Y,

11

e.
8
3c

32%

. 33-22. .
g \ 2‘ I,

7x 2:...
.JV

 

figfimny Ma? 5.
2.1.. .21..

awn»:

.7 4.3.

mmnwé.

2%,. :5»
“I... ‘. .1:
”mafia?

.

.... T‘ a...“ ..
‘ £21.” may .mkasbwfim
, E. » A. .4 a». 1 . .,

“rt-.... 5.. n

a? g

a . ”.23":

:3‘3 3
‘33:»

i?

gonna...

x, Lu .

pvt 3.6"
a. a

.2
§

. .
1,3... 2... ,
avifi 2......th
.Q gamu- .1.
. _ _ a
..
. d...u..é...§..: J ;
.... .1 .3 ,. z
‘ 2 a: .. ,nm" v
3:1,... 5.5.? -n. .6; ...!
. z :5 (2.2- r». .9 . u .
. .1' 1| '3- 'Tcl .0 ‘ til-s V
rrnmwinutjueiunnfzr m
, 36.1.3. 3‘. wit- 1 ..
1.... urn."- .§:«: u...”
f. u? 1:... I
LL’ 1. .5.
...: .ju‘vq...
. I....liJ-VC..HF. .1

1.,......::......ann.. ....
.i . ‘
2:!
I’D.
3|!

5:.
:U.Z(‘ .

1.1.xhn..._\

‘.. {at :2! 9...
. .V vl. ~ I!

(932‘?
.....f».i§l.€lt
!. 157......53915
vir
Ti:
3r...)
. .
4.103....
13;

'.II

“$117319

2<J”L’

 

LIBRARY

Micmgan State
University

 

 

 

This is to certify that the

dissertation entitled

1. Kinetic Evaluation of the Copolymerization
of Lactide with Substituted Lactides

II. Synthesis and Characterization of Polyacenes

presented by

Chun Wang

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Chemistry

 

a] ofessor

 

MS U is an Affirmatiw Action/Equal Opportunity Institution 0-12771

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DateDue.p65-p. 15

 

 

I. KINETIC EVALUATION OF COPOLYMERIZATION OF LACTIDE
WITH SUBSTITUTED LACTIDES

II. SYNTHESIS AND CHARACTERIZATION OF POLYACENES

By

Chun Wang

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

2001

ABSTRACT

I. KINETIC EVALUATION OF COPOLYMERIZATION OF LACTIDE
WITH SUBSTITUTED LACTIDES. II. SYNTHESIS AND
CHARACTERIZATION OF POLYACENES
By

Chun Wang

Because of its environmental and biodegradability, polylactide has
important applications as materials in medicine and as an environmentally
friendly commodity plastic. One major deficiency of polylactide is itS limited
range of physical properties. Copolymerization of lactide with substituted
Iactides is a strategy that can provide polylactides with new properties.
Evaluation of the kinetics of these copolymerizations is necessary since the
kinetics define the structure of polylactide copolymers, which in turn defines
the physical properties of the copolymers. A protocol was developed to
analyze the compositions of LA/substituted LA copolymers, which allowed us
to evaluate the kinetic parameters of the copolymerization of lactide with
ethylglycolide, isopropyl glycolide, and D-ethylglycolide. The data Show that
these copolymerizations can be described as ideal polymerizations, but due
to the monomer reactivity differences, poly(lactide-co-isopropylglycolide) and
Poly(lactide-co-D-ethylglycolide) tend to be blocky. These polymers Show
single glass transitions. A study of the degradation of poly(lactide-co-
isopropylglycolide) Showed that the degradation process can be described by

a random chain scission mechanism.

Polyacene is expected to have the lowest band gap among conjugated
polymers. Despite its promise as an electrical conductor, and as a material
for optics and battery electrodes, it has yet to be synthesized. We developed
the synthesis of polyacene materials through the acid-catalyzed condensation
of bis(methoxymethy|)benzenes. TGA, IR and Raman characterization
support the formation of an extended n-conjugated system. To improve
solubility, we developed a scheme to prepare n-octylated bis-
(methoxymethyl)benzene. Polymerization of this monomer yielded materials
with molecular weights as high as 7000. However, fluorescence data Show
that the acene segments have a maximum length of only ~5 units. We also
designed a synthesis of two-dimensional graphite by applying similar
chemistry to poly(p-phenylene). TGA, IR and Raman data support formation

of a graphitic structure after the material is heated to >500 °C.

To my Father and family

iv

ACKNOWLEDGMENTS

My first acknowledgment goes to my advisor: Professor Baker. He is not
the reason I came to MSU, however, it is because of him, I chose to stay as a
Spartan. Thank you very much for your guidance and attention in all these years,
I just wish I had accomplished more to deserve such a great mentor.

I would also like to thank Professor Borhan, Blanchard, Berglund, Reusch,
Scraton and. Chang. What I leamed from you will continue to influence me in the
future.

It’s time to say good-bye, my dear group. My special thanks goes to Cory,
who helped me turn into today’s “Chun”. Micah, Tara, Erin, Fadi, Gao, JB, Kirk,
Tianqi, Wenxi, Jingpin, Sue, Hou, Qiao, Yiyan, Mao. I learn a bit from each of
you. Thanks for your help and companionship, I have had a great time.

My friends, Don and Ellen, Xiaoxiao, Winnie, Zigang, Heather, Lisa and
Dan, Tong, Qingqing, Ning, Ying, for all the valleys and peaks that we have been
through, thanks for always being there with me. I appreciate your sincere
friendship.

The last toast goes to my family: Mom, for your eternal love, Dad, you
have always been my example in research, my dentist sister and brother-in-law,
for your sweet support, and Yijun, my husband, for your love and friendship.

Most important of all, all the success and glory attributes to my mighty

Father.

Table of Contents

List of Schemes xi
List of Figures xiv
List of Tables xviii

I. KINETIC EVALUATION OF COPOLYMERIZATION OF LACTIDE WITH

SUBSTITUTED LACTIDES
INTRODUCTION .............................................................................. 2
1. Structure and Physical Properties of Polylactide ..................... 2
2. Polylactide Degradation ................................................................................. 4
3. Applications of Polylactides ....................................................... 6
3.1. Biomedical Applications .............................................................................. 6
3.1.1. General Medical Applications .............................................................. 6
3.1.2. Controlled Drug Delivery ..................................................................... 6
3.1.3. Tissue Engineering .............................................................................. 7
3.2. Applications as Conventional Therrnoplastics ............................................. 7
4. Copolymerization ........................................................................ 9
4.1. Lactide Copolymerization .......................................................................... 10
4.1.1. Copolymerization of Lactide with Lactones and Carbonates ............. 11
4.1.1.1. Copolymerization with Caprolactone ................................ 11
4.1.1.2. Copolymerization with Other Lactones ............................. 12

vi

4.1.1.3. Copolymerization with Carbonates ................................... 12

4.1.2. Copolymerization with Glycolide ........................................................ 13
4.1.3. Copolymerization of Substituted Glycolides and Lactide ................... 15
4.1.4. Copolymerization with Lysine and Other Amino Acids ...................... 16
4.2. Block and Graft Copolymers ..................................................................... 18
4.2.1. Copolymerization of Lactide with Poly(ethylene glycol) ..................... 18
4.2.2. Graft Copolymers from Natural Polymers .......................................... 20

4.2.3. Lactide Copolymers with Bocks from Vinyl and Acrylic Monomers 21

4.2.4. Copolymerization with Other Comonomers ....................................... 23
4.3. Copolymerization Characterization ........................................................... 23
4.4. Kinetic Studies of Copolymerization ......................................................... 25

4.4.1. Monomer Reactivity Ratios ................................................................ 25

4.4.2. Types of Copolymerization Behavior ................................................. 27

4.4.2.1. Ideal copolymerization: m2 = 1 ........................................ 27
4.4.2.2. Alternating Copolymerization: r1 = r2 = 0 .......................... 28
4.4.2.3. Block Copolymerization: r1 >1, r2 >1 ................................. 28
4.4.3. Experimental Evaluation of Monomer Reactivity Ratios .................... 28

4.4.4. Monomer Reactivity Ratio Analysis of Lactide Copolymerizations 29

EXPERIMENTAL ............................................................................ 31
5.1. General ....................................................................................... 31
5. 2. Monomer Preparation and Synthesis ....................................................... 32
5. 3. Copolymerization ..................................................................................... 33
5.4. Copolymer Composition: Direct Determination ........................................ 33

vii

5. 5. Biodegradation ...................................................................... 34
RESULTS AND DISCUSSION ....................................................... 35

6.1. Copolymerization of Lactide with Ethylglycolide ................ 35

6.2. Kinetic Evaluation of the Copolymerization of Lactide with Ethylglycolide 40
6.2.1. Copolymer Composition Determination. ............................................ 40
6.2.1.1. ”C NMR ........................................................................ 40

6.2.1.2. Normal Phase High Pressure Liquid Chromatography

(HPLC) ............................................................................... 41

6.2.1.3. Reverse Phase HPLC ................................................... 44

6.2.1.4. Kinetic Evaluation Based on Reverse Phase HPLC Results

........................................................................................... 45
6.2.1.5. Direct Determination of Copolymer Composition. ............ 49
6.3. Kinetic Evaluation of the Copolymerization of Lactide with

lsopropylglycolide ........................................................................................... 60
6.4. Copolymerization of Lactide with R-ethylglycolide .................................... 69
6.5. Physical Properties and Structure of Copolymers ..................................... 76
6.6. Biodegradation of the Copolymers ............................................................ 81
BIBLIOGRAPHY .................................. _. ......................................... 87

ll. SYNTHESIS AND CHARACTERIZATION OF POLYACENES
INTRODUCTION ............................................................................ 94
1. Structure of Polyacene ................................................................................. 94
1.1. Chemical Structure ............................................................................... 94
1.2. Electronic Structure and Properties ...................................................... 95

viii

1.2.1. Electronic structure of unsubstituted polyacene ................. 95

1.2.2. Electronic structures of alkyl substituted polyacene ......... 102

1.2.3. Properties of polyacenes .................................................. 104

1 .3. Applications ........................................................................................ 104
1.3.1. Electronically conductive materials ................................... 104

1.3.2. Optical limiters .................................................................. 109

1.3.3. Electrodes for lithium ion batteries ................................... 111

1.3.3. Capacitors ........................................................................ 118

1.4. Synthetic Approaches to Polyacenes ................................................. 118
1.4.1. The Diels-Alder approach ................................................. 120

1.4.2. Topochemical polymerizations ......................................... 123

1.4.3. Acene-like materials ........................................................ 124

1.5.4. Soluble polyacenes .......................................................... 127

1.6. Characterization of Polyacenes .......................................................... 128
1.6.1. X-ray diffraction ................................................................ 129

1.6.2. ”C NMR ........................................................................... 129

1.6.3. ESR (Electron Spin Resonance) ...................................... 130

1.6.4. UV and fluorescence ........................................................ 130

1.6.5. Raman spectroscopy ........................................................ 131
EXPERIMENTAL ........................................................................... 132
2.1. General ............................................................................ 132
__2.2. Synthesis ............................................................................................. 133
RESULTS AND DISCUSSION ...................................................... 141

ix

3.1. Polyacenic materials. ......................................................................... 141

3.1.1. Simple routes to acene-like materials ............................... 141

3.1.2. Chemical dehydrogenation ............................................... 150

3.1.3. Polymerization in solution ................................................................ 150
3.1.4. Characterization of acene-like materials .......................... 154

3.1.4.1. Infrared ........................................................................ 154

3.1.4.2. Raman ......................................................................... 155

3.1.4.3. Elemental Analysis ...................................................... 155

3.2. Well-defined Soluble Polyacene ............................................................. 159
3.2.1. Monomer design and synthesis ......................................... 159
3.2.2.Polymerization condition development ............................... 163

3.2.3. Model cyclization study ...................................................... 164

3.2.4. Characterization ................................................................ 173

3.2.5. Discussion ......................................................................... 174

3.3. Acene/poly (p-phenylene) hybrids. ......................................................... 175
3.4. Conclusion .......................................................................................... 180
BIBLIOGRAPHY ........................................................................... 182

 

LIST OF SCHEMES - KINETIC EVALUATION OF THE COPOLYMERIZATION
OF LACTIDE WITH SUBSTITUTED LACTIDES

Scheme 1.
Scheme 2.
Scheme 3.
Scheme 4.
Scheme 5.
Scheme 6.
Scheme 7.
Scheme 8.

Scheme 9.

Scheme 10.
Scheme 11.

Scheme 12.

Scheme 13.
Scheme 14.
Scheme 15.
Scheme 16.

Scheme 17.

Scheme 18.
Scheme 19.

Preparation of polylactide.

Ring opening polymerization mechanism of lactide.
Copolymerization of lactide with e-caprolactone.
Copolymerization of lactide with TMC.
Copolymerization of lactide with glycolide.
Copolymerization of lactic acid with mandelic acid.
Copolymerization of lactide with BMD and DMD.
Polymerization of D,L—3~methylglycolide.
Synthesis of PLA-PEG copolymers.

Synthesis of poly(lactide-b-isoprene-b-lactide).
Synthesis of poly(HEMA-graft-lactide).

Chain propagation steps in the copolymerization of two
monomers.

Copolymerization of racemic lactide with ethylglycolide.
Reactivity differences leads to “blocky” copolymers.
Protocol for the determination of copolymer composition.

Copolymerization of racemic lactide with racemic
isopropylglycolide.

Chain propagation mechanism for lactide
copolymerization.

Copolymerization of lactide with R-ethylglycolide.

The relationship between monomer reactivity ratio and the
structure of copolymers.

xi

11
13
13
15
16
16
20
22
23
26

36
4O
52
60

68

69
80

LIST OF SCHEMES - SYNTHESIS AND CHARACTERIZATION OF
POLYACENES

Scheme 1 .

Scheme 2.

Scheme 3.

Scheme 4.

Scheme 5.

Scheme 6.

Scheme 7.

Scheme 8.

Scheme 9.

Scheme 10.

Scheme 11.
Scheme 12.

Scheme 13.

Scheme 14.

Scheme 15.

Polyacene oligomer synthesized by via Dials-Alder
adduction.

Polyacene oligomer synthesized via DieIs-Alder addition
and retro Diels-Alder elimination.

Topochemical polymerization of butadiyne and 1-buten-3—
yne to prepare polyacene.

Preparation acene-like material from phenol-formaldehyde
resin.

Acene-like material synthesized from styrene-
divinylbenzene copolymers and poly(p-phenylene).

Synthesis of highly alkyI-substituted pentacene derivatives
by homologation.

Acid catalyzed polymerization of methoxymethylbenzene.

Acid-catalyzed polymerization of bis(methoxymethyl)
benzenes.

Possible outcomes from the polymerization of
bis(methoxymethyl)benzenes.

The substitution pattern on the benzene ring leads to
different condensation products.

Alkyl substituted polyacene.

First approach to the synthesis of 1,4-bis-(methoxymethyl)-
2,5-dialkyl benzenes.

Modified synthesis of 1,4-biS-(methoxymethyl)-2,5-dialkyl
benzenes.

Control experiment used to develop polymerization
conditions.

Control experiment to prove BF3 caused side reactions.

xii

120

122

124

125

126

127

142
143

148

149

159
161

162

163

163

Scheme 16.
Scheme 17.
Scheme 18.
Scheme 19.

Scheme 20.
Scheme 21 .

Scheme 22.
Scheme 23.

Synthesis of a model compound for the cyclization study.
Model cyclization condition study.
Model dehydrogenation test.

lntramolecular cyclization vs. intermolecular bridge
formation.

Synthesis of alkylated polyacenes.

Proposed approach for the preparation of a soluble linear
polyacene.

Synthesis of a methoxymethyl-substituted polyphenylene.

Synthesis of sheet structures from methoxymethyl-
substituted polyphenylenes.

xiii

164
165
166
171

172
175

176
178

LIST OF FIGURES — KINETIC EVALUATION OF THE COPOLYMERIZATION
OF LACTIDE WITH SUBSTITUTED LACTIDES

Figure 1.
Figure 2.
Figure 3.
Figure 4.
Figure 5.

Figure 6.

Figure 7.

Figure 8.
Figure 9.

Figure 10.
Figure 11.
Figure 12.

Figure 13.
Figure 14.
Figure 15.

Figure 16.

Four generic copolymer architectures.

Examples of lactones that copolymerize with lactide.
A poly(lactide-co-L-Lysine) derivative.
Copolymerization of lactide with Hpr.

080 runs (second heating after flash quenching from 100
°C) for lactide and ethylglycolide copolymers. Heating rate:
10 °C/min under helium.

