by Sn(Oct)2 .......................................................................................... 17 Scheme 1.3. lntramolecular and intermolecular transesterification ................... 18 Scheme 1.4. Thermal degradation pathways for polylactide ............................. 20 Scheme 1 .5. Copolymers of lactide and carbonates ......................................... 35 Scheme 1 .6. Copolymerizations of lactide and caprolactone ............................ 38 Scheme 1.7. Copolymerization of substituted glycolides with lactide ................ 40 Scheme 1.8. Block copolymer of lactide and caprolactone ............................... 44 Scheme 1.9. Lactide and caprolactone triblock copolymers ............................. 45 Scheme 1.10. Poly(ortho ester) containing short polylactide blocks ................. 46 Scheme 1.11. Multiblock copolymer of lactide and dimethylsiloxane ................ 48 Scheme 1.12. Block copolymers of lactide with olefins ..................................... 49 Scheme 1.13. Block copolymers of lactide and ethylene oxide ........................ 51 Scheme 1.14. Graft copolymers of lactide and ethylene oxide ......................... 53 Scheme 1.15. Mechanism of copper-catalyzed ATRP ...................................... 56 Scheme 1.17. Block copolymers synthesized by end group transformation ..... 62 Scheme 1.18. Block copolymers of caprolactone and styrene synthesized by a difunctional initiator approach .......................................................... 64 Scheme 1.19. Graft polymers synthesized by the macromonomer approach... 66 Scheme 2.1. Early examples of the synthesis of polymandelide. ..................... 79 Scheme 2.2. Examples of the synthesis of mandelide copolymers .................. 80 xiv Scheme 2.3. Synthesis of mandelide from mandelic acid ................................. 82 Scheme 2.4. lsomerization of R,S mandelide to R,R/S,S mandelide ................ 83 Scheme 2.5. Racemization in polylactide and polymandelide .......................... 86 Scheme 3.1. Synthesis of end-capped polylactide initiator (RO-PLA-Br) ......... 135 Scheme 3.2. Synthesis of polyactide prepolymer (HO-PLA-Br) from a difunctional initiator ....................................................................... 136 Scheme 3.3. Synthesis of poly(lactide-b-MMA) copolymers using polylactide macroinitiators derived from end-capping polylactide (RO-PLA-Br) and from a difunctional initiator (HO-PLA-Br). ............................... 139 Scheme 4.1 . Synthesis of poly(lactide-b—OEGMA) ......................................... 170 Table 1.1. Table 1.2. Table 2.1. Table 2.2. Table 2.3. Table 2.4. Table 2.5. Table 3.1. Table 4.1. List of Tables Applications of polylactide as commodity polymers2 .......................... 4 Major companies involved in lactic acid and PLA biomedical fields2.. 8 Glass transition temperatures of commercial polymers. ................... 76 Melt polymerization of mandelide at 160 °C ..................................... 89 Poly(mandelide-co-rao-Iactide) copolymers prepared by bulk polymerization catalyzed by Sn(Oct)2 at 130 °C ............................ 109 Poly(mandelide-co-L-lactide) copolymers prepared by bulk polymerization at 160 °C ............................................................... 112 Weight and molecular weight change during hydrolytic degradation of polymandelide in phosphate buffer (pH: 7.4) at 55 °C .................. 119 Glass transition temperatures of amorphous poly(rao-lactide-bMMA) copolymers. ................................................................................... 139 Block copolymers of polylactide-b-POEGMA ................................. 169 AFM ATRP BBA bPY DSC DTC DTG EDTA ESEM GC/MS GPC °HTC MALDl-TOF MeeTREN MMA Mn MS Oct OEGMA PDI PDMS PE List of Abbreviations Atomic force microscopy Atom transfer radical polymerization t-Butyl benzyl alcohol 2,2’-bipyridine Differential scanning calorimetry 2,2-Dimethyl-trimethylene carbonate Differential therrnogravimetry (Ethylenedinitrilo)tetraacetic acid disodium salt dihydrate Environmental scanning electron microscopy Gas chromatography/mass spectroscopy Gel permeation chromatography 2,2-[2-Pentene-1,5-diyl]-trimethylene carbonate Matrix assisted laser desorption/ionization - time of flight Tris[2-(dimethylamino)ethy|]amine Methyl methacrylate Number average molecular weight Mass spectroscopy 2-ethylhexanoate Oligo(ethylene glycol) methacrylate Polydispersity index Poly(dimethyl siloxane) Poly(ethylene) xvii PEO PETE PMMA POEGMA PS PTFE PVC PLA PLLA Pmdl ROP Sn(Oct)2 TGA TMC Poly(ethylene oxide) Poly(ethylene terephthalate) Poly(methyl methacrylate) Poly[oligo(ethy|ene glycol) methacrylate)] Poly(styrene) Poly(tetrafluroethylene) Poly(vinyl chloride) Polylactide or poly(lactic acid) Poly(L-lactide) Polymandelide Ring opening polymerization Tin(ll) 2-ethylhexanoate or tin octoate Glass transition temperature Thermal gravimetric analysis Trimethylene carbonate xviii Chapter 1 Introduction Many synthetic polymers are produced and utilized because they are cheap, versatile and durable. Nearly all of today’s synthetic polymers are derived from oil and natural gas, which are finite resources that are diminishing in supply. Because of their durability and resistance to chemical and physical degradation, polymers tend to accumulate in what is today’s most popular disposal system, the landfill. According to a study by the US. Environmental Protection Agency, polymers account for 21% (by volume) of the 200 million tons of municipal waste produced each year in the US. Possible solutions such as recycling and incineration have proved either uneconomical or environmentally unfriendly. A better solution would be to tailor polymers from renewable resources to provide the necessary properties during use, and then have the polymers undergo degradation to non-toxic products leaving no hazardous impact on the environment. These environmentally friendly polymers could replace traditional polymers in single-use applications such as resins for packaging. Degradable polymers, especially biodegradable polymers, are well-suited for such applications. Biodegradable polymers are materials that are quantitatively converted either to 002 and H20 or to CH4 and H20 by microorganisms under aerobic or anaerobic conditions, respectively. Biodegradable polymers can be either natural (e.g. starch, cellulose, hemp) or synthetic materials. Generally speaking, synthetic polymers offer advantages over natural polymers in that they can be easily tailored to display a wide range of properties. Polylactide, poly(e-caprolactone), poly(fl-hydroxybutyrate) and poly(vinyl alcohol) are examples of synthetic biodegradable polymers. 1.1. History of polylactide Polylactide is not a new polymer. In 1932, Carothers synthesized a low molecular weight polylactide sample by heating lactic acid under vacuum. Historically, scientist have tried to make polymers resistant to environmental factors such as H2O, O2 and UV, and since polymers made from glycolic acid and other a-hydroxy acids were unstable toward hydrolysis, research on this class of polymers was discontinued in the first half of the 20th century. But this “instability" eventually found multiple applications in the medical field, beginning with the first biodegradable surgical sutures in the 1960s. Since the early 19803, there have been three generations of biodegradable polymers. The first two generations were starch-based polymer systems, and were only partially degradable or failed to provide the desired mechanical properties. Third generation biodegradable polymers are based on polylactide and polyhydroxybutyrate which combine reasonable biodegradability with good mechanical properties. 1 .2. Applications of polylactide and other biodegradable polymers 1 .2.1. Polylactide as a commodity polymer In many aspects, the basic properties of polylactide polymers lie between those of crystalline polystyrene and PETE. Some noteworthy properties include: o A flexural modulus > polystyrene . A resistance to fatty foods and dairy products equivalent to PETE . Excellent flavor and aroma barriers . Good heat sealability . A high surface energy allowing easy printability TheSe properties, in addition to its inherent biodegradability have made polylactide a promising candidate as a commodity polymer intended for single- use or limited use applications. Table 1.1 lists some of the applications of polylactide with its associated processing techniques. 1) Fibers for apparel Polylactide can be readily converted into various fiber forms using conventional melt-spinning processes. Compared to PETE/cotton, polylactide/cotton offers the following advantages: . An all-natural high performance fabric a physiological comfort due to improved thermal insulation and water vapor transport . Lower density 0 UV stability . Lower flammability and smoke generation Table 1.1 . Applications of polylactide as commodity polymers2 End-Products Fibers Clothing (active wear, sportswear, intimate apparel), carpet tiles m-woven fibers Personal hyfine, protective clothianiltration \Qriented films Container labels, tape Qwsion coating Dinnerware, food packaging, mulch film Flexible film Food wrap, trash bags, shrink wrap Cast sheet Deli trays lInjection molding Rigid containers, daily containers Foam Clam shells, meat trays 1' W- '95 1 . I 2) Personal hygiene and medicare Single use products based on biodegradable polymers have impartan, applications in personal hygiene and medical care. Single-use degradable P’Oducts such as baby diapers, surgical masks, blouses and compresses can limit contamination and secondary skin reactions. 3) Agriculture and horticulture The use of non-woven cloth allows natural cultivation of seeds without peStiCides or herbicides. The cloth easily allows air and rain to reach plants while preventing insects from penetrating. If made of polylactide, it degrades easily by hydrOIysis and the degradation product - lactic acid oligomers, were observed to pr or"‘Iote seed germination.1 4) p aper coatings Paper is coated with either wax or polymeric coatings for various reasons int-‘4 uding better water resistance and enhancement of gloss. A problem with the reCycling of coated paper is the disposal of the coatings liberated during the repulping process. Since current coatings are mostly made from polyethylene, \‘ney typically do not break down during the repulping process and cannot be recycled. Polylactide polymers have a high surface energy and easily form coherent, smooth and glossy surfaces with satisfactory printability. The low melt {\scosity and high polarity of polylactide is superior to polyethylene in terms of adhesion to the paper and compatibility with low temperature ext“.- s/on. During repulping, polylactide can hydrolyze to water soluble, non-toxic products and P036 no problem in waste water treatment. 1.2.2. Medical applications of polylactide and other biodegradable POMIIers Due to the current high cost of polylactide-based polymers, applications for these materials have been largely limited to biomedical fields. Table 1.2 shows the major companies involved in producing polylactide or polylactide b"-"SSCI-products. Examples of such products include sutures, implants and drug delivery matrices. The major advantage of using biodegradable implants over trad itional synthetic polymers, metals, or ceramics is that the device can degrade in Situ and a second operation is not necessary to retrieve the device. Biodegradable polymers offer other important features. For example, fractured bones fixated with a rigid, non-degradable stainless steel implant tend to re- fra<>ture upon removal of the implant since the regenerated bone tissue often d(Des not carry an appreciable load during the healing process. In contrast, a Carefully tailored degradable implant that degrades at an appropriate rate will s\owly transfer load to the damaged area, affording stronger bone tissue. Sutures were the first commercial product from biodegradable polymers and still account for 95% of all sales. The other 5% is attributed to orthopedic deVices in various forms such as pins, rods, tacks, staples and dental applications. SeVeral end products include Dexon®, Vicry® and Maxon® sutures, Lactomer® and Absolok® clips and staples; BiOfiX® and PhusmneQP/afas find I SCFSWS and Capronor® drug delivery devices. Table 1.2. Major companies involved in lactic acid and PLA biomedical fieldsz Company Country Lactic acid Lactide Polylactide End (co)po|ymers products Galactic Belgium .wtories X X x Mme Finland X Phusis France X X X Boehringer Germany X X %elheim Purac Netherlands X X X lCl U. Kingdom x BPl USA X Davis a. Geck USA x Etnor USA x Henley & USA X Johnson Johnson & USA X Johnson Medisorb USA X Technologies *- 1 41- Synthesis of polylactide I 1-3-1- Synthesis of lactic acid 0 O ”33%.... “3%.... 3 D-Lactic acid L-Lactic acid Figure 1.1. Structure of D,L-lactic acid Lactic acid is a chiral molecule that exists as two stereoisomers, L- and D- lactic acid (Figure 1.1). Lactic acid can be produced from petrochemical sources or by fermentation as shown in Figure 1.2.3 In the petrochemical route (Figure 1'2 A), ethylene is oxidized to acetaldehyde, and following treatment with HCN, the cyanohydrin is hydrolyzed to give racemic lactic acid (rec-lactic acid). PrQsentIy, Musashino in Japan is the only producer of raolactic acid. The fermentation process (Figure 1.2 B) produces almost exclusively L-lactic acid. Most major companies involved in the production of lactic acid such as Purac, cargilVDow, Galactic and ADM use fermentation processes to produce lactic acici a. . H ‘r i ‘u I 1-3-2. Manufacture of polylactide 1) Direct condensation Lactic acid is a difunctional molecule and can self-condense by \ntermolecular esterification to form polylactide (Scheme 1.1 A). One drawback Petrochemical feedstock Corn Ethylene Starch Oxidation i v Unrefined dextrose Acetaldehyde HCN ll . u Fermentation Lactonitrile v ii Racemic d,l-lactic acid 99.5% L-Lactic acid A B :39l'e 1.2. Petrochemical (A) and fermentation (B) routes for the synthesis of '0 acid Of this approach is the long reaction time, which is related to the equilibrium bet\Iveen the starting a-hydroxy acid, and the polyester product and water. High Molecular weight polylactide cannot be obtained unless water is efficiently rel'hoved from the reaction system to drive the reaction to completion. Besides LlSing high vacuum, researchers have used azeotropic distillation with a high “citing point solvent to remove water continuously. The effect of catalysts on direct condensation of lactic acid was also investigated. Protonic acids and tin compounds are effective at producing high molecular weight polymers at relatively low temperatures. Currently, Mitsui Toatsu4 utilizes a high boiling %Q\\Ient to produce high molecular weight polylactide. Another option is to react the end groups of low molecular weight polylactide with coupling agents such as 10 diisocyanates to yield high molecular weight polymer (Scheme 1.1 5/~ A drawback of this approach is the potential formation of branched or cross/inked MOlecuIar structures.5 2) RI'09 opening polymerization of lactide Ring Opening Polymerization (ROP) of lactide is the dominant route to POiylactide with molecular weights ranging from several thousand to several hundred thousand g/mol. CargilVDow adopted this route to synthesize PO'Ylactide on a large scale. In the ROP route shown in Scheme 1.1 C, lactic add is polymerized to afford a low molecular weight polylactide, which is then deF’Olymerized to give lactide, the cyclic dimer of lactic acid. Lactide can be furtl"|er purified by distillation under vacuum. Polymerization of lactide by ROP yie'ds polylactides, whose molecular weight can be easily controlled by varying the monomer to initiator ratio. The continuous process developed by CargilVDow is Shown in Figure 1.3.3 11 CH3 CH3 0 CH3 HO/Kn/OH Condensation : H 0%Ofi O/Kn/OH B 0 CH3 m 0 Low Mn Prepolymer Chain Coupling Agents Azeotropic Dehydrative CH 0 CH Condensation _ 3 O 3 OH ZnO H2O ’ H0 0 A A '\ 0 CH3 n 0 High Mn Polylactide M-OR C T): O O s°heme 1.1. Three routes for polylactide syntheses: A, azeotropic cohdensation; B, condensation followed by chain coupling; C, ring opening ptNymerization 12 Lactic Acid 4' \_ Lactic acid. Water Distillation ‘— —— o umn [.1 [:1 Lactide Reacto — __ Polymer Reactor Lactic acid — L—o Co ‘ polylactide Recycle PrejPonmer Reactor Additives po|y|actlde Producu Compound/mnisher Figure 1 .3. The CargilVDow continuous rocess for the - POlylactide. p synt'TSSIS Of 13 1.4. Catalysts and mechanism Of ring Opening polymerization 1.4.1 . Metal catalysts Many metal complexes have been Preposed as lactide polymerization catalysts. In the US. the FDA has approved the use of Sn(Oct)2 (Oct = 2- ethylhexanoate) for the synthesis Oi mater ials for surgical and pharmacological applications and zinc catalysts have been used industrially in France. Most catalysts fall into two categories: metal aikoxides and metal carboxylates. Aluminum alkoxides (A|(OR)3)_ belong to the first category. Ring opening Polymerization of lactide initiated by aluminum alkoxides is believed to proceed through a “coordination-insertion” I‘neChfill'lifiim-6 Coordination of the carbOt‘Y‘ oxygen of lactide to aluminum t0 fOIIOWGd by selective acyl-oxygen deavage leads to the formation ct linear polyesters. The alkoxide transferred “om aluminum ends up as an ester group at one end of the polyester chain. The use of functional aluminum alkoxides as initiators places the funCtional group at the end of the chain, and enables macromolecular engineering of POiyiaCtides as illustrated by the well-controlled SYcheSiS 0i macromonOmers and block COPOiyn'Iers,7'9 In recent years, alkoxy aluminum Schifi:s base complexes have Gen deVeioped for stereoselective polyme rizations.9 Spasskym and Coates" reported the stereoselective synthesis 01‘ iS°ta°fi° and SyndiOtaCfiC p°|ylactide from rat-\lactide and mesolactide respectively. Radano12 showed the first example of producing the po|y|actide stereocomplex from rac-lactide using racemic catalyst (Figure 1.4 Al- A5 smw" by Cameron,” adding electron- 14 ac t t w'th ' t k' 9 controlled amb,en ma In I I. - 6 practice tlde mor rizations of lac olyme rature p tempe 1 I B), . terns F e ' itiator sys um Schiff-base In in 1.4. Alum igur i for the ring I ca fates d eta ' ly use h st Ide 2 and alco o m the o ' esters. H ‘ one of 0 Che M cm is flactide and other y 0 ' ation o ' ly enz Opening po 3 either i” ' 'tiating spe the true rm ' ' erve as t as i puntres, s resen ' dded or p deliberately a an u 8 eq t ”d the . hiCh then pojymefize ' ecres, w tin alkoxrde sp te a to genera ater alcohol or w s 14 . din SCheme 2. me ' illustrate ” chanrsm tion insertion “ rdina - i i e coo lactide Via th The i “he Polylactide ' terrnmus o ter or acrd at the an es up forms droxy gro alkoxy or hy ' species ' tin alkoxlde tion of a pport the genera ’ o su ' dies als trcal stu . eore chain. 15 priol' to the ring opening polymerization of lactide by a “coordination-insertion” mechanism.15 Despite the wide use of aluminum and tin-based catalysts, AP“ is under suspicion as a potential player in Alzheimer’s disease and the use of tin derivatives in the biomedical field has been questioned despite the lack of any related acute problems in clinical applications. In response, some research 16,17 18-20 QFOUps have focused on developing magnesium, zinc, and iron“22 based Catalysts, since the ions of these metal participate in the normal metabolism of the human body and exhibit low toxicity. The development of lanthanide alkoxide catalysts has been limited by the to> OCt\Sn/07/U\OJ\"/O\R A/fkfsn‘O’R O O mash/WC O‘n n O - HOW/‘L‘O/lfikn n o n SCheme 1.2. Mechanism of ring opening polymerization of lactide catalyzed by Sfl(OCt)2 17 1.4-2. Transesterification Besides polymerization of monomers, metal catalysts can also catalyze side reactions such as inter— or intramolecular transesterification reactions (Scheme 1.3). Transesterification reactions can be identified by GPC, 130 NMR and MALDl-TOF analysis.2“'3‘°"33 lntramolecular transesterification, often termed “back-biting”, leads to cyclic structures and a decrease in the number average molecular weight. lnterrnolecular transesterification can cause redistribution of Polymer chain lengths and an increase in the polydispersity index. MALDl-TOF sPectra are particularly useful for detecting both cyclic and linear oligomers, since they allow the direct identification of mass-resolved polymer chains. Because Ce rtain stereosequences in the polymerization of rac- or mesa-lactide can only be c’t)t€iined by transesterification, 13C NMR ' spectroscopy can detect \thesterification reactions. M Intermolecular + .. _ > + w transesterification lntramolecular m transesterification O + ~ Scheme 1.3. lntramolecular and intermolecular transesterification 18 1 .5 - Thermal degradation and stability Polylactide belongs to a family of polymers with poor thermal stability. It can undergo slow thermal degradation at temperatures lower than the melting point of the polymer, but the degradation rate increases rapidly above its melting point.34 Thermal instability is a major limitation for some applications. For example, polylactide implants used in orthopedic surgery are supposed to Provide adequate strength, ductility, modulus, wear and fatigue resistance for il'ttemal fixation of bone fractures, and are expected to last until the new tissues are generated. However, if polylactide degrades during melt-processing (Compression, extrusion and injection-molding) or sterilization, the degradation profile and mechanical properties of the polylactide implants can be quite different from what is expected because the mechanical properties and in vivo 99 radation rates are largely dependent upon the molecular weight of the device. Polylactide samples from commercial suppliers are quite susceptible to extensive degradation after injection molding.35 For molding temperatures betWeen 130 °C and 215 °C and mold residence time of 12-16 seconds, the peak in the molecular weight distribution declined by 50-88% and the polydispersity increased. 