as .33..
.. 7:...
mmiunfia? .2...
«5.3.392». ,

I

E

. T: ..
i: .
. l

Sr». 2

a: Q... r
a ;

«£3.41.

3...}...
i: ..
53 ha;
.21?

.3... w in?” gammyjuaafl
, 3311f“ ham fa
$5.“. 3!

3n.
Nun's.“
. Zr «2
: . . ,
. .itsmwxh 5
‘ z
x 4 .
J an
4‘»

321.. v.

§fi%.a..
{

hwy

«afihqurwzfs .a. a, 2.?
m. T. . , ..

$1, fiimm.w‘fla§h..§mnz. .
1....“ E35:. 2... .. .21..

2 .9 2.14.9.5; if...
w .2. auwkna a. m ,
any». ‘ «a

:5:

.a’1

‘1-
A a
«#5?»
an. . .
§ . ‘
ARES

.. .3. I

I?

. 1.
a:
0nd: and"?
a rum?»
in». {fig
{1432 a
‘u!‘ .21”) n
Pfihhn‘r)
. at... .51
a ‘ .
,..._vv.ns

 

..t.§«f .
tit» 22;
t..... 5501.!
‘5}!
lie...
Alt: ‘3. A .
J A .41.»:

 

“1551::

2 ac (

This is to certify that the

dissertation entitled

SYNTHESIS AND CHARACTERIZATION OF ARGVIATIC

SUBSTITUTED PILYLACTIDES
presented by

Tara Simnons

has been accepted towards fulfillment
of the requirements for

Ph .D. degree in Chem stry

     

Major professor

Date Dec. 11; 2000

MSU is an Affirmative Action/Equa! Opportunity Institution 0- 12771

 

,' UBRARY 7
Michigan State
' University i

 

 

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

 

AER 2521732

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DatoOuopGS—sz

SYNTHESIS AND CHARACTERIZATION OF AROMATIC SUBSTITUTED
POLYLACTIDES

By

Tara Simmons

A DISSERTATION

Submitted to
Michigan State University
In partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

2000

L1.”

POI}
UDSL

l
SVTIII

for ti
COmp
Orth;
kinezi
dOmii

10mm

Ls
5

 

mew;
dew.
\‘u

addi:

 

SYNTHESIS AND CHARACTERIZATION OF AROMATIC SUBSTITUTED

POLYLACTIDES

ABSTRACT

Polylactides are important biodegradable and biocompatible polymers with a
variety of uses such as surgical sutures and bone fixation devices. However, its low glass
transition temperature (Tg) limits the available uses. A biodegradable analog of
polystyrene would be desirable for a number of packaging uses. Aromatic rings have
also been shown to increase the Tg in the structurally similar poly(hydroxyalkanoate)s.
Poly(Z-hydroxy-5-phenylvalerate) has a Tg of 13 °C compared to -—15 °C for the
unsubstituted poly(hydroxyvalerate).l A variety of aromatic substituted lactides were
synthesized based on phenyllactic acid, mandelic acid, and methylphenyllactic acid.

For all aromatic substituted polylactides, the polymerization rate is slower than
for the polymerization of lactide. In addition, the solution polymerizations do not reach
completion, indicating that there could be catalyst degradation, terminating the reaction,
or that the reaction follows an equilibrium mechanism. Both models can be fit to the
kinetic data obtained, but it appears both mechanisms take place, with equilibrium
dominating at lower reaction times, and catalyst degradation dominating the kinetics at
longer reaction times.

Degradation studies were carried out for poly(phenyllactic acid) and poly(p-
methylphenyllactic acid) and compared to the degradation of poly(lactic acid). The
degradation rate for the polymers were slower than polylactide, most likely because the

addition of an aromatic ring causes the polymer to be more hydrophobic. The

 

 

 

degré
510m
\\ ithi
pol-w
degra
incrc.

CODCt

\shici
gTCtilt
DIOUU

obnnr

 

 

 

degradation for poly(phenyllactic acid) and poly(p-methylphenyllactic acid) is initially
slower than the model because the hydrophobicity causes the concentration of water
within the polymer sample to be relatively low, decreasing the rate of hydrolysis. As the
polymer chains are hydrolyzed, the sample becomes more hydrophillic and the
degradation rate increases. As the reaction approaches completion, the degradation rate
increases. During the degradation reaction, the polymer sample aggregates,
concentrating the acidic sites within the polymer, and catalyzing the hydrolysis.

Polyphenyllactide was found to have a glass transition temperature of 50 °C,
which is nearly the same as polylactide. We believe that the methylene group allows for
greater flexibility and prevents the Tg from being higher. By removing the methylene by
producing polymandelide, we found a Tg of 96 °C. However, there were difficulties in
obtaining high molecular weight materials.

A number of substituted polystyrenes have been synthesized and the properties
varied by changing substituents. This idea was adapted to polyphenyllactide by
polymerizing methyl-substituted phenyllactic acids. 0-, m-, and p-Methylphenyllactic
acids were synthesized from the corresponding methyl-substituted benzaldehyde in
relatively low yields. Poly(p-methylphenyllactic acid) has the highest glass transition
temperature, 59 °C, and poly(m-methylphenyllactic acid) the lowest, 42 °C. This
demonstrates that poly(phenyllactic acid) can be modified through substituents on the
aromatic ring. By choosing the correct substituent, a polymer with a higher glass

transition temperature, like polystyrene’s Tg of 100 °C, should be possible.

1) Fritzsche, K.; Lenz, R. W.; Fuller. R. C. Makromol. Chem. 1990, 191, 1957-1965.

To my grandmother, Olga Simmons

iv

 

 

 

 

p05

dClI

(1..
(I)
"3

mi

and

grou
ling;
3DO\.
CDKA
Wald

misst

Blilr‘

SUPP
I

 

0i p-

\idat

 

Acknowledgements

I wish to thank Professor Greg Baker, without whom this work would not be
possible. I appreciate all of his help, encouragement, and assistance in organizing my
defense while I was in West Virginia. I would also like to acknowledge the work of the
department glass blowers, Manfred Langer and Scott Bankroff, for their assistance in
making (and repairing) the custom glassware used in this work. Thanks for their time
and patience in teaching me to make the sealed tubes used for the melt polymerizations.

I would like to acknowledge the assistance and support of current and past Baker
group members Erin Paske, Chun Wang, Kirk Olson, J. B. Kim, Tianqi Liu, Gao Liu,
Jingpin Jia, Susan Pasco, Cory Ruud, and Jun Hou. In addition to my friends listed
above, I would like to thank my good friends that have made my stay at Michigan State
enjoyable: Beth Gardner, Rob Gentner, Nadia Shams, Dalila Kovacs, Craig Shiner, Mike
Waldo, Marie Migaud, Nancy Lavrik, and Diane Frost. My deepest apologies if I've
missed anyone important.

I would like to thank the Corn Marketing Board of Michigan, the Crop and Food
Bioprocessing Center, and the Center for Fundamental Materials Research for financial
support for this research.

Lastly, and most importantly, I deeply appreciate the support and encouragement
of my family: Michalina, Millie, and Steve Ostrander. They believed in me when I

wasn't so sure of myself, and I could not have finished without them.

Table of Contents

List of Tables
List of Figures
List of Schemes
Chapter 1 — Introduction
1 Background
1.1 Polylactide Synthesis
1.2 Synthesis of Lactide
1.3 Properties of Polylactide
1.3.1 Poly(L-lactide)
1.3.2 Poly(D-lactide)
1.3.3 Syndiotactic Polylactide
1.3.4 Heterotactic Polylactide
1.4 Lactide Polymerization Mechanism and Catalysts
1.5 Equilibrium Polymerizations
1.6 Degradation of Polylactides
1.7 Methods for Controlling and Varying Physical Properties
1.7.1 Copolymerization
1.7.2 Blends
1.7.3 Additives
1.7.4 Substituted Lactic Acids

1.8 Goal of the Project

vi

ix

xii

10

17

22

25

25

26

27

28

3O

 

 

Chm

Chan

 

 

 

1.9 Works Cited
Chapter 2 — Phenyllactide
2.1 Introduction
2.2 Monomer Synthesis
2.3 Solution Polymerization
2.4 Melt Polymerization
2.5 Polymer Properties
2.6 Polymer Degradation
2.7 Summary
2.8 Works Cited
Chapter 3 — Mandelide
3.1 Introduction
3.2 Monomer Synthesis
3.3 Polymerization
3.4 Initial Polymer Properties
3.5 Copolymers
3.6 Summary
3.7 Works Cited
Chapter 4 — Methylphenyllactides
4.1 Introduction
4.2 Synthesis of Substituted Phenyllacic Acids
4.3 Synthesis of Phenyllactides

4.4 Polymerizations

vii

31

36

36

39

43

52

55

57

64

65

66

66

68

69

73

75

77

78

80

8O

82

83

84

 

 

 

 

Ch;

 

4.5 Polymer Properites

4.6 Degradation

4.7 Summary

4.8 Works Cited
Chapter 5 —- Conclusions

Experimental

viii

86

88

91

92

93

96

 

 

 

Ta

Ta

 

List of Tables

Table 2.1 Melt Polymerization of Phenyllactide with Various Catalysts 54
Table 3.1 Melt Polymerization of Mandelide with Various Catalysts 71

Table 3.2 Copolymerization of Mandelide with Phenyllactide 75

ix

 

 

 

 

 

Figt

Figt‘

List of Figures

Figure 1.1 Important biodegradable polymers

Figure 1.2 Synthesis of lactide via eEsterification

Figure 1.3 Synthesis of lactide via cracking

Figure 1.4 Catalyst used for the synthesis of syndiotactic PDLLA

Figure 1.5 Catalyst used for the synthesis of heterotactic PLA

Figure 1.6 Mechanism for the polymerization of lactide uUsing Sn(Oct)2

Figure 1.7 Mechanism for intramolecular tTransesterification

Figure 1.8 Mechanism for the polymerization of lactide using Al(OiPr)3

Figure 1.9 Aluminum porphyrin catalyst

Figure 1.10 Tolman's yttrium catalysts

Figure 1.11 The effect of monomer concentration on ceiling temperature

Figure 1.12 Structure of polyphenyllactide and polymandelide

Figure 1.13 Aromatic substituted lactides

Figure 2.] Structures of polystyrene, polylactide, and lactide-based
analogs of polystyrene

Figure 2.2 The effect of an aromatic sSubstitutent on T8

Figure 2.3 Solution polymerization of Pphenyllactide with Sn(Oct)2

Figure 2.4 Comparison of the equilibrium and catalyst degradation
models during the polymerization of phenyllactide

Figure 2.5 Plot of number average molecular weight vs. conversion

Figure 2.6 Solution polymerization of phenyllactide using Al(OiPr)3

11

12

14

15

17

19

29

3O

37

38

44

47

49

50

Figure 2.7 Melt polymerization of phenyllactide at longer reaction times

Figure 2.8 DSC scan of polyphenyllactide

Figure 2.9 TGA of polyphenyllactide

Figure 2.10 Degradation of polyphenyllactide and polylactide

Figure 2.11 Degradation of polyphenyllactide and polylactide

Figure 2.12 Weight loss during the degradation of polyphenyllactide

Figure 3.1 Melt polymerization of R,S-mandelide and phenyllactide

Figure 3.2 TGA of polymandelide

Figure 3.3 DSC of mandelide and phenyllactide copolymers

Figure 4.1 Melt polymerization of p-methylphenyllactide

Figure 4.2 TGA of poly(p-methylphenyllactide) compared to
polyphenyllactide

Figure 4.3 Weight loss during the degradation of poly(p-
methylphenyllactide)

Figure 4.4 Molecular weight decrease during the degradation of poly(p-

methylphenyllactide) and polyphenyllactide

xi

56

58

59

61

62

63

7O

74

76

85

87

89

90

‘7‘: ~ . . 9-1

 

List of Schemes

Scheme 2.1 Biosynthetic Pathway for Phenyllactide

Scheme 2.2 Synthesis of Phenyllactide

Scheme 2.3 Mechanism of Sn(0ct)2 Catalyzed Polymerization of
Phenyllactide

Scheme 3.1 Synthesis of Polymandelide

Scheme 3.2 Synthesis of Mandelide

Scheme 4.1 Synthesis of Methylphenyllactic Acids

Scheme 4.2 Synthesis of Methylphenyllactide

xii

40

42

51

66

68

81

83

 

 

AM

LA
3“)
Mn
mTLL
3k
oTLD

PDI

Al(OiPr)3
BBA
DSC
GPC

LA

MD

oTLD
PDI
PDLLA
PGA
PHB
PHBV
PhLA
PhLD
PLA
PLLA
PMA

PmTLD

List of Abbreviations

Aluminum isopropoxide

t-Butylbenzyl alcohol

Differential scanning calorimetry

Gel permeation chromatography

Lactic acid

Mandelide

Number average molecular weight
m-tolyllactide or m-methylphenyllactide
Weight average molecular weight
o-tolyllactide or 0-methylphenyllactide
Polydispersity

Poly(D,L—lactic acid) or Poly(D,L-lactide)
Poly(glycolic acid) or Polyglycolide
Poly(hydroxybutyrate)
Poly(hydroxybutyrate—c0-valerate)
Phenyllactic acid

Phenyllactide

Poly(lactic acid) or Polylactide
Poly(L-lactic acid) or Poly(L-lactide)
Poly(mandelic acid)

Poly(m-tolyllactide) or Poly(m-methylphenyllactide)

xiii

“I

 

PoTLD Poly(o-tolyllactide) or Poly(o-methylphenyllactide)

PPhLA Poly(phenyllactic acid)

PPhLD Poly(phenyllactide)

PpTLD Poly(p-tolyllactide) or Poly(p-methylphenyllactide)
pTLD p—tolyllactide or p-methylphenyllactide

Sn(0ct)2 Tin(II) 2-ethylhexanoate or Tin octoate

Tg Glass transition temperature

TGA Thermal gravametric analysis

xiv

 

 

 

 

 

 

 

 

 

hydrc
Biodc
either
Hydrt
photo.
combi

and p1

Chapter 1 -— Introduction

1 Background

Most commercial plastics are resistant to degradation and as environmental
concerns grow, there is an increasing emphasis on reducing waste through the use of
recyclable and degradable polymers. Degradable materials include biodegradable,
hydrolytically degradable, photodegradable, and oxidatively degradable materials.
Biodegradable polymers and plastics are quantitatively converted by microorganisms to
either C02 and H20, or to CH4 and H20 under aerobic or anaerobic conditions.
Hydrolytically degradable polymers and plastics degrade via hydrolysis, while
photodegradable polymers and plastics degrade by exposure to sunlight or by the
combined action of sunlight and atmospheric oxygen. Oxidatively degradable polymers
and plastics undergo an oxidative degradation process.1

These terms are often incorrectly applied to degradable polymers. For example, a
number of polymers were previously reported to be biodegradable, but the observed
degradation was due to the degradation of plasticizer1 in the polymer. Other polymers
such as polyethylene, polypropylene, polyacrylonitrile, and poly(vinyl chloride) undergo
degradation (photodegradation), but at a slow rate. Poly(vinyl acetate) and poly(vinyl
alcohol), having residual carbonyl groups, also photodegrade. Photodegradable polymers
often persist as non-degradable oligomers, eliminating the major benefits of degradation.

The first synthetic biodegradable polymer was poly(glycolic acid) (PGA),

reported in 1954.2 Little work was initially done with PGA because of its poor thermal

 

 

 

 

 

prop
degn
too 1
degr;

the fi

currei
mater
acid).
p01}'€:
polytl
these

PHB 1.
for pug
and st(
bacteri.

bJCIErj;

Iime 1
Infill)“

militia

properties and its low hydrolytic stability. PGA’s high crystallinity leads to a slow
degradation rate. Its high melting temperature makes processing difficult because it is
too close to the degradation temperature, causing the polymer to undergo thermal
degradation during processing. Despite these limitations, two decades later PGA became
the first biodegradable suture material that was unrelated to natural polymers.

Since then, a number of other synthetic polymers have been synthesized and are
currently referred to as biodegradable, bioabsorbable, bioresorbable and bioerodible
materials (Figure 1.1). The most important of these degradable polymers are poly(lactic
acid), also called polylactide. Others include starch-based materials and variety of
polyesters, such as polyglycolide, polycaprolactone, poly(hydroxybutyrate) (PHB) and
poly(hydroxybutyrate-co-hydroxyvalerate) (PHBV). There is commercial interest in
these materials for packaging, agricultural mulch films, fibers, and for medical uses.
PHB has polypropylene-like properties and was sold under the commercial name Biopol
for packaging applications. PHBs are unique in that a variety of microorganisms produce
and store the polymer as an energy reserve. Commercial production of PHBs involves
bacterial fermentation of alkanoic acids followed by harvesting of the polymer from the
bacteria. PHBs are degraded by a wide range of bacteria, fungi, and algae.