Glass transition temperatures of lactide and ethylglycolide
copolymers.

Mole fraction of lactide in the copolymer (F1) as a function
of the feed ratio (f1) during the copolymerization of lactide
with ethylglycolide as determined from reverse phase
HPLC

r1 and r2 determination by the Fineman-Ross method.

Mole fraction of lactide in the copolymer (F1) as a function
of the feed ratio (f1 during the copolymerization of lactide
with ethylglycolide as determined from 1H NMR.

r1 r2 determination by Fineman-Ross method.
r1,r2 Determination by the Telen-Tudos Method.

Mole fraction of lactide in the copolymer (F1) as a function
of the feed ratio (f1) during the copolymerization of lactide
with poly(lactide-co-raoisopropyllgycolide).

r1 r2 determination using the Fineman-Ross method.
r1,r2 Determination using the Kelen-Tudos method.

Melt polymerization of D-ethylglycolide at 130 °C catalyzed
by Sn(Oct)2 + t—BBA

Mole fraction of lactide in the copolymer (F1) as a function
of the feed ratio (f1) during the copolymerization of lactide

xiv

12
17
18
37

38

47

48
54

55
59
64

65
66
7O

72

Figure 17.
Figure 18.
Figure 19.

Figure 20.

Figure 21.

Figure 22.

Figure 23.

with D-ethylglycolide.
r1, r2 determination by Fineman-Ross method.
r1, r2 determination by Kelen-Tudos method.

Glass transition temperature of lactide and
isopropylglycolide copolymers.

The weight loss of polymer samples during hydrolytic
degradation.

The molecular weight loss of poly(LA-co-iPRGL) during
hydrolytic degradation.

Decrease in the molecular weight of P(LA-co-iPRGL)
during hydrolytic degradation fitted to the random chain
scission model.

The decrease in the degree of polymerization of polymer
samples during hydrolytic degradation.

XV

73
74

77

83

84

85

86

LIST OF FIGURES - -SYNTHESIS AND CHARACTERIZATION OF

POLYACENES

Figure 1. Chemical structure of polyacene.

Figure 2. Four possible double bond configurations for polyacene.

Figure 3. Empirical determination of the band gap (E9) of polyacene.

Figure 4. Six methyl-substituted polyacenes studied by Wang and
Hou.

Figure 5. Three possible isomers of structure 5.

Figure 6. Band gaps of conductive materials.

Figure 7. Molecular orbital energy diagram of polyacene and doped
polyacene.

Figure 8. Relationship between output fluence and input fluence of a
70% transmitting solution of a polyacene-based oligomer in
toluene.

Figure 9. Five-level energy diagram of polyacene.

Figure 10. Cell structure and electrochemistry of electrodes in PAS
banenes.

Figure 11. The CaLi configuration observed in the first stage Li-graphite
intercalation compound (Li-GIC).

Figure 12. An illustrative model of PAS.

Figure 13. Other acene derivatives made by Clardy’s group.

Figure 14. Therrnogravimetric analysis results for
poly(methoxymethylbenzene). Heating rate: 10 °C/min.

Figure 15. Thennogravimetric analysis results for
poly(methoxybenzene) and
poly(bis(methoxymethyl)benzene) when 02 is rigorously
excluded. Heating rate: 10 °C/min.

Figure 16. Therrnogravimetric analysis results for

poly(methoxymethylbenzene) and
poly(bis(methoxymethyl)benzene run under nitrogen

xvi

94
96
99
1 01

102
105
108

109

110

112

113

115
121
143

144

146

Figure 1 7.

Figure 18.

Figure 19.

Figure 20.

Figure 21.

Figure 22.

Figure 23.

Figure 24.

Figure 25.

contaminated with 10% oxygen. Heating rate: 10 °C/min.

Therrnogravimetric analysis results for poly (bis-1,2-
(methoxymethyl)benzene) in nitrogen. Heating rate: 10

°C/min.

Therrnogravimetric analysis results for poly(biS-1,4-
(methoxymethyl)benzene) in nitrogen. Heating rate: 10

°C/min.

Therrnogravimetric analysis results for poly(1,2-bis
(methoxymethyl)benzene) prepared via solution and bulk
reactions.

Infrared absorption spectra of the polybenzyl network
formed from 1,2-bis(methoxymethyl)benzene, taken after
being heated in nitrogen at 10 °C/min to 30 °C, 380 °C, 500
°C, 650 °C, and 800 °C

Raman spectra of polymers derived from ortho, para and
meta-bis(methoxymethyl)benzene after being heated to 800
O

C.

UV and fluorescence spectra of soluble polyacene products.

Therrnogravimetric analysis of methoxymethyI-substituted
polyphenylene.

Infrared of cross-linked and heat treated samples of a
methoxymethyl-substituted polyphenylene.

Raman Spectrum of a cross-linked and heat treated sample
of a methoxymethyl-substituted polyphenylene.

xvii

151

152

153

157

158

172

177

179

180

LIST OF TABLES - KINETIC EVALUATION OF THE COPOLYMERIZATION

OF LACTIDE WITH SUBSTITUTED LACTIDES

Table 1. Properties of copolymers.

Table 2. Determination of Conversion in the Copolymerization of
Lactide (LA) and Ethylglycolide (EG) by Normal Phase
HPLC.

Table 3. Monomer Determination by Reverse Phase HPLC.

Table 4. Copolymerization behavior of lactide with ethylglycolide.

Table 5. Comparison of the GC and 1H NMR methods for analyzing
the copolymer composition.

Table 6. Copolymerization behavior from Fineman-Ross method.

Table 7. Copolymerization behavior from Kelen-Tudos method.

Table 8. Comparison of copolymer composition analysis obtained
from parallel runs.

Table 9. Copolymerization behavior from Fineman-Ross method.

Table 10. Copolymerization behavior from Kelen-Tudos method.

Table 11. Copolymerization behavior from Fineman-Ross method.

Table 12. Copolymerization behavior from Kelen-Tudos method.

Table 13. Properties of P(LA-co-IPrGL).

LIST OF TABLES - SYNTHESIS AND CHARACTERIZATION OF
POLYACENES

Table 1. Electronic properties of structures 1-6.
Table 2. Concentration vs. yield in model cyclization study.
Table 3. Reaction temperature vs. yield in model cyclization.

Table 4. Reaction time vs. yield in model cyclization study.

xviii

39

103
167
167
168

Table 5.
Table 6.
Table 7.
Table 8.

Argon purge vs. yield in model cyclization study.
Acid ratio vs. yield in model cyclization study.
The rate of adding the acid vs. yield in model.

Different solvent vs. yield in model cyclization.

xix

168
169
169
170

I. KINETIC EVALUATION OF COPOLYMERIZATION OF
LACTIDE WITH SUBSTITUTED LACTIDES

INTRODUCTION

1. Structure and Physical Properties of Polylactide

Polymers based on lactic acid are an increasingly important class of
polymers. Because they are biocompatible and degrade to environmentally
benign products, applications for these polyesters are increasing rapidly.
Lactic acid polymers are prepared by the Ring Opening Polymerization (ROP)
of lactide, the cyclic dimer of lactic acid. As shown in Scheme 1, Iactides are
obtained by depolymerizing lactic acid oligomers, which in turn are produced
by the acid catalyzed polycondensation of lactic acid. This scheme is the
method of choice Since ROP leads to fast polymerization rates and excellent
control over the polymer architecture. The key step in initiating ROP of lactide
(Scheme 2) is the nucleophilic attack at the lactide carbonyl by a metal

alkoxide, formed in situ, or directly added to the polymerization. Propagation

H.O)K',OH ———> H~O O\n/Ko
CH3 0 x

CH3
oligomer
l ZnO
A
0 CH3 0 _
H 0 MOR H30 0 0
CH3 0 n o 0 CH3
polymer

Scheme 1. Preparation of polylactide

CH3 0 CH3 0

Scheme 2. Ring opening polymerization mechanism of lactide.

steps extend the polymer chain by an analogous process. The attack by the
alkoxide is affected by the steric hindrance posed by the lactide and the
nucleophilicity of the alkoxide. Although not proven, it is thought that attack by
the metal alkoxide is the rate determining step in lactide polymerization.‘

Lactic acid is chiral, and thus there are three lactide diastereomers, the
R,R (D-Iactide), S,S (L-lactide), and the RS or mesa forrn (DL-lactide). A 1:1
mixture of the R,R and 8,8 diastereomers iS often termed raolactide. L- and
D,L-Iactide are used almost exclusively for polymerization. The physical
properties of polylactides are sensitive to the stereochemistry of the monomers
and the stereoregularity of the resulting polymer chain. Polymerization of the
R,R or S,S diastereomer yields a crystalline polymer with a glass transition
temperature (T g) of 55-64°C,2 while rec-lactide yields amorphous materials“.

Poly(L-Iactide), PLLA, is a semi-crystalline polymer with a melting

temperature (Tm) of 180 °C and a crystallinity of about 70%.5 The T9 is reported

to be 55-64 °C,6'8 depending on the degree of crystallinity. PLLA has good

mechanical properties such as a high modulus and high strength, and for this
reason, PLLA was the material of choice for the development of practical
degradable polymers. For example, PLLA is widely used in orthopedic devices
since it can be processed by thermal processes such as injection-molding and
extrusion. However, due to its high crystallinity, PLLA shows the slowest
degradation rate of all resorbable polylactides.

Polymerization of the R,S diastereomer results in poly(D,L-lactide),
(PDLLA). Because of the random distribution of L- and D-units along the
polymer backbone, crystallization is inhibited and the polymer is an
amorphous, glassy material with a T9 of about 50-60 °C.9 Compared to
crystalline PLLA, the stiffness of PDLLA is Significantly lower, and because the
amorphous structure allows water to diffuse easily into the bulk polymer, its
degradation rate is much faster than PLLA.

2. Polylactide Degradation

The dominant degradation pathway for high molecular weight
polylactides is hydrolysis of the ester bonds along the polymer chain.10 Ester
hydrolysis is a well-known reaction in organic chemistry that can be catalyzed
by acid or base. In the case of polylactide, cleavage of ester bonds generates
carboxyl end groups that autocatalyze further degradation. The hydrolytic
degradation of polylactide is generally regarded as a homogeneous process,
rather than heterogeneous-surface erosion, and includes two stages. The first

stage is characterized by a decrease in molecular weight due to random

hydrolytic ester cleavage, while the second is the onset of weight loss once the
chains are short enough to be mechanically unstable or soluble.11

Several factors favor high degradation rates: a low Tg that facilitates
transport of water in the polymer matrix, a hydrophilic structure that enhances
the solubility of water in the polymer, and the presence of acid or base to
catalyze ester hydrolysis. Since degradation occurs fastest in amorphous
regions, the crystallinity and morphology of the polymer also help determine
the overall degradation rate. In amorphous samples, ester hydrolysis occurs
randomly along the polymer chain and the decrease in the degree of
polymerization can be fit to a random scission model.

Enzymatic degradation becomes important in glassy samples only at the
latter stages of degradation, when short, flexible chains become soluble and
are accessible to enzymes. In contrast, rubbery polymers seem to be able to
degrade polylactide much earlier via surface erosion. Proteinase K,""'13

pronase and bromelain13 were also reported to effect enzymatic degradation of

polylactide.

3. Applications of Polylactides

3.1. Biomedical Applications

3.1.1. General Medical Applications

The driving force for the use of PLLA in medical applications iS that PLLA
degrades in the body to benign products. Good mechanical properties make
polylactides the materials of choice for surgical fixation materials, such as:
sutures, clips, bone pins and plates.14 The degradation rates of polylactides must
be matched to their application so they persist in the body long enough to fulfill
their function as a support. Copolymerization of lactide with glycolide is
commonly used to increase the degradation rate. For example, most resorbable

sutures are either pure poly(glycolide) or glycolide-rich copolymers.

3.1.2. Controlled Drug Delivery

Biodegradable polymers such as polylactide are often used in controlled
drug release systems, often as carriers for drugs. Polymer-mediated controlled
release systems have many advantages of over conventional delivery
systems:15 1) agents can be delivered to tissues in a sustained, continuous and
predicable fashion; 2) they are well protected before being released; 3) site
Specific delivery (such as in brain) can be achieved by simple implantation or
direct injection; 4) repeated drug administration is not necessary.

Many classes of drugs benefit from time-release approaches to delivery.
Time release can provide a near-constant level of therapeutic agents,

maximizing the effectiveness of drugs and minimizing toxicity and adverse

reactions caused by “spikes” in concentration. Controlled delivery strategies
also have important applications in agrochemical fields with the timed-release
of nutrients and pesticides providing obvious advantages. Here again control
over the polymer degradation profile is critical, since it largely defines the rate

of release from the drug carrier.

3.1.3. Tissue Engineering

A major application of biodegradable polymers is tissue engineering, in
which they act as a scaffold for the regeneration of autografted new skin, wound
covers and delivery vehicles for transplanting cells."5'17 Tissue engineering
involves the creation of natural tissue by transplanting cells removed from the
patient or from a close relative and seeding the cells into a scaffold that serves
both as an adhesive substrate and a physical support for the cells. The design
and fabrication of suitable scaffolds from polylactide and lactide copolymers is a
Significant challenge and extensive studies are aimed at evaluating these
polymers as scaffolds for tissue growth.18'19 However, the degradation of the
scaffold must be tightly controlled, and the degradation profile of polylactide is
complicated, partly due to its crystallinity. Copolymerization of L-lactide with

l20 and blending of PDLLA with biodegradable rubbery materials“22 have

glycero
been proposed as partial solutions to this problem.
3.2. Applications as Conventional Thermoplastics

Polylactides are well-behaved therrnoplastics with a reasonable Shelf life

for applications ranging from fibers for clothing to packaging materials.

Because lactic acid is derived from the fermentation of starch, polylactides are

a renewable alternative to petroleum-based polymers. In even slightly moist
environments, PLLA degrades by hydrolysis to environmentally-friendly
products over a period of several months to a year, which compares to 100 to
500 years for conventional plastics such as polystyrene and polyethylene. Its
general physical properties and melt processing ease are similar to
conventional packaging resins. PLLA is a clear, colorless thermoplastic when
quenched from the melt and is similar in many respects to polystyrene. When
plasticized with its own monomer, PLLA becomes increasingly flexible so that a
continuous series of products can be prepared that can mimic polyvinyl
chloride, low density polyethylene, linear low density polyethylene,
polypropylene and polystyrene. Besides mechanical properties comparable to
traditional packaging plastics and biodegradability, another advantage of
polylactide is that it can be produced from renewable sources. Cargill-Dow is
commercializing polylactide production using dextrose from corn.23 Technology
to convert food wastes into polylactide has also been developed.24

To overcome the poor processability of PLLA and its poor impact
strength, the most common composition for high volume packaging resins is
the 90/10 copolymer of L-/D,L lactide. The addition of D,L lactide reduces the
crystallinity of the polymer and improves the impact strength. This material can
be readily polymerized in the melt, can be oriented above its T9, and is easily
melt processable. To modify the mechanical properties, lactide is often
copolymerized with glycolide and e-caprolactone.25 All of these materials are

biodegradable and compostable.

4. Copolymerization
Copolymerizations are polymerization processes in which a mixture of
two or more monomers is polymerized in the same reaction vessel to give a

copolymer, a polymer where each of the monomers is incorporated into the

mM1M1M2M2M2M2M1M2M1M,M1M2M2-vv~
random

WM1M1M1M1M1M1M1M2M2M2M2M2M2~IWW
block

WM1M2M1M2M1M2M1M2M1M2M1M2M1vvvvv
alternate

‘72

9% “6 9’1;
9% 9% 9%
graft

Figure 1. Four generic copolymer architectures.

polymer.26 Copolymerization is important in that it greatly increases the ability
of a polymer scientist to tailor-make polymer products with specific properties.
Copolymerization studies also provide important information about the
reactivities of monomers. AS illustrated in Figure 1, polymers can be
categorized into four general architectures: random, block, alternate and graft,
according to the distribution of the comonomer units along the copolymer
chain. In random copolymers, there is no preference for the placement of
monomers along the chain, and therefore random copolymers should be
homogeneous. Similarly, a perfectly alternating copolymerization yields a

homogeneous product. Both random and alternating copolymerizations

 

produce materials with properties intermediate between those of the parent
homopolymers. In contrast, block and graft copolymerizations generally result
in multiphase materials where each phase exhibits the properties of one of the
homopolymers. This provides an effective way of combining the individual
properties of each polymer in a single material. In addition to the four cases
shown in Figure 1, there are many intermediate cases. For example, most
“random” copolymerizations deviate from true random incorporation of
monomer units and show a preference for the incorporation of one monomer.
When the preference is high, the polymers resemble block copolymers and are
termed tapered block copolymers. In the limit of a large difference in reactivity,
the complete polymerization of one monomer is followed by the polymerization

of a second monomer, and one obtains a true block copolymer.

