. 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-..\Ԥ;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><ic:ity of the heavy metals” 31although these catalysts are extremely active
to“lard ring opening polymerization of lactide and other cyclic esters even at

"h b ient temperatures.

16

 

O

O
( A/j/‘kOLSn + ROH __-:_-_ /Vj/U\O’Sn‘OR + A/j/KOH

Sn(0ct)2

O
YLO
Ckn/L\ o
O —__.
O ( > 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><DC \
Sche ' ' ' we“
Me 1.10. Poly(ortho ester) containing short polylactide b .‘s
1 overcome tin
PO'aner matrix leading to a complicated drug release profi‘ «0' (scheme ‘30)
1.93
problem, self-catalyzed poly(ortho ester)s were synthesize d9 . Thus the
. . . 30!

by “hklng the poly(ortho ester)s with short segments of I acne (rolled

00"
degradation Products from lactic acid segments cc‘z—ltéalyza

risticS 0’

degradation of poly(ortho ester)s and the surface erosion Charade

p0'Y(0rtho ester)s are maintained.

46

 

 

‘35,, gopowmet‘s with rUbbery blocks

“a mechan‘w‘ properties of polylactide, eSpecial 1y
lactide), need to be (mproved for certain applications. For S«hemp
applications require 3“ improvement in impaCt I'e'~‘3i'5tance and i" polylactide's
elongation at break. Copolymerization 01‘ polylac'tide With elastomers can

address these probiems and lead to materials with tuned mechanical prOperties.
POIY(dimeth ylsiloxane) (PDMS), a biocom patib|e material with a low glass

' ‘ the
tranSItlon temperature and high oxygen and water permeability, was used as
t'de‘b'
rubbery b'ock '°’ molt/(L-ladldel COPO‘YWVS-“SS The multiblock polyp—"aw
. _ \\ omels
PDMS) was prep fired by polycondensatron 0f Difunctional polylactlde o g

P 0““5

° he
on w . ' e \n ‘
centa‘. b‘ k 8V3 S)it'lthesized by ring opening PO'Ymerization of L-lac’l‘d

and ROMS o\'\go mars (Scheme 1.11). Triblock copolymers with

presence 0‘ «hydroxy terminated PDMS macroinitiator. These materials

exhibited 900d elastomeric properties.

Diblock copolymers of lactide and bu‘ad‘energe is°preneg7 a"

why/ens”
(Scheme 1.12) were Synthesized by a combination Of living anionic
and rin
opening polymerizations. Poly(ethY‘ehe’b'L'laCtide) proved to be 9
a
900d

compatiblizer for poly(L-Iactide)/polyethylene blends.

47

 

 

 

O
O /\/o 4:0 0‘5‘
Nfi O N/\/\SI 5\ t“ I \
n / \
O O
multibloCK PO‘Y‘adide and PDMS copolymer

1‘ lactide and dimethylsiloxane

Scheme 1 .11- Multiblock copolymer 0

48

A M
‘L jet?
is

 

l /
.. of
/ OH O n
'“ polyisoprene—b-polylactide
lactide/_
m

 

AlEta
W“ ’W‘W/Loi
m
0 n
polybutadiene-b-polylactide

H
Ira/c.21003

polyethyleneb-polylactide

Scheme 1.12. Block copolymers of lactide with olefins

49

 

1.8-6. Block copolymers with poly(ethylene oxide) A

Poly(ethylene oxide) (PEO) is a water soluble and biocompatible , .
polyether, and has been approved by most countries as a food additive.
Because PEO is highly flexible and provides no binding sites for proteins,
polylactide-b-PEO copolymers show less protein and cell adhesion, which makes
specific attachment possible by intentional immobilization of bioactive factors.
Block copolymers of PEO and polylactide also represent a class of
biodegradable/biocompatible polymers with balanced hydrophobicity and
We! rophilicity.