Polylactide has been known to be bio— and environmentally degradable for some
time. Polylactide is structurally similar to polyglycolide, differing only in the added
methyl group Ct to the carbonyl. Depending on the stereochemistry of the polymer,

polylactides can be either semicrystalline or amorphous, allowing for greater flexibility in

 

\\/“\O//

‘ n

 

 

polyglycolide
poly(glycolic acid)

 

O
O /
n
poly(hydroxybutyrate)
O
O
polycaprolactone

Figure 1.1 Important biodegradable polymers

 

 

3%0’
. .1

polylactide

n

poly(lactic acid)

m

n

 

‘1 1’
I

 

' x
poly(hydroxybutyrate-co-valerate)

 

 

 

 

control
h}drol
by a

Pohla
pohla.
surgic;

poli'm

becau:
P0513
Dacha
Cmnn
addin
PIOCe

13015-1

controlling the degradation rate compared to PGA. Polylactides primarily degrade via
hydrolysis of the polyester to the a-hydroxy acid, lactic acid. Biodegradation catalyzed
by a variety of enzymes common in microorganisms has also been reported.2
Polylactides are biocompatible, and there is a low chance of rejection or infection when
polylactide is used inside the body. This has led to important medical applications as
surgical sutures, bone fixture devices, and drug delivery systems. Upon degradation, the
polymer forms lactic acid, which can be eliminated by the body.

Until recently, most uses for polylactide were limited to medical applications
because of the high cost of the polymer. However, methods were developed to produce
polylactide at lower costs, and polylactide can now compete with existing polymers as
packaging materials and fibers. The Dow Cargill joint venture has announced a
commercial polylactide plant with an announced capacity of 300 million 1b./yr. In
addition to reducing the cost of the polymer, improvements in the physical properties and
processing characteristics have minimized earlier problems associated with crazing and

polylactide’s high melting temperature.

1.1 Polylactide Synthesis

Polylactides are most commonly synthesized by polymerization of the cyclic
dimer of lactic acid. Direct polymerization of the the a-hydroxy acid by
polycondensation gives low molecular weight polymer due to the difficulty of removing

water during the reaction, which is necessary to drive the equilibrium-controlled reaction

 

to completion. The polycondensation reaction is also much slower than the
polymerization of the cyclic dimer. Polymerizations of the lactide dimer often have
many of the characteristics found in living polymerizations, most importantly the ability
to control the molecular weight by simple adjustments in the monomer-initiator ratio.

Lactide can be polymerized in solution or without solvent as a monomer melt.
Melt polymerization has a number of advantages, such as a faster reaction rate than those
in solution, due to both the higher reaction temperature and the higher monomer
concentration. In addition, there is no solvent to dispose of after the reaction, an
important consideration for commercial processes. There are also some disadvantages,
such as less control over the polymer molecular weight. Because of the higher reaction
temperature, transesterification reactions are more prominent, limiting the ability to
control the molecular weight and increasing the polydispersity. If reactions are run for
extended periods of time, molecular weights quickly decrease due to intramolecular
transesterification, forming cyclic oligomers (Figure 1.7).

Solution polymerizations of lactide are commonly run in toluene, but the use of
dichloromethane, THF, and a number of other solvents have also been reported. In many
cases, the choice of solvent depends on the catalyst used for the polymerization. The
largest disadvantage to solution polymerizations is the reaction time. Reactions that take

minutes for melt polymerizations often take days.

#713!

 

 

 

 

 

1.2 Syr

availab
dimer.

esterifit
method
product
of low

catalyst
cracking
lempem

drives [1

"1850 11K
Knowin,

m0nomc

1.2 Synthesis of Lactide

There are a number of methods

 

 

available for the preparation of the cyclic 0 Y‘L
H“ O
dimer. The simplest is acid-catalyzed self- OH 0%
OH
esterification (Figure 1.2), however the 0

method preferred for commercial Figure 1.2 Synthesis of lactide via
esterification.

production of lactide is thermal cracking
of low molecular weight polymer (Figure 1.3) in the presence of a transesterification
catalyst. Lactide is distilled out as it is produced, driving the reaction to completion. The
cracking method is faster than self-esterification because of the higher reaction
temperatures used and gives higher yields because removal of the lactide as it is produced
drives the equilibrium towards the dimer.

Thermal cracking is usually accompanied by some racemization of the lactic acid
stereocenter. Three lactide diastereoisomers are possible: (R,R), (S,S), and the (R,S) or
mesa lactide. A 1:1 mixture of the (RR) and (S,S) diastereomers is termed rac-lactide.

Knowing the stereochemistry of the monomer is important because of the effect that the

monomer stereochemistry has on the resulting polymer’s physical properties. For

 

 

O
0 o
H* catalyst O
OH 0 A ’ 0
OH
X 0

Figure 1.3 Synthesis of lactide via cracking method.

 

 

 

 

exam

acid)

(PLLA
(isotact
CC dep
involve
causing
dc‘grada
plasticiz
by inclu.
Problem
amoUnt (
feed (l‘u
Wilmen‘,

p0limer.

polymer 1'.

example, poly(L-lactic acid) (PLLA) is a semicrystalline polymer, where poly(rac-lactic

acid) is amorphous.

1.3 Properties of Polylactide

1.3.1 Poly(L—lactide)

L-lactic acid is formed via fermentation of starch, and thus poly(L-lactic acid)
(PLLA) is the most common polylactide produced. Because of its regular structure
(isotactic), PLLA is semicrystalline, with Tgs ranging from 55-64 °C and Tms of 159-215
°C depending on the degree of crystallinity in the polymer.3 There is some difficulty
involved in melt processing PLLA because the Tm is near the degradation temperature,
causing the polymer to begin to thermally degrade during processing. Thermal
degradation can produce small molecules such as dimer and oligomers that act as a
plasticizer and alter the polymer’s properties. Using a less crystalline copolymer, made
by including a small amount of R,R-lactide in the polymerization, frequently solves this
problem by decreasing Tm, allowing processing at a lower temperature and decreasing the
amount of thermal degradation. Increasing the amount of R,R—lactide in the monomer
feed (functionally equivalent to increasing the proportion of rac-lactide in a
polymerization) will decrease polymer crystallinity and eventually lead to an amorphous
polymer. Polymerization of rac-lactide gives poly(D,L-1actic acid) (PDLLA). The

polymer is amorphous with reported Tgs varying from 50-57 °C.3

1.3.2 Poly(D—lactide)

Poly(D-lactide) (PDDLA) is isotactic and has identical properties to PLLA.
However, since only L-lactic acid is typically obtained from fermentations, D—lactide is
relatively rare and there are few reports of PDDLA. One interesting derivative of
PDDLA is the polylactide stereocomplex, a 1:1 mixture of PLLA and PDDLA. The
mixture co-crystallizes and melts at 230 °C, more than 30° higher than either

homopolymers.

1.3.3 Syndiotactic Polylactide

The synthesis of syndiotactic polylactide

(RSRSRSRS...) requires the stereospecific GO (Q
polymerization of mesa-lactide. Coates4 :jméqpr

produced syndiotactic PLA by polymerizing 00

mesa—lactide with an optically active catalyst

Figure 1.4 Catalyst used for the

denved from BINAP (Figure 1-4)- A“ synthesis of syndiotactic PDLLA

enantiomorphic site-control mechanism is

assumed where the catalyst selectively attacks one lactic acid residue of the R,S-lactide
monomer, producing an enantiotopic selectivity of 96%. Syndiotactic PLA is crystalline
due to the high degree of stereoregularity and has a Tg of 34 °C and Tm of 152 °C. The
catalyst used by Coates is a variation of a catalyst used by Spassky and coworkers5 for

the enantioselective polymerization of D-lactide from rac-lactide. Spassky obtained

 

 

 

88%
polyl;
high
prefer-
monor
RIC-lat

catalys

1.3.4

P013’mer
Segment
i.e. RR
refemd

fOund th
COm’erM'C
I‘bUIQjU‘dQ
poll‘lactid
lactide Ste
$6160“.ng

where B D I

 

88% enrichment in D-units at low conversion, and at high conversions produced the
polylactide stereocomplex because the removal of D-units leaves L—rich monomer pool at
high conversions. The (-) form of Spassky’s catalyst (the form used by Coates)
preferentially polymerizes D-lactide and the (+) form preferentially polymerizes the L-
monomer, a property which has been used to produce the polylactide stereocomplex from
rac-lactide using a racemic mixture of the catalyst.6 These examples show that Coates’

catalyst should preferentially attack the D side of the mesa-lactide.

1.3.4 Heterotactic Polylactide

A number of authors have reported the

polymerization of lactide to produce enriched heterotactic 0
se ments7'9 where airs of stereo enic centers alternate N\
g P g , C /Zn—OiPr
N

i.e. RRSSRRSS..., a stereochemical sequence also
referred to as disyndiotactic. Kasperczek and coworkers 0

found that heterotactic structures were preferred at low Figure 1.5 Catalyst for the

synthesis of heterotactic

conversions when rac-lactide is polymerized using lithium PL A

t—butoxide, but the overall stereoregularity of the

polylactide was poor.10 This result illustrates the tendency for syndiotactic placement of
lactide stereocenters during most lactide polymerizations. Coates recently produced a
stereoregular heterotactic PLA using the sterically hindered zinc complex (BDI)ZnOiPr,

where BDI is a 2,6-diisopropylphenyl-substituted B-diimine (Figure 1.5).ll This hindered

catalyst accentuates the chain-end control nature of the polymerization mechanism,

enabling formation of the heterotactic polymer at room temperature.

1.4 Lactide Polymerization Mechanism and Catalysts

A number of catalysts are available for the polymerization of lactide, with tin(II)
2—ethylhexanoate (commonly called stannous octoate and abbreviated as Sn(Oct)2) used
most frequently. In recent years, there has been growing understanding of the
polymerization mechanism (Figure 1.6). Most believe that a hydroxy-containing
compound, such as water or an alcohol, first reacts with Sn(Oct)2 to form a tin alkoxide,
the active catalyst, and free 2-ethylhexanoic acid. After coordinating to the monomer, the
tin alkoxide reacts by attacking the carbonyl of the lactide ring. The lactide ring opens,
creating a new alkoxide. Other than polymerization, the growing polymer chain can
exchange with free 2-ethylhexanoic acid to re-form Sn(Oct)2 and the free polymer chain.
There has also been some indication of decomposition of the catalyst to form solid

precipitates,12 which will be discussed in more detail later.

10

Oct—Sn—OR + OctH

 

Oct—Sn—Oct

degradation 4 \ o
O
Oct—Sn—Oct
Oct ,0 OR
OCIH \Sn WAC/Sr
O

+
O
0 OR
H/ \(lkO/lWK o
o \(lLO
degradation 4 \
Oct—Sn—Oct 00 0 OR
t ,
+ _ OCIH \Sn’ j/iko t’
O
HJVO\l/U\OJ\N/VOR - ‘ n+1

- n+1

 

 

 

 

 

 

 

 

Figure 1.6 Mechanism for the polymerization of lactide using Sn(Oct)2.

11

Support for this mechanism has been recently published in the literature. Using
MALDI-TOF mass spectrometry, polymerization intermediates incorporating tin were
detected by Kowalski, et al. that are consistent with the initiating species being a tin
alkoxide.l3 Kricheldorf and coworkers have shown that reacting Sn(Oct)2 with benzyl
alcohol produces the free 2-ethy1hexanoic acid, also supporting the formation of tin
alkoxides.12 A number of other tin compounds can be used for lactide polymerization
including SnCl4, SnBrz, and San. Except for the tin alkoxides, these tin compounds

require an alcohol as an initiator.

/O O

['1/ \l/kO/KK VOXKO

I S

_ r (i R 010 2 ”\w 0 O

R O R—\O

//0\n/H 0&0 O \ x/O‘<\R
O O

._ O .1” O J R

'— RT sn R\(ko jTO
/O\”)\O/
- «n a t

Figure 1.7 Mechanism for intramolecular transterification.

S

 

 

 

 

 

 

12

In addition to tin compounds, aluminum alkoxides such as Al(OiPr)3 are
commonly used catalysts for lactide polymerizations. These reactions typically use
toluene as a solvent because more polar solvents decrease the rate of reaction,14 most
likely by competing with the monomer for coordination to the catalyst. There are some
advantages to solution polymerizations. Because they are usually run at lower
temperatures than melt polymerizations, there is less transesterification, and therefore
greater control over molecular weights. Solution polymerizations using Al(OiPr)3 are
reported to be living polymerizations.” The proposed mechanism shown in Figure 1.8 is
similar to that of tin-based catalysts, except there is no need to form an active catalyst
before polymerization can begin.

The number of isopropoxide groups that initiate chains varies with different
monomers. For example, only one isopropoxide group is active in e-caprolactone
polymerizations run at 0 °C. At 100 °C, the average number of active isopropoxides per
aluminum is 1.4. When lactide is polymerized at 70 °C, all three isopropoxides initiate
polymerization. Catalysis using aluminum isopropoxide is further complicated because it
exists in two forms, a trimer and tetramer. The trimer is more active in initiating
polymerization,l6 most likely due to reduced steric hindrance. It is believed that the
aggregates dissociate after initiation of polymerization.'7

Because the alkoxy group from the catalyst becomes the polymer end group,
changing that group will change the end group. This led to the use of catalysts with the

general form EtzAlOR, where the alkoxide group becomes the polymer end group and

can be varied to obtain polymers with different end groups. The ethyl groups are inactive

l3

 

 

O
O

 

0
F10\ ,

”k0 O “op
R0 1 l’

 

‘ n+1

‘Oi Pr

 

n+1

Figure 1.8 Synthetic scheme for the polymerization of lactide using Al(OiPr)3.

14

and do not initiate polymerization. Examples of such initiators include EtzAlOCHzBr,18
which has been used in the formation of star-branched polymers, and
EtzAlO(CH2)2CH=CH2,'9 and EtzAlOCHzOC(O)C(Me)=CH2,2° which leave a
polymerizable group at the end of the PLA chain, allowing it to be used as a
macromonomer for the formation of comb polymers.
Another class of aluminum catalysts
used for lactide polymerizations are aluminum
porphyrins,”23 commonly TPP-Al-OMe (TPP =
tetraphenylporphyrin). The advantage of these
catalysts is that the polymerizations are

“immortal,” which is related to living

 

polymerrzatrons. Immortal polymerizations Figure 19 Aluminum porphyrin

. . . . . catal st
are especrally resrstant to terrmnatron, even in y

the presence of water or alcohols, which commonly terminate lactide polymerizations or
cause chain transfer, decreasing molecular weights. Trofimoff reported that a lactide
polymerization at room temperature gave an Mn of 16400 with a PDI of 1.12. No rate

constant was given, but it was stated that the polymerization reached 94% conversion

21 21-23

after 96 hours. Aluminum porphyrin polymerizations typically use CH2C12 since
the higher solubility of the monomer in this solvent allows the use of lower reaction
temperatures.

Another important class of catalysts is based on lanthanides. Yttrium is one of the

most popular metals, and a variety of yttrium alkoxide species have been used to

15

polymerize lactide including Y5(O)(OiPr)13,24 Y(OCH2CH2NMe2)3,25'26 and Y(OAr)3,
where Am 2,6-di-tert--butylphenyl.27 Polymerizations using these catalysts are reported
to be much faster than aluminum-catalyzed polymerizations. Y(OAr)3 requires an
alcohol cocatalyst, such as iPrOH. An advantage of this system compared to
Y5(O)(OiPr).3 is that mononuclear Y(OiPr)3 is formed instead of the Y5(O)(OiPr)i3
cluster. Because of the lower steric hindrance in Y(OiPr)3, reaction times are reduced,
and there is greater control over molecular weight because all isopropoxide groups are
active toward polymerization, which is not the case for the cluster form. Polymerizations
using the cluster catalyst takes 10 hours to reach complete conversion, compared to less
than five minutes for Y(OiPr)3.27 Y(OiPr)3 also gives a polymer closer to the molecular
weight expected from the monomer to catalyst ratios. For example, a lactide
polymerizations using Y(OiPr)3 as the catalyst gave a Mlrl = 10,800 with a PDI of 1.14,
compared to the theoretical value of Mn = 7200. A lactide polymerizations using the
Y5(O)(OiPr).3 cluster gave Mn = 29300 and a PDI of 1.71, compared to 7600, the
expected Mn assuming each isopropoxide group initiates a polymer chain. The rate
constant for lactide polymerization catalyzed by Y(OAr)3 is 0.076 L mol'l min'1 at room
temperature with an initiator concentration of 5.8 mM and 0.22 L mol’l min’1 with an
initiator ratio of 0.58 mM.27

The Y(OCH2CH2NMe2)3 catalyst is among the fastest known in the literature,
with a rate constant of 30 min", however some difficulty is observed in controlling the
reaction due to slow initiation.25 This causes the plot of molecular weight vs. monomer/
initiation ratio to be curved and increases the average number of polymer chains formed

per initiator molecule.

16

 

Figure 1.10 Tolman’s yttrium catalysts.

Recently, Tolman et al. developed and evaluated new yttrium catalysts (Figure
1.10).28 The greatest difficulty encountered with these two catalysts was control over the
molecular weight as no control over the molecular weight was observed by changing the
monomer to catalyst ratio. One advantage of these catalysts is that structural variations
can be made to the catalyst to alter the polymerization rates and polymer molecular

weights.28

1.5 Equilibrium Polymerizations

Many polymerizations reach equilibrium concentrations of monomer and polymer
rather than 100 % conversion of monomer to polymer. In polymerizations, this can make
it difficult or impossible to reach high molecular weights, especially under conditions
such as high dilution which favor the reactants over polymer formation. An important

C).29‘30 At Tc there is no

concept related to this equilibrium is the ceiling temperature (T
progress in the polymerization because the rate of depolymerization is equal to the rate of

polymerization. Because TC will vary for different polymerization conditions (i.e.

monomer concentration, solvent, pressure, etc.), standard conditions for reporting ceiling

l7

temperatures are 1 atm. pressure and either pure monomer (solvent-free polymerizations)
or a 1 M solution of monomer.