19 transesterification' H \Okroko O intramolecular * volatile cyclic dimer or oligomers O A on I OH (ms/CH £yH dis-elimination _ o (B\ CH2 8 7 CH3 ' H OH coon ‘ojfié’o‘lfio’k ‘o’kgl' Wio’l‘k radio: reactions: C0, C02, aldehyde. cyclic oligomers Scheme 1.4. Thermal degradation pathways for polylactide 20 3-1. Thermal degradation pathways As shown in Scheme 1.4, there are three principal pathways for the armal degradation of polylactide. Path A illustrates thermal degradation by :ramolecular transesterification, also known as “back-biting”, which generates ilatile cyclic dimer or oligomers. Path B describes degradation by a cis- mination mechanism via the formation of a six-membered ring transition state. tth C shows degradation by radical pathways and the generation of volatile iall molecules. Ole-elimination is a concerted, un-catalyzed reaction. It can be the 'hinant pathway for esters with activated C-H bonds such as the poly(hydroxy tyrate)s which contain methylene hydrogens activated by an adjacent carbonyl bup.36 However, although it possesses three B-C-H bonds available for cis- hination, the methyl C-H bonds of polylactide are not activated and as a result, e contribution of the als-elimination pathway to the total thermal degradation of olylactide is trivial.“ lntramolecular transesterification34 is the dominant pathway for the thermal agradation of polylactide. Often catalyzed by residual metal catalyst, it is itiated from free hydroxy groups at the ends of the polymer chain. The cyclic mer, lactide, is the major degradation product from this pathway. The third pathway, radical reactions, is significant only at temperatures 200 °C.37 Homolytic cleavage of alkyl-oxygen bond or acyl-oxygen bond to rm macroradicals leads to the formation of volatile cyclic oligomers, CO, C02 id other small molecules. 21 The thermal degradation pathways and degradation products were investigated using several thermal analysis techniques including Then'nogravimetric Analysis (TGA, DTG), Differential Scanning Calorimetry (DSC), time and temperature resolved pyrolysis-MS and pyrolysis-GC/MS. Two distinct peaks (at 275 °C and 337 °C) were observed from DTG experiments indicating two dominant degradation mechanisms. GC/MS identified the low temperature peak as almost exclusively lactide, while the high temperature peak consisted of cyclic oligomers including lactide and other volatile small molecules.38 The products were consistent with the low temperature degradation dominated by depolymerization, while the high temperature degradation resulted from radical reactions. Lactide was produced in both low and high temperature degradation steps, but it was formed by different mechanisms.38 In low temperature degradation, it was produced as a result of depolymerization. In high temperature degradation, lactide as well as cyclic trimers up to pentamers were Procluced by a radical mechanism. 1.5-2. Factors affecting thermal degradation of polylactide ‘i Effect of polymer molecular weight Polylactide samples with different molecular weights show differences in theme] stability. Tests on highly purified samples (precipitation followed by Wafihing with dilute acid, and extensive drying) showed that the degradation \emperature initially increased sharply, and approached 353 °C as the viscosity 22 molecular weight rose to 100,000.39 This effect is related to the concentration of terminal hydroxy groups in the polymer sample, which decreases as the molecular weight increases. Since the terminal hydroxy groups initiate intramolecular transesterification, decreases in their concentration shift the onset f0r themtal degradation to higher temperatures. When the molecular weight of the polymer was high enough, the number of terminal hydroxy groups became negligible and any further reduction in their concentration leads to only small i"creases in the thermal degradation temperature. 2) Effect of residue metal catalysts on thermal degradation Polylactide samples that have not been scrupulously purified are usually contaminated with >100 ppm of metal catalyst residues. These metals can lower the polymer degradation temperature by coordinating with the carbonyl oxygen of esiel‘s and facilitate intramolecular transesterification. Several different organometallic compounds of Sn, Al were studied,4042 and in general, Sn(II) compounds were more active transesterification catalysts than Al compounds. The Sn(H) compounds are thought to interact more strongly with polylactide ester groups than Al(IH) compounds due to their larger ionic radius. 3) Effect of residual unreacted lactide or lactic acid on thermal degradation Flesidual lactide and lactic acid in polylactide samples can cause a weight loss at lower temperatures.39 However, the thermal degradation temperature of 23 polylactide samples is unaffected once lactide or lactic acid is removed by low temperature isothermal treatment. 4) Effect of additives Various additives have been used to enhance the thermal stability of DOIyIactide. For example, peroxides were added to stabilize polylactide '7lelts."’3'44 It was proposed that peroxides could deactivate residual metal Catalysts and introduce branches on polymer chains to counteract chain scission. When mixed with crude polylactide prior to processing, tropolone (2-hydroxy- 2:4.6-cycloheptatrienone), stabilized polylactide during melt processing by fO'Tirli ng chelating complexes with tin.45 24 1.6. Hydrolytic degradation The most attractive feature of polylactide is its degradability. It contains a high density of ester groups in the main chain and degrades through their hydrolysis. The hydrolytic degradation of polylactide has attracted much attention during the past two decades because of the polymer's potential as degradable medical and consumer products as well as the existence of many factors which Can influence the degradation process. Despite some important advances, some Controversies still exist in the literature. The hydrolytic degradation time for polylactide samples varies from a couple of weeks to several years depending on the polymer molecular weight, C’YStallinity, chemical composition, purity, size, additives (incorporated drugs), medi um pH, and temperature. surface erosion ”m"?! m. _ Hydrolytic degradation Enzymatic degradation in phosphate buffered solution Y-‘\9\)re 1.5. Bulk erosion and surface erosion in biodegradable polymers 25 1.6.1 . Degradation mechanisms There are two types of degradation processes: bulk erosion and surface erosion (Figure 1.5). Surface erosion is observed when the rate of water diffusion into the polymer matrix is lower than the rate of converting the polymer into water soluble oligomers. Polyanhydrides and polyorthoesters, which contain Chemical bonds highly sensitive to hydrolysis, are examples of materials that eXhibit surface erosion.”47 The hydrolytic degradation of polylactide follows a different mechanism - bulk erosion. Polylactide degrades through the hydrolysis 0’ ester bonds generating one carboxylic acid and one hydroxyl group for each ester hydrolyzed. The carboxyl groups thus formed catalyze the hydrolysis of Other ester bonds and increase the degradation rate, a phenomenon known as allt<><>atalysis. A feature of bulk erosion is that the molecular weight of the Sam ple decreases from the beginning of the degradation process, but weight loss 03“ only be observed after extensive cleavage of ester bonds and the formation of Water soluble oligomers. Small-sized polylactide particles and devices such as thin films and macro/nanospheres are thought to degrade homogenously, however, much faster degradation rates have been reported for the interior of large amorphous DOIYIactide samples."’8 The surface-interior differentiation was obvious due to the formation of a hollow structure. GPC data also revealed a bimodal molecular W6§§\\\ distribution. This degradation phenomenon is termed heterogeneous degradation and can be explained by a reaction-diffusion mechanism. Before degradation the polymer sample is homogeneous in terms of molecular weight 26 and molecular weight distribution. Once placed in a degradation medium, water penetrates into the polymer sample resulting in homogeneous hydrolytic cleavage of ester bonds throughout the sample. This macroscopically homogeneous hydrolytic degradation continues until water-soluble oligomers are generated. Those oligomers generated on or near the surface can escape from the matrix, dissolve in surrounding medium and are neutralized by the buffer. However, the oligomers generated inside the matrix cannot diffuse out of the sample, especially when the degradation temperature is lower than the glass transition temperature of the polymer. As degradation proceeds, acid-terminated o”gol‘ners accumulate at the core of the polymer leading to enhanced aUtocatalysis. Thus, the core of the polymer specimen degrades at a much faStS r rate than the shell, resulting in surface-interior differentiation. The bimodal mo‘ecular weight distribution reflects the existence of two polymer populations that degrade at different rates. 1.5.2- Factors affecting the hydrolytic degradation of polylactide 1) Monomer The effect of a small amount of monomer remaining in polymer samples on hydrolytic degradation was investigated by lkada.49 Accelerated hydrolysis of aS'POIyrnerized amorphous samples was observed compared to purified ones 6V%“ '\f the monomer content was as low as 5 wt%. During degradation experiments, residual monomers can be extracted from a matrix or hydrolyze to give a hydroxy acid. The hydrophilicity of the hydroxy acid not only enhances the 27 diffusion of water into the polymer matrix, but also catalyzes the hydrolysis of other ester groups. Reproducible degradation results require monomer-free samples. 2) Crystallinity In crystalline poly(L-lactide), hydrolytic degradation was found to occur Preferentially in amorphous regions. For samples of the same molecular weight in phosphate-buffered solution, samples with higher degrees of crystallinity showed faster declines in molecular weight, and the crystallinity of all films increased monotonically with hydrolysis.50 Due to looser chain packing, the diffusion coefficient of water is higher in the amorphous regions of a semi- c’l’sitatlline polymer than in the crystalline phase. There are two types of amC>I"phous regions, the amorphous region between the lamellae of spherulites, and free amorphous regions. Hydrolytic degradation occurs preferentially in the amorphous region near the surface of lamellae because of a high concentration of terminal carboxyl and hydroxyl groups, which are excluded from the crystalline feQion during crystallization. A higher initial crystallinity can introduce more defects in the amorphous region, thus leading to easier water penetration and faster degradation.50 3\ Kfid’fiives When drug delivery systems are considered, the effect of loaded compounds on the hydrolytic degradation of the polymer matrix is particularly 28 interesting. If an acidic compound is incorporated, it accelerates the degradation of the polymer matrix. However, a basic compound can act as either a catalyst or an acid neutralizer. lts effect on the degradation of the polymer depends on the relative importance of the two effects. For example, when coral was incorporated in polylactide for bony tissue regeneration, it mainly neutralized carboxyl end groups and slowed the degradation rate by eliminating the aUtocatalytic effect.51 However, when the base was caffeine, its effect on the degradation strongly depended on the loading concentration.52 4) Effect of copolymers Copolymers of polylactide, such as poly(lactide-co-glycolide)I were synthesized to tailor the rates of degradation. A copolymer of 50% rad-lactide and 50% glycolide degraded faster than both homopolymers and copolymers with other compositions. Surprisingly, there was no linear relationship between the cepolymer composition and the degradation rates. This effect may be related to crystallinity in the polymer since the homopolymers have a higher degree of crystallinity than the copolymers. 29 1.6.3. Degradation models Random chain scission and autocatalysis models have been cOhstructed for the molecular weight and sample weight change during hydrolytic degradation. The random chain scission model53 is based on two assumptions: each ester link has equal probability of being attacked by water, and dn/dt, the rate of breaking links is proportional to n, the number of links present in the system. ' dn/dt = kn The degree of degradation, 3, is defined as the number of broken links per chain deed by Po, the original degree of polymerization the chains. 8 =( Mn(0)/Mn(t)-1)/(Pa-1)/ M"(0) ‘8 the initial number average molecular weight and Mn(t) is the number average molecular weight at degradation time t. Thus, at any time during the degradation process, n = no - ano = non-a) and - d[no(1-a)]/dt = kno(1-a) '"tsgl’ating and using the approximation -In(1-a) z - a gives a = kt and when Po >>1 Mn(0)/M,,(t) - 1 = kPot eq. 1.1 30 in the autocatalysis model, the cleavage of ester links is catalyzed by carboxylic acid end groups in the system at a rate proportional to the concentration of acidic end groups (eq. 12).“55 d[COOH]/dt = k”[H20][ester][COOH] Where [COOH], [ester] and [H20] are the concentration of the terminal carboxyl QTOUps, ester groups and water in the system repectively. k” and the following k’ eq.1.2 and k are rate constants. When the number of chain scissions is small, both [H20] and [ester] can be considered to be constants and combined with k”. 80, d[COOH]/dt=k’[COOH] Since [COOH] a: 1/M,, ln[Mn(0)/Mn(t)1 = kl The Prout-Tompkins equation (eq. 1.3) was applied by Flamtoola to evaluate mass loss from poly(lactide-co—glycolide) particles.56 The original model was based on auto-catalytic thermal decomposition of potassium permanganate. The expression is as follows: eq. 1.3 = 'ktmax ln[x/(1-x)] = kt + m Where x is the fractional mass remaining at time t; k is the rate constant weight loss and tmax is the time to achieve 50% weight loss 31 1‘ 1.6.4. Enzymatic degradation The enzymatic degradation of polylactide follows a surface erosion mechanism because the size of an enzyme prevents it from diffusing into the POIymer matrix. Enzymes that degrade polylactide include pronase, proteinase-K and bromelain. Proteinase K preferentially degrades L-lactyl units, and the hydrolysis rate decreases for high concentrations of D-lactyl units, and when the distribution of the D and L monomer units becomes more random.”58 For Crystalline poly(L-lactide), enzymes selectively attack amorphous regions rather than crystalline regions,59 and as the degree of crystallinity increases, the enzymatic degradation rate decreases. The degree of crystallinity of poly(L- lactide) samples also increases upon degradation due to preferential degradation and Partial crystallization of the amorphous region. Due to the specificity of enzymes , a two-component blend, composed of poly(L-lactide) and poly(e- CaPVO'aCtone), can be selectively degraded to yield porous biodegradable polyester materials.60 1-6-5- Biodegradable polymers as drug carriers Drug release from biodegradable polymers is a complicated process, which occurs by several, often simultaneous, mechanisms such as diffusion ""0th intact polymers, diffusion through water-swollen polymers and surface layers’ or bulk erosion of polymers. The importance of each individual mec . . . h artism in drug release depends on the composrtion and molecular weight of 32 the polymer matrix, particle size, the nature and content of incorporated drugs as well as fabrication methods. The three general cases are diffusion control (polymer erosion slower than the diffusion processes), erosion control (polymer erosion is the fastest process), and control by swelling (diffusion of water into the polymer is faster than polymer erosion, but slower than polymer relaxation). In some studies,61 '62 biodegradable Polymer erosion was not observed during the period when drug release took Place, and the only advantage in these systems would be the eventual disappearance of drug carriers though degradation. Nanoparticles are defined as solid particles ranging from 1-1000 nm in size. Polymeric nanoparticles can be prepared by polymerization of reactive monomers in a dispersed phase or from preformed polymers. Drawbacks of the first strategy include the use of large volumes of organic solvents and the presence of residual monomers, catalysts, and solvents. The second strategy offers a more promising approach especially when biodegradable and biommF>etible polymers such as polylactide and its copolymers are used as the polymer rnatrices. Various methods have been used to prepare nanoparticles from Preformed polymers including emulsion-evaporation, solvent displacement (nano pr°°ipation), emulsification, solvent diffusion, and dialysis. These methods are Simi'a" in that they all require an organic solution containing the nanoparticle (:0me ments and an aqueous solution with or without stabilizers. At present, em - . “Iglon-evaporation is the most widely used method, but it poses problems 33 such as removal of solvent and surfactant residues due to their toxicities. ln addition, a homogeneous emulsion is required to produce nano-sized particles. The conventional procedure, ultrasonication, can sometimes induce chemical reactions or polymer degradation. Recently, a dialysis method using amphiphilic materials was developed for the preparation of nanoparticles with narrow size distributions.”65 It also proved to be a simple and effective preparation method for poly(lactide-co-glycolide) nanoparticles.66 F" _’-—' 1.7. Random copolymers of polylactide Despite the attractive properties of polylactide, it is difficult for polylactide to fulfill all applications due to its high crystallinity, hydrophobility and a lack of functional groups. Copolymerization of lactide with other monomers has been intensively investigated to better control the degree of crystallinity as well as its degradation behavior. it ° . U + \')J\o +OfioiOm/flm TMC Q trimethylene carbonate poly(lactide-co-TMC) oio + $ ————> +Ojfl‘1joj‘ofidm 56 ° DTC _ poly(lactide-co-DTC) 2 . 2 ~dimethyl-tnmethylene carbonate 0 JL 0 00 \(Uxo i WK 0 °HTC 2,2-[2-pentene-1 ,5-diyl]- poly(lactide-co-cHTC) t I'imethylene carbonate +O\g/L};{koj\o m Scheme 1.5. Copolymers of lactide and carbonates 35 1.7.1 Copolymerization with carbonates The carbonates that have been copolymerized with lactide include trimethylene carbonate (TMC), 2,2—dimethyl-trimethylene carbonate (DTC) and 2,2-[2-pentene-1,5-diyl]-trimethylene carbonate (CHTC) (Scheme 1.5). The carbonate linkage is more hydrophobic than an ester, and copolymers of carbonates and lactide are expected to be more stable toward hydrolytic degradation than polylactide. The homopolymer of TMC is an amorphous or poorly crystalline material With a glass transition temperature of ~ —18 °C. The melting temperature, crystallinity, and glass transition temperature of polylactide decreased with increasing TMC in the copolymerssm9 Copolymers with mechanical properties ranging from brittle and highly crystalline to rubbery and flexible, can be prepared by adiUSting the monomer feed ratio. For example, polyglycolide, an analog of PO'Y(L-la.ctide), is highly crystalline, stiff (melting point around 219 °C) and fails to meet the material requirements for surgical sutures. Copolymers containing TMC haVe been developed for flexible, strong and absorbable monofilament sutures- Although poly(lactide-co-TMC) was more stable toward in vitro hydro‘Ytic degradation conditions, in vivo degradation revealed a much faster degradation due to an enzymatic degradation process. Thus, incorporation of TMC in polylactide leads to increased shelf life and faster in viva degradation. The DTC homopolymer is crystalline (mp ~108 °C) with a glass transition tempe rature of ~ 27 °C. Poly(L-lactide-co-DTC) copolymers containing 11-88 mol'ry‘3 , . DTC are amorphous despite the fact that both homopolymers are 36 crystalline.7°'71 When the DTC content in the copolymer is higher than 50 mol%, the glass transition temperature is below normal body temperature (37 °C). This may have important implications for biomedical applications, since both mechanical properties and degradation rates change dramatically at the glass transition temperature. CHTC, a cyclic carbonate containing a cyclohexene moiety, was COpolymerized with L-Iactide to introduce unsaturated C=C double bond groups in the copolymer and provide opportunities for further modifications such as epoxidation.72'73 The incorporation of 0HTC decreased the glass transition and the melting temperature of poly(L-lactide) as in the above cases. 