4.1. Lactide Copolymerization

The copolymerization of lactide with other monomers has been
recognized as an important strategy to obtain materials with modified physical
properties and degradation rates that meet the needs of specific applications.
The four major parameters that can be varied to achieve an expected physical
property are: copolymer structure, molecular weight, chemical composition
(mole fraction of the monomers) and the chemical structure of the
comonomers. A variety of monomers have been copolymerized with lactide
and the next section provides a brief review of the major classes of lactide

copolymers.

10

4.1.1. Copolymerization of Lactide with Lactones and Carbonates

4.1.1.1. Copolymerization with Caprolactone

e-Caprolactone has been copolymerized with both L-lactide and D,L-
Iactide (Scheme 3).27'3° Poly(e-caprolactone ) (PCL) itself is biocompatible and
permeable to many drugs, however, the polymer is hydrophobic and its
hydrolytic degradation rate is very slow. PLA is more hydrophilic, and thus is
more permeable to drugs, but degrades rapidly. The major reason to
copolymerize lactide with e-caprolactone is to combine the permeability of PCL
with the rapid degradation rate of PLA to optimize properties such as
permeability, hydrophilicity, the T9 and crystallinity for use in drug delivery
devices.30 When caprolactone and lactide are added simultaneously to a

polymerization, the poly(lactide-co-e-caprolactone) that is obtained is rarely

a t}: —» ermh

Poly (Iactide-co-e-caprolactone)

Scheme 3. Copolymerization of lactide with e-caprolactone.

random because e-caprolactone is much more reactive than lactide. This leads
to non-random monomer incorporation and tapered copolymers. A well-defined
block structure can be achieved when these two monomers are added
sequentially.3‘Besides sutures and drug delivery systems,” one of the major

applications of poly(lactide-co-caprolactone) is for wound treatment, especially

1]

O O O
o (“\o i 0 Dick
OJ
Ph
A B C D

Figure 2. Examples of lactones that copolymerize with lactide.
for burns.”35 The T9 of poly(lactide) is 50-60 °C, which is too high for the

convenient application of PLA films on the surface of a wound. Incorporation of
a more flexible comonomer such as e-caprolactone is needed to obtain flexible
films with Tg’s near body temperature. It was reported that incorporation of 10-
20 mole °/o of e-caprolactone units in PLA reduced the T9 to 28-38 °C (poly(e-

caprolactone has a T9 around -50 °C).

4.1.1.2. Copolymerization with Other Lactones

Besides e-caprolactone, a number of other cyclic lactones are reported
to copolymerize with lactide (Figure 2) including: 5-phenyl-e-caprolactone
(A),36 1,5-dioxapan-2-one (B),37'38 yhbutyrolactone (C),39 and B-butyrolactone
(D).‘“"41 When polymerized, each of these polymers degrades to benign
products that are harmless to the environment. When used as comonomers,
each monomer tends to impart flexibility and permeability to hard polymers
such as polylactide. The copolymers of lactide with these cyclic lactones have

been proposed for various applications either in medicine or in packaging.

4.1.1.3. Copolymerization with Carbonates

12

The major advantage of copolymerizing lactide with cyclic carbonates is
to introduce the carbonate linkage into the polylactide chain. The carbonate
linkage is expected to be enzymatically hydrolyzable and more hydrolytically
stable than an ester linkage. The most common example is the
copolymerization of lactide with trimethylene carbonate (TMC).42 2,2-
Dimethyltrimethylene carbonate was also reported that have copolymerized

with lactide.”

O i 0
\t/uxo + o o WOWOWO+
O\"/'\ '\/I n O m
0 W C poly (LA-co-TMC)

Scheme 4. Copolymerization of lactide with TMC
4.1.2. Copolymerization with Glycolide
Glycolide and lactide are both based on the 1,4-dioxan-2,5-dione ring,
and structurally, they differ in that the methyl groups of lactide are absent in
glycolide. Because of reduced steric hindrance, glycolide is much more
reactive than lactide and the copolymerization rates of glycolide and lactide are

44.45 and

very different. Glycolide has been copolymerized with both L-lactide
D,L-lactide to give linear, branched and star-shaped copolymers.‘5"“""47 All of

these copolymers tend to be “blocky”.

0*" 75% Mini.

0
PLA-PGA

Scheme 5. Copolymerization of lactide with glycolide.

13

Pure poly(glycolide) is poorly soluble in all common organic solvents.
Similarly, the blocky segments in glycolide copolymer are sparingly soluble,
and thus the “blockiness” is the most important parameter to be controlled
during the copolymerization. The mole °/o of glycolic units in copolymers is
generally limited to ~50 °/o because higher contents render the copolymer
insoluble in common solvents.

Pure PLLA and PDLLA are not well suited for pharmaceutical
applications because their rate of degradation is too slow. However, the
absence of methyl groups on the glycolic units renders the polymer more
sensitive to hydrolysis and accelerates the degradation rate of the copolymer.
Copolymerization is a convenient method for controlling crystallinity in PLLA
and PGA. The melting temperature (Tm) of polyglycolide ranges from 225 to
230 °C48 while crystalline PLLA melts near 180 °C. Copolymerization of L-
lactide with glycolide tends to suppress crystallinity because of irregular
placement of monomer units in the backbone of the copolymer. Even more
effective is the use of DL-lactide as a comonomer. PDLLA is amorphous with
a T9 of about 50-60 °C, therefore the inclusion of DL-lactide can be used to fine
tune the crystallinity in poly(lactide-co-glycolide)s.

Poly(lactide-co-glycolide) copolymers undergo bulk degradation by
random hydrolytic scission of their ester bonds. The lactide segments yield
lactic acid that enters the tricarboxylic acid cycle, whereas PGA segments
produce glycolic acid which is either excreted in urine or enters the tricarboxylic

acid cycle. Because of their degradation to nontoxic products and their tunable

l4

mechanical properties, poly(lactide-co—glycolide) has been intensely studied for
medical applications. They are widely used as suture materials and have been
approved by the FDA for drug delivery.49 In addition, these copolymers are the
starting materials in the production of implantable medical devices, primarily in

orthopedic applications, and as scaffolds for tissue growth.9

4.1.3. Copolymerization of Substituted Glycolides and Lactide

There are only a few examples where lactide has been copolymerized
with substituted lactides or glycolides, partially because of the limited
availability of substituted lactides. The synthesis of a series of alkylated50 and
benzyl substituted51 glycolides were recently reported, and the
copolymerizations of these substituted lactides are being studied. Several

papers were published on the synthesis of poly(lactide-co-mandelide) by the

-H20
HO/lerH + OH I d ' O O
O HO p0 ycon ensatlon O
n O m

0

 

mandelic acid (MA) PLA-co—PMA
Scheme 6. Copolymerization of lactic acid with mandelic acid

O¢ Sn (0002
110°C {hire TAO]:

poly (lactide- --alt -glycolide)

 

Scheme 8. Polymerization of D,L-3-methylglycolide.

15

o o
o o
o O _. o 0
01% i coon n O
O m
o o coon

BMD (R = CHzph) PLA-co-PBMD
DMD (R = (31sz5) PLA-co—PDMD

Scheme 7. Copolymerization of lactide with BMD and DMD

direct condensation of lactic acid and mandelic acid (Scheme 6). Although the
molecular weight of these copolymers were low, inclusion of the mandelic acid
units led to a significant increase in T9 (from 57 °C for polylactide to 77 °C for
the copolymer) suggesting that copolymerization of mandelide with lactide may
be a successful strategy for obtaining polymers with high Tgs.

Another copolymerization example is the polymerization of lactide with
3-(S)-[(benzyloxycarbonyl)methyl]1,4-dioxane-2,5-dione (BMD) and 3-(S)-
[(dodecyloxycarbonyl)methyl]-1,4-dioxane-2,5-dione (DMD) (Scheme 7).
Treatment of the copolymers with alkali forms a carboxyl-functionalized surface
designed to enhance cell binding.52 D,L-3—Methylglycolide was synthesized by
Dong et aI. (Scheme 8).“5 Homopolymerization of this monomer resulted in a
copolymer with alternating Iactyl and glycolyl units in the polymer backbone.
An alternating structure for the polymer was claimed from an analysis of 130
NMR spectra of the polymer, but no physical properties or degradation data

were reported.

4.1.4. Copolymerization with Lysine and Other Amino Acids

l6

One limitation of polylactides in tissue engineering is that the adhesion
and growth of cells cannot be easily controlled since polylactides do not
contain functional groups (other than end groups) that would allow chemical
modification of its properties. Copolymerization of functionalized lactide
monomers can overcome this problem. Langer et al. prepared a monomer that

53'“ Homopolymerization

contained a lactic acid and a lysine residue.
incorporated lysine into the polylactide backbone (Figure 3), enabling a variety
of functionalities to be introduced onto the polymer through reaction at the 2-
amine group of the lysine units. These potential functionalities include peptide

sequences or growth factors that help to improve polymer-cell interactions.

NHCOOBZ

Figure 3. A poly(lactide-co-L-Lysine) derivative.

Another approach to improving polylactide-cell interactions is the direct
copolymerization of lactide with amino acids. Trans-4-hydroxy-L-proline (Hpr)55
was reported to copolymerize with lactide to give the copolymer structure

Shown in Figure 4.

l7

O O
O
IN\+
00 H H m

Cl '
PLA-co-PHpr

Figure 4. Copolymerization of lactide with Hpr.

4.2. Block and Graft Copolymers

Block and graft copolymers that contain polylactide segments can be
prepared by the sequential polymerization of two different lactide monomers, or
by isolating a polymer and using it to initiate a second polymerization. The first
approach is limited to monomers that polymerize by the same mechanism,
while the second allows for the use of two different polymerization
mechanisms. There are many examples where hydroxy terminated polymers
were used as macroinitiators for the polymerization of lactide to give block and
graft architectures. There also are reports of the polymerization of lactide-

based macromonomers to comb-like graft copolymers.

4.2.1. Copolymerization of Lactide with Poly(ethylene glycol)

Poly(ethylene glycol) (PEG), also referred to as poly(ethylene oxide)
(PEO) at high molecular weights, is commonly terminated with chain-end
hydroxyl groups which can serve as initiators for the ring-opening
polymerization of lactide. Using a PEG with only one end terminated with a
hydroxyl group yields a PLA/PEG diblock copolymer,56'61 while a PEG with
both ends terminated with hydroxyl groups yields a PLA/PEG/PLA triblock

copolymer (Scheme 9).62 Jeong et al. prepared thermosensitive PLA-PEG

18

hydrogels that have a temperature-dependent sol-gel transition. These
materials can be used in injectable drug delivery systems. Han and Hubbell
demonstrated the broad synthetic utility of PLA-PEG systems by attaching
acrylate moieties at the end of the chains to form cross-Iinkable systems.63
Biocompatibility is one of the most noted advantages of these materials.
PLA-PEG copolymer systems act as surfactants because the PEG block is
very hydrophilic while the PLA block is hydrophobic. Therefore, when a PLA-
PEG copolymer is subjected to a fabrication process that uses an aqueous
external phase, PEG enriches the surface. The inclusion of PEG in 00polymer
systems can impart extremely beneficial surface properties within the body
because of its ability to inhibit the adsorption of proteins to the polymer surface
and, as a result, prevents many polymer-cell interactions. Thus, PLA-PEG
copolymers are often the material of choice for scaffolds for tissue engineering

and drug delivery systems.

19

. ;
. .
‘19
l
_ :
-.
.
t‘
.. .
. .
~ .
.~£
..
r .
l"
. .
1.
, -
.6
..
».
...
. ..
t
. .-
...
. .
.
1 "
-:- a
.1. s.
, . ..
. .
..
4.;
. . .3
.- . ‘
~ . - _
T 4‘-
.‘d.
-
'O
s - .
. .
- . _
.. --
'1.-.
. - _
...,.,.'
~.' a
—.'.
g0.
. -
. . .
.1 .-
.3 -
. [a
-.4 ,
. .
. .
'AIVI.
. . -
.
4 -. o
i"
.2
I‘.
§.:'
.

I 1
‘Q T‘—
U

. ...
., ;~_.
‘ C

.a.::'.~...:r.rr.ttf.7
.iira-,» .‘r
!

.IIE;14.5[:,:601-

.‘hp-lt

wultimatumtransmitters-.

 

jgjljmzn‘muv- -' I”: _'
U

HOwOfVOVOCHB HO\/\O(/\/09/\OH
m m

 

O o 0
WWW ovwvwl
the is... ,.,
PLA—PEG PLA-PEG-PLA
O o
0 Z_\
0 > multi-block copolymer
0

Scheme 9. Synthesis of PLA-PEG copolymers

4.2.2. Graft Copolymers from Natural Polymers

Cellulose is the most abundant renewable biomass in the world,
however, its biodegradability and bioabsorbability is very poor. By using the
multiple hydroxyl groups of cellulose as initiators, this natural polymer can be
converted to poly(cellulose-g-lactide). Materials obtained from lactide and
cellulose combine the advantages of both and can be used in tissue

64,65

engineering. Similarly, poly(saccharide-g—lactide) materials were

synthesized and characterized.66 Polysaccharides have the advantage of

20

targeting carbohydrate-binding proteins located on mammalian cells.
Therefore, this type of copolymer has potential as gene delivery vehicles with
the capability to target specific cell types. Starch/lactide graft copolymers were
synthesized by You by reacting lactide with corn starch. Since both starch and
polylactide are degradable, this graft copolymer can be completely degraded
by acid, base and enzymes. Lactide also has been copolymerized with

comonomers such as xylofuranose,67 sorbitol,68 1,3-propanediol,22 and

mannitol.69

4.2.3. Lactide Copolymers with Backs from Vinyl and Acrylic Monomers
Using the macroinitiator approach, lactide has been copolymerized with
polybutadiene,7o polyisopropene,71 polystyrene,72 and poly(2-hydroxyethyl
methacrylate) (HEMA).73'75 These copolymers usually have either block (with
vinyl monomers) or graft architectures (with acrylic monomers). Scheme 10
illustrates the preparation of a poly(Iactide-b—isoprene-b-lactide) triblock
copolymer through a combination of living anionic polymerization (isoprene)
and controlled ring opening polymerization (lactide). Since the polyisoprene
segment is amorphous and has a low T9, PLA is toughened by the
polyisoprene. The resulting block copolymer is a degradable thermoplastic
elastomer that can be used in packaging applications. Using the same method,

poly(butadiene-b—Iactide)70 was synthesized for similar applications.

21

.. wafi- "2.1.431Q5'flgflm- ...;

 

.... ....t -. . . ... .. .~ . ... 2 . > - < ‘ . ’ ' _. . ~.
. ‘ - A... - ( ‘ -.. p , . » d“. _.. ...l .4. a." -u I . ' . ‘1 '. , W . _

It: '.I' r ..r 'rtzv..-...:.'.:t:;~.- mum r r' ' w 1.1: a W“." .. ..~:.'.- ' - ‘-. ,. _- .1 -

.J’o'I mm", JNO'IHII'53‘V.’ .1 fi'...’l.'¢‘l ‘I~ ' ' ‘-' fl” .‘ "fine" 1' w an. ‘K fd-Irda 0 «down ..uh "“‘ "‘5'" “ ‘

. 5,0 . , o I ' .. I t 0'" ‘Ju'C 9 Nil " C ‘ "9 1‘, ‘. “I. I . i ~.‘lk. N.“ .‘I'||!vi I.“'V!O' I"-« '

\Sl/ g
t‘BU/ \O/\/\ Li + /

TB DSOPLi

b) ethylene oxide

a) cyclohexane, 45°C, 4hr
l c) H+

TBDMSO-CHz-Pl-CHz-OH

l THF, TBAF, rt, 24h

HOWOH

HO-CHz-Pl-CHz-OH

a) toluene, AIEt3, rt, 2hr
b) lactide, 90 °C. 20hr

 

c) MeOH
O O
HFWNWMOJVQH
n m
PLA-PI-PLA

Scheme 10. Synthesis of poly(lactide-b-isoprene-b-lactide)

A common strategy used to prepare comb-like graft polymers is to use
an end-functionalized macromonomer to start a second polymerization. As
shown in Scheme 11, HEMA terminated PLA, a macromonomer, can undergo
radical initiated polymerization to yield poly(HEMA-graft—lactide). Since the
HEMA segments increase the hydrophilicity of the copolymer, this degradable
graft copolymer can be used as a hydrogel for the controlled delivery of

macromolecular protein drugs.73

22

2.0%... fa... amok M...