As the monomer concentration decreases, the ceiling temperature also decreases
because the initial monomer concentration is closer to the equilibrium monomer
concentration. The ceiling temperature can be calculated using the relationship that at
that the ceiling temperature, an equilibrium is established and the free energy equals zero:

AG=AH~TCAS =0, ( l- 1)
and
TC=AH/AS. (1-2)
Expressing the relationship in terms of an equilibrium constant gives
AG: -RT 1n =RT ln [M]eq, (1-3)
which can be rearranged to give the equation:
TczAll/(AS + R In [M]¢q). (1-4)
The effect of monomer concentration on ceiling temperature can be seen in figure 1.11
for the polymerization of a-methylstyrene in methylcyclohexane.3O At temperatures
greater than the Tc line, only monomer is present, while below Tc there is a polymer-
monomer equilibrium. The bottom curve shows the upper limit of monomer solubility.
Use of a better solvent, such as tetrahydrofuran, gives nearly the same polymerization
line, but lacks the region where the monomer is not soluble in the solvent.30

The equilibrium monomer concentration, [M]eq, varies with the solvent. For
example, [M]eq for the polymerization of trioxane at 30 °C in benzene, 1,2-
dichloroethane, and nitrobenzene, is 0.05 M, 0.13 M, and 0.19 M, respectively.31 The

concentration of dissolved polymer in the solvent can also affect [M]eq. For the cationic

18

 

     
 

 

 

 

300
£- 290.
”j ‘ monomer
% . monomer!
l— .
g J polymer equrl.
LlJ 280
0- I
2
“J 1
l— \
270. »
l D
l
‘ El
260 .

0.00 0055 0'10 #035 020 0.125 0'30 -035 0.40
MOLE FRACTION INITIAL MONOMER

Figure 1.11 The effect of monomer concentration on ceiling temperature

19

 

 

 

polyme
contain

propon

decreast
from it
polymer

The effe

T
less freql
the free I
materials
which ha
positive.
at [empm
Wil‘meriz

E.

 

dePOIImt
abOI'e [hi
dEgrade :

monomer-

 

polymerization of 1,3-dioxolane in CH2C12 at 60°C, [M]cq falls from 3.7 M for a solution
containing a small amount of polymer to 2.8 M for a solution containing a large
proportion of polymer.32

Pressure also can alter the ceiling temperature. Since increased pressure
decreases the initial entropy, pressure decreases the overall change in entropy in going
from monomer to polymer and increases Tc. The ceiling temperature for the
polymerization of tetrahydrofuran changes from 81°C at 1 atm. to 129°C at 2500 atm.33
The effect of pressure is given by

d In Tc/szAV/AH. (1-5)

The floor temperature, a condition similar to ceiling temperatures is encountered
less frequently. Instead of polymerization taking place below the temperature at which
the free energy equals zero, as in ceiling temperatures, floor temperatures are found for
materials that only polymerize above that temperature. Unlike most polymerizations
which have a negative AH, the AH for polymerizations with floor temperatures is
positive. For polymerization to occur, -TAS must be greater than AH, which is true only
at temperatures greater than the floor temperature. A common example is in the
polymerization of 83, which only polymerizes above 159°C.30

Establishment of an equilibrium in a polymer requires that the polymerization and
depolymerization reactions be kinetically accessible. Many polymers are useful materials
above their ceiling temperature because they are kinetically stable and usually will not
degrade to monomer. Many polymers, such as polyesters, only depolymerize to the

monomer in the presence of a catalyst, and not all polymers return to the monomer. For

20

example, polycaprolactone depolymerizes to cyclic oligomers from two to six repeat
units instead of 8—caprolactone, with the major fraction being the dimer.34
An advantage made available by a monomer-polymer equilibrium is the ability to
recycle the polymer to monomer via depolymerization. Ring-opening polymerizations
are frequently reversible, the exception being radical reactions. Polyesters are easily
depolymerized by intramolecular transesterification (also called backbiting reactions).
The catalyst used can affect the amount of transesterification, with catalysts such as
EtzAlOMe used to retard the formation of cyclic oligomers, and tin compounds or alkali
metals used to promote depolymerization.35
Lactide polymerizations are known to be reversible, and as described earlier, the
lactide dimer is frequently produced via depolymerization of low molecular weight
polymer. Another equilibrium effect in lactide polymerizations is unreacted monomer
present after solution36 and melt polymerization.” The calculation of the Tc is often
based on measurements of [M]eq at a number of temperatures. From this data, AH and AS
can be calculated, and since TC=AH/AS (equation 1-2), the ceiling temperature can then
be calculated. Duda calculated a TC of 641°C for melt polymerization of lactide by
extrapolating from the data obtained from solution polymerizations in dioxane. AH was —
22.9 kJ/mol and AS was —41.1 J/mol K36 The equation
[M].q = ltd/k. = UK... (MS)
can be rewritten as
[M]¢q = exp(AH/RT — AS/R). (1-7)
Using non-linear regression, Witzke used equation (1-7) to calculate the AH (—23.3

kJ/mol) and as (-220 J/mol to.” Using equation 1.2. this gives a Tc of 786 i 87°C.

21

 

 

 

 

 

 

 

Unfortunately the Tc cannot be measured directly because it is above the thermal

degradation temperature of polylactide.

1.6 Degradation of Polylactides

Several methods are used for measuring polymer biodegradation. Many are based
on changes in mechanical properties, morphology, and the chemical structure of the
polymer. Microbial growth, measured by the uptake of oxygen by microorganisms or the
evolution of carbon dioxide, has also been used. Various physical properties of
degradable polymers have been studied to find correlations between polymer structure
and degradation rates. Important properties include the glass transition temperature,
crystallinity, initial molecular weight, hydrophilic/hydrophobic interactions, sample size,
morphology, the availability of hydrolyzable groups, and surface area.

There are two limiting cases for the degradation of polyesters, surface erosion, and
bulk erosion. In the first, the surface layers of a polymer sample are degraded to low
molecular species by hydrolysis or enzymatic degradation, and as the molecular weight
of the surface layer decreases, the layer is lost and a fresh surface is exposed. In this
scheme, polymer samples slowly shrink in size as the samples degrade, while the overall
molecular weight of the sample remains nearly constant. In the second scheme, the
primary degradation reaction is the acid-catalyzed hydrolysis of polyester chains. The
chain ends of polylactides often are carboxylic acids. As water diffuses into the sample,
chain ends catalyze hydrolysis reactions which form an alcohol and a carboxylic acid.

Since the local concentration of acid groups increases, the degradation rate of polyesters

22

often increases because the acid groups formed catalyze further hydrolysis
(autocatalysis). Polymers that degrade by this mechanism show uniform decreases in
molecular weight, but little weight loss until the molecular weight reaches the point
where the polymer chains are too short to provide the needed mechanical properties.
Such samples disintegrate when the mechanical properties fall below a critical level.

Polymers degrade faster when there are at temperatures >Tg. Above Tg, the
polymer chains are more flexible, increasing the diffusion rate of small molecules, such
as water, and thus increasing the rate of hydrolysis in the interior of polymer samples.
Increased polymer chain flexibility also promotes enzymatic degradation at the sample
surface since the polymer chains can better adjust to an enzyme’s active site, increasing
the rate of biodegradation as well. The effect of Tg on the polymer degradation rate can
be seen in the temperature dependent degradation of poly(glycolide-c0-lactide) reported
by Reed and Gilding.38 They found that a plot of degradation (measured by tensile
strength) vs. temperature shows a discontinuity at Tg.

Crystalline regions of polymers are inflexible and ordered. Diffusion of small
molecules through a crystalline region generally is slow due to the high density of
crystalline phases. Not surprisingly, the amorphous regions in semicrystalline polymers
tend to degrade first, because their greater free volume enables diffusion of water into the
polymer to support hydrolysis, and the polymer chains in amorphous regions have the
flexibility to conform to an enzyme’s active site for biodegradation. Therefore, the
higher the crystallinity of a polymer, the lower the degradation rate.”40 Although PLLA

l

is completely degradable, highly crystalline samples can take years to degrade.4 Since

23

crystallites are tie together by amorphous regions, hydrolysis of the amorphous regions
leads to mechanical failure at low weight loss.

Other physical properties can also affect the sample degradation rate. Increasing
the surface area of the sample increases the degradation rate by exposing more polymer
exposed to either the water (hydrolysis) or enzymes (biodegradation). However, if the
degradation mechanism is by hydrolysis, large samples often degrade faster than small
samples due to autocatalysis of hydrolysis by the carboxyl groups that form as the
polymer chain is hydrolyzed.42 The initial molecular weight affects degradation by
increasing the number of chain scissions that must take place before the polymer reaches
the point where it loses mechanical stability. In addition, the number of carboxylic acid
chain ends available to catalyze the degradation is related to the molecular weight. The
diffusion rate of water into hydrophobic materials may be reduced and result in a slower
degradation rate.

In addition to the physical properties of the polymer sample, degradation rates are
strongly affected by degradation conditions. Important parameters include pH, ionic
strength, temperature, and for the biodegradation of medical implants, the implantation
site.43 Both high and low pH increase the degradation rate, since both acidic and basic
conditions catalyze hydrolysis. For the case of implants, it has been shown that certain
implant sites put a greater mechanical stress on the implant, giving them a higher

degradation rate.44

24

1.7 Methods for Controlling and Varying Physical Properties

1.7.1 Copolymerization

Polylactides have Tgs near 50 °C and melting transitions at 180 °C. From the
previous sections, it is clear that having control over the chemical structure of
polylactides could provide great flexibility in the physical properties of polylactides, and
their degradation behavior. The most common method for controlling and varying
physical properties is copolymerization. A number of monomers copolymerize with

51,52 and a

lactide, such as caprolactone, glycolidefs’47 carbonates,48 PEO,49'50 isocyanates,
lactic acid/lysine cyclic monomer53’54. The physical properties of random copolymers fall
between those of the parent homopolymers, and depend on the molar composition of the
copolymer. Copolymers frequently degrade faster than the homopolymers due a decrease
in crystallinity. For example the 50:50 copolymer poly(L-lactide-co-glycolide) degrades
faster than both PLLA and PGA.55 Both homopolymers are semicrystalline, but the
copolymer is completely amorphous.

In addition, polymers with alcohol end groups can be used as an initiator to form
block copolymers, which is frequently done with poly(ethylene glycol).50’56'S7 Block
copolymers are formed by sequential additions of monomer to a living polymerization.
These polymerizations can be carried out using catalysts such as AI(OiPr)3, however the
order in which the monomers are polymerized is sometimes limited. For example,

poly(caprolactone) can initiate lactide polymerization, but not vice versa.58 Comb block

copolymers can be formed through the polymerization of macromonomers. The use of

25

initiators like EtzAlO(CH2)2CH=CH2l9 places a polymerizable end group on the polymer
chain, that can be polymerized in a later step. The polymer chains in block copolymers
phase separate and thus have segments with properties of the homopolymers. Block
copolymers have two Tgs corresponding to the parent homopolymers.

Many block copolymers are useful for the formation of micelles and
microspheres. For example, a block copolymer of PLA and PEG forms a micelle with a
PEG exterior. This is useful in drug delivery systems because PEG inhibits protein

adhesion.59

1.7.2 Blends

Blending two or more polymers can give polymer mixtures with unique physical
properties. This method is rather limited since polymers usually phase separate instead of
forming homogeneous mixtures because the entropy of mixing for polymers is low.
Blending causes the rates of degradation of the paired polymers to become more similar.
For example, in a blend of poly(glycolide-co-lactide) and polycaprolactone, the rate of
degradation of the PGA/LA slows and that of the PCL increases to nearly match the new
rate of PGA/LA degradation.60 This is true even though DSC measurements show the
two polymers were phase-separated. The mechanism by which blending influences the
degradation rate is unknown, but it is believed to be related to changes in the morphology
and water content of the blend.“ For PEO/PLA blends, the degradation rate increases

with increasing amounts of PEO."2 The two reasons for this behavior are the decreases in

26

Tg and increases the hydrophilicity caused by PEO. The Tgs of the blends are

intermediate between the Tgs of PEO and PLLA.

1.7.3 Additives

Another method for changing polymer properties is the use of additives, such as
plasticizers. Additives are commonly used materials such as poly(vinyl chloride) (PVC).
PVC is a strong, but stiff and inflexible polymer, however, the addition of a plasticizer
(commonly dioctyl phthalate) makes PVC more flexible. Unplasticized PVC is
commonly used to form pipes used for plumbing (Tg = 80 °C), while plasticized PVC
(often called “vinyl”) has a Tg near 20 °C and is used for upholstery and car dashboards.63
Plasticizers make polymers more flexible by decreasing the Tg, reducing crystallinity or
decreasing the melting temperature. For PLA, common plasiticizers include unreacted
monomer, whether left present deliberately or because of an incomplete reaction, and
oligomeric PLA.“

In the case of degradable polymers, additives can also be added to change the
degradation rate. For polyesters, acids will catalyze hydrolysis, increasing the
degradation rate. Additives that decrease the Tg also increase the rate of degradation by
acting as plasticizers. Arrrino compounds have been shown to catalyze hydrolysis,“
however they also neutralize carboxyl end groups, and thus decrease the degradation
rate.2 This accounts for some of the contradictions in the literature about the role of a

number of additives. Because a number of drugs, such as methadone — a narcotic

analgesic, contain amino groups, the plasticizing effect of these compounds is especially

27

important when investigating materials for drug delivery systems. The increase in the
degradation rate depends primarily on the nucleophilicity of the amine and not the pKa or

its concentration in the polymer.65

1.7.4 Substituted Lactic Acids

A more direct route to control over the physical properties of polylactides is the
use of substituted lactides as monomer. In terms of physical properties, having only a
methyl group pendant to the polymer backbone is limiting, and to obtain a broader range
of physical properties for polyglycolides we can consider simple analogies based on
polyolefins. Relatively minor changes in the substituents attached to the polymethylene
chain of polyolefins drastically alter their physical properties, and thus analogous changes
to the polylactide backbone should provide degradable polymers with a broad range of
physical properties. For example, polystyrene has an aromatic ring directly attached to
the polymer backbone and because the steric hindrance of the benzene ring limits the
flexibility of the polymer backbone, polystyrene has a T3 of 109 °C. Having a
polylactide with a Tg >100 °C would enable the use of polylactides use in applications
such as disposable packaging.

Very few substituted poly(lactic acid)s are known in the literature, and those
reported usually are used in copolymers with lactide. Known poly(substituted lactic
acid)s include poly(phenyllactic acid) and poly(mandelic acid) (Figure 1.12). Although
mandelic acid is not really a substituted lactic acid, it is included in this list because of its

similarity in reactivity and polymerization mechanism. Phenyllactic acid has been

28

polymerized previously by direct polycondensation of the or-hydroxy acid. This method
only produces low molecular weight materials, and very little information is known about
the polymer other than it is degraded by chymotrypsin.66 L-Phenyllactic acid (L-PhLA)
has also been copolymerized with lactic acid by the same method and the copolymer
proven to be biodegradable by implantation in rats. The copolymers had a low molecular
weight (Mn = 3260, and Mw = 10400), and a T8 of 47 °C. A monomer feed ratio of 70
mol % LA to 30 mol % L-PhLA resulted in a copolymer composition of 65 mol % LA to
35 mol % L-PhLA.

Several groups explored the preparation of the mandelic acid homopolymer, but
the schemes reported to date resulted in polymers with number average molecular
weights <5000, too low for most practical applications. Probably due to their low
molecular weights, no glass transition temperatures were reported for these polymers.
Mandelic acid also has been reported to have been copolymerized with lactide by first
synthesizing the dimer, mandelide.67 No physical data was reported on the polymer

molecular weights, degradability, or other physical properties.

 

 

 

 

r ‘1
O O
\
\ I/
o/ o‘
4 n L - l n
polyphenyllactide polymandelide
poly(phenyllactic acid) poly(mandelic acid

Figure 1.12 Structure of poly(phenyllactic acid) and poly(mandelic acid)

29

Literature data for substituted PHBs can be used to predict how the properties of
polylactides might change with substitution. For alkyl-substituted
poly(hydroxyalkanoates), the crystallinity and glass transition temperature decreases as
the length of the sidechain increases. For example, poly(B-hydroxyvalerate) has a Tg of
—15 °C, but poly(B-hydroxynonate), which has four more carbons in its side—chain, has a

Tg of —37 °C.“ Slow crystallization rates are reported for long side-chain PHBs.43

1.8 Goal of the Project

The purpose of this project is to synthesize a variety of aromatic substituted
lactides in order to form new polymers and investigate the physical properties of those
polymers. Aromatic substituents were chosen in the hope of developing a biodegradable
substitute for polystyrene with a higher Tg than polylactide. Since a polymer softens and
loses mechanical stablity at the glass transition, the Tg defines the upper limit for
applications. This study will include polymers prepared from phenyllactide, mandelide,

and methylphenyllactide (Figure 1.13).

MO 0 m

phenyllactide mandelide methybhenyllactide

Figure 1.13 Aromatic substituted lactides.

30

1.8 Works Cited

1)

2)

3)

4)

5)

6)

7)

8)

9)

10)

11)

12)

13)

Chiellini, E.; Solaro, R. Adv. Mater. 1996, 8, 305-313.

Li, S.; Vert, M. In Degradable Polymers; Scott, G., Gilead, D., Eds.; Chapman &
Hall: London, 1995; pp 43-87.

Starkweather, H. W.; Avakian, P.; Fontanella, J. 1.; Wintersgill, M. C.
Macromolecules 1993, 26, 5084-5087.