Linear diblock or triblock copolymers with PEG as the central block have

Th _-‘=-

been made using hydroxy or dihydroxy terminated PEO as the macroinitiator
(scheme 1.13)."9'106 Triblock copolymers with polylactide as the central block

We re prepared by coupling two diblock copolymers with hexamethylene

 

diisocyante (HMDI) (Scheme 1.13).

L-lactide and ethylene oxide multiblock copolymers were synthesized
uSuing an initiator generated in-situ from AlFla and H20 (Scheme 1.13). Although
the ring opening polymerization of both monomers was carried out in one pot, the

c<>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><cretion of the degradation products.
Block copolymers of lactide and ethylene oxide with cross-linkable end
groups were also synthesized to form core-polymerized stable nanospheres112

and scaffolds for tissue engineering.

52

 

O O

m o o o o o m

 

WNW

iM WM 9%
3...... ME} 0%):

O 0
FAQMOMOAfi
0 p
_2\=°
PEO-g-polylactide 2 O

O

 

 

      

MN} ,+

 

my:

polylactide-g-PEO

Scheme 1.14. Graft copolymers of lactide and ethylene oxide

53

 

 

 

One application of block copolymers of lactide and ethylene oxide in drug
delivery systems has focused on developing systems that can form in situ drug
delivery matrices in the body. Although microspheres with encapsulated drugs
can be fabricated from these block copolymers and injected into the body, the
fabrication process is usually complicated and requires the use of organic
solvents, which sometimes cause denaturation of proteins incorporated in the
microspheres.

The notion of forming an in situ drug delivery system is based on the fact
tMat block copolymers of lactide and ethylene oxide can undergo reversible
te'11perature-dependent sol-gel transitions in aqueous environments. The
gellation temperature can be tuned by varying the composition of the block

cc>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=<Dupling or disproportionation with the rate constants of propagation and
termination being kp and k. respectively. In well-controlled ATRPs, the position of
the equilibrium between the dormant and active species favors the dormant

sDecies to ensure low radical concentrations and minimize bimolecular chain

t3 rmination reactons.

55

 

W

ATRP is a multicomponent polymerization System. The effect of each

 

. . . uld
component (monomer, Initiator, catalyst/ligand, solvent and temperature) Sho
be considered to ensure a well-controlled ATRP.
l - ka
R—X + CuY I ligand & R. + XCu“Y [\tgand
6am ‘
U ~“-th‘
monomet mtermination
Scheme 1.15. Mechanism Of COpper-catalyzed ATRP
1) Monomers «‘65
C)
. — «\0“
A vanety Of mo homers can be polymerlZed by ATRP TVp‘°a\ \afihdee-
. renes. ac l - . 9.0 \
indude methacrylates , aCrY‘ates: Sty ryon'tnles an d (meW‘) Q“\s \“3

The commO“ ieature of these monomers ‘8 that they all cohtain suD

, 1 16,1 17
stabttize the 9‘09 agati n g radicals -

However. there are two major types of monomers that cannot be
POWmei'V-ed successfully by AT RP' One type incl“ties halogenated alkenGSI
alkyl substituted olefins, and vinyl esters, all of which have intrinsically [ow
reactivity m free radica\ polymerizations. The 0th e r type inc’udes aCidic
monomers such as methacrylic acids, which poison transition m eta] cata/ys’
through formation of metal carboxylateS. In addition ’ acidic monomers s
protonate transition metal catalysts that contain nitrern ligands and in, can
with their complexation abimy. This problem can be overcome Griere
protected monomers. For example, by adiUSting the pH, SOdium meth:c:lsmg

ate

was polymerized in aqueous phase by a PEO maCI'Oinitiator and CuB
”hp 119

56

 

 

 

 

Alternatively, the acid ca
n be protected with an easy to remove ester, allowing

ethacrylate and

 

of '
ATRP protected methacrylic acids such as ti'imethylsl’Y' m

benzyl methacrylate.‘2°

2) initiators

Halo ena d akan ‘
g to I es, benzylic halides, u—haloesters. d—haloketoneS. 0"

halonitriles and sulfonyl halides have been Used as ATRP . 't' t Not all
ini ia ors.