 

 

HEMA polylactide macroinitiator
1 AIBN
PHEMAWPHEMAMPHEMA
PLA PLA PLA

%

poly (HEMA-graft-lactide)

Scheme 11. Synthesis of poly(HEMA-graft-lactide)

4.2.4. Copolymerization with Other Comonomers
Poly(lactide-co-amide)s are an emerging class of degradable graft
copolymers, which combine the mechanical properties of polyamides and the
biodegradability of the polylactides.76 Lactide polymerization was initiated with
N-isopropylacrylamide" and N,N-dimethylacrylamide.77 These copolymers
have been reported to be effective in drug delivery systems that target solid
tumors.
4.3. Copolymerization Characterization
The most important structural information needed to characterize
copolymers include the determination of the molar composition of the
copolymers and the distribution of the comonomers in the polymer backbone.

1H NMR is the most popularly used method to analyze the composition of

23

"" ""‘e!;'_>'13.fl‘

7".

'43..

“aflimw

~‘ -" "' .5 in”: I; ~‘ ":7? $3.23
i . l.
. '. ..é'xts’iif. - trfifigfl'fiflgw ”

 

fiiiilxali O , 0i"- :
’0 0|

~.-fuu u
a! I I I

03"” ’ ”2:,“
'f . ’. ” J3,

fir'.‘

lactide copolymers,:M‘r’J8 because the methyl and methine signals of the
lactide unit in 1H NMR spectra are generally well-separated from resonances
from most of the monomers that copolymerize with lactide. Near IR
spectroscopy has been used to determine the composition of poly(lactide-co-
glycolide),79 Raman spectroscopy for poly(lactide-co-s-caprolactone),80 and
HPLC and MALDl-TOF for poly(ethylene oxide-b-L-lactide).81 However, these
methods cannot be considered to be general analytical approaches since they
have only been applied to copolymers on a case by case basis.

Among all lactide copolymerizations, the most thoroughly analyzed
cases are copolymers of lactide with e-caprolactone and lactide with glycolide.
Quantitative 13C NMR is best suited for analyzing the sequence distribution of
these copolymers since the carbonyl signal is very sensitive to the sequence
effects. From high resolution 130 spectra, tetrad and even pentad peaks can be
clearly detected, and after assignment, the intensities of these peaks can be
used to calculate the average block length of each component. Poly(lactide-co-

”'83 are block or tapered

e—caprolactone)” and poly(lactide-co-glycolide)
copolymers due to the large difference in the monomer reactivity for the

lactide/caprolactone and lactide/glycolide pairs.

24

4.4. Kinetic Studies of Copolymerization

4.4.1. Monomer Reactivity Ratios

The composition of a copolymer is usually different from the composition
of the comonomer feed from which it is produced, since some monomers are
more reactive to copolymerization and some are less.26 Staudinger and
Schneiders observed that the relative tendencies for monomers to
copolymerize often bore little resemblance to their relative rates for
homopolymerization. Thus the composition of a copolymer cannot be
determined directly from the homopolymerization rates of the two monomers.
There are several models that can be used to analyze and predict copolymer
composition, with the best known being the first-order Markov or terminal
model of copolymerization. The principal assumption of this model is that the
chemical reactivity of the propagating chain in a copolymerization depends
only on the identity of the monomer unit at the end of the growing chain; the
chain composition preceding the last monomer unit does not influence the
reactivity. For the copolymerization of two monomers, M1 and M2, the first-
order Markov Model describes four possible steps for chain growth, since both
M1 and M2 can each add to a propagating chain ending in M1 or to one ending
in M2, as elucidated in Scheme 12. There are four rate constants for the
processes, km for adding monomer M1 to a propagating chain ending in
M(Mfi), k12 for adding monomer M1 to a propagating chain ending in M(M2*),

k22 for adding monomer M2 to a propagating chain ending in M(M2*), and k2, for

25

 

adding monomer M2 to a propagating chain ending in M(M1). The processes
with rate constants kn and k22 describe hom0polymerization or self-

propagation; while k12 and k2, describe cross-propagation.

M“ + M .——k—“—» M,*

k
M1* + M2 12 . M21:

k
M; + M —-?-’-> M1* (1)

k
M2,. + M2 —é2_> M2,!

Scheme 12. Chain propagation steps in the copolymerization of
two monomers.

From these four kinetic steps, the copolymerization equation can be

denved:

dlMil _ [M1](r1[M1]+[M2])

_ (2)
OIM2] [M2]([M1] + r2[M2])

 

 

where the parameters r1 and r2 are termed the monomer reactivity ratios of M1

['1 = k11/k12 and r2 = kgélk21 ‘3‘

and M2, and are defined as
The significance of the copolymerization equation is that it relates the
copolymer composition oIM1l/OIM2] to the concentration of the two monomers

in the feed, [M1] and [M2].

26

4.4.2. Types of Copolymerization Behavior
The monomer reactivity ratio (r) is very useful since it can be used to

define different types of copolymerization behavior.

4.4.2.1. Ideal copolymerization: r1r2 = 1

A copolymerization is termed as ideal when the r1 r2 product is unity. An
ideal copolymerization occurs when the two types of propagating species M1*
and M2* show the same preference for adding one or the other of the two
monomers. Under these conditions, the relative rates of incorporation of the
two monomers into the copolymer are independent of the identity of the unit at
the end of the propagating species. Most ionic copolymerizations are
characterized by the ideal type of behavior.

When r1 = r2 :1 the two monomers show equal reactivities toward both
propagating species. The copolymer composition is the same as the
comonomer feed with random placement of the two monomers along the
copolymer chain. This behavior is referred to as random or Bemoullian. For the
case where the two monomer reactivity ratios are different, that is r1 > 1 and r2
< 1 or r1 <1 and r2 > 1, one monomer is more reactive than the other toward
both propagating species. The copolymer will contain a larger proportion of the
more reactive monomer placed randomly along the chain. Therefore, a very
important practical consequence of ideal copolymerization is that it becomes
progressively more difficult to produce copolymers containing appreciable
amounts of both monomers as the difference in reactivities of the two

monomers increases. Only when r1 and r2 do not differ substantially will there

27

exist a large range of comonomer feed compositions, which yield copolymers

containing appreciable amounts of both monomers.

4.4.2.2. Alternating Copolymerization: r1 = r2 = 0

When r1 = r2 = 0, the two monomers enter into the copolymer in
equimolar amounts in a nonrandom, alternating arrangement along the
copolymer chain. This type of copolymerization is referred to as alternating
copolymerization. Many radical copolymerizations show a tendency toward
alternation. The behavior of most comonomer systems lies somewhere

between the two extremes of ideal and alternating copolymerization.

4.4.2.3. Block Copolymerization: r1 >1 , r2 >1

If both r1 and r2 are greater than unity there is a tendency to form a block
copolymer in which there are blocks of both monomers in the chain. This type
of behavior has been encountered in only a few c0polymerizations of olefins

initiated by Ziegler-Natta catalysts.

4.4.3. Experimental Evaluation of Monomer Reactivity Ratios

The experimental procedures for evaluating r1 and r2 involve
determining the copolymer composition for several different comonomer feed
compositions. Copolymerizations must be carried out to low degrees of
conversion (ca. 5%) to minimize deviation of the monomer ratio from its initial
value. The copolymer composition is determined either directly by analysis of
the copolymer, or indirectly by analysis of comonomer feed.

Several methods have been used to extract monomer reactivity ratios

from the copolymer composition data. Among them, the most widely used are

28

the Fineman-Ross84 and Kelen-Tudos methods;35'86 both are schemes to
linearize the experimental data. However, the more statistically sound methods
for analyzing composition data are nonlinear methods, which involve plotting
the instantaneous copolymer composition as a function of the comonomer feed
ratio for various feeds, and then using an iterative process to determine the r1
and r2 values that best fit the data.“88

In general, both the linearization and nonlinearization methods are
based on the differential form of the copolymerization equation (equation 2)
which requires that the conversion be held as low as possible to assure that
the monomer ratios does not differ much from the feed ratio. In practice, there
are limitations since one must be able to isolate a sample of the copolymer for
analysis. This requirement can be hard to meet, especially when the

reactivities of the two monomers are quite different. This issue will be

discussed further in the “Results and Discussion” section.

4.4.4. Monomer Reactivity Ratio Analysis of Lactide Copolymerizations
To date, three lactide copolymerization systems have been evaluated
kinetically, the copolymerization of lactide with glycolide,89 the copolymerization
of s-caprolactone with lactide,31 and the copolymerization of 3-methylglycolide
with lactide.90 The Fineman-Ross and Kelen-Tudos methods were used for
data analysis in these kinetic studies. For the copolymerization of lactide with
glycolide, the experimentally determined reactivity ratios were r1 (lactide) =
0.22, and r2 (glycolide) = 3.4. Therefore, lactide is less active toward chain

propagation as can be predicted by considering the steric hindrance caused by

29

 

 

the methyl group. In the copolymerization of lactide with 3-methylglycolide, r1
(lactide) = 0.92, and r2 (3-methylglycolide) = 1.71. Again a simple steric
argument provides a reasonable explanation of the data. Note that difference
between the r-values for these two comonomers is not as large as was seen
for the lactide/glycolide pair.

At first glance, the copolymerization of e-caprolactone and lactide might
seem confusing. Although the homopolymerization of s-caprolactone is much
faster than lactide (k22>k11), r1 (lactide) = 17.9 while r2 (caprolactone): 0.6. This
result can be explained by consideration of the growing polymer chain.
Compared to e-caprolactone, the alkoxide chain end derived from lactide is
less basic due to the proximity of the carbonyl group. Thus, the lactide chain
end should be less reactive. The slowest rate constant in the copolymerization
is km, the addition of caprolactone to the lactide chain end. This leads to the

overall value of r1 (k1 1/k12) being greater than r2 (k22/k21).

3O

 

EXPERIMENTAL

5.1. General

Unless otherwise specified, ACS reagent grade starting materials and
solvents were used as received from commercial suppliers without further
purification. Proton nuclear magnetic resonance (1H NMR) and carbon nuclear
magnetic resonance (“’0 NMR) analyses were carried out at room temperature
in deuterated chloroform (CDCI3) on Varian Gemini-300 and Varian Vax-500
spectrometers with the signals from residual protons in the solvent being used
as chemical shift standards. Mass spectral analyses were carried out on a VG

Trio-1 Bench top GC-MS.

The molecular weights of polymers were determined by gel permeation
chromatography (GPC) using a PLgel 20m Mixed A column at room
temperature with THF as the eluting solvent at a flow rate of 1 mL/min. The
peaks were detected using a Waters R401 Differential Refractometer and a
Varian 2050 Variable Wavelength detector set at 254 nm. The concentration
of the polymer solutions used for GPC measurements was 1 mg/mL, and the
GPC molecular weights are reported relative to monodisperse polystyrene
standards. Differential scanning calorimetry (DSC) analyses of the polymers
were obtained using a Perkin-Elmer D80 7. Samples were run under a helium
atmosphere at a heating rate of 10 9C/min, with the temperature calibrated with
an indium standard. Gas Chromatography (GC) was carried out on Perkin-

Elmer 8500 with a fused silica capillary column (SE-54-60W, J & W Sci. Inc.).

31

 

 

High Pressure Liquid Chromatography (HPLC) data was collected from Waters

600 system equipped with a Waters 996 Photodiode Array Detector

5. 2. Monomer Preparation and Synthesis
3,6-Diethyl-1,4-dioxane-2,5-dione (ethylglycolide). A mixture of 10 g
of 2-hydroxybutyric acid and 0.2 g of p-toluenesulfonic acid in 700 mL of
toluene was heated at reflux for four days, with the water removed
azeotropically using a Barrett trap. The toluene solution was then cooled,
washed with sat. NaHC03, and dried over MgSO4. After removing the toluene,
0.1 g of ZnO was added and the residue was distilled under reduced pressure
using a Kugelrohr distillation apparatus. Ethylglycolide was collected at 180 °C
(100 mtorr), and was dissolved in the minimum amount of ether needed to
dissolve the product. The solution was cooled to -30 °C, and petroleum ether
was added drop-wise until the solution turned cloudy. The colorless crystals
were collected by cold filtration and dried under vacuum to give 5.2 g (63%) of
ethylglycolide as colorless oil. 1H NMR indicates that the product is a statistical
mixture of diastereomers. 1H NMR (300 MHz, CDCla): 5 4.88 (dd), 4.83 (dd, 1H
total for the signals at 4.88 and 4.83), 2.08 (m, 2H), 1.15 (tt, 3H). “’0 NMR (75
MHz, CDCla) 8: 167.62, 166.58, 76.4, 75.6. mp 19.5 - 20.5°C.
3,6-DiisopropyI-1 ,4-dioxane-2,5-dione (isopropylglycolide).
lsopropylglycolide was synthesized by the procedure described by Yin.1 1H
NMR (300 MHz, CDCI3): 8 4.95 (d, 1H), 2.30 (m, 1H), 1.03 (d, 3H), 1.01 (d,
3H). "’0 NMR (75 MHz, CDCI3) 6166.38, 79.55, 29.36, 19.55, 15.81. MS (m/z)

201.4 (M+1), mp 137-138 °C (111.91 (136 °C).

32

w

l . I. '1 a. “uni-5‘. uni-'u’“ "-" ‘ “ ' ‘ n . - ..-.. .
”tr-“Mi: " ”3.4m 10 2' “-3: H." ' ‘. H """9'" w Jr~.r"“ A :lnn‘gnfot. "...".3. 1‘ Jv . - H... 1 .-. .
,g "11"” |,.,; ‘1‘}, ..un'.~. 4 .u . 'ucd.’ {Mann . ' "‘ "t 4‘“ a. . fl.“ ... 1.3.4. up» "1‘" .. ‘
.1 . 051 . .... - ”right-100! ‘ 4 I _. ' 4,, «1,1. .vt.o_.~oi«u-n.
. .

a! l."-
..u 0er
“I‘m 01

D P" ] -.,',,-¢3~ ‘yW ‘l. r". a." “1;— ‘ ""L‘”: f .r' ...A'IN.’I‘“;:.
W». , WWMe'fiffllmmlksfffiuym
1‘

 

3R,6R-Diethyl-1 ,4—dioxane—2,5-dione (R-ethylglycolide) R-
ethylglycolide was prepared by the procedure described by Yin.1 1H NMR (300

MHz, 00013): 6 4.85 (dd), 2.03 (m, 2H), 1.08 (tt, 3H). m.p. 120.5-122 °C.

5. 3. Copolymerization

General conditions for the bulk copolymerization of racemic lactide
with substituted glycolides. Solvent-free polymerizations were carried out in
sealed tubes prepared from 3/8 inch diameter glass tubing. Lactide and
substituted glycolides were added to the tubes to give the desired feed ratios,
and then were transferred into a dry box. Catalyst (Sn(Oct)2) and co-initiator
(tert-butyl benzyl alcohol (t-BBA)) were added to the tubes to at a mole ratio of
1 (catalyst) to 400 (monomer). The solvent was removed under vacuum, and
the tubes was sealed and immersed in an oil bath at 130 °C (at 165 °C for
copolymerization of lactide with isopropylglycolide). At the end of the
polymerization, each tube was cooled in ice, opened, and the polymer was
dissolved in CHCI3. A portion of the sample was evacuated to dryness and

analyzed by GPC, NMR and DSC.

5. 4. Copolymer Composition: Direct Determination

Copolymerizations were quenched at low conversion (<10°/o), and then
the mixture of monomer and copolymer was dissolved in CHC|3 and passed
through a short column prepared from a 9-inch long Pasteur pipette filled two-
thirds with silica gel. Pre-weighed 20-mL vials were used as receivers. The

eluent that collected in the vials was evaporated to dryness until the weight of

33

vials remained constant. 1H NMR showed that the recovered material was free
of polymer.

After the separation was complete, the copolymer adsorbed onto the
silica was transferred to a 25 mL round bottom flask. To this flask were added
Sn(Oct)2 (200 mg) and methanol (15 mL). A condenser was added and the
slurry was heated at reflux for 48 hrs. The reaction mixture was then
centrifuged for 40 minutes, and the clear solution was decanted into a 25 mL
round bottom flask. The flask was cooled in an acetonitrile-dry ice bath at -20
°C, and the methanol was evaporated under vacuum over a period of 2 hours.
The composition of the resulting methyl esters was then analyzed by GO or 1H

NMR.