Ovitt, T. M.; Coates, G. W. J. Am. Chem. Soc. 1999, 121, 4072-4073.

Spassky, N.; Wisniewski, M.; Pluta, C.; LeBorgne, A. Macromolecular Chemistry
and Physics 1996, 197, 2627-2637.

Radano, C. R; Baker, G. L.; Smith HI, M. R. J. Am. Chem. Soc. 2000, 122, 1552-
1553.

Kasperczyk, J. E. Macromolecules 1995, 28, 3937-3939.

Wisniewski, M.; LeBorgne, A.; Spassky, N. Macromolecular Chemistry and
Physics 1997, 198, 1227-1238.

Kasperczyk, J.; Bero, M. Polymer 2000, 41, 391-395.

Bero, M.; Dobrzynski, P.; Kasperczek, J. J. Polym. Sci, Pt. A: Polym. Chem.
1999, 37, 4038-4042.

Cheng, M.; Attygalle, A. B.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem. Soc.
1999, 121, 11583-11584.

Kricheldorf, H. R.; Kreiser-Saunders, 1.; Stricker, A. Macromolecules 2000, 33,
702-709.

Kowalski, A.; Duda, A.; Penczek, S. Macromolecules 2000, 33, 689-695.

31

14)

15)

16)

17)

18)

19)

20)

21)

22)

23)

24)

25)

26)

Ropson, N.; Dubois, P.; Jerome, R.; Teyssie, P. Macromolecules 1995, 28, 7589-
7598.

Dubois, P.; Jacobs, C.; Jerome, R.; Teyssie, P. Macromolecules 1991, 24, 2266-
2270.

Duda, A.; Penczek, S. Macromolecules 1995, 28, 5981-5992.

Ropson, N.; Dubois, P.; Jerome, R.; Teyssie, P. Macromolecules 1994, 27, 5950-
5956.

Tian, D.; Dubois, P.; Jerome, R.; Teyssie, P. Macromolecules 1994, 27, 4134-
4144.

Dubois, P.; Jerome, K; Teyssie, P. Polymer Bulletin 1989, 22, 475-482.

Dubois, P.; Jerome, R.; Teyssie, P. Makromolekulare Chemie-Macromolecular
Symposia 1991, 42-3, 103-116.

Trofimoff, 1L.; Aida, T.; Inoue, S. Chemistry Letters 1987, 991-994.

Aida, T.; Inoue, S. Acc. Chem. Res. 1996, 29, 39-48.

Endo, M.; Aida, T.; Inoue, S. Macromolecules 1987, 20, 2982-2988.

Simic, V.; Girardon, V.; Spassky, N.; Hubert-Pfalzgraf, L. G.; Duda, A. Polymer
Degradation and Stability 1998, 59, 227-229.

McLain, S. 1.; Ford, T. M.; Drysdale, N. E.; Jones, N.; McCord, E. F.; Shreeve, J.
L.; Evans, W. 1. Abstracts of Papers of the American Chemical Society 1994, 208,
360-POLY.

McLain, S. 1.; Ford, T. M.; Drysdale, N. E. Abstracts of Papers of the American

Chemical Society 1992, 204, 188-POLY.

32

27)

28)

29)
30)
3 1)

32)

33)

34)

35)
36)
37)
38)

39)

40)
41)

42)

Stevels, W. M.; Ankone, M. J. K.; Dijkstra, P. J .; Feijen, J. Macromolecules 1996,
29, 6132-6138.

Chamberlain, B. M.; Sun, Y.; Hagadorn, J. R.; Jemmesch, E. W.; Young Jr., V.
G.; Pink, M.; Hillmyer, M. A.; Tolman, W. B. Macromolecules 1999, 32, 2400-
2402.

Ivin, K. J. Makromol. Chem, Macromol. Symp. 1991, 42/43, 1-14.

Greer, S. C. J. Phys. Chem. B 1998, 102, 5413-5422.

Miki, T.; Higashimura, T.; Okamura, S. J. Polym. Sci, A-I 1970, 8, 157.

Kizub, L. I.; Markevich, M. A.; Berlin, A. A.; Yenikolopyan, N. S. Polymer Sci.
USSR 1968, 10, 2332.

Rahman, M.; Weale, K. E. Polymer 1970, 11, 122.

Nelissen, M.; Keul, I-I.; Hocker, H. Macromol. Chem. Phys. 1995, 196, 1645-
1661.

Hocker, H.; Keul, H. Macromol. Symp. 1995, 98, 825-834.

Duda, A.; Penczek, S. Macromolecules 1990, 23, 1636-1639.

Witzke, D. R.; Narayan, R.; Kolstad, J. J. Macromolecules 1997, 30, 7075-7085.
Reed, A. M.; Gilding, D. K. Polymer 1981, 22, 494-498.

Kanesawa, Y.; Tanahashi, N.; Doi, Y.; Saito, T. Polym. Degrad. Stab. 1994, 45,
179.

Kumagai, Y.; Doi, Y. Makromol. Chem. 1992, 193, 53.

Li, S. M.; Garreau, H.; Vert, M. J. Mater. Sci. Mater. Med. 1990, 1, 198-206.

Grizzi, 1.; Garreau, H.; Li, S.; Vert, M. Biomaterials 1995, 16, 305-311.

33

43)

44)

45)

46)

47)

48)

49)

50)

51)

52)

53)

54)

55)

Hammond, T.; Liggat, J. J. In Degradable Polymers; Scott, (3., Gilead, D., Eds;
Chapman & Hall: London, 1995; pp 88-111.

Therin, M.; Christel, R; Li, S. M. Biomaterials 1992, 13, 594-600.

Joziasse, C. A. P.; Veenstra, H.; Topp, M. D. C.; Grijpma, D. W.; Pennings, A. J.
Polymer 1998, 39, 467-474.

Grijpma, D. W.; Nijenhuis, A. J.; Pennings, A. J. Polymer 1990, 31, 2201-2206.
Grijpma, D. W.; Pennings, A. J. Polymer Bulletin 1991, 25, 335-341.

Schmidt, P.; Keul, H.; Hooker, H. Macromolecules 1996, 29, 3674-3680.

Li, S. M.; Rashkov, 1.; Espartero, J. L.; Manolova, N.; Vert, M. Macromolecules
1996, 29, 57-62.

Rashkov, I.; Manolova, N.; Li, S. M.; Espartero, J. L.; Vert, M. Macromolecules
1996, 29, 50-56.

Harkonen, M.; Hiltunen, K.; Malin, M.; Seppala, J. V. Journal of
Macromolecular Science-Pure and Applied Chemistry 1995, A32, 857-862.
Kylma, J.; Harkonen, M.; Seppala, J. V. Journal of Applied Polymer Science
1997, 63, 1865-1872.

Barrera, D. A.; Zylstra, E; Lansbury, P. T.; Langer, R. J. Am. Chem. Soc. 1993,
115, 11010.

Barrera, D. A.; Zylstra, E.; Lansbury, P. T.; Langer, R. Macromolecules 1995, 28,
425-432.

Miller, R. A.; Brady, J. M.; Cutright, D. E. J. Biomed. Mater. Res. 1977, 11, 711-

719.

34

56)

57)

58)

59)

60)

61)

62)

63)

64)

65)

66)

67)

68)

Jin, X. M.; Carfagna, C.; Nicolais, L.; Lanzetta, R. Macromolecules 1995, 28,
4785-4794.

Li, Y. X.; Kissel, T. Journal of Controlled Release 1993, 27, 247-257.

Jacobs, C.; Dubois, P.; Jerome, R.; Teyssie, P. Macromolecules 1991, 24, 3027-
3034.

Shields, R.; Harris, J .; Davis, M. Gastroenterology 1968, 54.

Cha, Y.; Pitt, C. G. Biomaterials 1990, 11, 108.

Pitt, C. G. In Biodegradable Polymers and Plastics; Vert, M., Feijen, J .,
Albertsson, A., Scott, G., Chiellini, E., Eds.; The Royal Society of Chemistry:
Cambridge, 1992; pp 7-17.

Nijenhuis, A. J .; Colstee, E.; Grijpma, D. W.; Pennings, A. J. Polymer 1996, 37,
5849-5857.

Sperling, L. H. Introduction to Physical Polymer Science; John Wiley & Sons:
New York, 1992.

Sinclair, R. G. Journal of Macromolecular Science-Pure and Applied Chemistry
1996, A33, 585-597.

Cha, Y.; Pitt, C. G. J. Controlled Release 1988, 7, 69.

Tabushi, 1.; Yamada, H.; Matsuzaki, H.; Furukawa, J. J. Polym. Sci, Polym. Lett.
Ed. 1975, 13, 447-450.

Kimura, Y. In Chem. Abst. 1993, 118, 255577r: Japan, 1993.

Curley, J. M.; Hazer, B.; Lenz, R. W.; Fuller, R. C. Macromolecules 1996, 29,

1762-1766.

35

Chapter 2 — Phenyllactide

2.1 Introduction

Polylactide is a versatile polymer whose properties can be modified through
control of the polymer stereochemistry and copolymerization. However, its low glass
transition temperature (25-50 °C) makes it unsuitable for applications where clear, glassy
materials are needed. Polymers such poly(methyl methacrylate) and polystyrene
currently are used for such applications, but neither is degradable nor derived from
natural sources. A degradable analog of polystyrene would be particularly attractive
since it could have important applications as disposable food containers and packaging.
Simple structural analogies to polystyrene can be made to try to identify the structural
modifications needed to convert polylactide to a material that has a Tg similar to
polystyrene’s 109 °C. As shown in Figure 2.1, the lactide and polystyrene systems differ
in the nature of their backbone and in the presence of an aromatic ring in the polystyrene
structure. Two structural analogs of polystyrene, poly(phenyllactic acid) and poly
(mandelic acid) are reasonable candidates. SciPolymer, a commercial software package,
predicts that the Tg of poly(phenyllactic acid) should be near 60 °C, while that of
poly(mandelic acid) should be near 100 °C. Both values are higher than that of
polylactide, and we set out to prepare the polymers to evaluate their potential to serve as
degradable analogs of polystyrene. In this chapter, we focus on the preparation and

characterization of poly(phenyllactic acid).

36

\

 

 

O O
\ O
o o/
l O
- n
polystyrene _ .

poly(phenyllactic acid)

 

 

I

 

 

O O o l o
O
O O
n
polylactide O O
r . n

poly(mandelic acid)

Figure 2.] Structures of polystyrene, polylactide, and the lactide-based
analogs of polystyrene, poly(phenyllactic acid) and poly(mandelic acid).

37

There are few reports of polymers based on phenyllactic acid. All known
examples of poly(phenyllactic acid) were formed by the direct polycondensation of the
a-hydroxyacid without a catalyst, a method that usually produces low molecular weight
materials (Mn<10,000).l Little information was published on the physical properties of
poly(phenyllactic acid), however, the polymers were shown to be biodegradable."2
C0polymerizations involving phenyllactic acid are more common, especially those using
lactic acid as the comonomer."2 A poly(lactic acid-co-phenyllactic acid) copolymer with
a 65% lactic acid content had a Tg of 47 °C.1 Aromatic rings have also been added to the
structurally similar poly(hydroxyalkanoate)s to alter their physical properties. As shown
in Figure 2.2, adding an aromatic ring to poly(hydroxyvalerate) increased the Tg from —
15 °C for the unsubstituted polymer to 13 °C for the aryl substituted polymer.3 The
polymer, poly(3-hydroxy-5-phenylvalerate) was produced biosynthetically by

Psuedomonas oleovorans from 5-phenylvaleric acid and had a molecular weight (M,,) of

350,000.

11°11
, o o
o o/

 

 

 

 

)/
, o 0/
l n _ - n
ol 3-h drox alerate
P Y( T911: 45ch ) poly(3-hydroxy-5-phenylvalerate)

Tg=13°C

Figure 2.2 The effect of an aromatic substituent on Tg.

38

A practical synthesis of high molecular weight poly(phenyllactic acid) will likely
mirror that of polylactide, and involve the ring opening polymerization of phenyllactide,
the phenyllactic acid dimer. As described below, the dimer is readily prepared using the
same route used to prepare lactide. Phenyllactic acid, the needed starting material, is
commercially available, but expensive (approximately $12/g from Aldrich). However,
Frost has shown that the biosynthetic pathway for the synthesis of phenylalanine (Scheme
2.1) can be harnessed for the bacterial production of phenyllactic acid.4 The normal
pathway leads from glucose to phenylpyruvic acid, which is converted by a transaminase
into the amino acid. Interception of the pyruvate and reduction of the carbonyl produces
phenyllactic acid from phenylpyruvic acid. Naturally occurring phenyllactic acid is
racemic, although it should be possible to direct the synthesis of a particular phenyllactic
acid enantiomer. Thus, fermentation may prove to be a low-cost source of phenyllactic

acid.

2.2 Monomer Synthesis

Because the polymerization of phenyllactic acid produces low molecular weight
material, we elected to prepare poly(phenyllactic acid) by ring-opening polymerization of
the phenyllactic acid dimer. The easiest method for forming the dimer is acid-catalyzed
self-esterification, removing the water as it is formed by azeotropic distillation. This

method gives a mixture of the cyclic dimer, along with low molecular weight linear

39

COzH -
COO

—_., NH+
glucose —> 0 L 33' 3
0 COzH
on

chorismic acid anthranilic acid
coon
o
HOZCZG' COzH
do“ m
:OH OH
prephenic acid hydroxyphenyl
l pyruvic acid
002 COzH C02H
NH: 0 OH
L - phenylalanine phenylpyruvic acid phenyllactic acid

Scheme 2.1 Biosynthetic pathway for phenyllactic acid

40

oligomers. The dimerization reaction is run in dilute solution to favor production of the
dimer over linear oligomers. This unfortunately causes the reaction to be rather slow,
taking about one week to reach 90% conversion. Despite the slow rate of dimer
formation, one advantage of this reaction is that there is no racemization of the
stereocenters, allowing the formation of both the L,L- and D,D-stereoisomers simply by
starting with the corresponding phenyllactic acid. The use of racemic phenyllactic acid
as the starting material resulted in a mixture of all possible isomers (D,D; L,L; and D,L).

The product mixture (dimer vs. linear products) depended on the amount of p-
toluenesulfonic acid used to catalyze the reaction. Increasing the amount of catalyst
increased the rate of conversion, but had little effect on the absolute yield of dimer
(typically near 40%) and increased the amount of oligomer formed. Interestingly,
catalyzing the reaction with 4A molecular sieves in place of p-toluenesulfonic acid gave
higher conversions but the dimer yield was <10 %, with the remaining product being
linear oligomers.

An alternative route to phenyllactide is the thermal cracking of low molecular
weight polymer (Scheme 2.2). This method is widely used to prepare lactide. Acid-
catalyzed condensation of phenyllactic acid at 120 °C without solvent produces low
molecular weight polymer in high yields. When the unpurified polymer is heated under
vacuum with ZnO, the dimer can be isolated by distillation. This method is faster and

gives a higher yield (80 to 90%), but unfortunately the lower volatility of phenyllactide

compared to lactide necessitates high temperatures (230 °C) during the cracking step, and

41

HO..,. +
._ OH _H_> HO, ,OH + o
90% 0 o
O
0 °
‘ 40% 8,8 dimer

50% low molecular
weight oligomer

21101 950/0
A

 

 

O

O
O

O

R,R/S,S 89%
R,S 11%

Scheme 2.2 Synthesis of phenyllactide

42

epimerization produces approximately 11% of the D,L-isomer and likely some amount of
D,D-phenyllactide (starting from 100% L-phenyllactic acid). A method for separating the
different isomers has not been found. As with polylactide, the properties of
poly(phenyllactide) are expected to depend on the stereoregularity of the polymer, and

thus having stereochemically impure monomer is not desirable.

2.3 Solution Polymerization

Solution polymerizations of lactide are slow, often taking days to reach high
conversion even at elevated temperatures (70 °C is typically used). Lactide
polymerizations are commonly run with a monomer concentration of 1 M, but
phenyllactide’s lower solubility forced the polymerizations to be run in dilute solution
(0.1 M), further increasing the reaction time.

Phenyllactide polymerizations were run in toluene using Al(OiPr)3 and Sn(Oct)2
as the catalysts. Because lactide polymerizations catalyzed by Sn(Oct)2 require an
alcohol co-catalyst or residual water for initiation,5 we used t-butylbenzyl alcohol as the
initiator in Sn(Oct)2 catalyzed polymerizations to maintain control over the molecular
weights. t-Butylbenzyl alcohol was chosen because of its low volatility and because the
t-butyl group provides a signal easily seen in lH NMR spectra.

Figure 2.3 shows typical results for phenyllactide polymerizations. The
polymerizations are slow, taking weeks to reach conversions above 70%. In addition, the
polymerizations typically did not reach high conversion but instead approached a limiting

value of conversion. Two possibilities can explain this trend: catalyst degradation or

43

100

90
g 80
.2 7O
2
2 60
5
o 50
E 40
0
2 30
8 20
10

Figure 2.3 Solution polymerization of phenyllactide with Sn(Oct)2 at various
temperatures. All reactions were carried out using 0.1M L,L-phenyllactide in toluene
with monomer to catalyst ratios of 100, and BBA as the initiator in a monomer to initiator

 

 

 

 

A500
070C
9900
0100C

 

 

 

 

time (hr)

ratio of 100. (Images in this dissertation are presented in color.)