1 .7.2. Copolymerization with caprolactone and its derivatives Poly(e—caprolactone) is a semi-crystalline biocompatible and biOdeQl'adable polyester with low melting temperature (63 °C) and low glass transition temperature (-60 °C). It degrades with a half life of one year in vivo and possesses higher permeability than polylactide, which is hardly permeable to most ciI’ugs. Thus, a wide range of drug delivery matrices with adjustable prope "ties can be achieved by combining the features of both polymers through coF’C’HII'nerization (Scheme 1.6).M76 Substituted caprolactone derivatives were also Qonlymerized with lactide.”78 37 O O O Ufixflloww O poly(lactide-co-caprolactone) Scheme 1.6. Copolymerizations of lactide and caprolactone 1.7.3. Copolymerization with morpholine-2,5-dione derivatives The hydrophobicity of polylactide and its lack of functional groups has made it unattractive as a carrier for water-soluble drugs such as peptides and proteins. One attempt to improve hydrophilicity and provide functional groups is the synthesis of polyesteramides from morpholine-2,5-dione derivatives that have amino, carboxylic and hydroxy side chain functional groups. a-Amino acids such as 'Ysine ,79'8" glutamic acid81 and aspartic acid82 have been copolymerized with lactide by an indirect method. After protecting their side chain functional groups, the am i no acids were condensed with 2—bromopropionyl bromide to give rnOrPh<>|ine-2,5-dione derivatives (Figure 1.6), which can be copolymerized with lactide and deprotected. These morpholine-2,5-diones polymerize poorly, giving low polymerization rates and low molecular weights. Copolymers with lactide were synthesized to o"°'<>Qme this difficulty and take advantage of the desirable physical properties Of leylactide. During homopolymerization and copolymerization, the ring op en ihg of the morpholine derivatives proceeded exclusively by the cleavage of the egter bond. 38 I U l -V' I The deprotected functional group can be further modified to improve polylactide-cell interactions. For example, Langer79'80 reported the synthesis of poly(lactide-co—lysine) and attachment of a cell adhesion promoting peptide to the copolymers primary amino group. O O R = R = R: H o 0 Ph /\/\/N OVPh /\n/ v \[r /\ o 0 Ph 0 Fig ure 1.6. Various morpholine-2,5-dione derivatives copolymerized with lactide 1-7-4- Copolymerization with glycolide and other substituted glch‘“leis Polyglycolide is highly crystalline with a low solubility in most organic solvents_ Copolymerization of glycolide with lactide provides a method for dismpting the crystallinity and tuning the degradation rate. The absence of methy| substituents on the glycolide ring makes the monomer mere reactive toward ring opening polymerization due to reduced steric hin drance- COpolymerS Of lactide with substituted glycolides such as ethylglycolide and isopropylglycolide have been studied by Yin and Wang (Scheme 1.7).‘33'84 Random copolymers with glass transition temperatures ranging from 15 — 66 °c were prepared by varying the feed ratio of lactide and ethylglycolide- 39 O O o . _._. (Owe n O m 0 0 poly(lactide-co-ethylglycolide) O O O O + 0 Cat. +0 0 0% OW m n 0 0 O poly(lactide-co-isopropylglycolide) Scheme 1 .7. Copolymerization of substituted glycolides with lactide 40 - — l I} E y 1.8. Block copolymers 1.8.1. General Block copolymers are macromolecules comprised of blocks or homosequences that are joined at their ends. Different block copolymer architectures can be realized by using synthetic procedures that control the connectivity of the blocks. The most common block copolymer architectures are AB diblock, ABA(C) triblock, comb, star and multiblock copolymers. Block copolymers are different from polymer blends in that the blocks are chemically linked. Besides displaying the properties of each block, block COPO'ymers often microphase separate and give rise to interesting physical behavior- In a heterogeneous polymer blend, polymers phase separate at the macrOSCOpic level which leads to domains >100 pm that can easily been seen under an optical microscope. Since the blocks of a block copolymer are Chemically joined to each other, as they phase separate they mus:t place the junction between the two blocks at the interface between the phases“ Thus me . to 59"8‘3\ domains must be small, on the order of several nanome‘e‘s micrometers. This microphase separation can lead morphologies and thus new properties. One of the most su Q06 d er'Ned “0‘“ . 9‘9 m'CTODhase separated block polymers is thermoplastic elastom rs where 8 7 0M" DOB/Styrene-polybutadiene-polystyrene ABA block cop ' 6 lybutadlen ' PO'YStyrene forms spherical domains in a continuous matrix of P 0 act 85 F’O'YStyrene has a glass transition temperature above 80 ac and can . . cK physucaj crosslmkers for the polybutadiene softblock Matrix. The MO 41 W , _ I copolymers are thus elastomers at room temperatures while still processible at temperatures higher than the glass transition temperatures of polystyrene because the polymers are not chemically crosslinked. Microphase separation behavior also leads to important applications such as adhesives, compatiblizers. For instance, diblock copolymers can be used to decrease the size of phase separated domains, decrease the interfacial tension and improve the mechanical properties of immiscible blends.8588 1.8.2. Phase separation and morphology When two polymers are mixed, more often than not, they are immiscible and Phase separate. The free energy of mixing AG... is given in eq. 1.4: AGM = AHM - TASM eq. 1.4 where AHM and ASM are the enthalpy and entropy of mixing respectively and T is temperature. Usually polymers have very small values of ASM due to their high . . . hereio'e’ molecular weight and ASM decreases as molecular weight Increases' T aka 56M . . \o m a 5"9htly positive enthalpy due to endothermic mixing is s1.|t't'\0‘e“x posmve. resulting in phase separation. as eepe“d‘“9 _ 0““ A Microphase separation leads to different classes of $1 bolh docks’ on the block cepolymer composition. For non-crystalizabl ¢ bases' If the and 3 blocks form random coils and segregate into separate P N - NA ‘ 5’ space requirement of the A blocks matches that of the B bloc/‘5' "a" . and B where N A and Ne are the number of monomer unlts in b/OCkS A reSPeCtiVely, then, lamellae with altemating A and B blocks will form (Figu’e 1'7 42 C). If NA << NB, packing in lamellae would either dissatisfy the requirement of a densest packing of segments, or lead to a large deviation from the unperturbed coil structure. Thus small A blocks will form spherical domains in the continuous matrix of B blocks (Figure 1.7 A). For larger NA (still NA < NE), A blocks will assemble into cylindrical domains in a continuous matrix of B (Figure 1.7 B). In addition to the ordered spheres, cylinders and layers, a bicontinuous structure exist in a narrow range of NA/ NB, between the cylindrical ad lamellar phases. B C Figure 1.7. Morphologies of AB block copolymers. White portions represent bl°°k A, while dark portions represent block B of the AB block copolymer 43 1 3.3. Block copolymers with poly(e-carprolactone) Block copolymers of lactide and caprolactone combine the good permeability of polycaprolactone with the relatively fast degradation rate of polylactide, providing controllable periods of biodegradation and drug release. They also act as blend compatiblizers for polylactide blends because Polycaprolactone is known to be miscible with many commodity polymers such as poly(vinyl chloride) and polycarbonates. <3 c... . i0 {gm/VOWWL O :0 polycapmlactone-b-polylactide Scheme 1 .8. Block copolymer of lactide and caprolactone Lactide and caprolactone can be polymerized by common metal catalysts h such as AI(OR)3 and Sn(Oct)2. The terminal hydroxy group 0‘ eac homopolymer can be considered as the initiator for the “e?“ ““9 tone ca9‘°\ac “e“ . DOB/me rization. In reality, block copolymers are obtained only is Poly . - .- - - 6°“ 60‘“ menzed and then used to Inmate lactide polymenza d (an Scheme 1.8.89 When the order of polymerization was COPOIYI'hers of lactide and caprolactone were obtained unde r D d' - 5 atiflb 00" 't'Ons. The formation of random copolymers Wa e is "‘0’ tranSGSterification reactions. The hydroxy chain and 01‘hob/carpfa/acmna _ . The reactive than that of polylactide due to both electronic and steric factors polylactide hydroxy chain end is less nucleophilic because it is a to the 3190170" 44 ”Terri withdrawing carbonyl group. In addition, the a—methyl group of the lactyl end group sterically hinder nucleophilic attack by the hydroxy and group. Since the occurrence of transesterification largely depends on the reactivity of the hydroxy end groups in the polymerization system, growing the polycaprolactone after lactide favors transesterification. To bypass the chain end reactivity issue and synthesize triblock COpolymers with polylactide as the center block, Song90 used a bimetallic catalyst to polymerize caprolactone followed by lactide. By extending the polylactide chain with ethylene oxide, they were able to obtain active initiating sites for ring opening polymerization of caprolactone (Scheme 1.9). O O - ‘ Teyssre Cat. A ethylene oxide_ ’(JK/VVO o : 2-4 0 O 0 Cat. (1‘) ll o O)“: MOWWW {6 Scheme 1.9. Lactide and caprolactone triblock copo I 31mg 45 1 8.4. Block copolymers with poly(ortho ester)s The poly(ortho ester)s are a family of hydrophobic biodegradable polymers, which under certain conditions, undergo hydrolytic degradation by a surface erosion mechanism. Because the ortho ester linkages in the polymer are very susceptible to acidic conditions, acidic additives are usually physically incorporated into the polymer to accelerate the rate of degradation. This approach can be problematic because the additives can diffuse out of the o o o o o OXP HWOH ‘ 0 0 p-TSA THF ll {4sz X :X |n|f°fioHoT>polymers apparently are bIocky, since it contained two distinct crystalline phases.107 Graft copolymers were prepared to promote hydrolytic degradation of polylactide. During hydrolytic degradation, the oligomers generated would reach 50 o 0 R1 Oviolkf 0% YL “Md H m n om/OK + m Cat- t polylactide-b-PEO Wage. O a} polylactide-b-PEO-b-polylactide “*‘K/toizw ——* Riwoiiringgugwglfiiiowr PEO-b-polylactide—b-PEO O AIR3°O.5 H o 0 \KILO + 8x 2 s {f MW WK ° n O p polylactide and PEO multiblock copolymers Scheme 1.13. Block copolymers of lactide and ethylene oxide 51 ,fifi‘ he \‘mes‘nofi ot water so\ub'\\'\ty much more rapidly because they are composed oi both PEG and po\y\act‘\de segments. ln graft copolymers, both polylactide and PEG can serve as the backbone or teeth. Graft copolymers with PEO as the backbone and either polylactide or polyglycolide as the teeth were prepared by condensation of a PEO oligomer bearing an epoxy group at each end with a second PEO oligomer terminated with two carboxylic acids. Ring opening POIyrnerization of lactide and glycolide was initiated from the pendent hydroxy QFOUps to give a graft copolymer (Scheme 1.14).”8 Graft copolymers with POlylactide as the backbone were prepared as shown in Scheme 1 .14.‘°9 One practical limitation in applications of poly(lactide-b-ethylene oxide) copolymers is the PEO block length. Although PEO is biocompatible, it is non- C'egradable, and due to its large hydrodynamic volume, PEO with molecular Weights >10,000 g/mol cannot be filtered through human kidney membranes and eliminated from human bodies. To solve this problem, star-shaped copolymers Were prepared"°'111 because their hydrodynamic volumes and solution Vi8cosities are lower than for linear copolymers with the same composition and molecular weight. Thus larger blocks of PEO can be incorporated into star- shaped polymers to increase the hydrophilicity of the system without affecting the e>polymer architecture as well as its concentration."3'114 For example, the Polymers can be dissolved into water to form a homogeneous solution (sol) at r(Dom temperature, and once the solution is injected into human bodies, it quickly gels in Situ to form a drug delivery matrix as shown in Figure 1.8. v Gel (37°C) Sol (r.t.) Figure 1.8. Sol-gel transition in lactide and ethylene oxide 54 ‘\ .9- Atom transfer radical polymerization (ATRP) The past few years have witnessed the rapid development of transition metal catalyzed atom transfer radical polymerization (ATRP). ATRP has been widely used in the literature to prepare polymeric materials with novel functionalities, compositions and architectures.115117 ATRP is one of the most versatile systems among the recently developed controlled radical polymerization methods.118 It is based on establishing a rapid Whamic equilibration between a low concentration of active free radicals and a large concentration of dormant species. The well accepted mechanism for ATRP is shown in Scheme 1.15. In the initiation and propagation steps, the radicals or active polymer sites are generated by a reversible transition metal mediated redox process, where X and Y are (pseudo) halogens. The transition metal (Cu for example) undergoes a one-electron oxidation with abstraction of a (pseudo) halogen atom X from the dormant species. The process is reversible with the rate constant of activation being km and the rate constant of deactivation being kdeact. The active species can grow by addition of monomers or terminate by c=§T° A 8 Figure 1.10. Difunctional monomers used for ring opening polymerization 4) Functional macromonomers Graft copolymers consisting of biodegradable and non-biodegradable components are interesting examples 0t polymers Whose physical and mechanical properties are controlled by composition, distribution of the comonomers in the chainS, as well as the chemical nature and length of the backbone and graft segments. A popular route to graft COPOtymers is to prepare Polymers with macromonomers. Using 2-nydroxyethyl (meth)acrylate as an initiator, glycolide, lactide, and caprolactone were oligomerized to give macromonomers.132.1ae Cepolymerization of the macromonomers with z‘hydVOXYothylmethacrylate, MM A, acrylates or itaconic anhydride by either free radical or controlled radical polymerization gave the graft architecture (Scheme 1.19), 65 or 0 MMA HEMA AlBN ATRP AlBN HO O l O O O 0 x y x X Y 0 yo 0 OZ 0 02 o 2 /0 polylactide polyla/ctide POlycaprolacone Scheme 1.19. Graft polymers synthesized by the macromonomer approach 66 References 1. 10. 11. 12. 13. 14. 15. 16. 17. Kinnersley, A. M.; Scott, T. C.; Yopp, J. H.; Whitten, G. H. Plant Growth Regal. 1990, 9, 137-146. Bogaert, J. C.; Coszach, P. Macromol. Symp. 2000, 153, 287-303. Lunt, J. Polym. Degrad. Stabil. 1998, 59, 145-152. Ajioka, M.; Suizu, H.; Higuchi, C.; Kashima, T. Polym. Degrad. Stabil. 1998, 59, 137-143. Woo, S. I.; Kim, B. 0.; Jun, H. 3.; Chang, H. N. Polym. Bull. 1995, 35, 415-421. Dubois, P.; Jerome, FL; Teyssie, P. Macromol. Symp. 1991, 42-3, 103- 116. Ropson, N.; Dubois, P.; Jerome, R; Teyssie, P. Macromolecules 1993, 26, 6378-6385. Dubois, P.; Jerome, R.; Teyssie, P. Polym. Bull. 1989, 22, 475-482. Leborgne, A.; Vincens, V.; Jouglard, M.; Spassky, N. Macromol. Symp. 1993, 73, 37-46. Spassky, N.; Wisniewski, M.; Pluta, C.; LeBorgne, A. Macromol. Chem. Phys. 1996, 197, 2627-2637. ‘ Ovitt, T. M.; Coates, G. W. J. Am. Chem. Soc. 1999, 121, 4072-4073. Radano, C. R; Baker, G. L.; Smith III, M. Fl. J. Am. Chem. Soc. 2000, 122, 1552-1553. Cameron, P. A.; Jhurry, 0.; Gibson, V. C.; White, A. J. P.; Williams, D. J.; Williams, S. Macromol. Rapid Commun. 1 999, 20, 61 6-61 8. Penczek, S.; Duda, A.; Kowalski, A.; Libiszowski, J.; Majerska, K.; Biela, T. Macromol. Symp. 2000, 157, 61-70. Ryner, M.; Stridsberg, K.; Albertsson, A. 0.; von Schenck, H.; Svensson, M. Macromolecules 2001, 34, 3877-3881. Kricheldorf, H. R; Lossin, M. J. Macromol. Sci-Pure Appl. Chem. 1997, A34, 179-189. Dunsing, R.; Kricheldorf, H. R. Polym. Bull. 1985, 14, 491-495. 67 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. Kricheldorf, H. Ft; Damrau, D. 0. Macromol. Chem. Phys. 1997, 198, 1753-1766. Schwach, G.; Coudane, J.; Engel, R.; Vert, M. Polym. Int. 1998, 46, 177- 182. Schwach, G.; Coudane, J.; Engel, Ft; Vert, M. Polym. Bull. 1994, 32, 617- 623. Kricheldorl‘, H. R.; Damrau, D. 0. Macromol. Chem. Phys. 1997, 198, 1767-1774. O’Keefe, B. J.; Monnier, S. M.; Hillmyer, M. A.; Tolman, W. B. J. Am. Chem. Soc. 2001, 123, 339-340. Simic, V.; Pensec, S.; Spassky, N. Macromol. Symp. 2000, 153, 109-121. Spassky, N.; Simic, V.; Montaudo, M. S.; Hubert-Pfalzgraf, L. G. Macromol. Chem. Phys. 2000, 201, 2432-2440. Yuan, M. L.; Xiong, C. D.; Li, X. H.; Deng, X. M. J. Appl. Polym. Sci. 1999, 73, 2857-2862. Deng, X. M.; Yuan, M. L.; Li, X. H.; Xiong, C. D. Eur. Polym. J. 2000, 36, 1151-1156. Shen, Y. 0.; Shen, Z. 0.; Zhang, Y. F.; Yao, K. M. Macromolecules 1996, 29, 8289-8295. Shen, Y. 0.; Zhu, K. J.; Shen, Z. 0.; Yao, K. M. J. Polym. Sci. Pol. Chem. 1996, 34, 1799-1805. Stevels, W. M.; Ankone, M. J. K.; Dijkstra, P. J.; Feijen, J. Macromolecules 1996, 29, 8296-8303. Stevels, W. M.; Ankone, M. J. K.; Dijkstra, P. J.; Feijen, J. Macromolecules 1996, 29, 6132-6138. Simic, V.; Spassky, N.; HubertPfalzgraf, L. G. Macromolecules 1997, 30, 7338-7340. Kowalski, A.; Duda, A.; Penczek, S. Macromolecules 2000, 33, 689-695. Kricheldorf, H. Ft; Eggerstedt, S. Macromol. Chem. Phys. 1999, 200, 1284-1291. Zhang, X. C.; Wyss, U. P.; Pichora, D.; Goosen, M. F. A. Polym. Bull. 1992, 27, 623-629. 68 35. 36. 37. 38. 39. 40. 41. 42. 43. 45. 46. 47. 48. 49. 50. 51. 52. 53. Gogolewski, S.; Jovanovic, M.; Perren, S. M.; Dillon, J. 6.; Hughes, M. K. Polym. Degrad. Stabil. 1993, 40, 313-322. Kopinke, F. D.; Mackenzie, K. J. Anal. Appl. Pyrolysis 1997, 40-1, 43-53. Wachsen, 0.; Reichert, K. H.; Krueger, R. P.; Much, H.; Schulz, G. Polym. Degrad. Stab. 1997, 55, 225-231. Kopinke, F. D.; Remmler, M.; Mackenzie, K.; Moder, M.; Wachsen, 0. Polym. Degrad. Stabil. 1996, 53, 329-342. Cam, 0.; Marucci, M. Polymer 1997, 38, 1879-1884. Degee, P.; Dubois, P.; Jerome, R. Macromol. Chem. Phys. 1997, 198, 1985-1995. Kricheldorf, H. Fl.; Berl, M.; Schamagl, N. Macromolecules 1988, 21, 286- 293. Noda, M.; Okuyama, H. Chem. Pharrn. Bull. 1999, 47, 467-471. Sodergard, A.; Nasman, J. H. Ind. Eng. Chem. Res. 1996, 35, 732-735. Carlson, D.; Dubois, P.; Nie, L.; Narayan, R. Polym. Eng. Sci. 1998, 38, 311-321. Wachsen, 0.; Reichert, K. H.; Kruger, R. P.; Much, H.; Schulz, G. Polym. Degrad. Stabil. 1997, 55, 225-231. Sparer, R. V.; Shih, C.; Ringeisen, C. D.; Himmelstein, K. J. J. Controlled Release 1984, 1, 23-32. Sanders, A. J.; Li, B.; Bieniarz, C.; Harris, F. W. Polym. Prepr. (Am. Chem. Soc, Div. Polym. Chem.) 1999, 40, 888-889. Li, S. M.; Garreau, H.; Vert, M. Joumal of Materials Science: Materials in Medicine 1990, 1, 198-206. Hyon, S. H.; Jamshidi, K.; lkada, Y. Polym. Int. 1998, 46, 196-202. Tsuji, H.; lkada, Y. Polym. Degrad. Stabil. 2000, 67, 179-189. Li, S. M.; Vert, M. J. Biomater. Sci-Polym. Ed. 1996, 7, 817-827. Li, S. M.; GirodHolland, S.; Vert, M. J. Control. Release 1996, 40, 41-53. Doi, Y.; Kanesawa, Y.; Kunioka, M.; Saito, T. Macromolecules 1990, 23, 26-31. 69 55. 56. 57. 58. 59. 60. 61. 62. 63. 65. 66. 67. 68. 69. 70. 71. 72. Cha, Y.; Pitt, C. G. Biomaterials 1990, 11, 108-112. Pitt, C. 6.; Cha, Y.; Shah, S. 8.; Zhu, K. J. J. Controlled Release 1992, 19, 189-199. ' Dunne, M.; Corrigan, 0. l.; Ramtoola, Z. Biomaterials 2000, 21, 1659- 1668. Reeve, M. 8.; McCarthy, 8. P.; Downey, M. J.; Gross, R. A. Macromolecules 1 994, 27, 825-831. Macdonald, R. T.; McCarthy, S. P.; Gross, R. A. Macromolecules 1996, 29, 7356-7361. Tsuji, H.; Miyauchi, S. Polym. Degrad. Stabil. 2001, 71, 415-424. Tsuji, H. l., Takeharu Macromol. Biosci. 2001, 1, 59-65. Gomer, T.; Gref, R.; Michenot, D.; Sommer, F.; Tran, M. N.; Dellacherie, E. J. Control. Release 1999, 57, 259-268. Lemoine, D.; Francois, C.; Kedzierewicz, F.; Preat, W.; Hoffman, M.; Maincent, P. Biomaterials 1996, 17, 21 91 -2197. Nah, J. W.; Jeong, Y. l.; Cho, C. S. J. Polym. Sci. Pt. B-Polym. Phys. 1998, 36, 415-423. Kwon, G. S.; Naito, M.; Yokoyama, M.; 0kano, T.; Sakurai, Y.; Kataoka, K. Phaim. Res. 1995, 12, 192-195. Lasic, D. D. Nature 1992, 355, 279-280. Jeong, Y. l.; Cho, C. S.; Kim, S. H.; Ko, K. S.; Kim, S. l.; Shim, Y. H.; Nah, J. W. J. Appl. Polym. Sci. 2001, 80, 2228-2236. Ruckenstein, E.; Yuan, Y. M. J. Appl. Polym. Sci. 1998, 69, 1429-1434. Zhu, K. J.; Hendren, R. W.; Jensen, K.; Pitt, C. G. Macromolecules 1991, 24, 1736-1740. Buchholz, B. J. Mater. Sci-Mater. Med. 1993, 4, 381-388. Schmidt, P.; Keul, H.; Hocker, H. Macromolecules 1996, 29, 3674-3680. Hori, Y.; Gonda, Y.; Takahashi, Y.; Hagiwara, T. Macromolecules 1996, 29, 804-806. Chen, X. H.; McCarthy, S. P.; Gross, R. A. Macromolecules 1998, 31, 662-668. 70 73. 74. 75. 76. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. Chen, X. H.; McCarthy, 8. P.; Gross, R. A. Macromolecules 1997, 30, 3470-3476. Song, C. X.; Sun, H. P.; Fang, X. D. Polym. J. 1987, 19, 485-491. Hill'anenVainio, M. P.; 0rava, P. A.; Seppala, J. V. J. Biomed. Mater. Res. 1997, 34, 39-46. Vion, J. M.; Jerome, R.; Teyssie, P.; Aubin, M.; Prudhomme, R. E. Macromolecules 1986, 19, 1828-1838. Shirahama, H.; Mizuma, K.; Umemoto, K.; Yasuda, H. J. Polym. Sci. Pol. Chem. 2001 , 39, 1374-1381 . Mecerreyes, D.; Humes, J.; Miller, R. D.; Hedrick, J. L.; Detrembleur, C.; Lecomte, P.; Jerome, R.; San Roman, J. Macromol. Rapid Commun. 2000, 21, 779-784. Barrera, D. A.; Zylstra, E.; Lansbury, P. T.; Langer, R. Macromolecules 1995, 28, 425-432. Barrera, D. A.; Zylstra, E.; Lansbury, P. T.; Langer, R. J. Am. Chem. Soc. 1993,115,11010-11011. Deng, X. M.; Yao, J. R.; Yuan, M. L.; Li, X. H.; Xiong, C. D. Macromol. Chem. Phys. 2000, 201, 2371 -2376. Veld, P. J. A.; Ye, W. P.; Klap, R.; Dijkstra, P. J.; Feijen, J. Makromol. Chem. 1992, 193, 1927-1942. Wang, C.; Mao, Y.; Baker, G. L. Abstr. Pap. Am. Chem. Soc. 2000, 220, 305-POLY. Wang, C.; Yin, E.; Baker, G. L. Abstr. Pap. Am. Chem. Soc. 2000, 219, 123-POLY. Retsos, H.; Margiolaki, I.; Messaritaki, A.; Anastasiadis, S. H. Macromolecules 2001, 34, 5295-5305. Chattopadhyay, 8.; Sivaram, S. Polym. Int. 2001, 50, 67-75. Kang, E. A.; Kim, J. H.; Kim, C. K.; 0h, 8. Y.; Rhee, H. W. Polym. Eng. Sci. 2000, 40, 2374-2384. Radonjic, G. J. Appl. Polym. Sci. 1999, 72, 291-307. Veld, P.; Velner, E. M.; VanDeWitte, P.; Hamhuis, J.; Dijkstra, P. J.; Feijen, J. J. Polym. Sci. Pol. Chem. 1997, 35, 219-226. 