5. 5. Biodegradation

About 50 mg of polymer sample was accurately weighed and chopped
into 1mm by 1mm pieces. The chopped sample was immersed in 15 mL of
0.1M phosphate buffer solution (pH=7.4). The degradation was carried out at
55 i 0.1 °C. After a predetermined period of time, samples were removed from
the buffer solution, washed thoroughly with distilled water and dried under
vacuum. The samples were weighed to determine the weight loss and the

molecular weight was determined by GPC.

34

..., (gun-r - u .. '1" -‘ ' f., "' .' -. .. . V
whom-001.511 ..' ur -‘- v . '4'," n .3.. ...-F . a. - .' -
5‘... . . u-n LU!" ‘ n3 "_“M’,” _ ..u. 4....1 4....

I .15. 1.. "'11. . rv uu‘o ... - ‘

I5‘01.“'lrft-ItyO'lr‘v1“It.:::I: '. . I, ‘ . . 1 .

..... 1...!
fi...o.<l 0'
' 3 MT]

. + ’1»? 73""“9-“11‘4diftfr'”:1t'nt‘bu '1 52:.-
halfalfzinsénmmr-s;s~5.6.581...»4.9.8,
c a . - - . 94 . . O I In 1 I

1
If.”

 

8.3918?

'inwlff-“P’I'Jér ”MWWW‘

RESULTS AND DISCUSSION

6.1. Copolymerization of Lactide with Ethylglycolide

3032 are an important class of

Poly(lactide), PLA, and its copolymers
biodegradable and biocompatible polymers with numerous applications in
medicine and as packaging materials, but a major deficiency of polylactide is
its limited range of physical properties. Therefore, an important area of
polylactide research is the preparation of copolymers and derivatives of
polylactide that extend the range of properties that can be obtained from
polylactide. A number of “random” and block copolymers have been prepared
to date, primarily with glycolide and 8-caprolactone. Yin explored another
strategy for obtaining new properties from polylactide, the synthesis and
polymerization of lactides derived from substituted lactic acids.50 Copolymers

prepared from these new monomers should provided new families of

copolymers and provide properties not currently available from PLA.

The copolymerization of lactide with ethylglycolide was briefly examined
by Yin.1 Since ethylglycolide (EG) and lactide (LA) have very similar structures,
their copolymers should retain the biodegradability of the homopolymers, but
have different physical properties. Yin used bulk polymerization at 130 °C to
synthesize a series of copolymers (EG/LA=1/5, EG/LA=1/3, EG/LA=1,
EG/LA=3, EG/LA=5) as illustrated in Scheme 13. The polymerizations were
catalyzed by Sn(Oct)2 with t-butylbenzyl alcohol (BBA) as the initiator. All of the

copolymers had Mn >40,000 g/mol. The thermal and molecular properties of

35

 

O 0
\(‘Lo V0 Sn(OCt)2
+ ¢ mixture of polymer and monomer

0% 0 t-BBA
o o 130 °c

 

rac-lactide rec-isopropylglycolide

Scheme 13. Copolymerization of racemic lactide with ethylglycolide

these polymers are summarized in Figures 5 and 6 and in Table 1. The DSC
plots (Figure 5) show one glass transition temperature (T9) for each
copolymer, an indication that the copolymers are not phase separated.
Considering the structural similarity of the monomers, it is not surprising that
copolymers are single phase. The glass transition temperatures listed in Table
1 fall between those of polylactide and poly(ethylglycolide) and increase with
the increase of the mole fraction of the lactide comonomer. The glass transition
temperatures of copolymers can be predicted using the Fox equation:
1 WA WI,

—=——+— (4)
T Tg, Tg,

x
where T9, T9,, and T93 are the glass transition temperature of the copolymer
and the pure homopolymers derived from monomers A and B respectively, and
w represents the mass fraction of the polymer. The Fox equation successfully
predicts the glass transition temperatures of random copolymers and
plasticized systems. An underlying assumption is that the polymer is
homogeneous in composition, and that the comonomers are distributed

randomly along the chain. As shown in Figure 6, the glass transition

36

 

 

Polylactide

_o——v—-'—’-'/—_'

EG/LA=1/5
_fi/fi—

EG/LA=1/3
d—ld/WN"

EG/LA=1 _ff/E—h

 

endo ——>

 

 

 

Poly(ethW
-50 O 50 1 00

Temperature ( °C)

Figure 5. DSC runs (second heating after flash quenching from 100 °C)
for lactide and ethylglycolide copolymers.
Heating rate: 10 °C/min under helium.

temperature of the copolymers fit the Fox equation fairly well. However,

whether the copolymer is truly a random copolymer or not must be defined by

l

.‘t

a characterization technique such as NMR,45 which is sensitive to structure at

M

the molecular level.

v-v-iu - 4' '~‘-r1" , ,'- '1 u‘ _ v'»; , . , .‘ _..,..:, . »‘;..‘¢1.1.Vu'."" .... -.. .
A mi “33: '1 u ' "V“ .31 i—‘c ~‘31 ‘3 z“; 3* 1‘ “ 1,1,”,37', :IT' ,. ".1.. ,,. .q, . .. .-. 1.1.. 1,. .1 { ‘ . ,. _,' ‘
' l‘ a h ‘ 0’ I '1 ‘ l. " V 1‘ an; ' ... I If) I ' P .1 . t'" '3'“ ’ """'” ' ' . " “I lit! ‘4': ’ 1
“”._.., ' ..., ’ ' I “N. ,y o, '. IO)! '4. - g .0 v o. 5' +. - "n- “ .' r." _.. 1 ,1 but ..-1 - v f.» I '
. «o' .. - ... .3... or 0401’! pl ill-'"'.1"‘ "" - ' "" ‘ ... 1. ”H” nt-t-S‘L'H” “

L
..l

l

.. -. man-:88answtififiiiittif-‘éfifiw

Based on copolymerizations of lactide and glycolide,”94 the distribution
of monomer units in the polymers is probably far from random because of

differences in reactivity for the monomers. As illustrated in Scheme 14, these

37

 

 

7O

 

 

 

 

50 ‘ curve from
A Fox equation
9
I
I-

30 -

1 O I I I I

0 0.2 0.4 0.6 0.8 1

W1/(W1-I-W2)

Figure 6. Glass transition temperatures of lactide and ethylglycolide copolymers.

effects are likely due to differences in the steric bulk at the carbon 01 to the
carbonyl. Lactide and glycolide share the same ring structure, but the absence
of lactide’s methyl groups makes glycolide far more reactive and thus “blocky"
polymers are obtained from copolymerizations. However, transesterification

reactions can increase the randomness of copolymers.

The results from the copolymerization of lactide with ethylglycolide show
that the physical properties of the copolymer can be readily tuned by adjusting

the monomer feed ratio. Nevertheless, these sterically hindered lactides should

38

be less reactive than lactide. Thus a batch copolymerization where all of the
monomer is charged into the polymerization at one time will give a polymer
whose composition varies with conversion. Special steps must be taken to
compensate for changes in the monomer feed ratio and maintain a constant
rate of addition of the comonomer into the polymer structure. To gain the
control over the copolymer structure and achieve certain physical properties in

the copolymers, we need to evaluate the kinetic parameters of these

 

 

 

 

 

 

 

 

copolymerizations.
Table 1. Properties of copolymers
Polymers Mnx10‘°' a Mw/Mn T9 °C b
Polylactide 35.2 1 .89 66
EG/LA=1/5 41.5 1.85 42
EG/LA=1/3 47.8 1 .76 38
EG/LA=1 41.7 1.88 30
EG/LA=3 50.8 1 .92 23
EG/LA=5 43.6 1.85 19
Poly(ethylglycolide) 45.6 1 .78 15

 

 

 

 

 

 

a. measured by GPC in THF using polystyrene as standard
b. measured by DSC under He at a rate of 10 °C/min

39

6.2. Kinetic Evaluation of the Copolymerization of Lactide with

Ethylglycolide

6.2.1. Copolymer Composition Determination.

As illustrated in the Introduction, a copolymerization system with two
comonomers can be defined by the reactivity ratios of the two monomers, r1
and r2. Experimental determination of these values requires precise
measurement of the copolymer compositions (F1 and F2) that result from
several different comonomer feed compositions. The copolymerizations must
be terminated at low conversions to avoid significant changes in the monomer
feed ratios, and thus the main experimental challenge is to accurately

characterize the composition of small polymer samples.

6.2.1.1. 13c NMR

13C NMR can be used to analyze the conversion for each monomer and

 

 

RO‘
less more
hindered hindered
=(— O O O
O O O

polymerizes

preferentially
L l l

glycolide rich lactide rich

Scheme 14. Reactivity differences leads to “blocky" copolymers.

40

41'. ., Ao"|.'c.l.vl1"a
pi“ (_ 1* . .1., .. gu r A" u ' '
" Colr' " I
‘ h I

w b , . . .5 .
'15:! 1w '.
”’ff":l;;3‘€f:

iii Nu

it"q’lj: .0:

"I A ""',:
1.1a

-' .‘f |;'ll ' ll Int-III "' .7
cu 'u-vv . .-
‘ 1

1i

_ ' ‘ 1 'w’. y
1' .

 

fl-

“8’11-

5' r:i..'. .3
’1' mix"

F")l

1
.15

1'2“» I r

y’QJI
1:." 0.1.51

the copolymer composition since the chemical shifts of the monomers and the
lactide and ethylglycolide residues in the copolymer are well-separated. The
principal drawback of this method is that the spin-lattice relaxation times (T 1) of
carbon atoms in polymers are usually long and the longest T1, that from the
carbonyl group, was measured as 30 sec. Since each NMR scan requires
5x30 sec of acquisition time, and the polymer solutions were relatively dilute,
8-hour runs were required to obtain a signal to noise ratio that would provide
reliable quantitative integration data. Given the large number of samples, this

proved to be impractical.

6.2.1.2. Normal Phase High Pressure Liquid Chromatography (HPLC)
HPLC is an alternative method for analyzing copolymer composition.
However, HPLC analysis is an indirect method in that comonomer composition
is obtained by analyzing the change in the monomer composition after
polymerization. The conversion must be maintained below 10 °/o to avoid
significant changes in the initial feed ratio during the reaction, and thus the
HPLC method is based on measuring the difference of two large numbers.
The best conditions for separating the monomers was to use CHCI3 as the
mobile phase at a flow rate of 1 mL/min. Biphenyl was used an internal
standard. The response factors for the monomers and the standard were
calculated from the results from a solution of lactide 7.6 mg (0.053 mmol) and
ethylglycolide 9.0 mg (0.052 mmol) dissolved in 1mL of a solution of biphenyl
in chloroform (6.75x10'3 M). Using the optimized conditions, four peaks were

observed, the two ethylglycolide diastereomers (4.6 min, 5.9 min) followed by

41

 

lactide (9.2 min). The results in Table 2 show that the HPLC method gives
unreasonably high conversions for both lactide and ethylglycolide, considering
the copolymerization time was limited to 30 seconds. One potential problem is
poor recovery; Le, a fraction of one or both monomers was retained on the
column or hydrolyzed and retained on the silica packing. To test this
hypothesis, we prepared a chloroform solution containing known amounts
lactide, ethylglycolide, and biphenyl. This sample was injected into the HPLC,
collected, and re-injected. After the 1st injection, the integrated areas of the
ethylglycolide and lactide peaks relative to the biphenyl standard were 0.79
and 0.98, respectively. After the second injection, the ratios were 0.07 and
0.20. Assuming no loss of biphenyl, the recovery of lactide and ethylglycolide
were 9% and 20%, respectively. Thus, normal phase HPLC is unsuitable for
analysis of the monomer composition. Since this problem is most likely caused
by the acidic hydroxyl group of the silica packing, we turned to reverse phase

HPLC.

42

f . ‘< '1 v‘ “m, b” ,y V 4"' .. "‘W “'Jv Ira ‘-
. . ' ”Q .J firth, -01).):IKHE )1
- . . -' J L_""1xi: ‘

 

 

1.13.; -

5.5-» -‘ 1.... '.1.-. v 0" ' "1' -“ _ '
3.1.;ilginr, .- u. n .. -o-.1‘“”':’ 'g' .° H ‘ “"‘ ’
- _ . .. . . ' ~ 1

" "" '-",1:" .....in g i..." "iuu- .

‘*' _..v' _- , -}1 .. ,af‘il.
(.“fiii‘étfi'-g$;id$}3{3 1&3: ' 1 ri'h_'-1:.3~'A-1»I:;=oi
a “UN "1‘1““: .10.». p.” '1‘ l )4 «g 5le ,

. «flue $53 Influx-1“ ‘4’ '

I m“

' ‘tgr "

“1
A“

to
K

Table 2. Determination of Conversion in the Copolymerization of Lactide

(LA) and Ethylglycolide (EG) by Normal Phase HPLC

 

 

 

 

 

 

 

 

 

 

 

 

 

feL:d LA ( LC) LA EG feed EG (LC) 56
(mg) (mg) “M" (mg) (mg) ”mm
7.6 4.9 35 1.0 0.14 86
2.7 1.1 19 0.7 0.03 95
10.4 7.5 28 1.4 0.16 85
23.8 12.7 47 6.3 0.23 96
26.5 10.8 59 14.0 0.20 98
9.4 7.3 23 9.1 0.40 97
6.7 2.6 60 11.9 0.31 98
5.6 .3 95 6.8 0.28 93
4.4 1.2 98 9.7 0.43 96
3.0 1.5 95 14.9 0.76 95

 

 

 

 

 

 

 

43

6.2.1.3. Reverse Phase HPLC

Reverse phase HPLC was run in 1:1 acetonitrile/HZO at a flow rate of
0.5 mUmin. The stationary phase was Nova—pak C18, and anisole was used
as the standard. Response factors were calculated from the results from an
injection of a standard solution of 10.0 mg lactide (0.069 mmol) and 19.2 mg
ethylglycolide (0.112 mmol) dissolved in 1 mL of a 0.012M solution of anisole
in 1:1 acetonitriIe/HQO. As expected, the compounds eluted in reverse order
compared to the normal phase HPLC results. The first compound eluted was
lactide (4.2 min), followed by the two ethylglycolide diastereomers (5.6 min, 6.1

min), and finally anisole (9.1 min). The results are summarized in Table 3.

Table 3. Monomer Determination by Reverse Phase HPLC

 

 

 

 

 

 

 

G

at...“ 5.29;”: 22.... 2...... 1 (w .. M)
3-4 0.6 4.6 7.8 9.8 0.14 0.12
5-4 1.2 2.1 14 14 0.41 0.43
54 1.8 1.9 17 6.3 0.53 0.66
7-4 4.1 3.7 13 7.5 0.57 0.70
8-4 2.8 1.9 13 5.7 0.64 0.77
9-4 4.5 1.2 11 8.5 0.82 0.88

 

 

 

 

 

 

 

 

 

 

6.2.1.4. Kinetic Evaluation Based on Reverse Phase HPLC Results

The premise of the first order Markov model for copolymerization is that
only the repeat unit at the end of the growing polymer chain plays a role in
determining the next monomer to be added to the chain. In a copolymerization
of lactide (monomer 1) and ethylglycolide (monomer 2), k" and k2,? are the rate
constants for homopolymerization, while km and k2, describe cross-
propagation rates, the addition of ethylglycolide to a lactide chain end, and the
addition of lactide to an ethylglycolide chain end. As shown earlier, this model

leads to the copolymerization equation:
2 2 (5)
f1 (1'2F1)/F1(1-f1) = r2 + [f1 (F1-1)/F1(1-f1) ]I'1

where f1 is the fraction of monomer 1 in the feed, and F1 is the fraction of
monomer 1 incorporated into the copolymer. A plot of [f12(F1-1)/F1(1-f1)2] (x
axis) vs. f1 (1-2F1)/F1(1-f1) (y axis) should yield a straight line with a slope of r1
(the monomer reactivity ratio of lactide, defined as k11/k12) and a y intercept
axis of r2 (the monomer reactivity ratio of ethylglycolide, defined as k22/k21).
This is known as the Fineman-Ross84 method for analyzing copolymerization
data. Since rate constants kn and k22 can be obtained from the
homopolymerization of lactide and ethylglycolide, respectively, we can
calculate the cross-propagation rate constants km and k2, from the
experimentally determined values of r1 and re. The kinetic results are

summarized in Figures 7 and 8 and Table 4.