 

 

equilibrium control of the polymerization. The former would result from slow
degradation of the catalyst by an unknown process with the conversion achieved defined
by the relative rates for propagation and catalyst degradation. Equilibrium control would
imply that the rates of propagation and depropagation would be nearly equal, especially
near room temperature. The solution polymerization data can be fit to equations that
correspond to these two cases.
For termination due to catalyst degradation, the expression for the rate of
propagation (equation 2.1):
dM

-_ : Milt
dt kpl NM] (2.1)

where kp is the rate constant for polymerization and [M*] is the concentration of active,
growing chains, can be modified to include the loss of active propagation sites. Equation

(2.2) assumes spontaneous loss of active sites,

_ d[M*]

[Mat ] -kdt
= 6
dt °

 

(2.2)
where kd is the rate constant of decomposition. Integrating (2.2) and solving for [M*]

gives equation (2.3),

* _ -kdt
[M j—Ae (2.3)

where A is a constant. Substituting (2.3) into equation (2.1) and integrating gives

equation (2.4)

_ [_M]_ =-kP_: -kdt
ln<[M]o> kd (e -1> (2.4)

45

The case of equilibrium control has been studied previously,6 and is described by

equations 2.5 and 2.6:
[Ml = [Mleq + <ch - [1V11<=ct)e'kp“]t (2.5)
x = (l-lMleq/[M]o)/(1-e"""]‘) (2.6)

where [M]; is the monomer concentration at time = t, [M]eq is the monomer concentration
at equilibrium, [M]o is the initial monomer concentration, kp is the polymerization rate
constant, [I] is the initiator concentration, and X is the conversion (1-[M],/[M]o).

The experimental data for Al(OiPr)3 and Sn(Oct)2 catalyzed polymerizations can
be fit to both the catalyst degradation and equilibrium models. As shown in Figure 2.4,
both models give reasonable fits to the data, but the catalyst degradation scheme seems
more plausible. In an equilibrium polymerization, the equilibrium monomer
concentration is expected to increase with increasing temperature,7 decreasing the percent
conversion at equilibrium. As shown in Figure 2.3, the opposite trend was observed, with
the highest conversions seen at the highest temperatures. In solution polymerizations of
alkyl-substituted lactides using the same Sn(Oct)2 catalyst system, a solid precipitate was
isolated which was proposed to be a 2:1 polymer formed from the alkyl-substituted lactic
acid and Sn(II).8 Similarly, a precipitate isolated from lactide polymerizations9 was
identified as a cyclic tin-lactic acid complex based on elemental analysis data. Loss of
catalyst from polymerizations by precipitation would be consistent with the observed

data. However, no precipitate was isolated from phenyllactide polymerizations.

46

 

.0
4:.

 

 

 

 

 

’1 §
“'5 0.3
E.
E‘ 0 2
57: 0.1
0 l l L
0 25 50 75 100

time (hr)

 

 

 

 

 

I:
.9
3
>
C
o
0
O 1 l 1
0 25 50 75 100
time (hr)

Figure 2.4 Both plots show the polymerization of phenyllactide in toluene solution
at 0.1M and a temperature of 50 °C. In the top graph, the data are fit to the catalyst
degradation model and, in the bottom graph, the same data are fit to the
equilibrium model.

47

Solution polymerizations of phenyllactide provided poor control over polymer
molecular weights. The extremely long reaction times (up to 600 h for Sn(Oct)2
catalyzed reactions) allowed for a greater amount of intra- and inter-molecular
transesterification to take place. Intermolecular transesterification tends to broaden the
molecular weight distribution, but has no effect on the number average molecular weight.
Intramolecular transesterification forms cyclic oligomers, which broaden the molecular
weight distribution and decrease the number average molecular weight. As shown in
Figure 2.5, the polymer molecular weights are lower than predicted by monomer/initiator
ratios, suggesting that intramolecular transesterification is an important side reaction. In
addition, intermolecular transesterification likely contributes to the polydispersities being
>1. Thus, even though polymerizations at 100 °C reach 80 % conversion (Figure 2.3),
the molecular weights remained below 8000 g/mol.

Aluminum isopropoxide is a better catalyst, in that it decreases the reaction time
to approximately 200 hours, however it reaches even lower conversions (Figure 2.6).
Increasing the temperature from 50 °C to 90 °C gives no reaction. This cannot be
because of ceiling temperature behavior because phenyllactide is polymerized by
Sn(Oct)2 at 90 °C, and the catalyst does not change the ceiling temperature. A more
likely reason is that the catalyst degrades quickly at the higher temperature leaving no
Al(OiPr)3 to catalyze the polymerization. The maximum conversion reached at 50 °C is

32 %, too low to have produced high molecular weights.

48

 

1 2000

 

 

 

 

10000 E
8000 l
i
a? 6000 E
t
4000 l
l
2000 i
I
0 . . . . . .
0 10 20 30 40
percent conversion

Figure 2.5 Plot of the number average molecular weight vs. conversion for
the Al(OiPr)3-catalyzed polymerization of L,L-phenyllactide in toluene (0.1
M) at 50 °C at a monomer to catalyst ratio of 304 (theoretical molecular
weight of 30,000 g/mol). The line shows the theoretical molecular weight as
a function of conversion.

49

 

 

percent conversion

 

 

 

0 r J r
0 25 50 75 100
time (hr)

Figure 2.6 Solution polymerization of L,L-phenyllactide using Al(OiPr)3. The
reaction was run at 0.1 M in toluene at a temperature of 50 °C with a theoretical
molecular weight of polymer of 30,000 g/mol.

50

 

Oct—S n—Oct Oct—Sn—OR + OctH

l o
l kdt Ph/fiio
kdz

degradation 4 \ y 0

Ph
0
O — — /
ct Sn Oct OctH Oct\Sn Ofio/i;(OR
+
0
Ph

 

 

 

 

 

 

 

 

 

 

 

 

 

Ph
0
0 OR
H/ o
o I 0
Ph Ph/\(“\O
n OWPh
kd I O
degradation 4 3 \
Oct—Sn—Oct _ 0 Ph-
+
0C1 ,0 /OR
_ _ OctH \Sn’ O/Ell/
Ph
0 o
{0 /OR L Ph " n+1
H’ O
O
_ Ph - n+1

 

 

Scheme 2.3 Mechanism of Sn(Oct)2 catalyzed polymerization of phenyllactide.

51

As with Sn(Oct)2 catalyzed polymerizations, transesterification occurs during the
reaction, limiting molecular weights to approximately 6000 g/mol. It also should be
noted that Al(OiPr)3 catalyzed reactions are much faster than the tin-catalyzed ones.
Sn(Oct)2 polymerizations were run both with and without alcohol cocatalyst. The
reactions with alcohol are more predictable since those initiated by residual water in the
monomer or catalyst solution lead to an unknown number of chains initiated. The
proposed mechanism for Sn(Oct)2-catalyzed polymerization is shown in Scheme 2.3.
Sn(Oct)2 first reacts with water or alcohol to form the active catalyst, which then reacts
with the monomer to begin the polymerization. Like solution polymerizations using
aluminum isopropoxide, transesterification becomes important at long reaction times or
low monomer concentrations, preventing the formation of high molecular weight
polymer. The Mn of polyphenyllactide obtained from solution polymerizations are
typically less than 8000 g/mol, regardless of the catalyst used. Since stannous octoate
catalyzed polymerizations are much slower than those using Al(OiPr)3, it is not

commonly used for solution polymerizations of lactide.

2.4 Melt Polymerization

Because melt polymerizations take place at much higher temperatures and have a
higher monomer concentration (no solvent), the reactions are complete in minutes instead
of hundreds of hours. Unlike solution polymerizations, the faster melt reaction allows
much higher monomer conversion, but a small amount of monomer (3% at 180 °C)

remains unreacted. Like the solution polymerizations, the kinetics can be fit to both

52

equilibrium and catalyst degradation models. Reaction of pure polymer with catalyst
produces the same amount of residual monomer as polymerization, providing support for
the equilibrium model. Lactide polymerizations are reportedly equilibrium controlled.“0

There are advantages and disadvantages to melt polymerizations. While the
higher temperature allows for faster reactions and not having solvent is preferred for
commercial processes in order to reduce cost and disposal of used solvent, there is greater
transesterification and epimerization. Epimerization gives less control over the polymer
structure, and therefore the final polymer properties. Transesterification reduces control
over the molecular weight and increases the polydispersity. Reactions at extended times
have decreasing molecular weights due to intramolecular transesterification, which forms
cyclic oligomers.

A variety of catalysts were tested for their efficacy in melt polymerizations (Table
2.1) at both two hours and 48 hours. All reached high conversions after 48 hours, but the
molecular weights were lower than calculated based on monomer to catalyst ratios. The
longer reaction time increases the amount of intramolecular transesterification (also
called backbiting) which forms cyclic oligomers, increases the number of polymer
chains, and therefore decreases the molecular weight. The decrease in number average
molecular weights at extended reaction times was monitored for phenyllactide
polymerizations using Sn(Oct)2 as the catalyst (Figure 2.7). The decrease in molecular

weight is consistent with random chain scission, described by the following equation,ll

53

Reaction Time

 

 

 

2h 4.8__h
ezLali/at °_/o_<29nv_ M. Ma eat m M. an 20.1
Sn(Oct)2 96 12.7 28.6 2.26 98 6.7 10.8 1.50
A1(OiPr)3 78 10.7 15.8 1.48 94 5.9 8.7 1.38
Ph4Sn 91 13.3 20.5 1.54 95 4.6 5.7 1.23
Bi(Oct)3 93 16.4 24.6 1.50 95 14.5 26.3 1.81
ZnO 16 2.8 3.5 1.25 88 11.0 19.5 1.77
SnO 24 5.6 8.0 1.43 90 9.7 16.6 1.71
PbO 96 11.9 26.7 2.25 95 6.7 9.6 1.42
FezO3 18 3.8 4.7 1.23 67 15.0 27.0 1.80
SnClz 88 12.0 17.0 1.42 96 4.7 5.5 1.16
SnBrz 75 35.8 52.3 1.46 96 14.6 29.2 2.00
SnBr4 26 6.3 7.9 1.26 97 11.4 22.1 1.94
Zn stearate 69 7.3 10.2 1.39 97 7.4 12.2 1.65

 

 

 

Table 2.1. Melt polymerization of phenyllactide with various catalysts. The first set
of data is for a 2 h reaction time, and the data on the right is for 48 h. All reactions
were carried out at 180 °C with monomer to catalyst ratios of 100.

54

1

up, = 1
k*t +

 

 

9P0 (2.7)
where DP) is the degree of polymerization, k is a constant, and DPo is the initial degree of
polymerization. A problem with this model and a reason that the fit is not ideal, is the
assumption that the product of the chain scission reaction cannot react further (other than
continued chain scission). This is not the case. For example, if the products of the chain
scission reaction are cyclic oligomers, these can be monomers for the polymerization
reaction. The amount of epimerization varied with the catalyst used but was less than
15%, with most samples being less than 10%, and was higher for longer reaction times

(up to 30%, with the majority of samples being greater than 10%).

2.5 Polymer Properties

Differential scanning calorimetry (DSC) shows that the polyphenyllactide
obtained from melt polymerizations is amorphous as demonstrated by its lack of a
detectable melting point (Figure 2.8). Typically, polylactide produced from L-lactide is
semicrystalline because of the regular tacticity. In the case of poly(L-phenyllactide)
approximately 5-15 % epimerization takes place, disrupting crystallization and leading to
an amorphous polymer. No DSC scans were run on the polymer produced by solution
polymerizations because of their low molecular weights.

The Tg of polyphenyllactide is 50 : 3°C, which is lower than expected when

compared to polystyrene, and surprisingly near that reported for low molecular weight

55

140
120
100
80
60
40
20

 

DPn

 

 

 

 

O 10 20 3O 40 50
time (h)

Figure 2.7 Melt polymerization of phenyllactide at longer reaction times. The
reaction was carried out at 180 °C with Sn(Oct)2 and BBA. The monomer to catalyst
and monomer to co-catalyst ratios were 100. The curve is based on equation 2.7.

56

poly(lactic acid-co-phenyllactic acid).I We believe this is due to the added flexibility in
the side-chain due to the methylene group between the main-chain and the aromatic ring,
making the polymer more similar to poly(allylbenzene) (Tg=60 °C)12 than polystyrene
(Tg=110 °C). Polyphenyllactide can be processed into films by solution casting from
cyclopentanone or melt pressed by heating above the Tg. The films are clear and
colorless.

Thermal gravemetric analysis indicates that the polymer degrades at
approximately 330 °C (Figure 2.9). The primary degradation pathway is believed to be
cracking back to the cyclic monomer. To confirm this, a sample of polymer was
degraded under vacuum and the products analyzed by 1H NMR. The major product of
this degradation (95 %) was phenyllactide, with 11% isomerization to produce the R,S
isomer. Based on these results, we believe the degradation temperature to be related to
the volatility of the lactide monomer. Its higher degradation temperature is consistent

with the lower volatility of phenyllactide compared to lactide.

2.6 Polymer Degradation

Hydrolytic degradation of phenyllactide was carried out at a pH of 7.4 at 55 °C,
the products being phenyllactic acid and low molecular weight oligomers. A number of
factors can affect the degradation rate, such as crystallinity, Tg, and hydrophobicity.
Since phenyllactide is amorphous, it has a faster degradation rate than poly(L-lactide),

but a slower degradation rate than poly(D,L-lactide). Because the reaction was run above

57

 

I Tg=50i3°C

_ \ .,
o i /
\/U\o O\.
s o
I O

-50 -25 0 25 50 75 100

heat flow

/

 

 

 

 

 

Figure 2.8 DSC scan of poly(L-phenyllactide). The sample was heated at a rate of
10 °C/min. in a helium atmosphere.

58

 

 

100 ~

percent weight
# O) on
O O O
1 T I 1

N
O
l

 

 

0 r 1 r l I I; r

0 100 200 300 400
tem p (°C)

 

Figure 2.9 TGA of polyphenyllactide. The sample was heated at a rate of
l 0°C/min in a nitrogen atmosphere.

59

the glass transition temperature of both polymers, this is not a factor in the degradation
rate, leaving polymer hydrophobicity as the major factor in the degradation rate.
Polyphenyllactide, having a relatively large, non-polar side-chain, is more hydrophobic
than polylactide, which explains the slower degradation rate. The calculated rate constant
for the degradation is 0.0037 s'I for polylactide and 0.00071 s'l for polyphenyllactide.

The molecular weight curve (Figure 2.10) also displays an apparent induction
time, and then follows the calculated curve based on random chain scission. The
induction time is related to the hydrophobicity of the polymer. For a hydrophobic
polymer, the concentration of water inside the polymer is relatively low, decreasing the
rate of hydrolysis. This phenomenon was reported for the degradation of lactic
acid/mandelic acid13 and lactic acid/ phenyllactic acid copolymers.l There is a noticeable
delay before the sample experiences weight loss (Figure 2.12). The molecular weight
must decrease enough for the oligomers formed to be soluble.

Near the end of the experiment, polyphenyllactide degraded faster than predicted
by the random chain scission model. Even though the samples were added to the tubes as
small pieces, the pieces aggregate into one large piece because the reaction temperature is
greater than the Tg. Because of this, once degradation begins, more acid end groups from
the polymer are concentrated in the center of the sample, where they can catalyze the
hydrolysis reaction and increase the degradation rate. Faster degradation in the center of
large specimens of polylactide has been reported previously with the same explanation

given.'4

60

 

 

 

 

 

 

 

 

 

 

100 £
C l O polylactide
80 I I polyphenyllactide
co .
E 60 r
E
40
20
0
O 20 40 60

time (days)

Figure 2.10 Degradation of polyphenyllactide and poly(lactide) 55 °C in a
phosphate buffer solution at a pH of 7.4.

61

 

0.18
0.16
0.14
0.12

 

e polylactide I
I polyphenyllactide

 

 

 

0.08
0.06

(DPoIDPt-1)IDP0

'2

A
Ilillrtrtlljrrfijr

0.02

 

   

 

0 1 O 20 30 40 50
time (days)

Figure 2.11 Degradation of polyphenyllactide and poly(lactide) 55 °C in a
phosphate buffer solution at a pH of 7.4. The data are the same as plotted in
Figure 2.10, but plotted to give a linear relationship.

62

 

 

 

 

 

1 00
I.
i: 80 :
.2 '
H )-
3 .
“I: 60 :
H
g i-
'9 40 P
0 -
g _ O
a\° C
20 :
O I. r 1 r 1 4L 1 r l m l J $
0 1 O 20 30 40 50 60

time (days)

Figure 2.12 Weight loss during the degradation of polyphenyllactide. The
degradation was run in a phosphate bufi‘er solution at a pH of 7.4 and a
temperature of 55 °C.

63

2.7 Summary

Solution polymerization of phenyllactide gives low conversions that depend on
the catalyst (Sn(Oct)2 or Al(OiPr)3) and reaction temperature. For both catalysts, the
conversion of monomer to polymer saturates due to catalyst degradation and the onset of
a monomer-polymer equilibrium. Because of the limiting behavior of solution
polymerization, melt polymerizations are the method of choice for obtaining high
molecular weight poly(phenyllactide). A variety of catalysts polymerize phenyllactide
under melt conditions, with repeatability and control over molecular weight best obtained
through the use of an alcohol initiator. Longer polymerization times reduce the
molecular weight through transesterification side reactions which broaden the molecular
weight distribution.