71 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. Song, C. X.; Feng, X. D. Macromolecules 1984, 17, 2764-2767. Schwach-Abdellaoui, K.; Heller, J.; Gumy, R. Macromolecules 1999, 32, 301 -307. N9, S. Y.; Vandamme, T.; Taylor, M. S.; Heller, J. Macromolecules 1997, 30, 770-772. Sintzel, M. B.; Heller, J.; N9, 8. Y.; Taylor, M. S.; Tabatabay, C.; Gumy, R. Biomaterials 1 998, 19, 791 -800. Zhang, S.; Hou, Z.; Gonsalves, K. E. J. Polym. Sci. Pol. Chem. 1996, 34, 2737-2742. Bachari, A.; Belorgey, G.; Helary, G.; Sauvet, G. Macromol. Chem. Phys. 1995, 196, 411-428. Wang, Y. B.; Hillmyer, M. A. Macromolecules 2000, 33, 7395-7403. Schmidt, S. C.; Hillmyer, M. A. Macromolecules 1999, 32, 4794-4801. Wang, Y. B.; Hillmyer, M. A. J. Polym. Sci. Pol. Chem. 2001, 39, 2755- 2766. Deng, X. M.; Xiong, C. D.; Cheng, L. M.;'Xu, R. P. J. Polym. Sci, Part C: Polym. Lett. 1990, 28, 411-416. Deng, X. M.; Xiong, C. D.; Cheng, L. M.; Huang, H. H.; Xu, R. P. J. Appl. Polym. Sci. 1995, 55, 1193-1196. Li, S. M.; Rashkov, l.; Espartero, J. L.; Manolova, N.; Vert, M. Macromolecules 1 996, 29, 57-62. Rashkov, l.; Manolova, N.; Li, S. M.; Espartero, J. L.; Vert, M. Macromolecules 1996, 29, 50-56. Cerrai, P.; Tricoli, M. Makromol. Chem, Rapid Commun 1993, 14, 529- 538. Deng, X. M.; Zhu, Z. X.; Xiong, C. D.; Zhang, L. L. J. Polym. Sci. Pol. Chem. 1997, 35, 703-708. Stevels, W. M.; Ankone, M. J. K.; Dijkstra, P. J.; Feijen, J. Macromol. Chem. Phys. 1995, 196, 3687-3694. Du, Y. J.; Lemstra, P. J.; Nijenhuis, A. J.; Vanaert, H. A. M.; Bastiaansen, C. Macromolecules 1995, 28, 2124-21 32. 72 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. Chen, X. H.; McCarthy, S. P.; Gross, R. A. Macromolecules 1997, 30, 4295-4301. Jeong, B.; Kibbey, M. R.; Bimbaum, J. C.; Won, Y. Y.; Gutowska, A. Macromolecules 2000, 33, 8317-8322. Cho, K. Y.; Kim, C. H.; Lee, J. W.; Park, J. K. Macromol. Rapid Commun. 1999, 20, 598-601. Choi, Y. R.; Bae, Y. H.; Kim, S. W. Macromolecules 1998, 31, 8766-8774. Li, Y. X.; Kissel, T. Polymer 1 998, 39, 4421 -4427. lijima, M.; Nagasaki, Y.; 0kada, T.; Kato, M.; Kataoka, K. Macromolecules 1999, 32, 1140-1146. Jeong, B.; Baa, Y. H.; Lee, D. S.; Kim, S. W. Nature 1997, 388, 860-862. Fujiwara, T.; Mukose, T.; Yamaoka, T.; Yamane, H.; Sakurai, S.; Kimura, Y. Macromol. Biosci. 2001, 1, 204-208. Coessens, V.; Pintauer, T.; Matyjaszewski, K. Prog. Polym. Sci. 2001, 26, 337-377. Patten, T. E.; Matyjaszewski, K. Adv. Mater. 1998, 10, 901 -91 5. Matyjaszewski, K. Chem-Eur. J. 1999, 5, 3095-3102. Matyjaszewski, K.; Xia, J. Chem. Rev. 2001, 101, 2921-2990. Ashford, E. J.; Naldi, V.; 0’Dell, R.; Billingham, N. C.; Arrnes, S. P. Chem. Commun. 1999, 1 285-1 286. Davis, K. A.; Charleux, B.; Matyjaszewski, K. J. Polym. Sci. Pol. Chem. 2000, 38, 2274-2283. Marsh, A.; Khan, A.; Garcia, M.; Haddleton, D. M. Chem. Commun. 2000, 2083-2084. Marsh, A.; Khan, A.; Haddleton, D. M.; Hannon, M. J. Macromolecules 1999, 32, 8725-8731. Alkan, S.; Toppare, L.; Hepuzer, Y.; Yagci, Y. J. Polym. Sci. Pol. Chem. 1999, 37, 4218-4225. Kajiwara, A.; Matyjaszewski, K. Macromolecules 1998, 31 , 3489-3493. Kim, J.-B.; Bruening, M. L.; Baker, G. L. J. Am. Chem. Soc. 2000, 122, 7616-7617. 73 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. Haddleton, D. M.; Kukulj, D.; Radigue, A. P. Chem. Commun. 1999, 99- 100. Zhang, 0.; Remsen, E. E.; Wooley, K. L. J. Am. Chem. Soc. 2000, 122, 3642-3651. Hedrick, J. L.; Trollsas, M.; Hawker, C. J.; Atthoff, B.; Claesson, H.; Heise, A.; Miller, R. D.; Mecerreyes, D.; Jerome, R.; Dubois, P. Macromolecules 1998, 31, 8691-8705. Bielawski, C. W.; Morita, T.; Grubbs, R. H. Macromolecules 2000, 33, 678- 680. Guo, Z. R.; Wan, D. C.; Huang, J. L. Macromol. Rapid Commun. 2001, 22, 367-371. Mecerreyes, D.; Atthoff, B.; Boduch, K. A.; Trollsas, M.; Hedrick, J. L. Macromolecules 1 999, 32, 51 75-51 82. Dubois, P.; Jerome, R.; Teyssie, P. Macromolecules 1991, 24, 977-981. Wallach, J. A.; Huang, S. J. Biomacromolecules 2000, 1, 174-179. Furch, M.; Eguiburu, J. L.; Femandez-Berridi, M. J.; San Roman, J. Polymer 1 998, 39, 1 977-1 982. Barakat, |.; Dubois, P.; Jerome, R.; Teyssie, P.; Goethals, E. J. Polym. Sci. Pol. Chem. 1994, 32, 2099-21 10. Gopp, U.; Sandner, B.; Hahne, B. Macromol. Symp. 2000, 153, 321-332. 74 Chapter 2 Polymandelide 2.1. General The glass transition temperature (T g) is an important physical parameter of polymers since it defines the maximum temperature for which a polymer is suitable for structural applications. Above Te, segmental movement of polymer chains is possible and polymers are rubbery and elastic. Below T9, polymers are stiff and hard. The glass transition temperatures of all known substituted polylactides are < 70 °C. This is an obvious limitation for applications such as disposable packaging materials, where mechanical rigidity is important. For example, a polylactide cup used for a hot beverage such as coffee would soften and lose its original shape since the temperature of the liquid is above the glass transition of the material. To increase the T9 of polylactides, we used a simple strategy based on the known structure-property relationships of commercial polymers as shown in Table 2.1. By varying the substituents attached to a polymer backbone of aliphatic carbon atoms, one can obtain polyethylene (PE), polypropylene (PP), polyvinylchloride (PVC) polystyrene (PS) and other polymers. The glass transition temperatures of these polymers range from —100 to 109 °C, and satisfy the requirements of a broad range of applications. Since the glass transition is related to mobility of chains, increasing the chain stiffness or intermolecular interactions between chains increases T9. Thus, either increasing the steric bulk of the substituents (e.g. methyl, and phenyl in the 75 Table 2.1. Glass transition temperatures of commercial polymers. polymer abbreviation Tg (°C) —CH2'?H— PE -125 H —CH2'CH— PP -8 (EH3 —CHz-CH— PVC 81 a —CH2'CH— PS 109 structures shown in Table 2.1 or incorporating dipoles (6.9. CI) increase T9 and the chain stiffness. We expect to see the same trends in substituted polylactides. Poly(phenyllactide), where the methyl group of polylactide is replaced by benzyl, was examined as a potential high Tg material. However, the glass transition temperature of poly(phenyl lactide) is 55 °C, which is comparable to that of polylactide. The relatively low Tg can be explained by the methylene unit that links the benzene ring to the polyester backbone. The methylene unit reduces the steric barrier for rotation around the polymer backbone, thus increasing the flexibility of the polymer. 76 A reasonable way to increase the T9 of polylactides would be to make a simple analogy to polystyrene, eliminate the flexible methylene unit and attach a benzene ring directly to the polyester backbone. The systematic name of the resulting polymer is poly[oxy(1-oxo-2-phenyl-1,2-ethanediyl)], but polymandelide or poly(mandelic acid), common names based on the monomers used to synthesize the polymer, are more convenient. The trivial name polymandelide will be used for this degradable polystyrene mimic. There has been minimal work on the synthesis of polymandelide and copolymers. All reported syntheses have produced low molecular weight polymers and the characterization of the polymers has been limited. Polymandelide was first obtained accidentally by 0kada and 0kawara1 from the pyrolysis of a phenyl-substituted trimethyltin bromoacetate (Scheme 2.1 entry 1). The IR and NMR spectral data for the resulting white solid are consistent with the polymandelide structure. No molecular weight data were provided, but the physical properties of the solid imply a low molecular weight product. In 1980, the reaction of the ol-keto acid phenylglyoxylic acid, with 2-phenoxy-4,4,5,5- tetramethyl-1,3,2-dioxaphospholane was used by Kobayashi2 to synthesize polymandelide (Scheme 2.1 entry 2). The deoxy-polymerization yielded a polymer with a number average molecular weight around 2,400 g/mol as determined by vapor pressure osmometry measurements. Tighe and Smith3 reported the first example of polymandelide synthesized by a ring opening polymerization scheme (Scheme 2.1 entry 3). 5-Phenyl-1,3-dioxalan-2,4-dione (the anhydrocarboxylate derivative of mandelic acid) was shown to undergo ring 77 opening polymerization in the presence of tertiary organic bases such as pyridine to generate polymandelide and 002. The degree of polymerization was reported to be 25-30. Pinkus4 (Scheme 2.1 entry 4) prepared polymandelide by the reaction of a-bromophenyllactic acid with triethylamine. GPC and viscosity measurements indicated a degree of polymerization around 12-20, which was comparable to that obtained by other methods. Domb5 (Scheme 2.1 entry 6) and Whitesell6 (Scheme 2.1 entry 5) prepared polymandelide by polycondensation of mandelic acid and by transesterification of the methyl ester of mandelic acid. In both methods, p-toluenesulfonic acid was used as the catalyst and either high vacuum or a Dean-Stark trap was used to drive the equilibrium toward polymer formation. The molecular weights of the polymers obtained from the polycondensations were below 3,000 g/mol. Low molecular weight poly(lactide-co-mandelide) was synthesized by a number of research groups."10 The polycondensation method (Scheme 2.2 entry 1) was used to modify the thermal and mechanical properties of polylactide as t" described well as to achieve desired degradation profiles. A Japanese paten the first example of using mandelide, the cyclic dimmer of mandelic acid, as a comonomer to prepare poly(lactide-co-mandelide) (Scheme 2.2 entry 2). Trans- 4-hydroxy-L-proline was melt condensed with lactic acid or mandelic acid (Scheme 2.2 entry 3) to obtain new biodegradable copolymers with pendant functional groups and improved degradability compared to pseudopoly(amino acid).12 Thermal analysis showed an increase in T9 with increased incorporation of mandelic acid. 78 i \Slk Br 170°C 15hr o 1) /| O ’ ~ + _SnBl' pseudocumene O D Q Ph 100°C _ 2) P‘Q + OH 30hr ' 0‘40 Ph . + ,P\ O o O-Ph n I 0 Ph N 3) \‘JK = O Rt. {/0 '1' C02 0‘ / \ o—t O O n .. 0 OH EteN L o e 4) ' {/0 + EteNH Br 0 n I H OH We melt 140 g 5) 0 vacuum ' {/0 + M90” 0 n H OH OH p-TSA > O + H20 6) 0 benzene O n Scheme 2.1. Early examples of the synthesis of polymandelide. 79 I ‘ Z “0 H0 P E o - H NO” ”2 s o + \v ‘ 0 e O 0 H20 0 x y 0 ° 0 Ph ' 1 OJIY GAY Cat. _ 0 hr" PW" i° Ari 0 x y .. .l O O ‘ Ph Hex—Non o 0 Ph HRHLOH Sn(Oct)2 t H iOMOJW’iO‘H N 0 x}-z z = CH3, 0CH2Ph 0 Scheme 2.2. Examples of the synthesis of mandelide copolymers 80 Copolymers of mandelic acid with poly(butylene succinate) and poly(ethylene adipate) were prepared by Yoon13 using mandelic acid and the corresponding diacid and diol. Increasing the mandelic acid content decreased the crystallinity and melting temperature of the polyesters, but increased the Te. As the mandelic acid content increased, mechanical properties such as elongation and tear strength were enhanced in the copolymers. The biodegradation rate of the poly(butylene succinate) copolymers also increased due to the disruption of crystallinity caused by incorporation of the mandelic acid monomer. Blends of polylactide and polymandelide were prepared to study the miscibility and the effect of the low molecular weight aromatic polyester on drug release.5 With triamcinolone (a steroid) as the model drug, the induction time for drug release from a polylactide and low molecular weight polymandelide blend decreased to half of that for pure racemic polylactide. 81 2.2. Monomer 2.2.1. Synthesis of mandelide OH | OH PTSA o gloro O 0 “items A O OREK. lo SWIG 91% R,S : RR, 8.8 = 1 1,," . 1 0 R,S O 0 ° 0 Scheme 2.3. Synthesis of mandelide from mandelic acid Previous literature examples of mandelide syntheses were based on acid catalyzed self-esterification reactions with reported yields < 20%. A likely cause of the low yields is a high concentration of mandelic acid, which favors the formation of linear oligomers. Mandelic acid was cyclized in the presence of p- toluenesulfonic acid to form mandelide by a route based on literature examples and results obtained by Simmons (Scheme 2.3).14 To favor intramolecular cyclization, the reaction was run in a dilute solution (< 0.1 mol/L). The reaction by—product, H20, was removed azeotropically using a Dean-Stark trap, and the conversion of mandelic acid could be roughly monitored by the volume of H20 82 collected in the Dean-Stark trap. Xylenes, toluene and benzene were investigated as solvents for the condensation, and xylenes gave the best results in terms of rate of the product formation and yield. Due to their low boiling points, toluene and benzene gave low yields (< 10%) of cyclic dimers even after 2 weeks at reflux. Using xylenes as solvent, mandelic acid was consumed within 3 days and gave a mixture of R,S and R,R/S,S mandelide in about a 1:1 ratio. However, when the reaction was allowed to continue for longer times (1 week) the R,S mandelide isomer slowly disappeared and the content of R,R/S,S isomers increased, eventually becoming the only cyclic dimers in the reaction system (Scheme 2.4). .0.” o o lsomerize O o o .m,, o o ——> J: heat + J; r o o O o o O o o R S . RR 8.8 moderate solubility low solubility, decompose m.p. = 137 °C before melting Scheme 2.4. lsomerization of R,S mandelide to R,R/S,S mandelide 83 2.2.2. Purification of mandelide After cyclization, the reaction mixture contained cyclic dimers, oligomers, p-toluenesulfonic acid and sometimes unreacted mandelic acid. Due to their different solubilities in cold xylenes, the less soluble R,R and 3,8 isomers precipitated when hot xylene solutions were cooled to room temperature, while the more soluble R,S isomer and other components were soluble in the xylene filtrate. Mandelic acid and cyclic oligomers were removed by washing with sat. NaHC03 solution. The majority of the remaining off-white powder was the R,S isomer, which was then washed with hexanes and ether followed by recrystallization from ethyl acetate. Another way to remove oligomers and acids was to filter a cold xylene solution through a short pad of silica gel to remove all components but the R,S isomer. Although this separation method is faster, the yield is usually lower due to the adsorption of mandelide on silica gel. 2.2.3. Physical properties of mandelide isomers All of the mandelides were obtained as white crystals. The R,S diastereomer melts at 137-138 °C while the R,R/8,8 isomers decompose around 210 °C without melting. Addition of R,R/8,8 mandelide to the R,S diastereomer decreases the melting point, as expected. The R,S isomer has relative good solubility in typical organic solvents such as THF, CH2CI2, CHCla, ethyl acetate and DMSO. In contrast, the R,R/8,8 isomers are poorly soluble in the same solvents, but do dissolve well in DMSO. Thus, NMR measurements were run in deuterated DMSO since it readily dissolves all the isomers. 84 2.3. Polymerization of mandelide 2.3.1. Melt polymerization Mandelide, the phenyl derivative of glycolide, can be polymerized by a typical catalyst used for the ring opening polymerization of glycolide and lactide - Sn(Oct)2. Due to the small amount of catalyst and initiator needed for most polymerizations, dilute stock solutions of catalyst and initiator were prepared. The desired aliquots were injected into the reaction vessels, and after removing solvents, the monomer/catalyst/initiator mixture was sealed in a tube under vacuum for melt polymerization. t-Butylbenzyl alcohol (BBA) was chosen as the initiator because the t-butyl group provides a distinct peak on NMR spectra which can be used for calculating number average molecular weights. The sealed tubes were put into a thennostatted oil bath set at 160 °C, above the melting point of the R,S isomer. After desired intervals, tubes were removed and quenched in ice water. Despite starting with pure R,S mandelide, epimerization in situ generated the R,R/8,8 isomers and the resulting polymandelide was amorphous. NMR analysis of partially polymerized samples showed that in the presence of Sn(0ct)2, the pure R,S isomer rapidly isomerized to the R,R/8,8 diastereomers. Thus melt polymerizations are complicated by the epimerization of R,S mandelide to the more stable R,R/8,8 diastereomers. Racemization of polylactide is not uncommon, especially when metal catalysts were present, and the rate of racemization increases dramatically with temperature. Kricheldorf and Serra16 screened approximately 70 L-Lactide 85 polymerization systems and found that racemization was related to the basicity of the catalyst. The proposed racemization mechanism is based on an ester- hemiacetal tautomerization which is favored by the acidity of the proton at to the carbonyl (Scheme 2.5). Rehybridization of the asymmetric carbon atom followed by racemization has also been proposed to explain the existence of more than two lactide diastereomers from the degradation of poly(L-lactide).17 ll“ ("3 R10\ /OH “3 o Rf‘O-(E-C—Ofiz = C/ =C\ = R1_O-C"’C—OR2 CH3 H3 092 H T n R10\ /OH fi’h (u) R1-O-?-C—OR2 = /C= \ —-~———'- R1-O‘fi3—C-OR2 Scheme 2.5. Racemization in polylactide and polymandelide Sn(0ct)2 catalyzed polymerizations show some of the lowest rates of racemization. However, the benzene ring increases the lability of the methine C- H bond, and mandelide racemizes rapidly. Compared to lactide, epimerization should be more significant and occur at lower temperatures. In addition, even purified Sn(Oct)2 contains residual ethylhexanoic acid, water and other impurities which may catalyze the racemization process. A control experiment shows the ease of racemizing mandelide. When pure R,S mandelide was heated at 160 °C under vacuum, the solid melted completely, and then slowly resolidified within 80 minutes to give a white solid that was only slightly soluble in CDCI3. NMR analysis showed that soluble 86 portion contained a 30/70 mixture of R,S mandelide and the R,R/8,8 isomers, while an NMR spectrum of the insoluble white solid measured in DMSO showed that it was solely the R,R/8,8 isomers. The R,R/8,8 isomers are the more stable mandelides, but pure R,R/8,8 mandelide cannot be melt polymerized by Sn(Oct)2 because they decompose at high temperature. Thus, either pure R,S mandelide or a low melting mixture of the R,R/8,8 and R,S isomers were used for both melt and solution polymerization. The R,R/8,8 isomers did not seem to interfere with the polymerization because they are soluble in the molten R,S isomer, and are consumed during the polymerization. The typical purification method for lactide (e.g. multiple recystallizations) are effective in that polymerizations using purified monomer provide good control over the molecular weight by simply varying the monomer to initiator ratio. But as shown in Table 2.2, more vigorous drying is needed for mandelide. The first two entries were polymerizations using mandelide dried by the protocol typically used for lactide monomers. Even when no BBA was added (Table 2.2 entry 2), Sn(0ct)2 catalyzed ring opening polymerization. When the ratio of Sn(Oct)2 to BBA was 1 (Table 2.2 entry 1), the conversion was higher, but the molecular weight was only half of what was expected, an indication of excess initiator in the polymerization. When the same catalyst and initiator solutions were used for lactide polymerizations, the expected molecular weights were obtained. Thus, the low molecular weight in mandelide polymerizations can only be caused by monomer impurities such as H20. 87 The single crystal x-ray structure of 8,8 mandelide shows that the methine proton is in close proximity to the carbonyl oxygen and C-HmO hydrogen bonds have been suggested”19 It is possible that mandelides have a strong affinity for water due to formation of H-bonds with H20. To obtain good control over the polymer molecular weight, mandelide monomers need to be scrupulously dried. The monomers used in entries 3-5 (Table 2.2) were dried under high vacuum (10’5 torr) at 40 - 45 °C, and the molecular weights obtained were close to the values predicted by the monomer initiator ratio. 7 The lower than expected molecular weight for entry 5 was can be attributed to transesterification reactions becoming more prominent as the polymerization reached completion. The increased polydispersity index (1.44) is consistent with that view. 88 Table 2.2. Melt polymerization of mandelide at 160 °C entry [BBA)/[Sn(0ct)z] time conversion Mn Mn PDI (min.) (expected) (GPC) 1a 1 3 93.1% 12,500 6,850 1.17 28 0 3 85.6% 11,470 10,400 1.19 3” 1 4 90.6% 12,060 11,480 1.26 4b 1 4 96.6% 12,950 11,430 1.29 5h 1 20 98.4% 13,200 8,830 1.44 (a): mandelide purified by recrystallization and drying overnight; (b): mandelide purified as for entries 1 and 2, but further dried under vacuum (10‘5 torr)at 40-45 °C. 89 2.3.2. Solution polymerization Solution polymerizations of mandelide were run either in toluene or CH30N. Because of the poor solubility of the monomer, solution polymerizations have slower rates and are more likely to suffer from problems associated with equilibrium polymerization. In CH3CN, R,S mandelide has a solubility of around 0.58 moVL at room temperature and 1.5 moVL at 50 °C. Although R,S mandelide is less soluble in toluene, the higher boiling point of toluene leads to faster reaction rates. The solubility of the R,R/8,8 mandelides is significantly lower (0.01 moVL in CH3CN at room temperature). Most solution polymerizations of mandelide were run in anhydrous CHgCN as shown below. Since, t-butylbenzyl alcohol and Sn(Oct)2 do not dissolve in CH30N even at 65 °C, their toluene solutions were used in the polymerizations. Given that mandelides readily epimerize, a mixture of the R,R/8,8 mandelides with the R,S diastereomer should polymerize. This proved to be true, and thus for solution polymerizations, the R,S mandelide was isolated from condensation of mandelic acid, but with no special steps taken to remove the R,R/8,8 mandelides. For the kinetic run shown in Figures 2.1 and 2.2, an R,S mandelide sample containing 26% of the R,R/8,8 diastereomers was used, and initially the polymerization solution was colorless and clear. However, within half an hour, the reaction became heterogeneous as indicated by the formation of a precipitate. 