45

Table 4. Copolymerization behavior of
lactide with ethylglycolide

 

 

 

 

Kpm (sec") x103
monomer lactide ethylglycolide
lactide 6.2 (km) 3.5 (km)
ethylglycolide 5.4 (R21) 4.8 (kg)

 

 

 

 

 

The shape of the f, vs F1 curve in Figure 7 resembles an ideal
copolymerization where the reactivities of the two monomers are similar.
Given the structural similarity of lactide and ethylglycolide, it is not surprising
that these two monomers incorporate randomly into the chain (ideal
copolymerization). These results are consistent with previous DSC data, which
showed one T9 for each copolymer, an indication that the copolymer is
homogeneous. Figure 8 shows that the Fineman-Ross determined reactivity
ratios are similar, with r1 = 1.78 and r2 = 0.92. From the experimentally
determined rate constants (Table 4), we find that lactide enters the polymer
chain at a faster rate, whether the chain end is a lactide or ethylglycolide unit.
This is reasonable, since the longer alkyl chain of ethylglycolide makes it a

more sterically hindered monomer.

However, the uncertainty of the data in Table 4 is large. The monomer
reactivity ratios obtained from the Fineman-Ross method were used to
generate the theoretical f1 vs. F1 curve for the copolymerization. We can see
that the experimental data shows some deviation from the theoretical f1 vs F1

curve (Figure 7). From repeat analyses, we estimate the error in determining

46

the monomer concentrations from HPLC could be as high as 20%. Since the
copolymer composition is indirectly determined by measuring the small change
in the concentration of monomer during copolymerization, the error in
determining the copolymer composition could be much higher. This problem
might be overcome by increasing the concentration of the sample, but the
solubility of lactide and ethylglycolide in the eluent (1:1 acetonitrile/water) is

limited. An improved protocol for determining the copolymer composition is

 

 

 

 

needed.
1
[E
1’ 0.8 7 Q
a)
g. o
o
o.
8 0.6 .
.E
S
“5 0.4 ~ ’
c
.9
‘6
S
2 0.2 ~
0
2 O
O l l l l
O 0.2 0.4 0.6 0.8 1

Mole fraction of LA in feed (f1)

Figure 7. Mole fraction of lactide in the copolymer (F1) as a function of
the feed ratio (f1) during the copolymerization of lactide with
ethylglycolide as determined from reverse phase HPLC

47

f1 (1 -2F1)/F1(1-f1)

 

 

 

 

-5 4 1 . . 9
-3 -2.5 -2 -1 .5 -1 -O.5 O

 

(112(F1-1)/F1(1-n)21

Figure 8. r1 and r2 determination by the Fineman-Ross method

48

6.2.1.5. Direct Determination of Copolymer Composition.

Considering the limitations of the NMR and HPLC approaches to
copolymer composition analysis, we devised a strategy that allows direct
determination of the copolymer composition (Scheme 15). After stopping the
copolymerization at the desired conversion, the copolymer was separated from
the monomer, and converted to small molecules via hydrolysis or
transesterification. Since quantitative analysis of this derivative by LC, GC or
NMR, is facile, the composition of copolymer can be directly obtained. There
are two critical steps in this method: the separation of monomer from the
copolymer, which must be complete and without side reactions such as
hydrolysis, and the hydrolysis/transesterification reaction, which must convert

the copolymer to the desired small molecules in quantitative yield.

Separation. Because of the low recoveries seen in our HPLC results
and our concern that silica might hydrolyze lactide and ethylglycolide, we
examined both silica gel and silica deactivated with 3% triethylamine as
stationary phases for the separation of the copolymer from the monomer.
Surprisingly, lactide quantitatively hydrolyzed to lactic acid and its linear dimer
on the treated column, while both lactide and ethylglycolide were stable to
untreated silica gel. The monomer and polymer can be cleanly separated on a
silica column using CHCI3 as the eluent provided the Mn of the polymer is >
8,000 g/mol. Assuming the conversion to polymer is limited to 10%, the
monomer to catalyst ratio must be at least 400 to 1. Two examples describe

the effectiveness of the method. A solution of lactide (22.0 mg) and polylactide

49

(19.0 mg) in CHCI3 was poured onto a silica column prepared from a 9 inch
Pasteur pipette. Rinsing the column with 15 mL of CHCI3 resulted in recovery
of 21.7 mg (98%) of lactide uncontaminated by polymer. Applying the same
method to a solution of 21.5 mg of ethylglycolide and 19.0 mg of polylactide
resulted in the recovery of 20.5 mg (95%) of ethylglycolide, again
uncontaminated by polymer. These results show that the monomer can be
cleanly separated from the polymer, and devising a method for recovering the

polymer from the silica would complete the separation scheme.

Transesterification. Transesterification is an attractive approach for
recovering the polymer since it allows direct conversion of the copolymer to
esters that can be easily analyzed by GC, HPLC, or NMR. After treating PLA
with conc. HCI and methanol for two days at reflux, 1H NMR showed 90%
conversion of PLA to lactic acid methyl ester, and 10 % to lactic acid. The
lactic acid is a problem as we were unable to devise a reverse phase HPLC
protocol to separate lactic acid methyl ester, lactic acid, ethylglycolic acid, and
its methyl ester. To overcome this problem, the mixtures of acids and esters
were treated with diazomethane. We wanted to avoid the routine use of
diazomethane, and we reasoned that this step probably can be omitted since

the rates of hydrolysis to lactic and glycolic acids should be similar.

We examined alternatives to the use of HCI, and Sn(Oct)2. the catalyst
commonly used to polymerize lactide, proved to be the most efficient catalyst
for converting polylactide to its methyl ester. The optimized transesterification

conditions, refluxing the polymer for 48 h with a large excess of methanol and

50

Sn(Oct)2 (30 eq. per polylactide repeat), gave quantitative yields of the methyl
esters. Thus, we developed a reliable protocol for obtaining the copolymer

composition

The analyses of the copolymers were completed by using GC, LC and
NMR to characterize the mixtures of methyl esters. The temperature program
used for GO analyses held the sample at 35 °C for 10 minutes, heated the
sample at 20 °C/min to 150 °C, and then held at that temperature for 10
minutes. Both the lactic acid methyl ester (b.p. 140-145 °C) and ethylglycolic
acid methyl ester (b.p. is 246 °C) are quite volatile, and the two compounds
were detected at 2.1 and 4 minutes, while o-chlorotoluene, the standard, was
detected at 8.7 min. GC-MS was used to prove the identity of the two peaks.
Preparation of the sample is critical. The methanol must be removed under
conditions that do not lead to loss of the esters. Evacuation of the methanol on
a vacuum line with the temperature kept below —10 °C and the pressure >300
mTorr gave good results. The GC analyses of samples prepared by
transesterification using HCI as the catalyst were complicated and contained
unexpected peaks, but those prepared using Sn(Oct)2 were clean and showed

only the two expected esters.

51

 

Sn(Oct)2 { LA + EG (90%)

 

 

 

 

LA + E6 0
130°C P(LA co-EG) (10 /o)
O
HO OMe
F Me
SD(OCI)2
——> P(LA-co-EG)
MeOH O
(f HOj/ILOMe
Et
0
LA + EG
GC j 1H NMR
GC ‘H NMR
ILA- {£6
1 ll
FLA! Fee FLAv FEG

Scheme 15. Protocol for the determination of copolymer composition.

Comparison of GC and 1H NMR. Both GC and 1H NMR are
convenient methods for analyzing the methyl esters of lactic acid and
ethylglycolic acid. Having two techniques is important, since copolymers
prepared from other pairs of comonomers may yield methyl esters that are
nonvolatile or have NMR spectra that cannot be integrated accurately due to
overlapping peaks. To assay which technique is best for the
lactide/ethylglycolide pair, we prepared and analyzed 4 samples by both

methods. The two methods gave comparable F1 values (Table 5), but 1H NMR

52

was the preferred technique because it is faster and more convenient. Figure
9 shows the copolymerization data obtained via the protocol and analysis by
NMR, and Figure 10 shows the data plotted in the Fineman-Ross format to
extract the values of r1 and r2. The rate constants for the copolymerization of
lactide and ethylglycolide are listed in Table 5. The data based on 1H NMR

analysis (Figure 10) are more consistent and have a lower uncertainty that

Table 5. Comparison of the GC and 1H NMR methods
for analyzing the copolymer composition

 

 

 

 

 

 

 

11 (LA) r=1 (LA) from so F1 (LA) from NMR
0.17 0.29 0.25
0.29 0.35 0.29
0.79 0.68 0.71
0.88 0.84 0.82

 

 

 

53

those obtained from the original HPLC scheme (Figure 7).

 

Mole fraction of LA in copolymer (F1)

0.8

0.6

0.4

0.2

 

I’1l’2 = 1.06

l l l I

O 0.2 0.4 0.6 0.8 1

 

 

 

Mole fraction of LA in feed (f1)

Figure 9. Mole fraction of lactide in the copolymer (F1)
as a function of the feed ratio (f1 during the
copolymerization of lactide with ethylglycolide as
determined from 1H NMR

54

f1 (1 -2F1)/F1(1-f1)

 

  

I
—L
I

['1 :158, r2 = 0.72

 

 

 

r1‘°-(1=,-1)/r=,(1-r,)2

Figure 10. r1 r2 determination by Fineman-Ross method

55

Table 6. Copolymerization behavior from Fineman-Ross method

 

 

 

 

Kpn] (sec'1) x103
monomer lactide ethylglycolide
lactide 6.2 (km) 3.9 (km)
ethylglycolide 6.7 (k2,) 4.8 (kg)

 

 

 

 

 

Recently, the Kelen-Tudos method”86 has emerged as a popular method for
extracting r1 and r2 from experimental data. Like the Fineman-Ross method,

this

method is based on the copolymerization equation:

 

0th _ [M1l(r1lM1]+lM2D ‘3’
(’1le - [M2]([M1]+72[M2])

in which d[M1], d[M2] are defined as the monomer concentration of monomer 1,
2 being incorporated into the copolymer at monomer feed concentrations of
[M1] and [M2]. if we further define x = [Md/[M2]. y = d[M1]/d[M2] and G = x (y-1),
than F = x2/y, g = F/(a + F), n = G/(a + F), and a = (FMaxFM,n)"2, the

copolymerization can be transformed into:
n = (r1+r2/0()8 -r2/0(

- Plotting n (y axis) vs 8 (X axis) yields a straight line with a slope of (r1+r2/01) and

an x-intercept of r2/01. Since 01 is known, r1 and r2 can therefore be calculated.

56

The kinetic data plotted according to this method are shown in Figure 11 and
Table 5. Comparing the kinetic results obtained from Fineman-Ross method

with those from Kelen-Tudos method, we can see that they are very similar.

Figure 9 shows the theoretical f1 vs F1 curve and the experimental data
obtained by the direct determination of copolymer composition. Compared to
the analogous data obtained from reverse phase HPLC, the data are more
consistent with theoretical values, which is consistent with the direct copolymer
determination protocol being a more accurate approach than the indirect

composition analysis based on reverse HPLC.

A copolymerization is termed as “ideal” when the product of mg is unity.
In the case of the copolymerization of lactide with ethylglycolide, mg is 1.06
(Kelen-Tudos method), which is very close to 1, so the polymerization can be
classified as ideal. Because the r1 and r2 values are not that different, both
monomers add to the growing polymer chain end randomly (Figure 5). An
important practical consequence of an ideal copolymerization is that it
becomes progressively more difficult to produce copolymers containing
appreciable amounts of both monomers as the difference in the reactivities of
the two monomers increases. However, because r1 and r2 do not differ
markedly in the copolymerization of lactide with ethylglycolide, there will exist a
large range of comonomer feed compositions that yield copolymers containing

appreciable amounts of both monomers.

Steric hindrance from the substituent on the glycolide ring plays an

important role in the polymerization rate,1 and not surprisingly, the

57

homopolymerization of lactide is faster than that of ethylglycolide (k1, > [(22).
Lactide, the more active monomer, adds faster to the growing chain end
whether the chain ends in lactide or ethylglycolide (k1, > k1,? , k2, > k2,? ).
Nevertheless, the data show that ethylglycolide is the more reactive chain end,
suggesting that the ethyl glycolide terminated chains are less stable, probably
for steric reasons. This effect should be magnified if the glycolide ring is made
more congested. The results for a glycolide with a more sterically demanding

alkyl group, isopropylglycolide, are summarized in the subsequent section.

Table 7. Copolymerization behavior from Kelen-Tudos method

 

 

 

 

Kpm (sec'1) x103
monomer lactide ethylglycolide
lactide 6.2 (k,,) 4.1 (km)
ethylglycolide 6.9 (km) 4.8 (kg)

 

 

 

 

 

58

1.5

0.5

 

  

_ r1: 1.53, r2 = 0.69

 

 

 

Figure 11. r1,r2 Determination by the Telen-Tudos Method

59

6.3. Kinetic Evaluation of the Copolymerization of Lactide with
lsopropylglycolide

The melting point of rad-isopropylglycolide (rac-IPrGL) is 161-163 °C,
and thus the temperature for the melt copolymerization of lactide with rac-
isopropylglycolide was set at 165 °C. Initially, the protocol used to analyze the
copolymer composition was identical to that used for ethylglycolide. However,
the transesterification reaction, the key step in the analysis of copolymer
composition, proved problematic. After 6 days, the conversion of poly(rao-
IPrGL) to the methyl ester of isopropylglycolic acid methyl ester was only 40%,
with the remainder consisting of the linear dimer (30%) and oligomeric
poly(rao-IPrGL) (30%). Switching the solvent to THF, a better solvent for the
polymer, did not improve the reaction rate. Thus, we resorted to analysis of the
un-polymerized monomer to calculate the copolymer composition. The
assumption that the decrease in monomer concentration is directly related to
the polymer composition is good (i.e. there are no side reactions), but
calculating the small change in monomer concentration requires subtraction of

two large numbers, and thus the error associated with the calculation is large.

0 O SD(OCI)2
\JLO o : I d
+ t-BB A po ymer an monomer
0% o
W 165 °C
0 o

rad-lactide rac-isopropylglycolide

 

Scheme 16. Copolymerization of racemic lactide with
racemic isopropylglycolide

60

To minimize the error, the monomer feed amount was increased to 500 mg. At
10% conversion for a feed of 90% lactide (45 mg) and 10% isopropylglycolide
(5 mg), the uncertainty in the weight of the isopropylglycolide should be :t 0.2
mg, and the uncertainty associated with measurement of the isopropylglycolide
should be i 4%, which should be acceptable. To test this idea, three sets of
experiments were run at different feed ratios, with each set consisting of a pair
of duplicate runs. After copolymerization, the compositions of the copolymers
were determined; the results of these parallel experiments are summarized in
Table 8. The data are consistent and thus this method was used for copolymer
composition analysis in the poly(lactide-co-rac-IPrGL) system. The results of
the kinetic study of the copolymerization of lactide with isopropylglycolide are
summarized in Figures 12, 13, and 14 and Tables 9, 10.

Table 8. Comparison of copolymer composition analysis
obtained from parallel runs

 

 

 

 

 

 

 

f1 (LA) F1 (LA)
0-1 0.28 0.63
0-4 0.28 0.63
5-1 0.55 0.94
5-2 0.59 0.97
8-3 0.80 0.97
8-4 0.81 0.99

 

 

 

 

 

61

The data points that define the f1~F1 curve for poly(LA-co-rao-IPrGL)
(Figure 12) show more scatter than seen in poly(LA-co-EG) (Figure 9). For
poly(LA-co-EG), copolymer composition was measured directly, but an indirect
approach, with its larger uncertainty, was used for poly(LA-co-rao-IPrGL). The
monomer reactivity ratios obtained from the Kelen-Tudos analysis of the data
were used generate the theoretical f1 vs. F1 curve for the copolymerization
(Figure 14). Except for slight deviation in the middle of the curve, the

experimental data (diamonds) and the calculated curve are consistent.

The overall shape of the f1~F1 copolymerization curves for lactide
copolymerized with ethylglycolide and isopropylglycolide are similar,
suggesting that both are “ideal”. However, the reactivity ratio product for
poly(LA-co-rac-IPrGL) (r1 r2 = 0.48) is farther from unity than that of poly(LA-co-
EG) (r1 r2 = 1.06), and therefore, we expect the microstructures of these two
copolymers to be quite different. In particular poly(LA-co-raorPrGL) should be

a somewhat “blocky” copolymer.

The kinetic rate constants derived from Fineman-Ross and Kelen-Tudos
analyses of the data are shown in Tables 9 and 10. Both methods gave similar
results. As would be expected from simple steric arguments, the
homopolymerization of LA is 8 times faster than that of IPrGL. Steric effects
also provide a plausible argument for the higher reactivity of lactide chain ends.
Chains ending in isopropylglycolide should be the least stable (most reactive)

and add lactide faster than chains terminated with lactide. Adding the sterically

62

hindered isopropylglycolide to a chain ending in isopropylglycolide should be

the slowest process. Therefore rates decrease in the order [(22 > k11> k1,? > k22.