Poly(phenyllactide) degrades at >300 °C to phenyllactide, which should enable
simple recycling of the polymer. Polyphenyllactide has no measurable crystallinity. It
can be processed into clear, colorless films by solution casting or by melt pressing the
polymer above the Tg. Unfortunately, the Tg of polyphenyllactide is too low for it to be a
suitable substitute for polystyrene. However, we expect that changing the substituents on

the aromatic ring of polyphenyllactide could provide the desired properties.

64

2.8 Works Cited

1)

2)

3)
4)

5)

6)
7)
8)

9)

10)

11)

12)

13)

14)

Fukuzaki, H.; Yoshida, M.; Asano, M.; M., K.; Imasaka, K.; Nagai, T.; Mashimo,
T.; Yuasa, H.; Imai, K.; Yamanaka, H. Eur. Polym. J. 1990, 26, 1273-1277.
Tabushi, 1.; Yamada, H.; Matsuzaki, H.; Furukawa, J. J. Polym. Sci, Polym. Lett.
Ed. 1975, 13, 447-450.

Fritzsche, K.; Lenz, R. W.; Fuller, R. C. Makromol. Chem. 1990, 191, 1957-1965.
Snell, K. D.; Draths, K. M.; Frost, J. W. J. Am. Chem. Soc. 1996, 118, 5605-5614.
Nijenhuis, A. J.; Grijpma, D. W.; Pennings, A. J. Macromolecules 1992, 25,
6419-6424.

Witzke, D. R.; Narayan, R.; Kolstad, J. J. Macromolecules 1997, 30, 7075-7085.
Ivin, K. J. Makromol. Chem, Macromol. Symp. 1991, 42/43, 1-14.

Yin, M.; Baker, G. L. Macromolecules 1999, 32, 7711-7718.

Kricheldorf, H. R.; Kreiser-Saunders, 1.; Stricker, A. Macromolecules 2000, 33,
702-709.

Duda, A.; Penczek, S. Macromolecules 1990, 23, 1636-1639.

Jellinek, H. H. G. In;; Jellinek, H. H. G., Ed.; Elsevier Sci. Rub. Co.: Amsterdam,
1978; pp 1-38.

Seefried Jr., C. G.; Koleske, J. V. J. Polym. Sci: Polym. Phys. Ed. 1976, 14, 663-
673.

Fukuzaki, H.; Aiba, Y. Makromol. Chem. 1989, 190, 2407-2415.

Li, S. J. Biomed Mater. Res. 1999, 48, 342-353.

65

Chapter 3 — Mandelide

3.1 Introduction

Like phenyllactic acid, there has been limited work on the synthesis of mandelic
acid polymers. Most syntheses have produced low molecular weight materials, with the
two most common methods being the polycondensation of mandelic acid,1 and the
transesterification of methyl mandelate2 with removal of methanol. Poly(mandelic acid)
was first synthesized by the pyrolysis of the trimethyltin ester of a-bromomandelic acid.3

Kobayashi, et al. used deoxypolymerization of phenylglyoxylic acid with cyclic

 

 

 

 

O O

1) Me3$n\0 P“ —> MegsnBr + ‘ 0/
Br Ph
2) >=° © ‘fi/Lo/ * 002
O
0 ‘_—_’ L Ph

0 O

3) PVKOH '1' EI3N —-> \\'/U\O//
Br Ph
_ .l

 

 

Scheme 3.1 Syntheses of poly(mandelic acid). 1) Okada,3 2)Smith,4 3) Pinkus5

66

phosphates to prepare poly(mandelic acid).4 Smith and Tighe used the pyridine-catalyzed
ring-opening polymerization of S-phenyl-l,3-dioxolane-2,4-dione to prepare
poly(mandelic acid), which despite having the advantage of being a faster reaction than
polycondensation, still gave molecular weights less than 4000 g/mol.5 Pinkus and
coworkers prepared poly(mandelic acid) by reacting a-bromophenylacetic acid with
triethylamine to give polymers with degrees of polymerization between 12 and 20.6

Mandelide, the cyclic dimer of mandelic acid, was synthesized previously.“8
The dimer structure was analyzed by x-ray crystallography,7 but was rarely used as a
monomer for polymerization. The only previous report of the polymerization of
mandelide was as a comonomer polymerized with lactide.9 The patent that describes this
work provides little information regarding the polymer properties.

Little work has been done exploring the properties of polymandelide. The
structures of polymandelide and poly(phenyllactide) differ only in that
poly(phenyllactide) has a methylene that links the aromatic ring to the polymer backbone,
and its removal should increase the T8 and give a biodegradable polymer with properties
similar to polystyrene. It was reported that low molecular weight poly(mandelic acid)
(2000 g/mol) has a r, of 75 °C.”)

Copolymers of mandelic acid and lactic acid are common, but all have relatively
low molecular weights (<10,000 g/mol). The glass transition temperature of the
copolymers increases with increasing mandelic acid content.ll These copolymers
degrade in vitro10 and in vivo,'2‘l3 but no degradation of poly(mandelic acid) segments

was detected during the length of the experiment (15 weeks).'0‘l2

67

3.2 Monomer Synthesis

Previously reported syntheses of S,S-mandelide were all based on acid-catalyzed
self-esterification reactionsl'g’9 The major differences between the methods were the
choice of solvent, mandelic acid concentration, and the choice of the acid catalyst. The

reported yields were poor, less than 15%. The result for the synthesis of R,R-mandelide

was similar, a 9% yield.14 There are no reported syntheses of R,S-mandelide. The

reactions described in the literature were carried out at relatively high concentrations,

 

leading to high yields of low molecular weight oligomers. Whitesell and coworkers

 

HOo Owiioll) ........ 0

OH HI 0
_____, (S,S) (R,R)
xylenes

O
(R and S) 0

Scheme 3.2 Synthesis of mandelide

produced mandelide unintentionally during the polymerization of mandelic acid by
polycondensation.l
When the same synthetic procedure used for the synthesis of phenyllactide was

used for the preparation of mandelide, there was little product after two weeks.

68

Increasing the reaction temperature by using a higher boiling solvent (xylenes, instead of
toluene) increased the rate of product formation, leading to a 50 % yield of mandelide as
a mixture of the R,R; S,S; and R,S diastereomers (Scheme 3.2). Reducing the mandelic
acid concentration decreased the proportion of oligomers formed, and provided much
greater yields of the cyclic dimer than had been previously reported. The R,S
diastereomer can be separated from the R,R and S,S isomers by recrystallization from
chloroform. Being less soluble, the R,R and S,S isomers crystallize leaving the R,S
diastereomer in solution. After removing the R,R and S,S isomers by filtration, the R,S

isomer is isolated by crystallization from ethyl acetate and hexanes.

3.3 Polymerization

Solution polymerizations of mandelide gave no polymer, most likely due to low
reaction rates, with additional problems caused by mandelide’s low solubility in toluene.
Solution polymerization of the less hindered phenyllactide also was slow, and
transesterification side-reactions prevented formation of high molecular weight products.
For these reasons, mandelide was polymerized exclusively using melt polymerization.
The R,R and S,S isomers of mandelide decompose at 210 °C instead of melting,8 making
them unsuitable for melt polymerizations. Therefore, all melt polymerizations of

mandelide used the R,S diastereomer, which melts at 137 °C.

69

The kinetics for the Sn(Oct)2/BBA catalyzed melt polymerization of mandelide
follow the same trends seen in phenyllactide polymerizations, and reach a limiting
conversion of 96%. Polymerizations were run at 180 °C in order to make direct
comparisons with phenyllactide polymerizations, although lower temperatures can be
used. As might be expected, the rates of mandelide polymerizations are slightly slower
due to increased steric hindrance (Figure 3.1). The rate constant for polymerization of
mandelide is about 90% of that for phenyllactide polymerization. In addition, a fraction
of the R,S monomer epimerized to the R,R and S,S form (approaching 50% of the
remaining monomer during the reaction), however, the presence of small amounts of the

R,R and S,S diastereomers does not appear to interfere with the polymerization because

 

 

 

 

 

 

 

 

 

1 00
80
I:
o
i!
2 60
c:
8
4o . O mandelide
l phenyllactide
20
0 L l 4 1
0 10 20 3O 40

time (min)

Figure 3.1 Melt polymerization of R,S-mandelide and S,S-phenyllactide at
180 °C using Sn(Oct)2 and BBA with monomer to catalyst and monomer to
cocatalyst ratios of 100. The lines are fits using the equilibrium model.

70

both isomers are soluble in the molten R,S-mandelide.

Unfortunately, while the reactions reach high conversions, the molecular weights
were low: typically less than 15,000 g/mol compared to 27,000 g/mol predicted from the
monomer to catalyst ratio. Excess initiator in the form of water or alcohol, or a chain
transfer step could increase the number of polymer chains formed. Monomer degradation
at the high reaction temperature (180 °C) could also lead to reduced molecular weights,
but no degradation products were detected by 1H NMR spectroscopy after polymerization
or when monomer was heated without catalyst. It is more likely that water is present as
an impurity, and it is more difficult to remove from mandelide than from phenyllactide.

Comparisons of a variety of polymerization catalysts gave similar results: the

molecular weights were well below the values expected from the monomer to catalyst

 

 

 

 

 

 

 

 

 

catalyst % conversion Mn Mw PDI
Sn(Oct)2 96 1200 1500 l .29
Al(OiPr)3 25 2400 3000 " 1 .26
Bi(Oct)3 64 7100 15,400 2.15
Ph4Sn - 1100 1500 1.33
PbO - 1100 1500 1.30
San - 1000 1400 1.31
SnBrz 76 15,800 42,700 2.69
Zn stearate 51 5400 7400 1.38

 

 

 

 

 

 

 

Table 3.1 Melt polymerization of R,S-mandelide with a variety of catalysts.
All reactions were carried out at 180 °C for two hours with monomer to
catalvst ratios of 100.

71

ratio (Table 3.1). The results parallel those obtained when the same catalysts were used
to polymerize phenyllactide. Similar conversions were obtained, except that Al(OiPr)3
was a less efficient catalyst for converting mandelide to polymandelide. It is possible
that steric hindrance has a greater effect on Al(OiPr)3 catalyzed polymerizations than
those catalyzed by Sn(Oct)2, because Al(OiPr)3 exists as an aggregate in solution.
However, the state of aggregation is not known for melt polymerizations. One run with
SnBrz gave the highest molecular weight, but attempts to repeat that result were
unsuccessful. This would support the assumption of water in the monomer controlling
initiation.

Repeat polymerization reactions at the same monomer to catalyst ratios using the
catalysts that gave the best results in terms of molecular weight (Sn(Oct)2, Bi(Oct)3,
SnBrz, and Zn stearate), gave higher molecular weights for Sn(Oct)2, Bi(Oct)3, and Zn
stearate, but lower molecular weights for SnBrz. These results show that the results vary
for reactions run under the same conditions. Polymerizations catalyzed by Sn(Oct)2 were
repeated at different monomer to catalyst and monomer to initiator ratios. While larger
ratios did produce higher molecular weight materials, the increase in molecular weight
was not linearly related to the change in ratios. In other words, doubling the monomer
initiator ratio did not double the molecular weight of the polymer (Table 3.2)

The inconsistent results, while frustrating, would be expected if the monomer was
impure. The quantity of the impurity could vary from one batch of monomer to the next,
giving inconsistent polymerization results. Evacuating the polymerization samples

overnight before use showed no significant improvement in molecular weight, but later

72

 

work by others showed that higher molecular weights are obtained with monomer dried

under vacuum for a period of many days.

3.4 Initial Polymer Properties (based on low molecular weight materials)

DSC measurements on a low molecular weight sample of polymandelide indicate
a Tg of 96 °C. This increase of 46 °C compared to polyphenyllactide is most likely due to
the removal of the methylene group between the polymer chain and the aromatic ring,
which reduces some of the flexibility of the polymer backbone. Later work on high
molecular weight samples gave a Tg of 100°C. The polymer is amorphous, as expected
because epimerization during polymerization renders the polymer atactic.

TGA measurements of polymandelide show that the polymer degrades at a
slightly lower temperature than polyphenyllactide. The degradation pathway is believed
to. be the same for both polymers: intramolecular transesterification (backbiting) to give
the cyclic monomer, as demonstrated in the controlled thermal degradation of
polyphenyllactide.

The degradation temperature likely is related to the relative volatility of the
monomer. Since mandelide is more volatile than phenyllactide, as expected due to its
lower molecular weight, the onset for the degradation of polymandelide is ~30 °C below

that for polyphenyllactide.

73

 

 

 

 

 

 

 

 

 

 

 

 

 

100 v
I.
)-
80 -
5 t
8 60
3. r
E I-
a, .
'5 40 r
3 ; -—-polymandelide
- -—polyphenyllactide_j
20 -
l
, w
o r l L L 1 L L

O

100 200 300 400
temperature (C)

Figure 3.2 TGA of polymandelide and polyphenyllactide. Both scans were carried
out under a nitrogen atmosphere at a heating rate of 10 °C/min.

74

3.5 Copolymers

The principal limitation of polymandelide is its low molecular weight. In hopes of
increasing the molecular weight, copolymers of D,L- mandelide and L,L-phenyllactide
were synthesized via melt polymerization using Sn(Oct)2 as the catalyst. As expected,
the molecular weight of the copolymers decreased with increasing proportions of
mandelide in the copolymer (see Table 3.2). Since polyphenyllactide has a lower Tg than
polymandelide, the Tg of the copolymers decreased with increasing amounts of the
phenyllactide comonomer. By synthesizing copolymers, we were able to obtain materials

containing mandelide with higher molecular weights, but they were still below 20,000

 

 

 

 

 

 

 

g/mol.
PhLDzMD Mn Mw PDI Tg
0:100 6600 9400 l .42 87
10:90 7600 11,400 1.50 86
25 :75 8700 14,200 1.63 79
50:50 8300 13,100 1.57 45
75:25 11,100 19,600 1.77 42
100:0 21,300 46,800 2.20 50

 

 

 

 

 

 

 

Table 3.2 Copolymerization of R,S-mandelide. All reactions were carried out at 180

°C using Sn(Oct)2 and BBA with monomer to catalyst and monomer to cocatalyst ratios
of 100.

75

 

 

 
 

 

—PhLD:MD = 10:90
— PhLDzMD = 25:75
—PhLD:MD = 50:50
— PhLDzMD = 75:25

   

 

 

 

 

 

1mrmlrrrrlrrrrlrerlrrrglrrrurrrr

-50 -25 0 25 50 75 100 125
temperature (°C)

 

Figure 3.3 DSC of copolymers of R,S-mandelide and S,S-phenyllactide. All were
run under He atmosphere with heating rates of 10 °C/min.

76

3 .6 Summary

The properties of polymandelide indicate that it could be a biodegradable
substitute for polystyrene. Modification of polyphenyllactide by removing the methylene
between the aromatic ring and the polymer chain increases the glass transition
temperature by approximately 50 °C, however there are consistent difficulties in
obtaining high molecular weight polymandelide. This led us to explore modification of

polyphenyllactide by adding substituents to the aromatic ring, which is covered in

Chapter 4.

77

3 .7 Works Cited

1)

2)

3)

4)

5)

6)

7)

8)

9)

10)

11)

12)

13)

Whitesell, J. K.; Pojman, J. A. Chemistry of Materials 1990, 2, 248-254.

Domb, A. J. Journal of Polymer Science Part a-Polymer Chemistry 1993, 31,
1973-1981.

Okada, T.; Okawara, R. J. Organometal. Chem. 1973, 54, 149-152.

Kobayashi, S.; Yokoyama, T.; Kawabe, K.; Saegusa, T. Polymer Bulletin 1980, 3,
585-591.

Smith, I. J.; Tighe, B. J. Makromol. Chem. 1981, 182, 313-324.

Pinkus, A. G.; Subramanyam, R.; Clough, S. L.; Lairmore, T. C. Journal of
Polymer Science; PartAsPolymer Chemistry 1989, 27, 4291-4296.

Lynch, V. M.; Pojman, J.; Whitesell, J. K.; Davis, B. E. Acta. Cryst. 1990, C46,
1125-1127.

Schoberl, A.; Wiehler, G. Ann. Chem. 1955, 595, 101-130.

Kimura, Y. In Chem. Abst. 1993, 118, 2555 77r: Japan, 1993.

Fukuzaki, 11.; Aiba, Y. Makromol. Chem. 1989, 190, 2407-2415.

Kylma, J.; Harkonen, M.; Seppala, J. V. Journal of Applied Polymer Science
1997, 63, 1865-1872.

Fukuzaki, H.; Yoshida, M.; Asano, M.; M., K.; Imasaka, K.; Nagai, T.; Mashimo,
T.; Yuasa, H.; Imai, K.; Yamanaka, H. Eur. Polym. J. 1990, 26, 1273-1277.
Imasaka, K.; Yoshida, M.; Fukuzaki, H.; Asano, M.; Kumakura, M.; Mashimo,
T.; Yamanaka, H.; Nagai, T. International Journal of Pharmaceutics 1992, 81,

31-38.

78

14) Toniolo, C.; Perciaccante, V.; Falcetta, J.; Rupp, R.; Goodman, M. J. Org. Chem.

1970, 35, 6-10.