1H NMR showed that by the time the polymerization reached 24% conversion, the soluble mandelides had epimerized to roughly a 50:50 mixture of R,S and R,R/8,8 mandelide, while the precipitate corresponded to the R,R/8,8 90 isomers. As the polymerization proceeded, the 1:1 ratio was maintained in solution. When the conversion reached around 70%, the solution became clear, homogeneous, and viscous. Thus the super-saturated mandelide solution provided a constant supply of the R,R/8,8 isomers, which either were directly incorporated into the polymer, or epimerized and polymerized. The linear relationship shown in Figure 2.1 is consistent with the polymerization being first order with respect to the monomer concentration. Compared to lactide under the same polymerization conditions, the rate is 4 times slower, as would be expected from the larger steric bulk of the phenyl group compared the methyl group of lactide. Figure 2.2 shows that the molecular weight of the polymer increased linearly with conversion, and the PDI decreased with conversion, which indicates the “living” character of the polymerization. However, when the reaction reached completion (97% conversion) and was allowed to run longer times (5 days), the molecular weight decreased, and the PDI increased to 1.5. This is consistent with intra and intermolecular transesterification becoming more prominent as the available monomer diminishes. Some discoloration of the polymer was observed for long polymerization times, but the degradation pathway was not identified. 91 1.4 |n([M]o/[M]1) 20 Time ( hr) Figure 2.1. Kinetics of solution polymerization of mandelide in CH30N at 70 °C under argon. [mandelide]:[Sn(Oct)2]:[BBA] = 100:1:1; [mandelide]: 0.93 moVL (75% R,S mandelide and 25% R,R/8,8 mandelide) 92 Mn (x 10'3) ES 0. 10 20 30 40 50 60 70 ° % Conversion Figure 2.2. Molecular weight versus conversion during solution polymerization of mandelide in CH30N at 70°C under argon. [mandelide]:[Sn(Oct)2]:[BBA] =100:1:1; [mandelide]: 0.93 mol/L(75% R,S mandelide and 25% R,R/8,8 mandelide) 93 2.3.3. Purification of polymandelide The most widely used polymer purification methods are dissolution- precipitation schemes, where a solution of the polymer in a good solvent is slowly dripped into a non-solvent. Ideally the polymer precipitates into thread-like pieces of polymer that are easily collected by filtration, while the impurities remain in solution. The solubility of polymandelide is similar to that of polystyrene, and polymandelide dissolves in THF, toluene, Cchlg, chloroform, ethyl acetate and DMSO. Non-solvents for polymandelide include hexanes, ether and methanol. The initial dissolution-precipitation scheme was based on a CHZClzlmethanol solvent pair. The precipitation experiments resulted in milky solutions, regardless of molecular weight of the polymers, and the polymer was collected by centrifugation in low yield. If the milky solution was allowed to stand for two days, thin white films of polymandelide formed as the solvent slowly evaporated. As the amount of CHzclz or toluene needed to dissolve the crude polymandelide was much larger than for crude polylactide samples of similar weight, the concentration of polymandelide in CHzclz or toluene may be too low to form good precipitates. Further investigation showed that residual R,R/8,8 mandelides complicate the precipitation scheme. A crude polymandelide sample was washed with a small amount of CHQClz. NMR analysis showed that the CHzclz solution contained monomer and polymer, while the CH2012 insoluble portion consisted only of the RR/SS mandelide. Thus, the excess CH20I2 needed to dissolve the crude polymer was due solely to the presence of unreacted R,R/8,8 mandelide. It is important to remove residual R,R/8,8 94 mandelide from incomplete polymerizations in order to recover purified polymandelide in reasonable yields. By using less CHQClz, pro-filtering to remove the R,R/8,8 mandelide followed by normal precipitation, the yields of recovered polymandelide were more than 80% and the polymer was recovered in a form that could be easily collected by filtration. The precipitated polymer was then heated to 60-90 °C under vacuum until a constant weight was obtained. Methanol and residual water were removed under vacuum, and residual monomer can be further removed by sublimation under high vacuum. Using the above protocol, NMR analyses showed that we isolated polymandelide free of monomer. However, based on the characterization of polylactide purified by precipitation alone, other impurities such as residual metal catalysts may still be present in the polymer samples. As described in the Introduction, residual metals can catalyze transesterification reactions leading to the formation of cyclic dimers or oligomers. For practical purposes, it is also essential to remove metal residues from products intended for medical applications. A common way to remove metal residues from polylactide samples is to dissolve the polymer in an organic solvent and wash the organic solution with dilute HCI. The same method was applied to polymandelide. A solution of polymandelide in methylene chloride or toluene was washed several times with dilute HCI, followed by washes with distilled water until the water layer was neutral. The resulting solution was then treated as usual to afford white polymer samples. GPC analysis of the polymer before and after the extractions showed no change in molecular weight, confirming that no significant chain scission 95 occurred during the process. Therefore, extraction using dilute acid is a safe method for removing residual metals from polymandelide. 2.3.4. Characterization of polymandelide Polymandelide can be cast from toluene or THF to give clear, colodess films. Melt pressed films prepared at ~140 °C are clear but have a light yellow color. The density of the polymer, obtained by flotation measurements of films in aqueous salt solutions was ~1.25 g/cma. Polymandelide was characterized by FT-IR, NMR, DSC, X-ray powder diffraction and TGA. The IR spectrum (Figure 2.3) obtained on a polymandelide sample spin cast on' a gold-coated silicon substrate shows two bands that are diagnostic for esters, a strong band at 1766 010'1 (0:0 stretching) and a broad band around 1200 cm". Weak absorptions around 1456, 1498, 1605 and 3055 cm'1 are characteristics of aromatic compounds. Given the mixture of diastereomers present in a polymerization of mandelide, it is not surprising that 13C NMR spectrum of polymandelide is complicated (Figure 2.4). Broad peaks resulting from complex polymer tacticities were observed for each carbon resonance. Multiple peaks were also observed in 1H NMR for methine and aromatic protons. Since no authentic samples of stereoregular polymandelide are known, no attempt was made to assign the stereochemical sequences. DSC analysis of polymandelide samples shows only a single glass transition, and no crystalline transitions. Powder X-ray diffraction experiments 96 Absorbance ii I). L 4000 3500 3000 2500 2000 1500 1000 500 Wavenumber (cm 1) Figure 2.3. FTIR of a polymandelide film spin-coated on a gold-coated silicon wafer. 97 TTYTITTITTIIfTfiI—Y—TT—rTT—rTrTTTTIITTj—Trj—TII]11VYIYTIYTVI 75.2 75.0 74.8 74.6 74.4 74.2 74.0 73.8 73.6 73.4 73.2 11IIITITrrIT—TIIIIIIIIIIllTfiIrijr‘rllllllllITTTTTIIIIIITFTIITTTTIITIIIITI1 133.0 132.5 132.0 131.5 131.0 130.5 130.0 129.5 129.0 128.5 128.0 127.5 127.0 126.5 Figure 2.4. 13C NMR of polymandelide in d6-DMSO 98 ITITTTTfiTTIIIIIITTTfiTTTTITroVYITTTTIT 167.20 167.00 166.” 166.60 166.40 166.20 166.” 1654” Figure 2.4. 13C NMR of polymandelide in d6-DMSO (cont’d) concur, and show only amorphous scattering and no evidence of crystallinity. Smith and Tighe also reported that their low molecular weight polymandelide samples produced from either racemic or optically active precursors were also amorphous.3 Like other polymers, the T9 of polymandelide depends on the molecular weight of the polymer, and eventually becomes independent of chain length at high molecular weights. The shift in T9 is related to the concentration of chain ends in the polymer. Chain ends have larger degrees of freedom compared to other chain segments, and because the chain end concentration decreases with increased molecular weight, T9 increases and then plateaus at high molecular weights. According to the literature] a polymandelide sample with Mn ~ 1,100 g/mol afforded a T9 of ~ 75 °C. In a higher molecular weight polymandelide sample (Mn: 16,000 g/mol) the T9 shifted to ~ 95 °C, and it 99 eventually reached 100 °C when Mn: 60,000 g/mol (Figure 2.5). This value is similar to that of polystyrene (109 °C) making the analogy between polymandelide and polystyrene even stronger. The thermal decomposition of polymandelide was characterized by TGA under nitrogen. As mentioned in the Introduction, the decomposition of polylactide in the presence of residue metal catalysts is considered to be a series transesterification reactions that generate volatile cyclic dimers or oligomers. After removing catalyst residues from polylactide samples, the onset for decomposition shifts to higher temperatures. Presumably, polymandelide should undergo the same degradation processes, and show a similar dependence of the thermal stability on purity. Polymandelide and polylactide samples were purified by simple precipitation into a non-solvent. For these samples, the onset for thermal degradation of polymandelide occurred at higher temperatures than that of polylactide (Figure 2.6). However, after both samples were purified by washing with dilute HCI, the order was reversed (Figure 2.7). In terms of the onset of decomposition, no significant change was found for polymandelide samples before and after acid treatment (Figure 2.8), but the stability of the polylactide sample improved significantly. If the two polymers degrade by the same depolymerization mechanism, one would expect that polymandelide would have an onset for degradation at higher temperatures due to the lower volatility of mandelide compared to lactide. 100 Heat flow endo —> S l l l l l 1 60 70 80 90 100 110 120 130 T(°C) Figure 2.5. DSC of polymandelide samples, showing the molecular weight dependence of T9. A: Mn = 60,000 g/mol; B: 16,000 g/mol. Samples were heated at 10 °/min under helium. 101 It is possible that the methine protons in polymandelide are more labile and radical pathways are more favorable in the mandelide system. A less plausible explanation would be that catalyst is removed far more efficiently from polymandelide than polylactide. To test these possibilities, a polymandelide sample free of monomer was sealed in a glass ampoule and heated in a 200 °C oil bath for 24 hours. NMR confirmed that the polymer had not degraded significantly, and there was no evidence for the formation of mandelide. However, GPC traces confirmed a large decrease in molecular weight and a singlet in the 1H NMR at 10.0 ppm suggested the formation of benzaldehyde, which can be formed from the radical cleavage of the polymer backbone. Smith and Tighea reported that the onset for polymandelide decomposition occurred at about 205 °C and carbon monoxide and benzaldehyde were the principal decomposition products seen in their therrnogravimetric experiment. 102 100 80 r 35 E 60 - U) ’5 A 3 Q - B Q- .— E 40 (U (D 20 " 0 1 1 1LLr-‘L‘ffi-‘l 0 100 200 300 400 500 600 700 T (°C) Figure 2.6. Thermal Gravimetric Analysis of polylactide (A) and polymandelide (B) before washing with dilute HCI. Samples were heated at 40°C/min under N2 103 100 80- 60- 20L 0 1 l 0100200300400500600700 Figure 2.7. Thermal Gravimetric Analysis of polymandelide (A) and polylactide (B) after washing with dilute HCI. Heating rate: 40°C/min. under N2 104 100 80 - :\°‘ 5 5° ’ g / after 33 before 3' 4o - / <0 (I) 20 - L“ O 1 I L l L r: 0 100 200 300 400 500 600 700 T (°C) Figure 2.8. Thermal Gravimetric Analysis of polymandelide before and after washing with dilute HCI. Samples were heated at 40°C/min under N2 105 2.4. Copolymerization of mandelide with lactide To expand the range of end use temperatures available from biodegradable polymers, a series of random copolymers were prepared by copolymerizing mandelide with rec-lactide and L-lactide. Polymerizing rec-lactide with mandelide should provide a series of glassy materials, while incorporation of mandelide in poly(L-lactide) should affect the crystallization of poly(L-lactide). Depending on the degree of crystallinity in the copolymer, these materials may mimic various toughened therrnoplastics and thermoplastic elastomers. Poly(rac-lactide-co-mandelide) copolymers were prepared by bulk copolymerization at 130 °C using Sn(0ct)2 as the catalyst and BBA as the initiator. The polymers were purified as described earlier. The molar composition of the copolymers as determined by 1H NMR had mandelide to lactide mole ratios of 11:89, 25:75, 45:55, 75:25 and 89:11 (Table 2.3), which were close to the feed ratios. DSC measurements of the copolymers showed a single glass transition temperature for each copolymer. As shown in Figure 2.9, the T95 range from 48 °C to 100 °C, with the lower and upper limits corresponding to the T9 of the L- lactide and mandelide homopolymers respectively. The dramatic increase in the glass transition temperature with mandelide content is caused by the introduction of bulky phenyl substituents on the polymer backbone that reduce chain mobility. Since the polymers are homogeneous (single T9), the glass transition temperatures should follow the Fox equation, (Eq. 2.1). 106 1fT=W1IT1 + wlez Eq 2.1 where T, T1 and T2 are the glass transition temperature of the copolymer, polymandelide and polylactide homopolymers respectively, and W1 and W2 are the weight fractions of two components in the copolymer. A good fit to the Fox equation was observed, with the primary deviation apparently coming from a “too low" value for the T9 of polylactide (Figure 2.10). This behavior has been observed previously, for other lactide copolymers, but its origin is unclear. L-lactide and mandelide were copolymerized to study the effect of mandelide content on the crystallization of polylactide. Copolymers were synthesized with mandelide to lactide mole ratios of 2:98, 5:95, 12:88, 20:80 and 45:55 (Table 2.4.). As the polymerizations were allowed to run to high conversions at 160 °C, the ratio of mandelide to lactide in each copolymer was close to the feed ratio. DSC experiments were run on purified copolymers. Only one glass transition temperature was observed for each copolymer, and the T9 increased as the mandelide content increased. The first scan (10 °C/min) used polymer directly after precipitation and drying, and only two copolymers (mandelidezlactide = 2:98 and 5:95) afforded a melting peak. As the mandelide content in the copolymers increased, the melting temperature decreased from 172, to 160 and to 151 °C (Figure 2.11). Despite annealing poly(L—lactide-co-mandelide) (mandelidezlactide 12:88) in the DSC for 18 hours at 130 °C, no melting peak was detected when the sample was heated to 185 °C. The result is reasonable when compared to 107 data for poly(L-lactide). Polylactide derived from > 93% L-lactic acid usually crystallizes, while polylactide prepared from 50-93% L-lactic acid is generally amorphous. In the lactide case, R-lactic acid residues in the polymer chain act as defects that interfere with crystallization. Mandelide serves the same function in poly(L-lactide-co-mandelide). A recent report also described similar data; incorportation of > 10 mol% mandelic acid in L-lactide polymerizations gave amorphous materials.10 Based on the thermal degradation results described earlier for polymandelide and polylactide, the degradation of poly(lactide-co-mandelide) as measured by TGA should be sensitive to impurities in the polymer. For samples contaminated with catalyst residues, the onset temperature for weight loss did not correlate with polymer composition. However, when the samples were treated with dilute HCI, the onset temperature increased as expected as the fraction of lactide in the copolymer increased (Figure 2.12.). 108 Table 2.3. Poly(mandelide-co-rac-lactide) copolymers prepared by bulk polymerization catalyzed by Sn(0ct)2 at 130 °C Mandelide/lactide Entry (molzmol) Mn‘b’ PDl‘b) 'r9 (°C)‘°’ feed copolymer ratio ratio“) 1 100:0 100:0 68,000 1.63 100.3 2 90:10 89:11 37,000 1.45 94.5 3 75:25 75:25 42,000 1 .60 90.0 4 50:50 45:55 58,000 1 .47 82.1 5 25:75 24:76 80,000 1.65 66.9 6 10:90 11:89 98,000 1.65 61.3 7 0:100 0:100 20,000 1.47 46.5 (a): determined by ‘H NMR; (b): determined by GPC in THF and reported relative to polystyrene standards; (c): measured at 10 °C/min under helium. 109 P09...»— PmdlzPLA = 8 : 1 PmdlzPLA = 3: 1 r PmdlzPLA = 45 : 55 PmdlzPLA = 1 : 3 PmdlzPLA =1 :8 *— _...-——~/ Heat Flow ——> 0 25 50 75 100 125 T (°C) Figure 2.9. Glass transition temperatures of poly(mandelide-co-rec-lactide) copolymers. Samples were heated at 10 °C/min under helium 110 100 80 6‘ l-—°° 60 —— Tgcal. from Fox Eq. 40 _ ' T9 measured by DSC 20 0 0.2 0.4 0.6 0.8 1 Weight Fraction of Lactide Figure 2.10. Glass transition temperatures of poly(mandelide-co-rac-lactide) copolymers fitted to the Fox Equation 111 Table 2.4. Poly(mandelide-co-L-lactide) copolymers prepared by bulk polymerization at 160 °C Mandelide:L-lactide E... ....l'"°'";":;i..ym.. W“: 1113221) ratio ratio“) 1 02100 02100 21,800 1.47 172 50.7 2 2:98 2298 66,000 1.20 160 31.2 3 5295 5295 48,100 1.25 153 23.8 4 10290 12:88 59,900 1.47 — -— 5 20280 20280 48,150 1 .23 — -— 6 45255 49151 42,100 1.23 — — (a): determined by 1H NMR; (b): determined by GPC in THF and reported relative to polystyrene standards; (0): measured at 10 °C/min under helium. 112 PLLA L PLLA:pde= 98:2 PLLAzpmdl= 95:5 PLLAzpmdl= 88:] 2 Heat flow endo -——> PLLAzpmdl= 80:20 l l l l l 0 30 60 90 120 150 180 T(°C) Figure 2.11. Thermal properties of poly(mandelide—co-L-lactide) copolymers. Samples were heated at 10 °C/min under helium 113 100 80- Sample weight (%) 0 l 100 200 300 400 500 600 T (°C) Figure 2.12. Thermal Gravimetric Analysis of poly(mandelide-co-rao-lactide) c0polymers. A: polymandelide; B: mandelidezlactide = 3:1; C: mandelidezlactide = 1:3; D: polylactide. Samples were washed with dilute HCI after precipitation and heated at 40 °C/min under N2 114 2.5. Hydrolytic degradation of polymandelide One of the most interesting and important features of biodegradable polymers is their degradability. Like polylactide, polymandelide contains hydrolyzable ester linkages in the polymer backbone and differs only in that the pendant methyl groups of polylactide are replaced by phenyl groups. The aromatic rings make the polymer more hydrophobic than polylactide and should lead to a slower degradation rate. Copolymers of L-lactic acid and mandelic acid obtained via polycondensation schemes have been subjected to in vitro7 and in viva“9 degradation studies. In vitro studies show that the mandelic acid content has a large affect on the degradation profile. As the mandelic acid content increased, the profile shifted from being parabola-like, characterized by an initial rapid degradation followed by gradual erosion of the polymer, to an “S”-type degradation profile, which is characterized by initial swelling followed by degradation of the ester linkages in the swollen state. Similar changes were observed in vivo. To date, no data had been obtained on the hydrolytic degradation of high molecular weight polymandelide. The rate is expected to be slow, since a low molecular weight polymandelide sample (1,300 g/mol) showed no weight loss during 15 weeks of hydrolytic degradation. The conditions used to study the degradation of polymandelide (phosphate buffered solution at pH 7.4 and 55 °C) were identical to those used to characterize the degradation rates of other substituted polylactides. Carrying out the degradation at 55 °C allows for completion of the degradation in several 115 months. In addition, the degradation rates of poly(L-lactide) measured in phosphate buffered solutions have been shown to mirror those measured in ViVO.20'21 Powdered samples (~ 1 mm in size) were allowed to age in the phosphate buffer without stirring to simulate the low flow rates of body fluids in smooth and hard tissues.22 The initial Mn of the polymandelide sample was 78,200 g/mol. The molecular weight decrease fits the random chain model with a rate 1/120th that of lactide degraded under the same conditions. Weight loss began after 80 days, and thus the random chain scission is not relevant after 80 days (Table 2.5). The delay in the onset for weight loss relative to the loss in M. is consistent with a bulk erosion mechanism, where carboxylic acid groups generated by ester hydrolysis catalyze further degradation of the polymer. Carboxylic acids near the surface of the sample can be neutralized by the phosphate buffer, but acid end groups inside the sample cannot escape or be neutralized, leading to faster degradation in the core of the material. An alternative mechanism, surface erosion, would require that sample weight loss precede substantial loss in molecular weight. A slight shoulder appeared in the GPC trace of the polymer sample after 97 days of degradation as shown in Figure 2.13. The shoulder grew more prominent with time, until a bimodal molecular weight distribution became obvious. Such a distribution is characteristic of heterogeneous degradation, in which the surface and core of the sample degrade at different rates, resulting in two distinct molecular weight distributions. This surface-core differentiation with 116 its characteristic biomodal molecular weight distribution is a common feature in the degradation of polylactide samples > 50 um in size.23 However, a bimodal distribution was not observed for polylactide and polyphenyllactide degraded at 55 °C, presumably because the degradation temperature was higher than the T9 of the polylactide and nearly identical to the T9 of polyphenyllactide. For both cases, the polymer chains should have enough mobility to allow low molecular weight oligomers bearing acid end groups to diffuse out of the sample, especially as the polymers partially hydrolyze and become more hydrophilic. As shown in Figure 2.14 the molecular weight of polymandelide was plotted against degradation time according to the random chain scission model. A linear trend was observed before any significant weight loss (up to 97 days). After 97 days, the data dramatically deviated from the random chain scission model due to the heterogeneous nature of polymandelide’s degradation. The rate of polymandelide hydrolytic degradation before 97 days was calculated to be ~ 120 times slower than amorphous polylactide under identical degradation conditions. The result can be explained by the large difference in T9. A large drop in molecular weight in parallel with a constant sample weight has been observed for polylactide and other substituted polyglycolides. For autocatalyzed degradation, significant weight loss requires the formation of water soluble oligomers, which only occurs after extensive hydrolysis of the polymer chains. This behavior can be fit by the Prout Tompkins model described in the introduction (Figure 2.15) which was based on autocatalytic thermal 117 decomposition of potassium permanganate and has been applied to evaluate mass loss from po|y(lactide-co-glycolide) particles.24 118 Table 2.5. Weight and molecular weight change during hydrolytic degradation of polymandelide in phosphate buffer (pH: 7.4) at 55 °C Degradation Weight percent time (days) of remaining Mn(GPC) PDI polymer 0 100 78,200 1 .61 12 96.0 73,800 1.62 20 0.964 66,860 1 .64 40 98.6 58,900 1 .67 64 96.1 49,100 1.67 97 92.4 33,800 1 .80 161 78.3 13,300 2.48 188 62.8 6,980 4.19 225 51 .2 4,590 4.84 304 26.2 —a —a a: the molecular weight was too broad to be determined 119 304 days /V\ 22.55188 /\\ AJ\ 188 days 161 days _/\_ 91 days _ A .264 days- -_ k 40 days ‘ k 20 days k 12 days 8 910111213141516171819 Elution Volume (mL) Figure 2.13. GPC traces of polymandelide samples during hydrolytic degradation at pH=7.4 and 55 °C 120 0.12 0.003 0.10 - E 0-002- .2? 