One way to understand the kinetics of these two copolymerizations is to
think in terms of the rate-determining step of the polymerization. According to
Yin,1 two steps in the chain propagation need to be considered (Scheme 17),
nucleophilic attack (km) and ring opening (kpg). For chains ending in hindered
monomers such as isopropylglycolide, the intermediate is relatively unstable
and the activation energy to form products is decreased. Such chains should
react rapidly, especially with reactive monomers like lactide. Chains ending in
lactide units are more stable, and their reaction with hindered monomers is
slow. This process is analogous to that seen for the copolymerization of vinyl
monomers such as styrene (stable chain end) and vinyl acetate (unstable

chain end).

63

 

Mole fraction of LA in copolymer (F1)

 

 

 

 

O l l 1 1
0 0.2 0.4 0.6 0.8- 1

Mole fraction of LA in feed (f1)

Figure 12. Mole fraction of lactide in the copolymer (F1) as a
function of the feed ratio (f1) during the copolymerization of lactide
with poly(lactide-co-rao-isopropyllgycolide)

f1(1-2F1)/F1(1-f1)

0.5

 

  

r1: 5.49, r2: 0.11

 

 

 

f12(F1-1)/F1(1-f1)2

Figure 13. r1 r2 determination using the Fineman-Ross method

65

4.5

3.5

2.5

 

h—u

)—

 

f1: 5.39, r2 = 0.09

 

 

Figure 14. r1,r2 Determination using the Kelen-Tudos method

66

0.6

0.8

 

Table 9. Copolymerization behavior from Fineman-Ross method

 

 

 

 

 

 

Kpu] (sec") x103
monomer lactide isopropylglycolide
lactide 40.3 (km) 7.3(k12)
isopropylglycolide 53.6(k27) 4.8 (k22)

 

 

 

Table 10. Copolymerization behavior from Kelen-Tudos method

 

 

 

 

 

Kp[l] (sec") x103
monomer lactide isopropylglycolide
lactide 40.3 (km) 7.5(k12)
isopropylglycolide 58.8 (km) 4.8 (k22)

 

 

 

 

67

Scheme 17. Chain propagation mechanism for lactide copolymerization.

68

6.4. Copolymerization of Lactide with R-ethylglycolide

Our results show that steric factors play an important role in determining
the microstructure of lactide copolymers, and it would be interesting to explicitly
add another complication to the copolymerization, stereochemistry.
Ethylglycolide has been obtained as a single diastereomer, D-ethylglycolide
(R,R). The physical properties of D-ethylglycolide are quite different from the
ethylglycolide monomer mixture used earlier, which was composed of a 1:1
mixture of meso and rao-ethylglycolide. For example, the melting point of D-
ethylglycolide is 122 °C, while the monomer mixture used previously melts at

room temperature.

 

O O
O O SH(OCI)2
O + O > polymer and monomer
t-BBA
O O 130 °c
rac-lactide D-ethylglycolide

Scheme 18. Copolymerization of lactide with R-ethylglycolide

The homopolymerization rate for D-ethylglycolide has not been
reported, so to obtain k22, D-ethylglycolide was polymerized at 130 °C. Judging
from the propagation rate constant kp[l] (Figure 15), the polymerization rate of
the pure diastereomer is slower than that of the ethylglycolide mixture (3.2 vs
4.8), which is consistent with literature.”96 Yin’s study of the polymerization of

rao-lactide with L-lactide showed that depending on the polymerization

69

conditions, rad-lactide polymerized 1.7 to 3 times faster than L-lactide.1 This
phenomena is termed “syndiotactic preference”, a bias toward adding a
monomer of opposite stereochemistry to the end of the growing chain. The
difference in the activation energies for these two copolymerizations was
ascribed to chain end control, where differences in the steric hindrance for the
addition of monomers of different stereochemistry lead to preferential insertion

of one stereoisomer.

The study of the copolymerization of lactide with D-ethylglycolide

followed that used for poly(lactide-co-ethylglycolide) except that the

 

 

 

 

2
A 1.5 L
o\0 Kp[|]=3.2
>
5
9 1 L
‘2’
' 0.5 L
O 1 l
0 200 400 600
time (sec)

Figure 15. Melt polymerization of D-ethylglycolide at
130 °C catalyzed by Sn(Oct)2 + t-BBA

70

transesterification time was increased from two to three days. The
copolymerization behavior of lactide with D-ethylglycolide is summarized in
Figures 16, 17 and 18. In the f. vs F1 plot, most data are consistent with the
theoretical curve (generated from the r1 and r2 values derived from the Kelen-
Tudos method). As expected, lactide is the more reactive monomer, and
therefore we observed k11> km and k21> kgg. As to the chain end reactivity, we
observed the same trend as seen in the copolymerization of lactide with
isopropylglycolide; lactide adds faster to an isopropylglycolide chain end than
to a lactide chain end (k2, > kn). When the incoming monomer is. D-
ethylglycolide, it is the lactide chain end that is more reactive (R12 > kgg).
Therefore, we conclude that effect of stereochemistry on the copolymerization

can largely be explained by steric arguments.

71

 

I

0.2

 

 

Mole fraction of LA in copolymer (F1)
0

 

O l l l J
0 0.2 0.4 0.6 0.8 1

Mole fraction of LA in feed (f1)

Figure 16. Mole fraction of lactide in the copolymer (F1) as a function
of the feed ratio (f1) during the copolymerization of lactide with D-
ethylglycolide

72

f1(1-2F1)/F1(1-f1)

 

I
O)
l

 

 

 

 

f12(F1-1)/F1(1-f1)2

Figure 17. r1, r2 determination by Fineman-Ross method

73

 

 

 

 

Figure 18. r1, r2 determination by Kelen-Tudos method

74

 

Table 11. Copolymerization behavior from Fineman-Ross method

 

 

 

 

Kpm (sec'1) x103
monomer lactide D-ethylglycolide
lactide 6.2 (km) 4.3 (km)
ethylglycolide 8.0 (k2,) 3.2 (kg)

 

 

 

 

 

Table 12. Copolymerization behavior from Kelen-Tudos method

 

 

 

 

Kp[l] (sec") x103
monomer lactide D-ethylglycolide
lactide 6.2 (km) 4.8 (R12)
ethylglycolide 7.3 (km) 3.2 (’92)

 

 

 

 

 

75

6.5. Physical Properties and Structure of Copolymers

Prior work suggested that poly(lactide-co-ethylglycolide)s were
homogeneous materials with the two comonomers distributed randomly along
the polymer chain. Differential scanning calorimetry (DSC) analyses of the
poly(lactide-co-isopropylglycolide)s (Table 13) show that these polymers also
are single-phase materials, since all of the copolymers exhibit one Tg that falls
between that of the polylactide and poly(isopropylglycolide) homopolymers.
The change in the T9 of random copolymers with composition can be predicted
by the Fox equation (Figure 19). The good fit to the Fox equation (1 2 °C
deviation) is consistent with the poly(LA-co-rac-IPrGL)s being homogeneous.
However, this is not a rigorous test in that the Tgs of the two homopolymers are
similar, making it hard to detect deviations from the behavior expected for a

random copolymer.

Table 13. Properties of P(LA-co-IPrGL)

 

 

 

 

 

 

Polymers Mn*10‘3 ° MJMn T9 °C ”
Polylactide 35.4 1 .78 56
LA/iprGL = 3/1 27.8 1.82 48
LA/iprGL = 1/1 25.7 1.76 48
LA/iprGL = 1/3 28.2 1.90 43
Poly (isopropylglycolide) 32.6 1.74 42

 

 

 

 

 

 

a. measured by GPC in THF using polystyrene as standard
b. measured by DSC under He at a rate of 10 °C/min

76

T9 (00)

75

65

 

35

25

 

 

1 l l I

 

 

0.2 0.4 0.6 0.8 1
W1/(W1+W2)

Figure 19. Glass transition temperature of lactide and
isopropylglycolide copolymers.

77

Although the kinetic data indicate that both poly(LA-co-EG) and poly(LA-
co-rao-IPrGL) are random copolymers, there are differences in the extent of
“randomness” of these two copolymers. When lactide copolymerizes with
ethylglycolide, the reactivities of these two monomers are very close due their
structural similarity (r1 = 1.53, r2 = .69). Therefore, even though lactide is more
reactive than ethylglycolide, these two monomers have nearly the same
opportunity to be added to the copolymer. As seen from the results of the
kinetic study (Figure 9), the ma product is 1.06, very close to unity, and thus is
a random copolymer. On the other hand, the reactivity difference in the
copolymerization of lactide with isopropylglycolide (r1 = 5.39, r2 = .09, mg =
0.48) results in a strong preference for the polymerization of lactide. Therefore,
lactide would be consumed first during the copolymerization, and as the
concentration of lactide decreases in the monomer feed pool,
isopropylglycolide enters the polymer chain at an increasing rate. The ma
product is a signature that this copolymerization is less random than poly(LA-
co-EG), and there may be small blocks of polylactide and
poly(isopropylglycolide) randomly distributed along the polymer chain.
However, these blocks are apparently not long enough the generate phase
separation. Thus, this copolymerization resembles the copolymerization of
lactide with glycolide.94 Although the physical properties of the poly(lactide-co-
D-ethylglycolide)s were not studied, the kinetic parameters of this
copolymerization indicate this copolymer is a random copolymer similar to

poly(LA-co-EG) rather than a “blockY’ random copolymer, such as poly(LA-co-

78

GL) and poly(LA-co-iprGL). Nevertheless, kinetic evaluation and the study of
the physical properties of copolymers provide only a partial analysis of the
copolymerization. A full characterization of the polymer microstructure would
require a full sequence analysis by NMR. The relationship between the
monomer reactivity difference and the copolymer structure is illustrated in

Scheme 19.

79

P (LA-co-GL)
1’1 = 3.4, r2 = .22
r1r2 = .75

“blocky" random

P (LA-co-EG)
r1 = 1.53, r2 = .69

“ideal” random

P (LA-co-IPrGL)
r1 = 5.29, r2 = .09

“blocky" random

P (LA-co-DEG)
r, = 1.39, r2 = .39
r1r2 = 1.06

“ideal” random

0%.:0

polymerizes
preferentially

glycolide rich lactide rich

 

O
O 0Q— o
ofi
l erizes POIYmerizes
.U' gummy J1 randomly

0.000000000...
l
a» 2:0
O O 0 O
0%: 0%:
polymerizes 1
preferentially

lactide rich isopropylglycolide rich

 

 

lymerizes
polymerizes . po
U randomly U randomly

 

Scheme 19. The relationship between monomer reactivity
ratio and the structure of copolymers.

6.6. Biodegradation of the Copolymers

We studied the hydrolytic degradation of a series of lactide and
isopropylglycolide copolymers at pH 7.4 at 55 °C. The samples were 1 mm by
1 mm pieces of polymer, and at predetermined times, samples were removed
from the degradation experiment and analyzed for loss in mass and molecular
weight. The size of the polymer samples must be small to avoid the
autocatalytic effects. Because the degradation product is acidic (carboxyl
group formed by ester hydrolysis), acid groups localized in large samples
cause an increase in the local degradation rate. For small samples, the
diffusion rate of the aqueous buffer solution in the polymer will be higher than

the polymer chain scission and the hydrolytic degradation proceeds uniformly.

The two most characteristic features of polymer hydrolytic degradation
are mass loss and a decrease in molecular weight during degradation process.
The weight loss process is associated with the formation of short polymer
chains that are soluble in the buffer solution. The degradation data are
summarized in Figures 20 and 21. For all samples, there was an initial drop in
molecular weight, but the mass remained constant, and then declined. This
behavior is consistent with a bulk degradation mechanism. It is also clear that
the composition of the copolymer strongly influences the degradation profile.
Since lactide is more hydrophilic than isopropylglycolide, the higher the lactide
content in a copolymer, the faster the degradation rate in terms of both mass

and molecular weight loss. The extreme case is that the polylactide

81

homopolymer totally degrades in 30 days, whereas poly(isopropyllgycolide)

samples under the same conditions lost only 2% of their weight in 60 days.

To further understand the degradation of these copolymer samples, we
tried to fit a random chain scission model to the data in Figure 21. In the
random scission model, the average number of bond cleavages per polymer

molecule (N) was calculated according to:

N=[Mn(0)/Mo(t)- 7]= KdPn(0)t

where MM and MO) are the number-average molecular weights of polyesters
at time 0 and time t, kd is the rate constant for hydrolytic degradation, and Pn(0)
is the number-average of degree of polymerization at time 0. kd was obtained
from the slope from Figure 22. As expected, polylactide degrades the fastest,
as the lactide content decreases, the degradation rate slows.

Poly(isopropylglycolide) has the lowest degradation rate.

In an earlier study, we found that alkyl substituted lactide showed
constant weight1 until the degree of polymerization (Pn) dropped to 30. The
data for the copolymers show a similar trend (Figure 23); all of the polymer
samples started to lose weight when the Pn drop to 30-40. This “turning point”
in the degradation process must be related to the structure of the polylactide
backbone, and perhaps corresponds to the entanglement limit for these
polymers. Solubility arguments cannot explain the data since the same

phenomena is seen for polymers of different degrees of hydrophobicity.

82

 

Remaining Weight Fraction

 

 

 

 

1 ~ I g A
0 . I A
0.8 . '
o
o A P(iprGL)
0.6 ~ . I LA/iprGL=1/3
o LA/iprGL=1/1
o
04 _ . . LA/ipl’GL=3/1
o
o
02 - ’
o
e
O l 1 1 4
0 20 40 60 80 100
Time (days)

Figure 20. The weight loss of polymer samples during hydrolytic degradation

83

Mn(t)/Mn(0)

 

1.2

   
 
 
 
   
 

0 LA/iPRGL=3/1
O LA/iPRGL=1/1
1 A LAfIPRGL=1/3
I P(rac-iPRGL)
I
0.8 -

 

 

 

 

 

time (days)

Figure 21. The molecular weight loss of poly(LA-co-
iPRGL) during hydrolytic degradation.

84

 

 

 

 

 

0.1 .
.WIPRGL21/1
ALA/iPRGLz1/3
0'08 ” . .P(iPRGL)
A oLA/iPRGL=3/1
9:, .P(racLA)
D.
’2‘ 0.06 7
E
g 0.04 ~
5.
>.
0.02 l "
.7 ‘
l A
O 05", 4 ________ A
0 10 20 30 40 50 60
time (day)

Figure 22. Decrease in the molecular weight of P(LA-co-
iPRGL) during hydrolytic degradation fitted to the random
chain scission model

85

 

 

 

 

12
1 ‘ Q r
C
.52
8 0.8 ~ ,
E A P(lprGL)
IE .
g) 06 _ ILA/iprGL-1/3
3 OLA/iprGL=1/1
8’
I: 0.4 . o LA/iprGL=3/1
(U
E
G)
”5 0.2 ~
0 l 1
10

1000 100

Pn

 

Figure 23. The decrease in the degree of polymerization
of polymer samples during hydrolytic degradation.

86

BIBLIOGRAPHY

1

10

11

12
13
14
15

16

17

Yin, M. Synthesis and Characterization of Substituted Polylactides,
Michigan State University, 2000.

Gilding, D. K.; Reed, A. M. Polymer1979, 20, 1459-1464.
Jamshidi, K.; Hyon, S. H.; lkada, Y. Polymer1988, 29, 2229-2234.

Varmanair, M.; Pan, R.; Wunderlich, B. J. Polym. Sci. Pt. B-Polym.
Phys. 1991, 29, 1107-1115.

Helder, J.; Feijen, J.; Lee, S. J.; Kim, S. W. Makrom. Chem-Rapid
Commun. 1986, 7, 193-198.

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

Nijenhuis, A. J.; Grijpma, D. W.; Pennings, A. J. Polym. Bull. 1991, 26,
71-77.

Migliaresi, C.; Cohn, D.; Delollis, A.; Fambri, L. J. Appl. Polym. Sci.
1991, 43, 83-95.

Bendix, D. Polym. Degrad. Stabil. 1998, 59, 129-135.

Scott, G.; Gilead, 0.; Editors Degradable Polymers: Principles and
Applications, 1995.

Pitt, C. G.; Gratzl, M. M.; Kimmel, G. L.; Surles, J.; Schindler, A.
Biomaterials 1981, 2, 215-220.

Williams, D. F.; Mort, E. J. Bioeng. 1977, 1, 231-238.
Williams, D. F. Eng. Med. 1981, 10, 5-7.
Holland, S. J.; Tighe, B. J. Adv. Pharm. Sci. 1992, 6, 101-64.

Joziasse, C. A. P.; Veenstra, H.; Topp. M. D. C.; Grijpma, D. W.;
Pennings, A. J. Polymer1998, 39, 467-474.