79

Chapter 4 — Methylphenyllactides

4.1 Introduction

A degradable glassy polymer with pr0perties similar to polystyrene would be a
valuable addition to the family of environmentally degradable polymers. Previously, the
similarities in structure between polystyrene and poly(phenyllactide) prompted us to
investigate poly(phenyllactide) as a potential polystyrene mimic. The polymer proved to
be glassy, but the Tg of the polymer was 50 °C, less than the 88 °C predicted by computer
modeling and nearly 50 ° below that of polystyrene. One strategy used to raise the Tg in
styrenic polymers is to add substituents to the aromatic ring. The addition of polar
groups such as Cl and CN tend to increase the Tg of the polymer by increasing dipole-
dipole interactions between chains. Small alkyl groups also increase the Tg, in this case
by increasing the rotational barriers and decreasing the flexibility of the polymer
backbone. Thus, adding a methyl group to the ortho position of the polystyrene ring
increases the polymer Tg from 109 °C to 136 °C, while substitution at the meta and para
positions has little effect on Tg. Similar trends in Tg might also be obtained in the
polyphenyllactide system. Following this analogy, poly(o-methylphenyllactide) (or
poly(o-tolyllactide) and abbreviated as PoTLD) would have the highest Tg.

However, it is possible that the methylene group that connects the aromatic ring
and the main polymer chain may minimize the steric effects of ortho substitution,
resulting in small increases in Tg. Ideally the best analogy to the polymers of substituted

phenyllactic acids would be substituted poly(allylbenzene)s, but these polymers have not

80

yet been synthesized. Several aryl-substituted substituted poly(hydroxyalkanoate)s have
been reported. Poly(3-hydroxy-5-phenylvalerate) (PHPV) has a Tg of 19 °C and is
amorphous even though the polymer backbone is stereochemically regular.l Its methyl
substituted analog, poly(3-hydroxy-5-(4’-tolyl)valeric acid) (PHTV) has a Tg of 17 °C,1
demonstrating that a para methyl group has little effect on the glass transition
temperature. However, unlike PHPV, PHTV is semicrystalline with a Tm of 95 °C.1
Based on this example, stereoregular PpTLD might exhibit some crystallinity, even
though poly(L-phenyllactide) is amorphous.

In this‘ chapter, the preparation of ortho, meta, and para-substituted
poly(methylphenyllactide) is described. Like other lactide polymerizations, the polymers
were synthesized by the Sn-catalyzed ring opening polymerization of the corresponding
cyclic dimer. The monomer synthesis, polymerization, polymer properties, and

preliminary degradation data for the polymers will be described.

 

O
O
”“2““ Q \
> O
NaOAc/AOZO \
CH3 CH3N\<

3M HCI ZntHg)
3M HCl

Scheme 4.1 Synthesis of methylphenyllactic acids.

81

4.2 Synthesis of Substituted Phenyllactic Acids

Methyl substituted phenyllactic acids have not been reported, but Wong and
coworkers2 described a general route to hydroxy and methoxy substituted phenyllactic
acids that can be adapted to the synthesis of methyl-substituted phenyllactic acids.
Starting from the corresponding benzaldehydes, we prepared the oxazolones using N-
acetylglycine (Scheme 4.1). The yields were modest, with the highest yield for the para
substituted oxazolone (46%) and lowest for the ortho derivative (16%). These yields are
considerably lower than Wong reported for the methoxy and hydroxy substituted
compounds (64-90%). Since the formation of oxazolone involves nucleophilic attack at
the benzaldehyde carbonyl, any electron-donating group on the aromatic ring would
make the reaction more difficult. While methoxy, hydroxy, and methyl groups are all
electron donating, the acidic conditions of the reaction could protonate the hydroxy and
methoxy groups. Once protonated, these groups would now be electron withdrawing and
make the addition more favorable. Since the methyl group cannot be protonated, it
remains electron donating and causes the lower yield. The low yields for the ortho
substituted product were likely due to the steric hindrance of the ortho methyl group.
During purification, the ortho and meta oxazolones often hydrolyzed to give the a-keto
acid, the product of the next step in the reaction scheme.

Hydrolysis of the oxazolone in refluxing 3 M HCl overnight gave the acid. For
these compounds, the enol is more stable than the keto form because the enol double
bond is conjugated with both the aromatic ring and the acid carbonyl. Clemmensen

reduction of the crude hydrolyzed product gave the methyl-substituted phenyllactic acid

82

 

 

,
O
O s o//
H+
OH ,
OH
H3O
H3C _ n
ZnO
A
O
O
O
H C CH
3 O 3

Scheme 4.2. Synthesis of methylphenyllactide via cracking

in high yield (80-90%). The meta and ortho-substituted compounds were difficult to
purify due to formation of oligomers (frequently near 50 %). These compounds were

used for the formation of the cyclic dimer without further purification.

4.3 Synthesis of Phenyllactides

There are two general routes to the dimer, acid-catalyzed condensation of the 01-
hydroxyphenyllactic acid and thermal cracking of low molecular weight oligomers and
isolation of the more volatile cyclic dimer by distillation. The former method is slow but
can be carried out without loss of stereochemical purity, while the latter is fast but is
always accompanied by some degree of racemization. The starting 0t-
hydroxyphenyllactic acids were racemic, and thus avoiding racemization was not a

consideration in this case. This also allowed us to use unpurified methylphenyllactic acid

83

for the preparation of the monomers because the major impurity was the corresponding
oligomer.

The hydroxyacid was oligomerized (Scheme 4.2) by stirring in 2-5 mL of 3M HCl
at 120 °C overnight. ZnO was added to the oligomer, and after heating the mixture at
220 to 240 °C under vacuum, the crude lactide was isolated by distillation/sublimation as
it formed. The crude lactide was purified by washing with aq. NaHCO3 and
recrystallized from ethyl acetate/hexanes. Purification of the ortho and meta-
methylphenyllactides required column chromatography using silica gel eluted with 75/25
hexanes/ethyl acetate to produce dimer suitable for polymerization. Typical yields were
50 % for pTLD, with lower yields for oTLD and mTLD (20-40%) because of the extra
purification step.

Self-esterification, used previously to prepare phenyllactide and mandelide, was
also used for the synthesis of the para-substituted dimer. Refluxing the hydroxyacid for
seven days in xylenes with a catalytic amount of p-toluenesulfonic acid gave the dimer in
36% yield, compared to a 50 % yield via the thermal route. Monomer from both methods

behaved identically in polymerizations.
4.4 Polymerizations

Since only melt polymerizations of phenyllactide gave high molecular weight
materials in reasonable times, melt polymerization was used exclusively for the

polymerization of the methylphenyllactides. Initially, the molecular weights of the

polymers were consistently lower than those obtained for phenyllactide under the same

84

 

 

 

 

 

 

percent conversion

 

 

 

 

time (min)

Figure 4.1. Melt polymerization kinetics of pTLD using Sn(Oct)2 and BBA at 180 °C.
The monomer to catalyst and monomer to cocatalyst ratios were 100.

85

conditions. Phenyllactide prepared via the oxazolone (beginning with benzaldehyde) also
gave low molecular weights, indicating residual impurities from the monomer synthesis.
After purification by column chromatography, the monomers gave polymers that were
close to the molecular weights predicted by the monomer/ initiator ratios.

Kinetic results for the melt polymerization of pTLD using Sn(Oct)2 and BBA are
shown in Figure 4.1. The propagation rates were extracted from pTLD and phenyllactide
polymerizations run under identical conditions using an equilibrium model for the
polymerization, and indicate that the polymerization rate for pTLD is 60% of the rate for
phenyllactide. In addition, the reaction equilibrates at 90% conversion, compared to 97%
with PhLD. Kinetic experiments were not run for the ortho and meta
methylphenyllactides, but we expect similar behavior, with slower rates likely for the

ortho compound because of increased steric hindrance.

4.5 Polymer Properties

Differential scanning calorimetry shows that all three poly(methylphenyllactide)s
are amorphous, which is expected because the polymers were made from racemic
lactides. Polylactide made from racemic monomer is also amorphous. PpTLD has the
highest Tg at 59 °C, followed by PoTLD (51 °) and PmTLD (42 °C). Having the methyl
group in the para position inhibits bond rotation the most, which is why the para
substituted polymer has the highest Tg.

Therrnogravimetric analysis shows that PpTLD has nearly the same degradation

temperature as PPhLD (Figure 4.2). Since the degradation mechanism for degradation of

86

 

 

 

 

 

100 - - j
80 - ,
E
.9
Ci .
3 60
‘E
a:
2 40 l-
8.
20 I-
O 4 L 1 l r l 4
0 100 200 300 400

mp (° C)

Figure 4.2. TGA of poly(methylphenyllactide) (blue) and polyphenyllactide (red).
The TGA scans were run under nitrogen with a heating rate of 10 °C/min.

87

substituted lactides is depolymerization, the onset for degradation should be related to the
volatility of the monomers. Phenyllactide and methylphenyllactide have similar

volatilities, and thus have similar degradation profiles.

4.6 Degradation

Degradation experiments were run on poly(p-methylphenyllactic acid), and reveal
similar degradation rates for PPhLD and PpTLD. The data were fit to a degradation
model based on random chain scission, and the degradation rates were extracted from the
fit. The relative rates were 9.8 x 104 for PPhLD and 1.2 x 10'3 for PpTLD. The results
of the weight loss experiments were very different. The weight of the PPhLD samples
decreased to nearly 0 at 60 days, while the PpTLD samples retained >80% of their
weight. These results can be explained by the additional methyl group of PpTLD, which
increases the hydrophobicity of the polymer and decreases the solubility of low molecular
weight oligomers in the buffer solution. In addition, the initial molecular weight of
PpTLD was higher than the initial molecular weight of PPhLD, meaning that PPhLD
would require fewer breaks in the polymer chain to become soluble in the buffer solution.
The initial Mn of the PpTLD was 63,600 g/mol and the initial Mn of PPhLD was 46,500
g/mol. Thus, the mass losses are smaller for PpTLD even though the rates for the
decrease in molecular weight of PpTLD and PPhLD are similar. It is expected that the
ortho and meta substituted polymers would degrade faster than the para substituted

because of the lower glass transition temperatures. The Tg of the para substituted

88

 

 

 

 

 

 

 

 

100 g I . _ I I.
: 9 I I -
.5 8° i .
‘6 I
.13 60 -
i... ..
S .
.9 .
g 40 I 0%wtfraction 0
g . (PPhLA)
F I%wtfraction
2° f (PpTLA)
t
0 1 1 r l r r r l 1 r r Y
0 20 40 60
time (days)

Figure 4.3. Weight loss during the degradation of PpTLD compared to PPhLD. Both
degradations were carried out in a phosphate buffer solution at a pH of 7.4 and a

temperature of 55 °C.

89

 

 

 

 

 

 

 

 

 

100 1

: I polyphenyllactide

80 ' Apoly(p-tolyllactide)
5 60 Z
5 -
E :
E 40 :
20
0 .

0 20 40 60

time (days)

Figure 4.4. Decrease in molecular weight decrease during the degradation of PpTLD
and PPhLD. The degradations were carried out at 55 °C in phosphate buffer solutions
at a pH of 7.4.

90

polymer is slightly above the temperature of the experiment, causing the degradation to
occur more slowly than polyphenyllactide. These factors combined cause PpTLD to lose
weight more slowly than PPhLD (Figures 4.3). The rates of molecular weight decrease

for both polymers are nearly the same (Figure 4.4).

4.7 Summary

Based on this work with methyl-substituted phenyllactides, we have demonstrated
that substitutions on the aromatic ring can be used to modify the physical properties of
poly(phenyllactic acid). In addition, the choice of the position for the substituent also
affects properties. Unlike poly(methylstyrene), the ortho substituted polymer has the
lowest Tg instead of the highest. The additional methylene between the ring and the
chain changes how the substitution location affects physical properties, such as glass

transition temperatures.

91

4.8 Works Cited

1) Curley, J. M.; Hazer, B.; Lenz, R. W.; Fuller, R. C. Macromolecules 1996, 29,

1762-1766.

2) Wong, H. N. C.; Xu, Z. L.; Chang, H. M.; Lee, C. M. Synthesis 1992, 793-797.

92

Chapter 5 — Conclusions

This work presents modifications on the basic structure of polylactides by adding
substituents to the side chain in order to modify the physical properties. By choosing
aromatic substituents, we expect to be able to obtain a biodegradable polymer with
properties similar to polystyrene. The basic property we have measured is the glass
transition temperature, although the degradation rates were also studied. The Tg has a
large effect on the potential uses for a polymer. For example, if one wishes to use
polylactide in place of a Styrofoam cup, the addition of a hot beverage such as coffee will
increase the temperature of the cup above the T8 of polylactide, causing the cup to soften
and be useless. If the addition of an aromatic group to polyethylene increases the T8 for
polystyrene, one would expect that an aromatic group added to polylactide would also
increase the Tg.

The polymer initially chosen was polyphenyllactide. The major difference
between our analogy using polyethylene and polystyrene is that there is a methylene
group between the aromatic ring and the main polymer chain. This makes the polymer
more similar to poly(allyl benzene). While polystyrene has a Tg of 100 °C, poly(allyl
benzene)’s is 60 °C. Based on this information, it is not surprising that the Tg of
polyphenyllactide (50 °C) is not substantially different than that of polylactide.

What may not have been expected is that poly(L-phenyllactide) is amorphous,
while poly(L-lactide) is crystalline. This is explained by the small, but significant,
amount of epimerization occurs during the polymerization of phenyllactide. Less

epimerization typically occurs during the polymerization of L-lactide because the

93

reaction is run at a lower temperature. The high melting point of L-phenyllactide
necessitates a higher temperature during polymerization (180 °C), increasing the amount
of epimerization.

The next logical action in developing a polystyrene-like lactide-based polymer
would be to remove the methylene between the aromatic ring and the polymer chain,
producing polymandelide. While polymandelide does have a higher glass transition
temperature of nearly 100 °C, there was some difficulty in obtaining high molecular
weight materials.

To work around the problems with mandelide polymerizations and to also give
greater possibilities for modification, we chose to work with substituted phenyllactides.
The addition of a methyl group to the aromatic ring provides a slight modification,

producing relatively minor changes to the Tg relative to that of polyphenyllactide (i 10

°C). We also determined that the position of the methyl group was important in
determining the Tg. Poly(o-methylphenyllactide) has a glass transition temperature
nearly the same as polyphenyllactide, while poly(m-methylphenyllactide) has a Tg of 42
°C and poly(p-methylphenyllactide) has a Tg of 59 °C. While not quite as high as
polystyrene or polymandelide, this proves that the addition of substituents to the aromatic
ring of polyphenyllactide can be used to further modify the properties of polylactide.
Based on this work, one would expect that continued modification by the addition
of a larger group, such as t-butyl would increase the glass transition even further.
Because substituted phenyllactides are relatively easily synthesized, they are likely to
have the most flexibility for modification in order to obtain a variety of polymers with

varying properties. Another possibility is the use of reactive groups, such as bromine,

94

which can be used to modify the material after polymerization. For example, one
possibility would be to copolymerize p-bromophenyllactic acid with lactic acid, then
replace the bromine with a crosslinkable group to produce an even wider variety of
materials. A large number of options for modifications to the basic structure of

polyphenyllactide are available.

95

 

 

Experimental

General. L-Phenyllactic acid was purchased from Aldrich and used as received.
THF and toluene were distilled over CaHz, then distilled under nitrogen from sodium
benzophenone ketyl. CHzClz was distilled over CaHz. All reactions requiring anhydrous
or inert conditions were carried out in oven-dried glassware under a positive atmosphere
of argon or nitrogen. Solutions or liquids were introduced using oven-dried syringes or
cannula through rubber septa. All reactions were stirred magnetically using Teflon-
coated stir bars unless otherwise noted. For reactions requiring heating, electrically
heated silicon oil baths were used, and the stated reaction temperature is either the
temperature of the bath or the reflux temperature. In the cases requiring -78°C cooling,
the reactions were chilled with a dry ice/acetone bath. Organic solutions obtained after an
aqueous work-up were dried over MgSO4. Removal of solvents was accomplished using
a Bi'lchi rotary evaporator at water aspirator pressure. Proton nuclear magnetic resonance
spectra ('H NMR) and carbon nuclear magnetic resonance spectra (13C NMR) were
recorded on a Gemini 300 or VXR 5008 spectrometer. 1H NMR spectra were recorded at
300 and 500 MHz. Chemical shifts are reported in parts per million (ppm) relative to
internal tetramethylsilane (Me4Si, 6: 0.00 ppm) with CDC13 as the solvent. 13 C NMR
spectra were recorded at 75 MHz and chemical shifts are reported (as ppm) relative to
CDC13 (6 = 77.0 ppm). Melting points were taken using an Electrotherrnal Melting Point
Apparatus. Low resolution electron impact (EI) mass spectra (MS) were recorded on a
Finnigan 4000 mass spectrometer and high resolution mass spectra (HRMS) were
recorded on Kratos MSSO or MS25 mass spectrometers. Molecular weights were

obtained by gel permeation chromatography (GPC) using a PLgel 20p. Mixed A column

96

from Polymer Laboratories with THF at a flow rate of 1.0 mL/min and are reported
relative to polystyrene standards. Detection was by a Waters R401 Differential
Refractometer, Waters 996 Photodiode Array, or Waters W410 Differential
Refractometer operating at 35 °C with a column oven heated to the same temperature.
Thermogravimetric analysis (TGA) results were obtained from a Perkin Elmer TGA7
with a heating rate of 10 °C/min in a nitrogen atmosphere unless otherwise stated.
Differential scanning calorimetry (DSC) was performed in aluminum pans under a helium
atmosphere on a Perkin Elmer DSC7 with a heating rate of 10 °C/min. The instrument
was calibrated with indium and cooled with liquid nitrogen.