5% 0.001 » 0.08 - 2 § Q 0.000 . ‘7 0 50 100 5:2 0_06 Degradation Time (days) 2 A Q 2: 0.04 A 0.02 A A 0.00 ' ‘ ' 0 50 100 150 200 250 Degradation Time (days) Figure 2.14. Molecular weight change of polymandelide (A) and polylactide (A) during hydrolytic degradation in phosphate buffer at 55 °C. The lines are fit to a random chain scission model. The inset shows the molecular weight data before 97 days. 121 Remaining Weight Fraction 0.2 - Weight loss fit to Prout-Tompkins model O 4 1 1 J 1 L 0 50 100 150 200 250 300 350 Degradation Time (days) Figure 2.15. Weight loss during hydrolytic degradation of polymandelide in phosphate buffer (pH=7.4) at 55 °C 122 2.6. Experimental section General. Unless otherwise specified, ACS reagent grade starting materials were used as received from commercial suppliers. THF was distilled over CaHg, and then was distilled from sodium benzophenone ketyl under nitrogen. Toluene was freshly distilled from sodium benzophenone ketyl under nitrogen. Anhydrous acetonitrile was obtained from Aldrich and used as received. Characterization 1H and ”C NMR analyses were performed at room temperature in CDCI3 on a Varian Gemini-300 spectrometer using TMS as the chemical shift standard unless otherwise specified. Reflectance FTIR spectra were obtained under nitrogen using a Nicolet Magna—560 FT IR spectrometer containing a PIKE grazing angle (80°) attachment. Typically, 256 scans were collected using a MCT detector. Polymer molecular weights were measured by gel permeation chromatography at 35 °C in THF using a Plgel 20p. Mixed column at a flow rate of 1 mL/min. Two detectors were used, a Waters R410 Differential Refractometer and a Waters 996 Photodiode Array. The concentration of the polymer samples was 1 mg/mL, and each solution was filtered through a Whatman 0.2 pm PTFE filter before injection. The molecular weights are reported relative to monodisperse polystyrene standards. Differential scanning calorimetry (DSC) data were obtained with a Perkin Elmer DSC 7 instrument calibrated with indium and hexyl bromide standards. The samples were placed in aluminum pans, and were heated at 10 °C/min under a helium atmosphere. Liquid nitrogen was used as the coolant. Thermogravimetric analysis (TGA) data 123 mn- were obtained from a Perkin Elmer TGA 7 instrument at a heating rate of 40 °C/min under nitrogen. Reported melting points were measured with an Electrotherrnal Melting Point Apparatus and are uncorrected. The densities of solutions were measured using a series of hydrometers (Curtin Matheson Scientific. Inc). Synthesis of mandelide Racemic mandelic acid (6.03 g, 39.7 mmol) and a catalytic amount of p—toluenesulfonic acid (0.20 g, 1.2 mmol) were dissolved in xylenes (600 mL). The solution was refluxed for 3 days and water was removed via a Dean Stark trap. The conversion of the reaction was monitored by NMR and by the amount of water that collected in the trap. The solution was allowed to cool to room temperature and most of the R,R/8,8 mandelide precipitated from solution and was collected by filtration to give 1.3 g (47%) of a 1:1 mixture of the R,R and 8,8 isomers (mp 193 °C (decomposes)). The filtrate was washed three times with saturated aqueous NaHCOa and the solvent was dried and removed by rotary evaporation. The crude mixture of R,S, R,R and 8,8 mandelide was recrystallized three times from ethyl acetate to give 1.5 g (53%) of R,S- mandelide, mp 137 °C. The R,S-mandelide could also be purified by passing the crude filtrate through a layer of silica gel, followed by solvent removal and recrystallization. This method gave a lower yield of product. R,S mandelide: 1H NMR (300 MHz, d6-DMSO) 5 6.44 (s, 1H), 7.35-7.62 (m, 5H); “‘0 NMR (75 MHz, d6-DMSO) a 164.71, 133.23, 129.56, 128.94, 127.56, 77.56. 124 R,S/S,S mandelide: ‘H NMR (300 MHz, d6-DMSO) 5 6.61 (s, 1H), 7.35- 7.58 (m, 5H); 13C NMR (125 MHz, dG-DMSO) 8 166.47, 132.32, 129.28, 128.35, 128.29, 77.47 Melt polymerization of mandelide. Stock solutions of Sn(Oct)2 and BBA in anhydrous toluene were prepared in a dry box, fitted with vacuum adapters and removed from the box. Mandelide was loaded into small glass ampoules (~ 3 mL) with stir bars and connected to a vacuum line through a vacuum adapter. After evacuating the ampoule for 2 hours, the ampoule was backfilled with argon, and a syringe was used to add a predetermined amount of the Sn(Oct)2 and BBA solutions to ampoules though the adaptor. After solvent removal, the glass ampoules were sealed under vacuum. The ampoules were added to a therrnostatted silicon oil bath, and after desired time intervals, ampoules were removed from the bath and were quenched with ice water. The ampoules were then broken and the residue was extracted with methylene chloride or THF. Filtration and removal of the solvent in vacuo gave cmde polymandelide as an off-white or light brown colored sold. The conversion of the polymerization was measured by 1H NMR. Purification of polymandelide Crude polymandelide was dissolved in methylene chloride and the insoluble portion (R,R/S,S mandelide) was removed by filtration. The polymer solution was concentrated to ~ 10 wt%. and was slowly dripped into a well-stirred cold methanol solution. The polymer precipitate was collected on a fritted glass funnel and was dried under vacuum at 60-70 °C. If necessary, the precipitation procedure was repeated. 125 Polymandelide: 1H NMR (300 MHz d6-DMSO): 6 6.0626 (m, 1H), 7.42- 7.9 (m, 5H); “’0 NMR (125 MHz, dS-DMSO): 6166.5, 132.5, 129.3, 128.4, 127.5, 74.3. Solution polymerization of mandelide Mandelide (2.50 g, 9.32 mmol) was placed in a Schlenk flask and dried under vacuum. After transferring the flask into a drybox, anhydrous acetonitrile (10 mL) was added to the flask. The flask was then connected to a vacuum line and heated to 70 °C under argon. Sn(Oct)2 (93.2 mol) and BBA (93.2 umol) solutions were added into the flask by syringe under argon to initiate polymerization. At predetermined intervals, a syringe was used to remove aliquots of the reaction solution, which were analyzed by NMR and GPC to determine the conversion and molecular weight of the polymer. Density measurement. Polymandelide powder obtained from precipitation of the polymer was melt pressed at 140 °C. Chunks of polymer that were free of air bubbles were selected for density measurements. The samples were added to a graduated cylinder filled with distilled water. The polymer sample sank to the bottom, and NaCl was added to increase the density of the solution to the point that the polymandelide remained suspended in the solution. The density of the solution (equivalent to the density of the polymer) was measured by a hydrometer. Hydrolytic degradation of polymandelide A pH 7.4 phosphate buffer solution was prepared by adding dilute NaOH solution into commercially available phosphate buffer (pH=7.0) at 55 °C. Around 50 mg of the polymer 126 power (~ 1 mm in size) was placed inside a test tube with a screw cap and 15 mL of pH of 7.4 phosphate buffer solution was added. Multiple samples were prepared and placed into a water/ethylene glycol bath therrnostatted at 55 :l: 0.2 °C. At desired times, the test tubes were removed from the bath. The solutions were then filtered through a pre-weighed fritted glass funnel, and the collected polymer powder was rinsed repeatedly with a large amount of distilled water. The polymer and the funnel were dried under vacuum at 70 °C until constant weight was obtained. 127 References 10. 11. 12. 13. 14. 15. 16. 17. 0kada, T.; Okawara, R. J. Organometal. Chem. 1973, 54, 149-152. Kobayashi, S.; Yokoyama, T.; Kawabe, K.; Saegusa, T. Polym. Bull. 1980, 3. 585-591 . Smith, I. J.; Tighe, B. J. Macromol. Chem. Phys.1981, 182, 313-324. Pinkus, A. G.; Subramanyam, R.; Clough, S. L.; Lairrnore, T. C. J. Polym. Sci. Pol. Chem. 1989, 27, 4291 -4296. Domb, A. J. J. Polym. Sci. Pol. Chem. 1993, 31, 1973-1981. Whitesell, J. K.; Pojman, J. A. Chem. Mat. 1990, 2, 248-254. Fukuzaki, H.; Aiba, Y.; Yoshida, M.; Asano, M.; Kumakura, M. Makromol. Chem. 1989, 190, 2407-2415. Fukuzaki, H.; Yoshida, M.; Asano, M.; M., K.; lmasaka, K.; Nagai, T.; Mashimo, T.; Yuasa, H.; lmai, K.; Yamanaka, H. Eur. Polym. J. 1990, 26, 1273-1277. lmasaka, K.; Yoshida, M.; Fukuzaki, H.; Asano, M.; Kumakura, M.; Mashimo, T.; Yamanaka, H.; Nagai, T. Int. J. PhaIm. 1992, 81, 31-38. Kim, W. J.; Kim, J. H.; Kim, S. H.; Kim, Y. H. Polym.-Korea 2000, 24, 431- 436. Kimura, Y. ln Chem. Abst. 1993, 118, 2555771: Japan, 1993 Lee, R. S.; Yang, J. M. J. Polym. Sci. Pol. Chem. 2001, 39, 724-731. Jin, H. J.; Lee, B. Y.; Kim, M. N.; Yoon, J. S. J. Polym. Sci. Pt. B-Polym. Phys. 2000, 38, 1504-151 1. Simmons, T. Dissertation, Michigan State University, 2000, 66-78. Cerrai, P.; Tricoli, M. Makromol. Chem, Rapid Commun. 1993, 14, 529- 538. Witzke, D. R.; Narayan, R. Abstr. Pap. Am. Chem. Soc. 1998, 216, U11- U12. Kopinke, F. D.; Remmler, M.; Mackenzie, K.; Moder, M.; Wachsen, O. Polym. Degrad. Stabil. 1996, 53, 329-342. 128 18. 19. 20. 21. 22. 23. 24. Lynch, V. M.; Pojman, J.; Whitesell, J. K.; Davis, B. E. Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1990, C46, 1125-1127. Taylor, R.; Kennard, O. J. Am. Chem. Soc. 1982, 104, 5063-5070. Matsusue, Y.; Yamamuro, T.; Oka, M.; Shikinami, Y.; Hyon, S. H.; lkada, Y. J. Biomed. Mater. Res. 1992, 26, 1553-1567. Therin, M.; Christel, P.; Li, S. M.; Garreau, H.; Vert, M. Biomaterials1992, 13, 594-600. Li, S. J. Biomed. Mater. Res. 1999, 48, 342-353. Grizzi, l.; Garreau, H.; Li, S.; Vert, M. Biomaterials 1995, 16, 305-311. Dunne, M.; Corrigan, O. l.; Ramtoola, Z. Biomaten'als 2000, 21, 1659- 1668. 129 Chapter 3 Block copolymers of lactide and methyl methacrylate (MMA) 3.1. General Due to its biocompatibility, biodegradability and non-toxicity, polylactide is an important material for medical applications such as surgical sutures,1 implants,2 tissue scaffolds3 and drug delivery matrices.4 A more recent research emphasis has been the development of polylactide as a commodity polymer for packaging materials, coatings and fibers.5 For many applications, the properties of polylactide need to be fine-tuned. For example, better control over its degradation rate is needed for medical applications,6 and improved impact strength is needed to overcome polylactide’s brittleness. Blendsf”10 ""3 and composites,14 were synthesized to extend the properties of copolymers, polylactide and the range of applications for which polylactide is suitable. For example, polylactide has been blended or copolymerized with 8-caprolactone,15 16 glycolide, and ethylene glycol”18 to achieve a wider range of degradation rates, while block copolymers that combine polylactide with a rubbery block such 193° or polybutadiene21 lead to toughened materials. Interesting as polyisoprene polymer architectures were prepared by using a hydrophilic polymer, poly(2- hydroxyethyl methacrylate) (polyHEMA), as a polymer initiator from which lactide was polymerized to give a comb polymer with polylactide teeth.22 Matyjaszewski also reported the polymerization of methacrylate-tenninated polylactide to give poly(methyl methacrylate)-g-poly(lactic acid).23 130 Poly(methyl methacrylate) (PMMA) is a commodity polymer with good optical and mechanical properties. It also has been used in medical implants,“ 26 drug delivery systems27 and hard contact lenses due to its biocompatibility. Polylactide and PMMA have been combined in various ways to yield new materials with unique properties. Composite biomaterials28 that combine good mechanical properties with partial biodegradability were prepared by the free radical polymerization of MMA in the presence of a—Al203 and low molecular weight crystalline polylactide. These composite materials were considered for structural applications in orthopedic surgery. Poly(fl-hydroxybutyric acid) is significantly toughened when it is “reactive blended” with PMMA,29 but the same blending strategy applied to crystalline polylactide gave a highly interconnected network structure.31 A useful biocompatible and partially biodegradable system would be a toughened polylactide prepared by the combination of polylactide and PMMA in a block copolymer architecture. Because the synthesis of the two blocks are “mechanistically incompatible”, the polymerization mechanism must be transformed to chemically connect the two blocks. In order to obtain well-defined block copolymers with low polydispersities, “living” or “controlled” polymerization methods are preferred. Polylactide can be prepared by the direct condensation of lactic acid or by ring opening polymerization (ROP) of lactide, the cyclic dimer of lactic acid. Most research has focused on ROP because it offers a high polymerization rate and easy control over molecular weights ranging from several thousands to several 131 millions.32 Atom Transfer Radical Polymerization (ATRP) of MMA has been Intensively investigated since it leads to well-defined polymers with low polydispersitiesf"3 The characteristic feature of ATRP is a fast dynamic equilibrium between the active and dormant species34 which ensures a low radical concentration, and minimizes bimolecular termination reactions. There are examples of complex polymer architectures that have been synthesized via a combination of ROP and ATRP.35 Interesting star and graft copolymers of poly(caprolactone) and PMMA were prepared by end-group transformation and by the use of monomers that contain an ATRP initiator.36 Shell cross-linked nanoparticles with poly(caprolactone) as the core and poly(acrylic acid) as the shell were synthesized using a difunctional aluminum catalyst.” A difunctional initiator, 2,2,2—tribromoethanol,38 has been used in a one-step (simultaneous) block copolymerization of caprolactone and MMA. In this study, we employed both end-group transformation and the use of a difunctional initiator to synthesize block copolymers of lactide and MMA. Different polymer properties were observed for copolymers synthesized by the two approaches. The miscibility of the two blocks and the effect of the amorphous PMMA block on the thermal properties and crystallization of polylactide were investigated. 132 3.2. Results and discussion 3.2.1 Synthesis of polylactide macroinitaitors Lactide and MMA are polymerized by different mechanisms, the former by ring opening polymerization and the later by radical or anionic polymerization. To synthesize well defined PMMA/polylactide block copolymers, one must employ two different “living” or “controlled” polymerization methods. In the work we describe here, we used ring opening polymerization to prepare polylactide blocks and ATRP for the PMMA blocks. The polymerization mechanisms were combined in two different ways, which differ in the way the polylactide block is synthesized. In the first approach, polylactide was synthesized using a Sn(0ct)2 catalyst with t-butyl benzyl alcohol as the initiator, and the resulting polymer was capped with an a-bromoacyl bromide that can be used to initiate ATRP. In the second approach, the polylactide block was synthesized with Sn(Oct)2 and a difunctional initiator that can initiate both ROP and ATRP. Using the first strategy, lactide was polymerized in toluene using Sn(Oct)2 as the catalyst and t-butylbenzyl alcohol as the initiator. t-Butylbenzyl alcohol was chosen as the initiator because the t-butyl group provides a distinct 1H NMR peak useful for and group analysis. The polylactide chain was then converted to an ATRP macroinitiator using wbromoisobutyryl bromide (Scheme 3.1). The 1H NMR spectrum of the macroinitiator is shown in Figure 3.1. Two signals are diagnostic for the end groups, two peaks at 1.94 ppm (c) due to the diastereotopic methyl groups on the carbon or to the carbonyl group at the capped end of the polymer chain, and a singlet at 1.3 ppm (8) due to the t-butyl 133 group. A quartet at 5.15 ppm (d) from the methine proton on the polylactide chain, and a doublet at 1.6 ppm (b) serve as markers for the polylactide chain. The Mn of the macroinitiator was calculated from the NMR data in two ways, from the integration ratio of protons d and c, or from the ratio of the d and a protons. Both methods gave comparable results ((d/c): Mn = 5,500; (d/a): Mn = 5,250), which suggests a high end-capping efficiency. The second strategy uses a difunctional initiator that can initiate both ROP and ATRP. As shown in Scheme 3.2, the initiator is readily synthesized from a difunctional alcohol and an a-bromoacyl bromide.39 In principle, this initiator could support simultaneous ATRP of MMA and ROP of lactide, but in this study, lactide was polymerized first to afford a polylactide macroinitiator. Thus, the two approaches yield two different polylactide macroinitiators for ATRP of MMA. The key difference between the two macrominitiators is that the macroinitiator prepared by the end capping approach has an a-bromo ester group as the chain end, while the macroinitiator prepared from the difunctional initiator is terminated with a hydroxy group. As seen later, the difference in the chemical nature of the chain ends lead to dramatic differences in the thermal properties of the polymer. 134 :1 C CH 1°10 o o 0 5 My“). B > Ro-PLA—on/Ofl o RO-PLA-Br CE Scheme 3.1. Synthesis of end-capped polylactide initiator (RO-PLA-Br) 4.5 4.0 Figure 3.1 . bromide 1H NMR of polylactide end capped with a-bromoisobutyryl 135 HO BI“ N8HCO3 A O + Br glyme 7 HO 0 Br OH O if ”29*“ O o . Ho-—PLA—o/7<\0Jl\fiBr 0 Sn (0%) HO-PLA-Br 2 Scheme 3.2. Synthesis of polyactide prepolymer (HO-PLA-Br) from a difunctional initiator 136 3.2.2. Synthesis of polylactide-b-PMMA The two polylactide macroinitiators were used to initiate the solution ATRP of MMA to give the corresponding linear diblock copolymers. The solvents used were toluene and anisole, with the more polar anisole being the preferred solvent for macroinitiators based on crystalline poly(L-Iactide). ATRP of MMA was carried out at ambient and elevated temperatures to study the effect of polymerization temperature on the final composition of the block copolymers. For ATRP at high temperatures (e.g. 70 °C), the catalyst was CuBr with bipyridine as the ligand (Scheme 3.3). ATRP at ambient temperature required a more active catalyst, CuBr paired with MersTREN.4o Both sets of conditions afforded soluble block copolymers with good control over molecular weight and polydispersity. However, a disadvantage of ambient temperature polymerization was significantly longer polymerization times, even when run in bulk, and as a result, most polymerizations were run at 70 °C using bipyridine as the ligand. At the end of the polymerization, the reaction vessel was opened to the air, and the polymerization solution changed from dark red to green as Cu(l) oxidized to Cu(ll). The catalyst was removed by treatment with activated carbon followed by filtration. Residual copper catalyst in the organic layer could also be removed by extraction with an aqueous solution of EDTA (ethylenediaminetetraacetic acid disodium salt). After the extraction, the organic layer was colorless. The usual method for removing the copper catalyst is by filtration through silica gel and neutral alumina,41 but when this method was used, the yield of recovered 137 polymer was low, possibly due to strong adsorption of the polymer to silica or alumina. Figure 3.2 shows typical GPC traces for the polylactide macroinitiator and the block copolymer. Both “controlled” polymerization products had monomodal distributions with low polydispersities, and there was no evidence that the block copolymer was contaminated by polylactide or PMMA homopolymer. The polymer characterization data appear in Table 3.1. The molecular weights of the block copolymers were obtained by two methods, from the GPC analysis and by directly calculating Mn from the 1H NMR data. Shown in Figure 3.3 is a typical 1H NMR spectrum of the block copolymer. For the case shown in Figure 3.2, Mn for the starting polylactide was 4,740 g/mol. Comparing the integrated intensities for peaks a and e gives 1:2.2 as the molar ratio of the two blocks, which corresponds to Mn = 19,200 g/mol. Mn calculated from the GPC data was 20,500 g/mol. The reflectance FTIR spectrum of the block copolymer, measured for a film spin-coated on a Au-coated silicon wafer, is shown in Figure 3.4. As expected for block copolymers, the spectrum appears as the linear sum of the two homopolymer spectra. C=O peaks were detected at 1767 cm'1 and 1745 cm' 1, corresponding to polylactide C=O and PMMA C=O stretching respectively. These data and the monomodal GPC traces further confirm the formation of the block copolymer. 