initial
ors work well for all monomers. The basic requirement for a good initiator i
that it should be at least as reactive as the subsequenuyt wing chains-
. ' - OrmOd gro
If the reactlwty IS too low, polymers with low po'YdiSpers'- be obtained
ities cannot \i the

(\°“°
.16
O‘yme“ («“9 0‘ \“e

re ' ' ' too ’ .
activity is high, 100 many radicals are generated at 1369‘“
the

because new chains can form at very late stages of th
e D

\'\l\9-
2x 00“?

powmeflzaion leading to the loss 0f i"matiNg s 0
Decles d"
by re

chosen tor a particular monomer. The most pap“,
ar hal . o ' ' tors
Qgen atoms in lnltla

are Br and Cl. Since radical polymerizations can t
Olerate manyf '
UflCthHBI

grOUpS, initiators with functional groups or macroinitiat
ors can also b
9 used to

prepare functional polymers‘2"‘ 23

or block copolymers_ 1 1 9 124

 

 

3) Transition metal catalysts and ligands

ts. First. the metal

There are a few prerequisites for efficient ATRP catalys
rated by one

ble to switch between two oxidation states sepa
toward halogen

[6 to

catalyst must be a

Second, the catalyst should have suitable atiinity

e\ectron.
atoms. Third, the metal should be able to expand “S coordinaflon sphe
a°°°mm°date a (PS°“d°)ha'°ge"' Fourth. the \igand should complex the meta'
ty Of meta| cata'YSts based on niCKel, ruthenium. “'

t
red by mos

relatively strongly. A varie

and copper meet these requirements, but 00p per catalysts are favo

’ ALA.
N N

“—-

 

\ I \ /
. - , . , bi ridin N.N, ~ . n- n‘ .
“mmme 4’4 “$253? W e rhyid?leiN ’Nnfiiarr‘ “‘9
{both ( (PMD €123
/ \ / N32 \ \
N N / T\
1.1.4.7’ioAO-hexamethyl tnsiz-(dimethylamino) N‘(2- . lmeth _
thethytehetetraamme ethyllamine pYrid fiyggtihanimgiéa
(HMTETA . .‘P‘GBTREN’ (PMPMI)
researchers, due ‘0 “‘9'" versatility and low Cost.

Figur91'9' Examples of nitrogen-containing "Qands used in ATRP

58

 

 

The catalyst activity is strongly dependent upon the ligand. Various
bidentate and multidentate nitrogen-containing ligands were developed (Figure
1.9) to tune the activity of copper halide catalysts. Tetradentate ligands such as
MeoTREN form more active catalyst/ligand complexes and lead to well controlled
ATRP even at room temperatures.125 Ligands also help solubilize transition
metal halides in monomers or solvents. Long chain alkyl groups on bipyridine
”"93 can improve the SOIUbi'ity 0* copper catalysts and lead to homogenous
polymerization conditions.

Ligands also play an important role in catalyst removal and recycling.
Schiit base ligands were covalently bound to silica gel and cross-linked
polystyrene beads for ease in catalyst separation and recyclingiza

Upon abstraction of one halogen atom by a Cu(l) complex, the

coordination number as well as the coordination structure changes. The
determination of the active catalyst structures remains a Challenge, and the exact
structure of the active species in the Cu/bpy system is not yet completely clear.
In the pr0posed Scheme 1.16, CU(|)(bPY)2 exists in a tetrahedral geometry, and
upon halogen abstraCtion, the coordination sphere of copper expands to

accommodate the extra halogen by adopting a trigonal bipyramidal geometry

4) Solvents
ATRP can be carried out in bulk, in solution or in heterogeneous SYStems
such as emulsions and suspensions. Generally, solvents are used for better

heat transfer and minimization of viscosity problems. If the resulting polymer is

59

 

not soluble in its monomer, as in the case of polyacrylonitfile, a solvent is
indispensable. The solvent should be chosen to minimize chain transfer and

other solvent assisted reactions.