Synthesis of L,L-Phenyllactide. L-Phenyllactic acid (3.69 g, 12.5 mmol) and a
catalytic amount of p-toluenesulfonic acid (0.02 g, 0.1 mmol) were dissolved in toluene
(600 mL). The solution was refluxed for eight days, removing the water by a Dean Stark
trap filled with 4A molecular sieves. After washing three times with saturated aqueous
NaHCO3, the solvent was removed and the crude crystals were recrystallized twice from
ethyl acetate and hexanes to afford the dimer (1.40 g, 42%) as white crystals. 1H NMR:
3.0 (dd, J=8, 15 Hz, 1 H), 3.4 (dd, J=4, 15 Hz, 1H), 5.0 (dd, J=4, 8 Hz, 1 H), 7.2-7.4 (m, 5
H). ”C NMR: 36.6, 76.6, 127.5, 128.7, 129.7, 134.6, 165.5. MS m/z (rel inten) El 91
(100), 104 (15), 131 (11), 148 (25), 161 (10), 296 (M+, 10), 297 (M+H+, 6); HRMS (EI)
calcd. for C13H1504 (MI) 296.1049, found 296.1048. mp 165 °C.

Catalyst solutions. Commercially available (Aldrich) aluminum isopropoxide
(Al(OiPr)3) was distilled under reduced pressure and then dissolved in freshly distilled
toluene. Tin(II) 2-ethylhexanoate (Sn(Oct)2) (Aldrich) and bismuth(lll) 2-ethylhexanoate

(Bi(Oct)3) (Alfa Aesar) were dissolved in freshly distilled toluene before use.

97

 

Tetraphenylporphyrin aluminum methoxide was prepared following the procedure
outlined by Endo, Inoue, and Aida.l 4-t-butylbenzyl alcohol (BBA) was purchased from
Aldrich and dried over 4 A molecular sieves, then dissolved in distilled toluene before
use.

Solution Polymerization - General Procedure. To a custom-made flask
containing a glass-covered stir bar, L,L-phenyllactide (0.50 g, 1.7 mmol) was added and
the apparatus was assembled and evacuated overnight with a diffusion pump. The empty
flask was filled with argon and 17 mL of freshly distilled toluene was added. Styrene (0.5
mL, 4 mmol) was added and polymerization initiated with 0.25 mL of 1.6 M nBuLi (0.40
mmol). After one hour, the solution was degassed and the solvent was vacuum
transferred to the flask containing the phenyllactide monomer. Upon completion of the
transfer the solution was placed under argon and heated to 50 °C. Al(OiPr)3 (207 pL of a

27.0 mM solution in toluene) was added by syringe. The polymerization was terminated

with an excess of 2 N HCl (1 mL, 2.0 mmol). The polymer solution was washed with

water until neutral and the solvent removed. The polymer obtained had too low a
molecular weight to be precipitated. 1H NMR: 3.0 (dd, J=8, 14 Hz, 1 H), 3.2 (dd, J=4, 14
Hz, 1 H), 5.3 (dd, J=4, 8 Hz, 1 H), 7.0-7.3 (m, 5 H).

For solution polymerization in THF, a similar procedure was followed except the
styrene was polymerized at -78 °C for 2 hours.

Melt Polymerization - General Procedure. A Schlenk flask containing L,L-
phenyllactide (0.25 g, 0.85 mmol) was evacuated and refilled with argon three times, and
then Sn(Oct)2 (49 pL of a 51.5 mM solution in toluene) was added. The flask was placed

under vacuum and heated to 180 °C. Upon completion of the reaction, the flask was

98

 

 

allowed to cool and the polymer dissolved in toluene (2 mL), then precipitated twice into
cold methanol to give 0.23 g (92 %) of polyphenyllactide.

Melt Polymerization - Catalyst study. L,L-Phenyllactide (100 mg, 0.34 mmol)
was added to glass tubes sealed at one end. Inside a drybox, catalyst was added to give a
monomerzcatalyst mole ratio of 100: 1. Vacuum adapters were connected to the tubes, the
tubes were connected to a vacuum line, and evacuated. The tubes were then sealed and
placed into an oil bath heated to 180 °C. After two hours, the tubes were removed and
placed in cool water to quench the reaction. The tubes were broken open and the contents
dissolved in THF. The solvent was then removed to yield the polymer.

Melt Polymerization - Kinetics. The procedure described in the catalyst study was
followed using 50 mg of L,L-phenyllactide (0.17 mmol) and Sn(Oct)2 as the catalyst for
all samples in a molar ratio of 100:1 (monomer: catalyst). A tube was removed at the
desired times and handled as outlined previously.

Solution Polymerization with Added Alcohol. The same procedure as outlined
above was followed, except that a solution of 4-t-butylbenzyl alcohol (BBA) in toluene
was added to the reaction flask before the addition of catalyst to give a monomer: alcohol
mole ratio of 100: 1.

Melt Polymerization with Added Alcohol - Catalysts. The procedure described
above was followed, with a solution of 4-t-butylbenzyl alcohol in toluene added to give a
monomer: alcohol mole ratio of 100: l. The reaction tube was then evacuated and sealed.

Melt Polymerization with Added Alcohol - Kinetics. The procedure described

above was followed, with 4-t-butylbenzyl alcohol at amolar ratio of 100:1 (monomer:

99

alcohol). A solution of the alcohol in toluene was added to the reaction tube, then
evacuated and sealed.

Synthesis of Mandelide. R,S-Mandelic acid (6.03 g, 39.7 mmol) and a catalytic
amount of p-TsOH (0.20 g, 1.2 mmol) were dissolved in xylene (600 mL). The solution
was refluxed for six days, removing the water by a Dean Stark trap filled with 4A
molecular sieves. The solution was then washed three times with saturated aqueous
NaHC03 and the solvent removed. The crude solid was recrystallized repeatedly from
ethyl acetate and hexanes to separate the diastereomers. Combined yield of the R,R and
S,S isomers : 1.3 g (47 %). mp 193 °C (decomp). 1H NMR: 6.1 (s, 1 H), 7.2-7.3 (m, 5
H). Yield of R,S isomer: 1.5 g (53 %). mp 135 °C. 1H NMR: 5.9 (s, 1 H), 7.4-7.5 (m, 5
H).

Melt Polymerization of Mandelide - Catalyst Study. The procedure used for L,L-
phenyllactide was followed.

Melt Polymerization of Mandelide - Added Alcohol. To a Schlenk flask
containing R,S-Mandelide (0.5075 g, 1.89 mmol), SnBrz was added to give a monomer to
catalyst ratio of 400:1. One equivalent of distilled methanol was then added. The flask
was moved to a vacuum line, placed under a positive pressure of argon, and heated with
an oil bath at 180 °C. After four hours, the flask was removed and allowed to cool. The
product was analyzed without purification.

Melt Polymerization of Mandelide — Kinetics. The procedure used for

phenyllactide with t-butylbenzyl alcohol was followed.

100

 

Synthesis of (T PP)Al-0CH 3 Initiator. Tetraphenylporphyrin was obtained from
OK. Chang and was used without further purification. The catalyst was prepared
following the procedure outlined by Endo, Inoue, and Aida.1
Synthesis of Oxazolones.

2-Methyl-4m-methylbenzal)-5-oxaazolone. p-Tolualdehyde (12.5 g, 104 mmol),
N-acetylglycine (8.26 g, 71.2 mmol), acetic anhydride (17.55 g, 172.1 mmol), and sodium
acetate (4.94 g, 60.2 mmol) were added to a 250-mL round bottom flask fitted with a
condenser. The mixture was heated at 100 °C in an oil bath for 5 hours, and then 50 mL
of ice water was added. The crude product was collected by filtration, washed four times
with 50 % aq. ethanol, and dried overnight under vacuum. After recrystallization from
acetone, chloroform and petroleum ether, the oxazolone was obtained in 46.1 % yield
(6.60 g). mp 135-136 °C (tit17 133-135 °C). ‘H NMR: 8 2.38 (s, 6H), 7.11 (s, 1 H), 7.23
(d, J=8 Hz, 2 H), 7.26 (d, J=8 Hz, 2 H). 13C NMR: 5 15.7, 21.7, 129.7, 130.5, 131.7,
132.2, 142.0, 165.5, 168.0. Elemental analysis: calc’d 71.33% C, 5.49% H, 6.94% N.
Found: 71.16% C, 5.75% H, 6.75% N.

2-Methyl-4(m-methylbenzal)-5-oxazolone. Yield 36 %. mp 106-108 °C IH NMR:
6 2.39 (s, 6 H), 7.10 (s, 1 H), 7.23 (d, J=7 Hz, 1 H), 7.32 (t, J=8 Hz, 1 H), 7.87 (d, J=9
‘ Hz, 2 H). ”C NMR: 8 15.7, 21.4, 128.8, 129.4, 131.8, 132.1, 12.3, 132.7, 133.0, 138.6,
165.9, 167.9. Elemental analysis: calc’d (for the hydrolyzed oxazolone) 65.72% C,
5.98% H, 6.39% N. Found: 66.19% C, 6.02% H, 6.20% N.

2-Methyl-4(o-methylbenzal)-5-oxazolone. Yield 16.3 %. mp 106-108 °C. 1H
NMR: 6 2.39( d, J=1 Hz, 3 H), 2.47 (s, 3 H), 7.20-7.31 (m, 3 H), 7.42 (s, 1 H), 8.51-8.54

(m, 1 H). 13C NMR: 5 15.7, 20.0, 126.5, 128.4, 130.7, 131.0, 131.6, 131.7, 132.5, 139.6,

101

166.3, 167.9. Elemental analysis: calc’d (for the hydrolyzed oxazolone) 65.72% C,
5.98% H, 6.39% N. Actual: 66.73% C, 5.92% H, 6.39% N.
Synthesis of Methylphenyllactic Acids.

p-Methylphenyllactic acid. 2-Methyl-4(p-methylbenzal)-5-oxazolone (13.76 g,
68.5 mmol) was added to a l-L round bottomed flask and 3 M HCl (500 mL) was added.
A condenser was attached and the reaction refluxed overnight. Zn (62.0 g, 0.95 mol) and
HgClz (10.00 g, 36.8 mmol) were added to a separate flask, 3 M HCl (12 mL) and water
(100 mL) were added and the mixture was stirred for five minutes to form the Zn(Hg)
amalgam. The solution was drained off and the amalgam rinsed with water. The
amalgam was the added to the hydrolyzed oxazolone and refluxed for 4 hours. The
reaction mixture was filtered while hot into a separatory funnel, extracted five times with
100 mL portions of ethyl acetate, and the solvent removed. Recrystallization of the crude
product from ethyl acetate and hexanes yielded 10.60 g of p-methylphenyllactic acid
(86.0 %). mp 98-100 °C. 1H NMR: 5 2.31 (s, 3 H), 2.95 (dd, J=7, 14 Hz, 1 H), 3.15 (dd,
J=4, 8 Hz, 1 H), 4.48 (dd, J=4, 7 Hz, 1 H), 7.12 (s, 4 H). 13C NMR: 8 21.1, 39.7, 71.0,
129.3, 129.4, 132.5, 136.8, 178.2.

m-Methylphenyllactic acid. mp 70-72 °C. 1H NMR: 8 2.32 (s, 3 H), 2.94 (dd,
J=7, 14 Hz, 1 H), 3.16 (dd, J=4, 14 Hz, 1 H), 4.49 (dd, J=4, 7 Hz), 1 H), 7.02-7.22 (m, 4
H). 13C NMR: 8 21.4, 40.1, 71.0, 126.4, 128.0, 128.5, 130.3, 135.7, 138.3, 178.4.

o-Methylphenyllactic acid. mp 70-73 °C. 1H NMR: 6 2.45 (s, 3 H), 2.94 (dd,
J=9, 14 Hz, 1 H), 3.26 (dd, J=4, 14 Hz, 1 H), 4.47 (dd, 4, 9 Hz, 1 H), 7.13-7.21 (m, 4 H).

13C NMR: 8 19.6, 34.5, 70.6, 126.1, 127.2, 130.0, 130.6, 134.5, 136.9, 178.6.

102

Synthesis of Methylphenyllactides.

p-Methylphenyllactide. p-Methylphenyllactic acid (0.515g, 2.86 mmol) was
added to a 50-mL round bottom flask and heated to 120 °C in an oil bath for 2 days. The
flask was removed from the oil bath and ZnO (0.05 g, 0.61 mmol) was added. The flask
was connected to a Kugelrohr distillation apparatus and heated to 220 °C. The crude
product was dissolved in toluene and washed three times with aqueous NaHCO3, dried
over MgSO4 and the solvent removed. The residue was recrystallized twice from ethyl
acetate and hexanes to give 0.233 g (50.2 %) of the product as a mixture of the RR, SS,
and RS isomers. mp 157-164 0C. 1H NMR: 8 2.30 (s, 6 H), 2.31 (s, 6 H), 2.93 (dd, J=8,
15 Hz, 2 H), 3.12 (d, J=5 Hz, 4 H), 3.29 (dd, J=4, 15 Hz, 2 H), 4.24 (t, J=5 Hz, 2 H), 4.97
(dd, J=4, 8 Hz, 2 H), 6.99-7.12 (m, 16 H). 13C NMR: 8 21.1, 36.2, 38.0, 76.7, 76.9,
129.4, 129.6, 129.8, 130.6, 131.6, 137.1, 137.6, 165.0, 165.6. Elemental analysis:
74.04% C, 6.22% H. Actual: 74.14% C, 6.33% H.

m-Methylphenyllactide. m-Methylphenyllactide was prepared and purified
following the procedure used for the para compound except that before recrystallization,
the crude lactide was further purified using a silica gel column eluted with 25% ethyl
acetate/75% hexanes. mp 121-123 °C. lH NMR: 8 2.27 (s, 6 H), 2.32 (s, 6 H), 2.90 (dd,
J=8, 15 Hz, 2 H), 3.12 (t, J=4 Hz, 4 H), 3.30 (dd, J=4, 15 Hz, 2 H), 4.24 (t, J=5 Hz, 2 H),
5.00 (dd, J=4, 8 Hz, 2 H), 7.02-7.22 (m, 16 H). 13C NMR: 8 21.4, 36.7, 126.7, 128.2,
128.6, 130.4, 134.6, 138.4, 165.5. Elemental analysis: 74.04% C, 6.22% H. Actual:
74.16% C, 6.46% H.

o-Methylphenyllactide. o-Methylphenyllactide was synthesized and purified

following the same procedure as that used for the para compound, except that before

103

recrystallization, the crude lactide was purified further using a silica gel column eluted
with 25% ethyl acetate/75% hexanes. mp 127-131 °C. 1H NMR: 8 2.23 (s, 6 H), 2.32 (s,
6 H), 3.02 (dd, J=9, 15 Hz, 2 H), 3.12 (dd, J=6, 15 Hz, 2 H), 3.32 (dd, 4, 15 Hz, 2 H),
3.44 (dd, J=4, 15 Hz, 2 H), 4.35 (dd, J=4, 6 Hz, 2 H), 4.98 (dd, J=4, 9 Hz, 2 H), 7.14-7.24
(m, 16 H). 13C NMR: 8 19.5, 19.6, 33.6, 35.2, 76.4, 76.8, 126.3, 126.5, 127.6, 127.7,
130.2, 130.3, 130.6, 130.9, 132.5, 133.3, 136.7, 137.2, 165.2, 165.8. Elemental analysis:
74.04% C, 6.22% H. Actual: 73.91 % C, 6.50% H.

Melt Polymerization - Catalyst study. Methylphenyllactide (100 mg) was added
to glass tubes sealed at one end. Inside a drybox, the catalyst was added to give a molar
ratio of 100:1 (monomer: catalyst). Vacuum adapters were connected, the tubes
connected to a vacuum line, and evacuated. A solution of t-butylbenzyl alcohol (BBA) in
toluene was added to give a monomerzalcohol mole ratio of 100:1, and the solvent was
removed by vacuum. The tubes were then sealed and placed into an oil bath heated to
180 °C. After two hours, the tubes were removed and placed in cool water to quench the
reaction. The tubes were broken open and the contents dissolved in THF. The polymer
was isolated by removal of the solvent.

Melt Polymerization - Kinetics. The procedure used for the catalyst study was
followed using 50 mg of monomer, Sn(Oct)2 as the catalyst,and t-butylbenzyl alcohol as
the cocatalyst for all samples at a molar ratio of 100:1 (monomer: catalyst/ monomer:
alcohol). A tube was removed at the desired times and handled as outlined previously.

Melt Depolymerization. A sample (100 mg) of well-purified polyphenyllactic
acid (containing no residual monomer as detected by 1H NMR) was added to glass tubes

sealed at one end. Vacuum adapters were connected, the tubes moved to a vacuum line,

104

and evacuated and filled with argon three times. Sn(Oct)2 and BBA solutiOns were added
to give the equivalent of a monomer to catalyst ratio of 100, assuming all of the polymer
is converted to monomer. The tubes were then sealed and placed into an oil bath heated
to 180 °C. Tubes was removed at the desired times and handled as outlined in the
polymerization procedure.

Hydrolytic Degradation of Polymers. A sample of approximately 50 mg of the
polymer that had been well purified was accurately weighed and the weight recorded.
The sample was then placed inside a test tube with a screw cap and 15 mL of pH of 7.4
phosphate buffer solution was added. The degradation was carried out at 55 i 0.2 °C.
At the desired time, the sample was removed, rinsed repeatedly with distilled water and
dried under vacuum overnight. The sample was weighed to determine the weight loss

and the molecular weight was determined by GPC.

105

Works Cited

1) Endo, M.; Aida, T.; Inoue, S.; 1987, 20, 2982-2988.

106