138 0 Br RO-PLA—OJKrofi o CuBr, pr, 70 °C RO-PLA-Br , O CuBr, MeBTREN, r.t. Ho—Pm-O’X‘okfiar HO-PLA-Br BDY Scheme 3.3. Synthesis of poly(lactide-b-MMA) copolymers using polylactide macroinitiators derived from end-capping polylactide (RO-PLA-Br) and from a m /N N N\_ MesTREN difunctional initiator (HO-PLA-Br). Table 3.1. Glass transition temperatures of amorphous poly(rao-lactide-b-MMA) PLA -b- PMMA CF copolymers. Copolymer compos't'on 100:0 67:33 31:69 0:100 (molar ratio) lactide: MMA Mn 20,000 31,700 17,900 12,000 PDI 1 .47 1 .52 1 .30 1 .04 Tg(°C) 47 61 90 110 1 39 Elution volume (mL) Figure 3.2. GPC traces of the polylactide macroinitiator and the resulting block copolymer. A: RO-PLA-Br (Mn: 7,200, PDI=1.24); B: poly(lactide-b-MMA) (Mn: 20,500, PDI=1.31). ,CH, 9 o 8 0 CH3 0 ”2 CH3 0 C [TTIIII111I11TIITTTTITTTTITTTTITTITTTTTTITTITITTTIITTTTITI’TT 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 ppm Figure 3.3. ‘H NMR of poly(lactide-b-MMA) 140 polylactide M A o O C (U E 8 D PMMA < A M POIYIactide-b-PMMA M w k f 4000 3500 3000 2500 2000 1500 1000 Wavenumber (cm") Figure 3.4. FT-IR spectra of polylactide, PMMA, and poly(lactide-b-MMA) films spin-coated on gold-coated silicon wafers. 141 3.2.3. Kinetics of the ATRP of MMA using polylactide macroinitiators A crystalline polylactide macroinitiator synthesized from the difunctional initiator was used to study the kinetics of the ATRP of MMA. Anisole was used as the solvent since the crystalline polylactide, the copper catalyst, and the block copolymer all have good solubility in anisole. The kinetic data in Figure 3.5 show a linear relationship, confirming that the ATRP of MMA initiated by polylactide was first order with respect to the monomer concentration as expected, and that the concentration of the active chains in the reaction was constant. Mn for the copolymers was calculated from the NMR data as described earlier, and as shown in Figure 3.6, a plot of the molecular weights of the. copolymers vs. conversion is linear and parallels the corresponding GPC data. The slight off-set for the two sets of data reflects the fact that GPC is a relative method and the molecular weights are calibrated using polystyrene standards. The linearity of the molecular weight vs. conversion data shows that ATRP initiated by the polylactide initiator is a “controlled” process. It is possible that the polymeric nature of the initiator may affect the polymerization kinetics and lead to slower ATRP than is seen for low molecular weight initiators. However, a control experiment designed to test for this effect gave unexpected results. When ethyl-2-bromoisobutyrate, a common ATRP initiator, was used to polymerize MMA under the same conditions used for the macroinitiator, we observed poor control over the molecular weight and a slow polymerization rate. Two additional control experiments helped explain this apparent contradiction. When a small amount of Sn(Oct)2 was added to the 142 system, we observed a color change from dark red to orange-red, and fast consumption of MMA. When polylactide homopolymer (synthesized from BBA and Sn(0ct)2 and not capped with an ATRP initiator) was added in the reaction as a spectator, we again observed a good control over the kinetics (Figure 3.7). Both observations point to Sn(Oct)2 as a modifier of the ATRP reaction. Aluminum isopropoxide has been used as a modifier in ATRP41'42 to ensure good kinetic control. It was proposed that aluminum isopropoxide, a Lewis acid, could coordinate with the initiator or the dormant polymer species and lower the dissociation energy of the carbon-halogen bond, thus facilitating the halogen transfer in the ATRP equilibrium. In our case, the same effect could result from tin catalyst residues in the macroinitiator and the spectator polylactide that were not removed by precipitation into cold methanol. Washing the polymer with dilute acid is more efficient at removing catalyst residues, but it is not favored by most researchers due to possible chain scission. A polylactide sample was washed with dilute HCI and then water. After precipitation into methanol, the dried polymer was added to an ATRP of MMA. After 28 hours, the conversion of MMA to polymer (15%) was comparable to a parallel polymerization run in the absence of added polylactide. 143 1.2 In ([Mlo/[Mltl 0.4 - 0.2 - 0 2 4 6 8 10 12 Time (hr) Figure 3.5. Semi-logarithmic kinetic plot of the solution ATRP of MMA in anisole at 70 °C using a polylactide macroinitiator. [MMA]=0.474 moVL, [MMA]: [polylactide]: [CuBr]: [Bipy] = 300 : 1 : 1 :2.5 144 2.0 4.0 . 11.8 g 3.5 g - 1.6 .7 30 E 52 « 1.4 X 2 2.5 .12 2.0 . . . 1.0 010 20 30 40 50 60 70 Conversion (%) Figure 3.6. Molecular weight of block copolymers vs. monomer conversion. .: Molecular weight measured from GPC using polystyrene as standard; A: PDI of copolymers; — : Molecular weight of copolymers calculated from NMR integration; -— —— : a guide to the eye drawn line parallel to the solid line 145 0.8 In ([Mlo/[erl Time (hr) Figure 3.7. ATRP kinetics of MMA in anisole initiated by ethyl-2- bromoisobutyrate with polylactide (Sn(0ct)2 and BBA) as the spectator. [MMA]=0.474 moVL, [MMA]:[ethyl-2-bromoisobutyrate]:[CuBr]:[Bipy] = 300:1 :1 :2.5; M..(polylactide)= 21 ,4009/mol, PDI = 1.47 146 3.2.4. Thermal properties of polylactide macroinitiators and block copolymers Macroinitiators synthesized by the end-capping scheme and by using a difunctional initiator have different thermal properties. In the Sn(Oct)2 catalyzed ROP of lactide, the alcohol initiator forms an ester at one end of the chain, with the other end terminated in a tin alkoxide. During workup in protic solvents, the tin alkoxide is usually hydrolyzed, and thus the polylactide macroinitiator synthesized from a difunctional initiator has a hydroxy chain end. In the end- capping approach, the macroinitiator is not terminated with a hydroxy group since tin alkoxide is replaced by an a-bromo ester. The difference in end groups is manifested in the thermal properties of the two polylactide macroinitiators. As shown in Figure 3.8, the onset for thermal degradation of end-capped polylactide is ~100 ° higher than for the hydroxy terminated macroinitiator. As described in the Introduction, a variety of degradation mechanisms have been proposed for polylactide, including intramolecular transesterification, cis-elimination and radical chain scission. lntramolecular transesterification to give volatile cyclic dimers or oligomers is generally accepted as the dominant thermal degradation pathway since both cis-elimination and homolytic cleavage reactions have higher activation energies and should become significant only at higher temperatures.“"“"5 Further favoring intramolecular transesterification are catalyst residues that often contaminate polylactides, even after washing with dilute HCI."’6 However, if hydroxy chain end of polylactide is capped, then intramolecular transesterification is kinetically inaccessible and polylactide will 147 remain intact until the onset of alternate degradation pathways at much higher temperatures. Other research groups have observed similar enhancements in the thermal stability of polylactide by blocking or capping hydroxy end groups.“ 49 The thermal degradation profiles of the polylactide macroinitiators are transferred to the corresponding block copolymers. As shown in Figure 3.9, the copolymer derived from an end-capped polylactide macroinitiator had a higher onset for thermal degradation and degraded in a single step. In contrast, the copolymer derived from the difunctional initiator showed an earlier degradation and displayed a stepwise weight loss. These steps were shown to correspond to the degradation of polylactide and PMMA respectively by heating a sample to 390 °C, and then analyzing the residue by 1H NMR. Only PMMA was present in the colored residue and no polylactide resonances were detected. Thus the hydroxy end groups facilitated polylactide degradation by intramolecular transesterification at low temperatures, which was followed by the degradation of PMMA at a higher temperature by a radical pathway. 148 100* 80 ’ RO-PLA-Br 9‘ / :7 60 ~ .C .9 i3 Q) _ 40 I. 3 / «‘3 HO-PLA-Br 20 - 0 1 1 L - L =i 100 200 300 400 500 600 T (°C) Figure 3.8. Thermal Gravimetric Analysis of polylactide macroinitiators synthesized by the and capping (RO-PLA-Br) and difunctional initiator strategies (HO-PLA-Br). Heating rate: 40 °C/min. under N2. 149 Sample weight (%) Figure 3.9. Thermal Gravimetric Analysis of poly(lactide-b-MMA) copolymers prepared from macroinitiators: a, HO-PLA-Br; b, RO-PLA-Br. Heating rate: 40 °C under N2 150 -n'-l !"Z‘u all-LI. 3.2.5. Miscibility and crystallization of poly(L-lactide) blocks in the copolymers. Poly(rac-lactide-b—MMA) and poly(L-lactide-b-MMA) copolymers were synthesized to study the miscibility of polylactide and PMMA blocks, and the effect that the PMMA block has on the crystallization of poly(L-Iactide). As shown in Table 3.1, DSC runs detected only one T9 for each poly(rao-lactide-b- MMA) copolymer, and the Tg increased as the PMMA block length increased. Thus, the two blocks are miscible, consistent with the data on blends of polylactide and PMMA.50 For poly(L-lactide-b-MMA), three different compositions were synthesized (molar ratio: 72:28, 54:46, 32:68) with the crystalline polylactide block constant in molecular weight (MIn = 21,800). The semicrystalline polylactide macroinitiator was used as a control. Both the pure poly(L-lactide) and the 72:28 sample readily crystallized under a variety of conditions. The DSC scans of Figure 3.10 are second heating scans recorded at 10 °C/min after each sample was first melted at 180 °C, and cooled to —20 °C at a rate of 10 °C/min. Samples 54:46 and 32:68 failed to crystallize under these conditions. However, after annealing sample 54:46 overnight at 130 °C, the sample did crystallize as shown in Figure 3.11. The 32:68 sample has the lowest poly(L—lactide) content and DSC did not detect crystallinity even after 36 hours of annealing. However, a low degree of crystallinity was observed by polarized optical microscopy. Despite the differences in crystallinity, all samples displayed a single glass transition that increased as the content of PMMA increased, which is consistent with a two- 151 'I t. . phase mixture of crystalline poly(L-lactide) and a homogeneous mixture of PMMA and polylactide. Polarized optical microscopy was used to study the crystalline morphology as well as the relative crystallization rate. Pure poly(L-lactide) and sample 72:28 were melted at 180 °C and then annealed at 140 °C. As shown in the micrographs Figures 3.12 and 3.13, pure poly(L-Iactide) had the highest crystallization rate. Within ten minutes, two spherulites formed in the field of view and grew rapidly until they impinged upon each other. Sample 72:28 crystallized more slowly, and consistent with a slower growth rate, more spherulites nucleated in the optical field. Despite difference in the rate of crystallization from the melt, both samples were highly crystalline and eventually the entire field of view was filled with spherulites. The different rates of crystallization can be understood in terms of kinetic barriers to crystallization caused by the methacrylate block. Two factors are at play. First, the polylactide block has the same length in all of the poly(L-Iactide-b-MMA) copolymers, and the longer chains in the 72:28 polymer should lead to increased chain entanglements and a slower crystallization rate. In addition, the crystallization process must exclude the PMMA chains from the crystal lattice, imposing another restraint on the growth rate. (Wide-angle X-ray scattering (Figure 3.14) showed that poly(L- lactide) block in the copolymer (72:28) had the same diffraction pattern as pure poly(L-lactide). The most intense peaks at 20 values of 16.3 and 18.7 ° agree with those reported for the or from of optically pure poly(L—lactide).51 These effects also show up in the DSC runs shown in Figure 3.10 as a shift of the 152 crystallization exotherrn for the 72:28 sample to a higher temperature than for pure poly(L-lactide). Because the PMMA block is chemically bonded to the poly(L-lactide) block and polarized optical microscopy shows no evidence of PMMA aggregation (dark regions), the PMMA blocks in 72:28 must be located between lamellae in the spherulites. 153 Heat flow endo —> I' 0 30 60 90120150180 T(°C) Figure 3.10. DSC second heating scans of poly(L-lactide-b-MMA) copolymers taken after cooling from 180 °C at 10 °C/min. Samples were heated at 10 °C/min. under helium. a: PLLA 100 (pure PLLA); b: PLLA 72 (PLLA:PMMA= 72:28); 0: PLLA 54 (PLLA:PMMA= 54:46); d: PLLA 32 (PLLA:PMMA= 32:68) 154 '8 c’ C S" E ’33 it: :1: c A m 0 50 100 150 T (°C) Figure 3.11. Normalized DSC heating scans of poly(L-lactide-b-MMA) (54:46). Samples were heated at 10 °C/min. under helium. c: after cooling at 10 °C/min from 180 °C; 0’: taken after annealing overnight at 130 °C 155 10 min 25 min 500 gm Figure 3.12. Optical micrograph of pure poly(L-lactide) (Mn: 21,800) annealed at 140 °C and observed through cross polarizers (black regions were due to air bubbles) 156 Figure 3.13. Optical micrograph of PLLA 72 (PLLA: PMMA: 72:28) annealed at 140 °C and observed through cross polarizers(black regions were due to air bubbles) 157 3500 3000 - 2500 - 2000 *- 1500 r 1000 500 “ Intensity 51015 20 25 30 35 40 45 50 20 5 Intensity .§§§§§§ 5101520253035404550 28 Figure 3.14. Wide-angle X-ray diffraction pattern of poly(L-lactide) and poly(L- Iactide-b-MMA) (72:28) after annealing at 130 °C overnight. (A) as precipitated from solution (B). — polylactide; — block copolymer 158 3.3. Experimental Section General. Unless othenNise specified, ACS reagent grade starting materials were used as received from commercial suppliers. Toluene was freshly distilled from sodium benzophenone ketyl under nitrogen. Methyl methacrylate (MMA) was distilled over KOH and powdered calcium hydride and was stored in a freezer at —17 °C in a drybox. 1,2-dimethoxyethane (glyme) was distilled from calcium hydride. Racemic and L-lactide were recrystallized three times from ethyl acetate before use. Characterization 1H and ”C NMR analyses were performed at room temperature in CDCla on a Varian Gemini-300 spectrometer using TMS as the chemical shift standard unless othenrvise specified. Reflectance FTIR spectra were obtained under nitrogen using a Nicolet Magna-560 FT IR spectrometer containing a PIKE grazing angle (80°) attachment. Typically, 256 scans were collected using a MCT detector. Polymer molecular weights were measured by gel permeation chromatography (GPC) at 35 °C in THF using a Plgel 2011 Mixed column at a flow rate of 1 mL/min. Two detectors were used, a Waters R410 Differential Refractometer and a Waters 996 Photodiode Array. The concentration of the polymer samples was 1 mg/mL, and each solution was filtered through a Whatman 0.2 pm PTFE filter before injection. The molecular weights are reported relative to monodisperse polystyrene standards. Differential scanning calorimetry (DSC) data were obtained with a Perkin Elmer DSC 7 instrument calibrated with indium and hexyl bromide standards. The samples were placed in aluminum pans, and were heated at 10 °C/min under a helium 159 atmosphere. Liquid nitrogen was used as the coolant. Thermogravimetric analysis (TGA) data were obtained from a Perkin Elmer TGA 7 instrument at a heating rate of 40 °C/min under nitrogen. Optical microscopy experiments were carried out on a Nikon OPTIPHOT-2-POL microscope equipped with a Mettler FP82-HT hot stage. Synthesis of 2,2-DImethyl-3-hydroxypropyl asbromoisobutyrate. To a well-stirred suspension of 3.5 g of sodium bicarbonate in 19.5 g of dry glyme was added 18.75 g (0.18 mol) neopentyl glycol. a-Bromoisobutyryl bromide (9.2 g, 0.04 mol) was added to the mixture dropwise and stirring was continued for 20 minutes. The mixture was filtered through filter paper, and the filtrate was concentrated using a rotary evaporator. Ether (100 mL) was added, and the milky solution was washed with water several times to remove excess neopentyl glycol. The ether layer was dried with sodium sulfate and concentrated under vacuum. Vacuum distillation of the residual oil afforded 5.3 g (52%) of the difunctional initiator as a colorless liquid: bp 76 °C (0.1 mm); IR bands at 3400 and 1720 cm"; 1H NMR (300 MHz 60013): 6 0.94 (s, 6H), 1.91 (s, 6H), 3.36 (s, 2H), 4.0 (s, 2H), 5.2(b, 1H); ”C NMR (75 MHz CDCla): 6172.04, 70.76, 68.07, 55.82, 36.59, 30.69, 21.33; MS (EI) m/z=253.0 (M); Anal. Cal. for C9H17Br03: C, 42.7; H, 6.8. Found: C, 41.75; H, 6.85. Bulk ring opening polymerization of lactide using the difunctional Initiator Racemic or L-lactide was dried under vacuum overnight before ring opening polymerization. A predetermined amount of lactide (5.00 g, 347 mmol) was added to a Schlenk flask. The flask was connected to a vacuum line and 160 was evacuated and refilled with argon three times. Toluene solutions of Sn(Oct)2 (1.60 mL, 0.217 moVL) and the difunctional initiator ( 2.82 mL, 0.123 moi/L) were added to the monomer by syringe. After removing the toluene, the Schlenk flask was refilled with argon and was heated at 140 °C in a thermostatted oil bath. After one half hour, the flask was removed from the oil bath and cooled to room temperature. A small amount of toluene or dichloromethane was used to dissolve the polymer sample, and the polymer solution was added drop-wise to a large volume of well-stirred cold methanol. After filtration, the polylactide sample was dried under vacuum at 50-60 °C until it reached constant weight. Synthesis of end-capped polylactide macroinitiators. Racemic or L- Iactide (4.32 g, 0.030 mol) was dissolved in 40 mL of toluene under argon at 90 °C. Toluene solutions of Sn(Oct)2 (4.29 mL, 0.233 mol/L) and t-butyl benzyl alcohol (BBA) (9.42 mL, 0.106 moVL) were added to start the polymerization. Using a syringe, small samples were removed and characterized by NMR to monitor the conversion of monomer to polymer. Upon reaching completion, the polymer solution was cooled to room temperature, and a-bromoisobutyryl bromide (3.71 mL, 0.027 mol) and pyridine (2.45 mL, 0.030 mol) were added. After stirring for half an hour, the mixture was filtered and the filtrate was concentrated using a rotary evaporator. Then the polymer solution was precipitated into a large amount of cold methanol, filtered and dried under vacuum at 50-60 °C before use. Synthesis of polylactide-b-PMMA using CuBr catalyst Polylactide prepared from the difunctional initiator and end-capped polylactide were used as 161 ATRP initiators. CuBr/bpy was used as the catalyst at 70 °C, and CuBr/MeaTREN was used for ambient temperature ATRP. ATRP was run either in bulk or in solution (toluene, anisole) in helium filled drybox. In a typical run, polylactide macroinitiator (0.63 g, 0.029 mmol) was first dissolved in 10 g of anisole at 70 °C, followed by CuBr (0.0090 g, 0.063 mmol) and bipyridine (0.0224 g, 0.143 mmol) to start the polymerization. The conversion of the polymerization was monitored by NMR. The polymerization was stopped by taking the reaction flask out of the drybox and opening it to the air. The polymer solution was diluted with toluene or dichloromethane and activated charcoal was added. After filtration, the polymer solution was concentrated and precipitated into a large amount of cold methanol. The purified block copolymer was then filtered and dried as usual. 162 References 10. 11. 12. 13. 14. 15. 16. 17. 18. Jain, R. A. Biomaterials 2000, 21, 2475-2490. Middleton, J. C.; Tipton, A. J. Biomaterials 2000, 21, 2335-2346. Sheridan, M. H.; Shea, L. D.; Peters, M. C.; Mooney, D. J. J. Control. Release 2000, 64, 91-102. Uhrich, K. E.; Cannizzaro, S. M.; Langer, R. S.; Shakesheff, K. M. Chem. Rev. 1999, 99, 3181-3198. Drumright, R. E.; Gruber, P. R.; Henton, D. E. Adv. Mater. 2000, 12, 1841- 1846. Bergsma, J. E.; Debruijn, W. C.; Rozema, F. R.; Bos, R. R. M.; Boering, G. Biomaterials 1 995, 16, 25-31. Wachsen, O.; Platkowski, K.; Reichert, K. H. Polym. Degrad. Stabil. 1997, 57, 87-94. Gogolewski, S.; Jovanovic, M.; Perren, S. M.; Dillon, J. G.; Hughes, M. K. Polym. Degrad. Stabil. 