\ l /\ / kact ~ 1 M‘s" e

t? CK

kdea ~ 107 M‘s“
m a U \\
/ / /

kp~103MH1S
P— P

" “—IG :76

 

 

m

Scheme 1.15. Proposed Cu(l) and Cu(ll) species using bpy as the ligand

60

 

1.10. Comblnatlon of ring opening polymerization and (controlled) radical
polymerization

Ring opening polymerization and (controlled) radical polymerization are
both widely used in polymer synthesis. A combination of both polymerization
methods can lead to interesting polymer structures and properties. ln general,
four strategies, and group transformation, the use ot ditunctional initiators,

difunctional monomers and functional macromonomers have been employed to

combine these two polymerization methods.

1) End group transformation

End group transformation converts the propagating center from the first
polymerization to the initiating species needed for the Second polymerization.
This technique has been used to prepare block copolymers127 and dendrimers128

using two or more mechanistically incompatible monomers, For example. ring
opening polymerization of caprolactone by Al(Oi-Pr)3 placed an ester group on
one end and a hydroxy end grOUp on the other end of the PO'Ycaprolactone
chain. The hydroxy chain and was transformed into an ATRP initiator moiety by
reaction with 2-bromoisobutyrylbomide, and then used to initiate the ATRP of t-

butylacrylate by a nickel or capper catalyst (Scheme 1.17).

2) Difunctional initiators

Difunctional initiators are molecules with two initiating sites. The use of

difunctional initiators can avoid problems encountered in end group

61

 

transfom ation such as a low concentration of the active species and time-
consuming separation and purification steps. The greatest advantage of such a
strategy is that block copolymers can be prepared in a minimum of steps without
intermediate functionalization reactions. in cases where the two polymer

mechanisms are compatible, one-pot synthesis of block copolymers are

 

 

possible.129
0
1) IPrOH ‘Xl/KB
' Toluene r
A\Et3 -7 5 00 4:113 DWO}H‘ Br L
2) caprolactone O " Pyridine
a) 1N HCI 3 THF

 

CE ( \/U\O’Bu _
Br N'B"2(Pph3)2 90 °C

or CuBr, diHbipy 110 °C

Scheme 1.17. Block copolymers synthesized by and group transformerion

Several research groups have applied this strategy to dual radical and ring
opening pOIymerizations. For example. the product of the alcohol exchange
reaction between Al(O-1Pr)3 and benzopinacol can be used as a difunctional

initiator since the substituted tetraphenylethane moiety in the initiator thermally

62

decomposes to initiate the radical polymerization of MMA and styrene, while the
AI—alkoxide bond can initiate the ring Opening polymerization of caprolactone.“30
In a preparation of shell-cross-linked nanoparticles, WooleY 9’ al. used
another difunctional initiator, AI(OCHZCBr3)3, to initiate the ring opening
polymerization of caprolactone and the ATRP of t-butylacrylate.‘27 Combined
ring opening polymerization of caprolactone and nitroxide-mediated “living”

radical polymerization of styrene were realized by employing a difunctional

initiator bearing a primary alcohol group and a nitroxide moiety. (Scheme 1.18)

3) D'liunctional monomers
Ditunctional monomers provide a new way to manipulate the structures
and properties of polymers through controlled b|00k or graft polymerizations.
Hedrick at al. reported the synthesis of graft hyperbranched and oroSSlinked
polymer systems using functional monomers via a combination of ring opening
polymerization and controlled radical polymerizations.

The ring opening copolymerization of monomer A in Figure 1.10 with
lactide yielded an aliphatic polyester containing pendent acrylate groups. UV or
radical initiated polymerization of the resulting polymer gave a crosslinked
system." Functional monomer B in Figure 1.10 was copolymerized sequentially
with caprolactone and MMA in a one reaction. It acted as both a monomer and

an initiator for ring opening polymerization and ATRP respectively.131

63

 

ES?

1
O \
2
(3 123°C

iwot. 3'?

"00'"

Scheme 1.18. Block copolymers of caprolactone and styrene synthesized b a
difunctional initiator approach y

O 0

\° >§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