1 993, 40, 31 3-322. Cleek, R. L.; Ting, K. C.; Eskin, S. G.; Mikos, A. G. J. Control. Release 1997, 48, 259-268. Tsuji, H.; Mizuno, A.; lkada, Y. J. Appl. Polym. Sci. 1998, 70, 2259-2268. Yin, M.; Baker, G. L. Macromolecules 1999, 32, 7711-7718. Wang, D.; Feng, X. D. Macromolecules 1997, 30, 5688-5692. Joziasse, C. A. P.; Grablowitz, H.; Pennings, A. J. Macromol. Chem. Phys. 2000, 201, 107-112. Kasuga, T.; Ota, Y.; Nogami, M.; Abe, Y. Biomaterials 2001, 22, 19-23. Qian, H. T.; Bei, J. 2.; Wang, S. G. Polym. Degrad. Stabil. 2000, 68, 423- 429. Li, Y. X.; Kissel, T. Polymer 1998, 39, 4421-4427. Chen, X. H.; McCarthy, 8. P.; Gross, R. A. Macromolecules 1997, 30, 4295-4301. Celikkaya, E.; Denkbas, E. B.; Piskin, E. J. Appl. Polym. Sci. 1996, 61, 1439-1446. 163 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. Frick, E. M.; Hillmyer, M. A. Macromol. Rapid Commun. 2000, 21, 1317- 1322. Schmidt, S. C.; Hillmyer, M. A. Macromolecules 1999, 32, 4794-4801. Wang, Y. B.; Hillmyer, M. A. Macromolecules 2000, 33, 7395-7403. Barakat, I.; Dubois, P.; Jerome, R.; Teyssie, P.; Goethals, E. J. Polym. Sci. Pol. Chem. 1994, 32, 2099-21 10. Shinoda, H.; Matyjaszewski, K. Macromolecules 2001, 34, 6243-6248. Mousa, W. F.; Kobayashi, M.; Shinzato, S.; Kamimura, M.; Neo, M.; Yoshihara, S.; Nakamura, T. Biomaterials 2000, 21, 2137-2146. Shinzato, S.; Kobayashi, M.; Mousa, W. F.; Kamimura, M.; Neo, M.; Kitamura, Y.; Kokubo, T.; Nakamura, T. J. Biomed. Mater. Res. 2000, 51, 258-272. Miller, D. J.; Lang, F. F.; Walsh, G. L.; Abi-Said, D.; Wildrick, D. M.; Gokaslan, Z. L. J. Neurosurg. 2000, 92, 181 -1 90. Vallet-Regi, M.; Granado, S.; Arcos, D.; Gordo, M.; Cabanas, M. V.; Ragel, C. V.; Salinas, A. J.; Doadrio, A. L.; San Roman, J. J. Biomed. Mater. Res. 1998, 39, 423-428. RodriguezLorenzo, L. M.; Salinas, A. J.; ValletRegi, M.; Roman, J. S. J. Biomed. Mater. Res. 1996, 30, 515-522. Avella, M.; Errico, M. E.; lmmirzi, B.; Malinconico, M.; Falcigno, L.; Paolillo, L. Macromol. Chem. Phys. 1998, 199, 1901-1907. He, Y.; Shuai, X.; Cao, A. M.; Kasuya, K.; Doi, Y.; lnoue, Y. Macromol. Rapid Commun. 2000, 21 , 1277-1281. Avella, M.; Errico, M. E.; lmmirzi, B.; Malinconico, M.; Falcigno, L.; Paolillo, L. Macromol. Chem. Phys. 2000, 201, 1295-1302. Hyon, S. H.; Jamshidi, K.; lkada, Y. Biomaterials1997, 18, 1503-1508. Coessens, V.; Pintauer, T.; Matyjaszewski, K. Prog. Polym. Sci. 2001, 26, 337-377. Greszta, D.; Mardare, D.; Matyjaszewski, K. Macromolecules 1994, 27, 638-644. Hedrick, J. L.; Trollsas, M.; Hawker, C. J.; Atthoff, B.; Claesson, H.; Heise, A.; Miller, R. D.; Mecerreyes, 0.; Jerome, R.; Dubois, P. Macromolecules 1998, 31, 8691-8705. 164 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. Mecerreyes, D.; Trollsas, M.; Hedrick, J. L. Macromolecules 1999, 32, 8753-8759. Zhang, 0.; Remsen, E. E.; Wooley, K. L. J. Am. Chem. Soc. 2000, 122, 3642-3651. Mecerreyes, D.; Moineau, G.; Dubois, P.; Jerome, R.; Hedrick, J. L.; Hawker, C. J.; Malmstrom, E. E.; Trollsas, M. Angew. Chem-Int. Edit. 1998, 37, 1274-1276. Newman, M. S.; Kilboum, E. J. Org. Chem. 1970, 35, 3186-3188. Ciampolini, M.; Nardi, N. Inorg. Chem1966, 5, 41-44. Guo, J. H.; Han, Z. W.; Wu, P. P. J. Mol. Catal. A-Chem. 2000, 159, 77- 83. Schubert, U. S.; Hochwimmer, G.; Spindler, C. E.; Nuyken, O. Macromol. Rapid Commun. 1999, 20, 351 -355. Kopinke, F. D.; Remmler, M.; Mackenzie, K.; Moder, M.; Wachsen, O. Polym. Degrad. Stabil. 1 996, 53, 329-342. Kopinke, F. D.; Mackenzie, K. J. Anal. Appl. Pyrolysis 1997, 40-1, 43-53. Wachsen, O.; Reichert, K. H.; Kruger, R. P.; Much, H.; Schulz, G. Polym. Degrad. Stabil. 1997, 55, 225-231. Cam, D.; Marucci, M. Polymer1997, 38, 1879-1884. Lee, W. K.; Losito, I.; Gardella, J. A.; Hicks, W. L. Macromolecules 2001, 34, 3000-3006. Lee, S. H.; Kim, S. H.; Han, Y. K.; Kim, Y. H. J. Polym. Sci. Pol. Chem. 2001, 39, 973-985. Jamshidi, K.; Hyon, S. H.; lkada, Y. Polymer 1988, 29, 2229-2234. Eguiburu, J. L.; Iruin, J. J.; Femandez-Bern'di, M. J.; Roman, J. S. Polymer 1998, 39, 6891 -6897. lkada, Y.; Jamshidi, K.; Tsuji, H.; Hyon, S. H. Macromolecules 1987, 20, 904-906. 165 Chapter 4 Block copolymers of lactide and OEGMA 4.1. General Polylactide is the most prominent biodegradable material in the packaging, pharmaceutical and medical fields. However, polylactide is brittle with a low impact resistance, and articles made from polylactide tend to shatter. Another limitation of polylactide is its hydrophobic nature, which leads a to slow biodegradation that is undesirable in some medical applications. In drug delivery systems, carriers with a better hydrophobic/hydrophilic balance are desired to achieve faster water uptake and more rapid drug release at early stages in the degradation process. Amphiphilic copolymers systems that include lactide and ethylene oxide have been synthesized to address both problems. Methoxy-capped Oligo(ethylene oxide) methacrylate (OEGMA) is a commercially available hydrophilic monomer. Polymerization of OEGMA gives a polymer with the same backbone as PMMA, but with hydrophilic side chains composed of ethylene oxide oligomers. The side chains bear some of the same characteristics of PEO, such as hydrophilicity, biocompatibility and good resistance to protein adsorption and cell adhesion."3 Poly(OEGMA) (POEGMA) has been widely used in coatingsz'4 hydrogels5 and drug delivery nanospheres6 as a hydrophilic and biocompatible material. Block polymers of lactide and OEGMA are a new amphiphilic biocompatible polymer system. Previous studies of lactide/PEO block copolymers were limited to the PEO block lengths that are available from commercial suppliers. In lactide/OEGMA 166 block copolymers, the length of each block can be controlled simply by adjusting monomer to initiator ratio and the reaction time since two “controlled” polymerization methods can be applied, just as in the case of lactide/methyl methacrylate block copolymers. Another advantage of using OEGMA is that each monomer contains an average of 5 ethylene oxide units, which means that many ethylene oxide units are incorporated into copolymers even for relatively short OEGMA block lengths. In aqueous solutions, OEGMA is known to polymerize rapidly and in a controlled fashion using ATRP since both monomer and polymer are soluble in water. POEGMA also has good solubility in typical organic solvents. In the scheme described here, polylactide was polymerized first using a difunctional initiator for ease of molecular weight control. Then the macroinitiator was used to polymerize OEGMA in toluene or anisole in the presence of CuBr and bipyridine. 4.2. Results and discussions 4.2.1 . Polymerization of lactide and OEGMA As shown in Scheme 4.1, lactide was polymerized by Sn(Oct)2 and a difunctional initiator, 2-bromo-2-methylpropionic acid 3-hydroxy-2,2-dimethyl- propyl ester. The resulting macroinitiator was purified by dissolution in CHzClg, and precipitation into methanol, followed by drying under vacuum at 50 °C. For ATRP, a predetermined amount of macroinitiator, CuBr/bipyridine, OEGMA and 167 toluene were used in the system. The polylactide macroinitiator was first dissolved in toluene at 70 °C, and then OEGMA with the CuBr/bipyridine catalyst were added to start the polymerization. Parallel reactions were set up that could be stopped at desired intervals and conversions to give copolymers of different composition. As seen from entries 2-4 in Table 4.1, the degree of polymerization increased and the polydispersity index decreased as the POEGMA block length increased. This behavior is consistent with well-controlled ATRP, where the polydispersity index decreases as the conversion increases.7 Block copolymer formation was confirmed by GPC and spectroscopic data for the copolymers. GPC measurements show a single peak for the polylactide block that shifts to higher molecular weights (Figure 4.1). There is no evidence of polylactide or POEGMA homopolymer in the GPC trace of the block copolymer. FTIR (Figure 4.2.) and ‘H NMR (Figure 4.3.) spectroscopy of polylactide, POEGMA and the block copolymers confirmed the presence and composition of both blocks in the resulting copolymers. 168 Table 4.1. Block copolymers of polylactide-b-POEGMA ratio of two blocks M..(calc.from entry M"(PLA) PLAzPOEGMA NMR ratio) M" (GPC) PD' 1 18,7008 100:45 53,760 54,750 1.65 2 30,000” 100:12.5 45,600 58,900 1.59 3 30,000b 100235 73,750 70,900 1.40 4 30,000b 100:40 80,000 76,530 1.33 a: obtained by NMR, M, = 25,500, PDI: 1.55 (by GPC) b: obtained by NMR, M. = 46,500, PDI: 1.58 (by GPC) B A 10.0 12.0 14.0 16.0 18.0 Figure 4.1. GPC traces of polylactide macroinitiator (A: Mn: 26,150, PDI: Elution Volume (ml) 1.47) and polylactide-b-POEGMA (B: Mn: 45,790, PDl= 1.39) 169 Sn(Oct)2 O 2 H 11,/0‘ O BNOXOMOA} o o " 0 LA PLA H2 CuBr 0:320 pr anisole 70 °C V H C M °\><’°1‘1/L01 H Br O?_0 m 0 O n \3 H30 “'5 Poly(lactide-b-OEGMA) Scheme 4.1. Synthesis of poly(lactide-b—OEGMA) 170 POEGMA A Q 8 3 8 polylactide-b-POEGMA g J’L actide 4000 3500 3000 2500 2000 1500 1000 Wavenumber (cm") Figure 4.2. FTIR spectra of polylactide, POEGMA and polylactide-b- POEGMA) films spin coated on gold-coated silicon substrates. 171 I w b e , J1. .3 1.1.3.2. IYTYYITYTVlY—firTIYTTfIYY YYIVT YjYTTYTYTT—rYl[IVTYIYYIIIVY'VI'TYY TVI'V‘VITTTTIYV IVIY 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 Figure 4.3. NMR spectrum of polylactide-b-POEGMA 172 4.2.2. Purification of block copolymers Upon exposure to air, the polymer solutions changed from the brownish red of a typical ATRP system to green, paralleling the oxidation Cu(l) to Cu(ll) by oxygen. Activated carbon failed to remove residual copper catalysts despite lengthy filtrations using 0.2 pm filters. Centrifugation gave clear but tinted solutions. EDTA is an excellent chelating agent for heavy metal ions, and washing the polymer solution with saturated aqueous EDTA solution proved effective at removing residual copper. After two or three extractions, the blue color of the copper containing organic phase was completely clear. Washing with distilled water followed by removal of the solvent yielded the copolymers as white solids. The copolymers were purified by removing residual OEGMA from the crude polymer samples. OEGMA is soluble in hexanes, but the copolymers are insoluble. NMR spectra of the copolymer showed that washing the crude polymer with hexanes several times gave OEGMA-free polymer. L- Iactide/OEGMA block copolymers were also purified by dissolution in toluene or methylene chloride followed by precipitation into cold methanol. 173 4.2.3. Thermal properties of lactide and OEGMA copolymers POEGMA, poly(rao-lactide), and poly(rao-lactide)-b—poly(OEGMA) copolymers containing 9 and 28.5 mol % OEGMA were analyzed by DSC (Figure 4.4). The polylactide block in both copolymers and the polylactide homopolymer had the same length (polylactide M" = 30,000, PDI = 1.58). POEGMA, a viscous liquid at room temperature, has the lowest T9 -50 °C, while that of polylactide is the highest (48 °C). The T95 of the block copolymers fall between those two extremes. The copolymer containing 28.5% OEGMA showed a broad glass transition temperature below 0 °C, which is consistent with partial miscibility of the two blocks. The narrower transition for the 9% OEGMA copolymer (T9 = 25 °C) suggests improved miscibility for shorter OEGMA block lengths. Both block copolymers are elastic rubbery materials, but the block copolymer consisting of 9% OEGMA is stiffer. 174 Heat flow endo —> u: l 1 l l l l I -100 -75 -50 -25 0 25 50 75 100 T(°C) Figure 4.4. Differential Scanning Calorimetry of poly(rao-lactide), poly(OEGMA), and poly(rao-lactide)-b-poly(OEGMA) copolymers. A: polylactide; B: poly(rac-lactide)-b-poly(OEGMA) (9% OEGMA); C: poly(rac lactide)-b-poly(OEGMA) (28.5% OEGMA); D: poly(OEGMA) 175 100 .o' 30 h - E 60 —- .° \ E o 9 . ‘0 40 - . 20 0 l I; l l 0 10 20 30 40 50 Time (s) Figure 4.5. Strain recovery of poly(rac-Iactide)-b-poly(OEGMA) (9% OEGMA) at 37 °C. Stress was applied for the first 22 seconds. The mechanical properties of the block copolymer containing 9% OEGMA was characterized by a strain recovery test. A circular shaped sample (~ 1” in diameter) was sheared axially in a meometer at 37 °C to 100% strain in 22 seconds. The load was removed, and the strain recovery was measured during the next 22 seconds. The strain recovery plot, (Figure 4.5) shows that ~90% of the strain was recovered leaving only 10% residual strain. The residual strain is irreversible deformation that typically results from viscous flow in the sample. Experiments run at 30 °C, closer to the T9 of the polymer, showed more rubbery behavior, with almost complete strain recovery. A shear modulus of 1.6 x 104 Pa 176 was calculated from the stress-strain curve for the copolymer measured at a frequency of 1 Hz at 37 °C (Figure 4.6), while common rubbery materials have average shear moduli of ~105 Pa at 25 °C. At 30 °C, the modulus increased to 2.2 x 104 Pa. 45.0 40.0 35.0 30.0 25.0 20.0 Stress ( Kpa) 15.0 10.0 5.0 0.0 ‘ ‘ 0.00 1.00 2.00 3.00 Strain (%) Figure 4.6. Stress-strain curve of p0ly(rao-Iactide)-b-poly(OEGMA) (9% OEGMA) measured at a frequency of 1H2 at 37 °C 177 4.3. Preparation of polylactide-b-POEGMA nanoparticles Polylactide-b-POEGMA copolymers contain both hydrophilic and hydrophobic blocks, and thus they should be ideal candidates for nano-sized drug carriers. The hydrophobic block can incorporate hydrophobic drugs while the hydrophilic block stabilizes the nanoparticles in the aqueous phase and prevents coagulation. Because block copolymers of rac-lactide and OEGMA are rubbery around room temperature, copolymers of L-Iactide and OEGMA were used because their higher T9 leads to nanoparticles that better maintain their shape. A poly(L-lactide) macroinitiator with a molecular weight of 18,900 glmol was used to initiate the ATRP of OEGMA in anisole at 70 °C. Poly(L-lactide)-b- poly(OEGMA) (6 mol% OEGMA) was used to form nanoparticles. Due to its simplicity, the dialysis method was used to prepare the nanoparticles. Block copolymers were dissolved in dioxane, loaded in a dialysis tube with a molecular weight cut off of ~ 3,000 glmol, and then the dialysis tube was immersed in a water bath for 24 hours (Figure 4.7). The clear block copolymer solution became translucent as water diffused into the tube and dioxane out of the tube. A portion of the translucent solution was then dripped onto a gold-coated silicon wafer and freeze dried. Imaging by AFM and ESEM gave images with poor resolution, but SEM measurement of uncoated nanoparticles taken at a high scan rate revealed spherical sub-micron particles (Figure 4.8). The nanoparticles were not robust in the electron beam and readily degraded, especially at slower scan rates. 178 Dialysis tube (molecular weight cutoff 3,000 glmol) Block copolymer in dioxane Figure 4.7. Preparation of poly(L-lactide)-b-poly(OEGMA) nanoparticles via dialysis. 179 Figure 4.8. SEM photographs of poly(L-lactide)-b-poly(OEGMA) particles freeze-dried on a gold substrate 180 Preliminary results on drug loading Lidocaine (Figure 4.9) is one the most accurately documented local anesthetics and is a widely used model compound for the encapsulation and delivery of hydrophobic drugs. The dialysis procedure described earlier was used to incorporate lidocaine in poly(L-Iactide)-b-POEGMA nanospheres. Equal weights of lidocaine and the copolymer were loaded into a dialysis tube and tin/NA (‘1 o ) lidocaine dialyzed against water. Figure 4.9. The chemical structure of lidocaine The amount of lidocaine incorporated into the nanospheres must be determined to measure the drug loading efficiency in this system. The lidocaine concentration can be quantified by UV/vis spectroscopy once the molar extinction coefficient, 8, is known. Methanol solutions of lidocaine with known concentrations were prepared and the spectrum of each solution was measured. A plot of absorbance at 262 nm vs. concentration gave 8:210 L mol'1 crn'1 (Figure 4.10). To determine the loading, a known amount of drug-loaded nanospheres were dried under vacuum to remove water. To avoid interference from the polymer, the nanospheres were extracted with methanol, a non-solvent for the polymer. NMR analysis confirmed that only lidocaine was extracted into 181 the solvent. The methanol solution of lidocaine was then filtered through a 0.2 pm filter to remove dust, concentrated, transferred to a 50 mL volumetric flask and then diluted to give exactly 50 mL. The UV absorbance at 262 nm was measured and the concentration determined. For 100 mg of copolymer, 100 mg of lidocaine and 50 mL of dioxane, the drug loading efficiency (wt%) (= [(amount of remaining drug in nanoparticles)/(initial feeding amount of drug)] x 100%) was 13%. Drug loading efficiency is affected by many factors such as preparation methods, block length, molecular weight of the (co)polymer, size and distribution of the nanoparticles as well as the hydrophobic/hydrophilic nature of the drug incorporated. Although the loading efficiency obtained was not superior to those obtained in literature, the advantages of having this amphiphilic block copolymer as a drug delivery matrix are obvious: the use of a stabilizer (e.g. poly(vinyl alchol)) and low boiling point halogenated solvents can be avoided by choosing the proper preparation method. Higher loading efficiencies can be expected for a more hydrophilic drug. 182 3.5 O 1 1 0 0.005 0.01 0.015 c (moVL) Figure 4.10. UV absorbance vs. concentration of lidocaine in methanol (71.: 262 nm) 183 4.4. Experimental Section General Unless otherwise specified, ACS reagent grade starting materials were used as received from commercial suppliers. Toluene was freshly distilled from sodium benzophenone ketyl under nitrogen. Oligo(ethylene oxide) methacrylate (OEGMA) (Mn average 300 glmol, Aldrich) was purified by passing the neat monomer through basic alumina and was stored in a freezer at -17 °C in a helium-filled drybox. Racemic and L-lactide were recrystallized three times from ethyl acetate before use. Dialysis tubing (flat width: 50 mm; vol/cm: 7.94 mL; wall thickness: 30 um; molecular weight cutoff: 3,000 glmol) used for the preparation of polymer nanoparticles was obtained from Fisher Scientific. Characterization 1H and ”C NMR analyses were performed at room temperature in CDCI3 on a Varian Gemini-300 spectrometer using TMS as the chemical shift standard. Reflectance FT IR spectra were obtained under nitrogen using a Nicolet Magna-560 FTIR spectrometer containing a PIKE grazing angle (80°) attachment. Typically, 256 scans were collected using a MCT detector. Polymer molecular weights were measured by gel permeation chromatography (GPC) at 35 °C in THF using a Plgel 20p Mixed column at a flow rate of 1 mL/min. Two detectors were used, a Waters R410 Differential Refractometer and a Waters 996 Photodiode Array. The concentration of the polymer samples was 1 mg/mL, and each solution was filtered through a Whatman 0.2um PTFE filter before injection. The molecular weights are reported relative to monodisperse polystyrene standards. Differential scanning calorimetry (DSC) data were obtained with a Perkin Elmer DSC 7 instrument calibrated with indium 184 and hexyl bromide standards. The samples were placed in aluminum pans, and were heated at 10 °C/min under a helium atmosphere. Liquid nitrogen was used as the coolant. Strain recovery experiments were run on a Paar-Physica UDS- 200 stress-controlled rheometer equipped with a forced-air oven. A 25 mm diameter cone-and-plate fixture was used for the measurements. Synthesis and purification of polylactide-b-POEGMA using CuBr catalyst Polylactide prepared from the difunctional initiator was used as the ATRP initiator. CuBr/bpy was used as the catalyst and anisole or toluene as the reaction solvent. In a helium filled drybox, the polylactide macroinitiator was dissolved in anisole or toluene at 70 °C, and then predetermined amounts of catalyst, ligand and monomer were added to start the polymerization. The conversion of monomer to polymer was monitored by NMR. The polymerization was stopped by taking the reaction flask out of the drybox and opening it to air. The polymer solution was diluted with toluene and then was washed twice with aqueous EDTA to remove residual copper catalyst. After filtration, the polymer solution was concentrated and extracted with hexanes. The purified block copolymer was dried under vacuum until it reached constant weight. Preparation of polylactide-b-POEGMA nanoparticles using dialysis Poly(L-lactide)-b-POEGMA (20 mg, 6.4 mole °/o OEGMA) and 20 mL of dioxane were loaded of into a dialysis tube. The tube was immersed into a 2L water bath filled with Milli-O water and equipped with a stir bar and a stopcock on the bottom. The Milli-Q water was replaced continuously for the first 2 hours and 185 every 6 hours afterwards. Within half an hour, the solution inside the dialysis tube became translucent and the dialysis was stopped after 24 hours. Preparation of polylactide-b-POEGMA nanoparticles loaded with Lidocaine using dialysis The same procedure was applied with the only change being that lidocaine was added in the dialysis tube as well. 186 References 1. Okamoto, H.; Kano, Y.; Nakashima, S.; Kotaka, T.; Takahashi, S.; Kasemura, T.; Nozawa, Y. J. Adhes. 1999, 71, 263-278. Zhang, F.; Kang, E. T.; Neoh, K. G.; Huang, W. J. Biomater. Sci, Polym. Ed. 2001, 12, 515-531. Fujimoto, K.; lnoue, H.; lkada, Y. Cent. Biomed. Eng. 1993, 27, 1559- 1567. Zhang, F.; Kang, E. T.; Neoh, K. G.; Wang, P.; Tan, K. L. Biomaterials 2001, 22, 1541-1548. Ward, J. H.; Peppas, N. A. J. Control. Release 2001, 71, 183-192. Spamacci, K.; Tondelli, L.; Laus, M. J. Polym. Sci. Pol. Chem. 2000, 38, 3347-3354. Davis, K. A.; Paik, H. J.; Matyjaszewski, K. Macromolecules 1999, 32, 1767-1776. 187