Mooney, D. J.; Baldwin, D. F.; Suh, N. P.; Vacanti, L. P.; Langer, R.
Biomaterials 1996, 17, 1417-1422.

Chaput, C.; Yahia, L. H.; Landry, D.; Rivard, C. H.; Selmani, A. Biomat-
Living Syst. Interact. 1995, 3, 19-27.

87

18

19

20

21

22

23
24
25
26

27

28

29

30

31

32

33

34

Freed, L. E.; Marquis, J. C.; Nohria, A.; Emmanual, J.; Mikos, A. G.;
Langer, R. J. Biomed. Mater. Res. 1993, 27, 11-23.

Organ, G. M.; Mooney, D. J.; Hansen, L. K.; Schloo, B.; Vacanti, J. P.
Transplant. Proc. 1992, 24, 3009-3011.

Bergsma, J. E.; Rozema, F. R.; Bos, R. R. M.; Boering, G.; Debruijn, W.
C.; Pennings, A. J. Biomaterials 1995, 16, 267-274.

Cordewener, F. W.; Rozema, F. R.; Joziasse, C. A. P.; Bos, R. R. M.;
Boering, G.; Pennings, A. J. J. Mater. Sci-Mater. Med. 1995, 6, 561-
568.

Arvanitoyannis, l.; Nakayama, A.; Kawasaki, M; Yamamoto, N. Polymer
1995, 36, 2271-2279.

Roper, H.; Koch, H. Starch-Starke 1990, 42, 123-130.
Lindsay, K. F. Mod. Plast. Int. 1989, 19, 110.
Chiellini, E.; Solaro, R. Adv. Mater. 1996, 8, 305-313.

Odian, G. Principles of Polymerization; 3rd ed.; Wiley-lnterscience: New
York, 1991.

Deng, X. M.; Yuan, M. L.; Xiong, C. D.; Li, X. H. J. Appl. Polym. Sci.
1999, 71, 1941-1948.

HiljanenVainio, M. P.; Orava, P. A.; Seppala, J. V. J. Biomed. Mater.
Res. 1997, 34, 39-46.

Vion, J. M.; Jerome, R.; Teyssie, P.; Aubin, M.; Prudhomme, R. E.
Macromolecules 1986, 19, 1828-1838.

Sawhney, A. S.; Hubbell, J. A. J. Biomed. Mater. Res. 1990, 24, 1397-
1411.

Vanhoome, P.; Dubois, P.; Jerome, H.; Teyssie, P. Macromolecules
1992, 25, 37-44.

Deng, X. M.; Zhu, Z. X.; Xiong, C. D.; Zhang, L. L. J. Polym. Sci. Pol.
Chem. 1997, 35, 703-708.

Kricheldorf, H. R.; KreiserSaunders, l. Macromol. Symp. 1996, 103, 85-
102.

Song, C. X.; Feng, X. D. Macromolecules 1984, 17, 2764-2767.

88

35

36

37

38
39

40
41
42

43
44
45

46

47
48

49

50
51

52

53

Veld, P.; Velner, E. M.; VanDeWitte, P.; Hamhuis, J.; Dijkstra, P. J.;
Feijen, J. J. Polym. Sci. Pol. Chem. 1997, 35, 219-226.

Apokan, N. K.; Yilmaz, O. K.; Baysal, K.; Baysal, B. M. Polymer 2001,
42, 4109-4116.

Gruvegard, M.; Lindberg, T.; Albertsson, A. C. J. Macromol. Sci-Pure
Appl. Chem. 1998, A35, 885-902.

Stridsberg, K.; Albertsson, A. C. Polymer 2000, 41, 7321-7330.

Yamaoka, T.; Takahashi, Y.; Fujisato, T.; Lee, C. W.; Tsuji, T.; Ohta, T.;
Murakami, A.; Kimura, Y. J. Biomed. Mater. Res. 2001, 54, 470-479.

Abe, H.; Doi, Y.; Hori, Y.; Hagiwara, T. Polymer1998, 39, 59-67.
Hiki, S.; Miyamoto, M.; Kimura, Y. Polymer2000, 41, 7369-7379.

Matsumura, S.; Tsukada, K.; Toshima, K. Int. J. Biol. Macromol. 1999,
25, 161-167.

Schmidt, P.; Keul, H.; Hocker, H. Macromolecules 1996, 29, 3674-3680.
Li, Y. X.; Kissel, T. Polymer1998, 39, 4421-4427.

Kricheldorf, H. R.; Jonte, J. M.; Berl, M. Makromol. Chem. Phys. 1985,
25-38.

Dong, C. M.; Oiu, K. Y.; Gu, Z. W.; Feng, X. D. J. Polym. Sci. Pol.
Chem. 2000, 38, 4179-4184.

Li, Y. X.; Nothnagel, J.; T., K. Polymer1997, 38, 6197-6206.

Agrawal, C. M.; Athanasiou, K. A.; Heckman, J. D. In Porous Materials
For Tissue Engineering 1997; Vol. 250, p 115-128.

Jain, R.; Shah, N. H.; Malick, A. W.; Rhodes, C. T. Drug Dev. Ind.
Pharm. 1998, 24, 703-727.

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

Simmons, T. Synthesis and Characterization of Aromatic Substituted
Polylactides, Michigan State University, 2000.

lmasaka, K.; Yoshida, M.; Fukuzaki, H.; Asano, M.; Kumakura, M.;
Mashimo, T.; Yamanaka, H.; Nagai, T. Int. J. Pharm. 1992, 81, 31-38.

Hrkach, J. 8.; Ou, J.; Lotan, N.; Langer, R. Macromolecules 1995, 28,
4736-4739.

89

54

55
56

57

58

59

60
61

62

63

65

66
67
68

69
70
71

72

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

Lee, R. S.; Yang, J. M. J. Polym. Sci. Pol. Chem. 2001, 39, 724-731.

Fujiwara, T.; Miyamoto, M.; Kimura, Y. Macromolecules 2000, 33, 2782-
2785.

Fujiwara, T.; Miyamoto, M.; Kimura, Y.; Sakurai, S. Polymer 2001, 42,
1515-1523.

Chen, X. H.; McCarthy, 8. P.; Gross, R. A. Macromolecules 1997, 30,
4295-4301.

Nagasaki, Y.; Okada, T.; Scholz, C.; lijima, M.; Kato, M.; Kataoka, K.
Macromolecules 1998, 31, 1473-1479.

Han, D. K.; Hubbell, J. A. Macromolecules 1997, 30, 6077-6083.

Otsuka, H.; Nagasaki, Y.; Kataoka, K. Curr. Opin. Colloid Interface Sci.
2001, 6, 3-10.

Dunn, R. L.; English, J. P.; Strobel, J. D.; Cowsar, D. R.; Tice, T. R.
Prog. Biomed. Eng. 1988, 5, 149-60.

Han, D. K.; Hubbell, J. A. Macromolecules 1996, 29, 5233-5235.
Yoshioka, M.; Hagiwara, N.; Shiraishi, N. Cellulose 1999, 6, 193-212.

Chen, D. P.; Sun, B. Q. Mater. Sci. Eng. C-Biomimetic Supramol. Syst.
2000, 11, 57-60.

Benns, J. M.; Kim, S. W. J. Drug Target. 2000, 8, 1-12.
Chen, X. H.; Gross, R. A. Macromolecules 1999, 32, 308-314.

Arvanitoyannis, l.; Nakayama, A.; Psomiadou, E.; Kawasaki, N.;
Yamamoto, N. Polymer1996, 37, 651 -660.

Dittrich, M.; Hampl, J.; Soukup, F. J. Microencapsul. 2000, 17, 587-598.
Wang, Y. B.; Hillmyer, M. A. Macromolecules 2000, 33, 7395-7403.

Frick, E. M.; Hillmyer, M. A. Macromol. Rapid Commun. 2000, 21, 1317-
1322.

Zalusky, A. S.; Olayo-Valles, R.; Taylor, C. J.; Hillmyer, M. A. J. Am.
Chem. Soc. 2001, 123, 1519-1520.

90

73

74

75

76

77

78

79

80

81

82

83
84
85
86

87
88

89
90

Lim, D. W.; Choi, S. H.; Park, T. G. Macromol. Rapid Commun. 2000,
21, 464-471.

Barakat, l.; Dubois, P.; Grandfils, 0; Jerome, R. J. Polym. Sci. Pol.
Chem. 1999, 37, 2401-2411.

Cadee, J. A.; De Kerf, M.; De Groot, C. J.; Den Otter, W.; Hennink, W.
E. Polymer 1999, 40, 6877-6881 .

D’Angelo, S.; Galletti, P.; Maglio, (3.; Malinconico, M.; Morelli, P.;
Palumbo, R.; Vignola, M. C. Polymer 2001, 42, 3383-3392.

Kohori, F.; Sakai, K.; Aoyagi, T.; Yokoyama, M.; Yamato, M.; Sakurai,
Y.; Okano, T. Colloid Surf. B-Biointen‘aces1999, 16, 195-205.

Thakur, K. A. M.; Kean, R. T.; Hall, E. S.; Doscotch, M. A.; Munson, E.
J. Anal. Chem. 1997, 69, 4303-4309.

Thosar, S. S.; Forbess, R. A.; Kemper, M.; Shukla, A. J. J. Pharm.
Biomed. Anal. 1999, 20, 107-114.

Kister, G.; Cassanas, G.; Bergounhon, M.; Hoarau, D.; Vert, M. Polymer
1999, 41, 925-932.

Lee, H.; Lee, W.; Chang, T.; Choi, S.; Lee, D.; Ji, H.; Nonidez, W. K.;
Mays, J. W. Macromolecules 1999, 32, 4143-4146.

Kreiser-Saunders, l.; Kricheldorf, H. R. Macromol. Chem. Phys. 1998,
199, 1081-1087.

Nieuwenhuis, J. Clin. Mater. 1992, 10, 59.
Fineman, M.; Ross, 8. D., . J. Polymer Sci. 1950, 5, 259-62.
Kelen, T.; Tudos, F. J. Macromol. Sci, Chem. 1975, A9( 1), 1-27.

‘ Kelen, T.; Tudos, F.; Braun, 0.; Czerwinski, w. K. 1990, 191, 1863-

1869.
Bataille, P.; Bourassa, H. J. Polym. Sci. Pol. Chem. 1989, 27, 357-365.

Polic, A. L.; Duever, T. A.; Penlidis, A. J. Polym. Sci. Pol. Chem. 1998,
36, 813-822.

Leicht, R.; Fuhrrnann, J. J. Polym. Sci. Pol. Chem. 1983, 21, 2215-2224.

Dong, C. M.; Qiu, K. Y.; Gu, Z. W.; Feng, X. D. Chin. Chem. Lett. 2000,
11, 815-818.

91

91
92
93
94

95

96

Schmidt, E.; Sachtleben, R. Justus Liebigs Ann. Chem. 1878, 193, 112.
Sinclair, R. G. J. Macromol. Sci-Pure Appl. Chem. 1996, A33, 585-597.
Breitenbach, A.; Kissel, T. Polymer1998, 39, 3261 -3271.

Liu, 6.; Fang, Y. E.; Shi, T. Y. Chem. J. Chin. Univ-Chin. 1997, 18, 486-
488.

Thakur, K. A. M.; Kean, R. T.; Hall, E. S.; Kolstad, J. J.; Munson, E. J.
Macromolecules 1998, 31, 1487-1494.

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

92

ll. SYNTHESIS AND CHARACTERIZATION OF
POLYACENES

93

INTRODUCTION

1. Structure of Polyacene

1.1. Chemical Structure

Polyacenes are the long chain members of the family of edge-fused
aromatic compounds that includes benzene, naphthalene, anthracene, and
higher homologs. As shown in Figure 1, acenes can also be viewed as one-
dimensional analogs of graphite, a ribbon taken from a single graphene sheet.

The valence bond structure of true acenes is ladder-like with the rungs of the

0 000

Edge Fused Aromatic Compounds

. n

Polyacene

 

Figure 1. Chemical structure of polyacene.

94

ladder connecting two trans-polyene chains. Theory predicts that these planar

polyacenes should have small band gaps (E9), about 0.4 eV.1

1.2. Electronic Structure and Properties

1.2.1 . Electronic structure of unsubstituted polyacene

Long chain polyacenes have not yet been synthesized, but the
electronic structure of the infinite polyacene has long been the object of
theoretical interest. There are four possible configurations for polyacene2 as
shown in Figure 2. In each unit cell, there are four sp2 carbon atoms, each
contributing one it electron and one hydrogen, thus giving rise to four 1: bands,
two of which are filled, which are termed valence bands, and two unfilled
bands, denoted as conduction bands. One of the most important features of
the electronic structure is the band gap, the difference in energy gap between
the highest energy valence band or highest occupied molecular level (HOMO)
and the lowest energy conduction band or lowest unoccupied molecular level

AUMO)

95

(a

V

©©©©©

non-altemant (aromatic)

altemant (quinoid) cis

CCCCC

altemant (quinoid) trans

(9 00000

single-site defect in one of the lateral chains

(b

V

(e

V

Figure 2. Four possible double bond configurations for polyacene.

Configuration a is a non-altemant structure; there is no alternation of
single and double bonds and all of the bonds has the same bond length and
bond energy. The band gap of this configuration is zero, the same as that of a
metal. However structure a is not the most stable configuration because it is
susceptible to a Peierls distortion that lowers the overall energy of the system
by forming alternating single and double bonds. There are two principal ways
for non-altemant polyacenes to undergo a periodic distortion. One leads to the
cis conformation b while the other leads to trans conformation c.
Configurations b and c, are termed “quinonoid” structures, and both have a
larger band gap than the nonaltemant polyacene. Structure d results from a

defect breaking the pattern of alternation between the single and double bonds

96

in one of the lateral chains, dividing the polyacene into a cis structure to the
left, and a trans stmcture to the right. According to our knowledge, there has
been no work on such defects, which should be involved in the conduction

process at low doping levels?"6

The first group to study the electronic structure of polyacene was Salem
and Longuett-Higgins in 1960.7 In their pioneering tight-binding study of the
electronic structure of polyacene, they concluded that the ground-state
configuration of the polymer should correspond to the non-altemate structure
a. Later, using the same calculational method, Boon pointed out that Salem
and Longuett-Higgins failed to take into account the plane of symmetry that
corresponds to the molecular plane. Inclusion of that symmetry gave the trans
form c as the most stable structure.8 Whangbho et al. included long-range
interactions through an extended-Hiickel-crystalline-orbital (EHCO) method to
compute the band structure of polyacene,9 and their results also pointed to the
alternate trans form as the most stable configuration with a band gap on the
order of 0.45 eV. Based on ab initio calculations, Cioslowski also concluded
that the trans-distorted structure has the lowest energy per unit cell, and
therefore was the most favored.10 A Complete-Neglect-of-Differential-Overlap-
Crystalline-Orbital (CNDO CO) study by Tanaka et al. agreed,11 with the band
gap estimated to be 0.504 eV for c. He also suggested that because of the

small band gap, polyacene should be a promising electrical conductor.

There is no experimental data for the band gap of polyacene. The next

best data are extrapolations of F.g obtained from the electron absorption

97

spectra of finite polyacenes. Similar plots of polyenes and polyaryls give
reasonably good estimates of £9.12 A plot of the Eg for polyacene oligomers up
to hexacene vs. the reciprocal of the number of rings (1/n) in the acene is
shown in Figure 3. The intercept with y-axis is 0.32 eV, close to the band gap
predicted for the trans form, providing circumstantial evidence that the trans

form is the most stable polyacene configuration.

98

 

©©©©

     
  

tetracene

0000
000000

E, (8V)
to

   

 

 

 

0 0.1 0.2 0.3 0.4
1/n

Figure 3. Empirical determination of the band gap (E9) of polyacene.

Polyacene is structurally related to polyacetylene in that it can be
regarded as resulting from the lateral interaction between two parallel trans-
polyacetylene chains. Studies of the electronic structure of polyacene and
polyacetylene within a first-neighbor tight-binding Hamiltonian model led to the
conclusion that the electronic structure of polyacene closely resembles that of
two interacting polyacetylene chains.7 However, the band gap of trans-

polyacetylene is about 1.4 eV, much larger than that of polyacene. The fused

99

ring structure of polyacenes force the top and bottom polyacetylene strands to
be in the same plane. Therefore the rt-band of the top strand interacts with
that of the bottom strand, decreasing E9. Defects on a single strand polymer,
such polyacetylene, degrade the conductivity. However, for a ladder or double
stranded polymer such as polyacene, charge carriers have the opportunity to
bypass a defect on the other polyene chain. In fact, polyacene meets
simultaneously the four criteria recently enumerated by Bredas for a polymer to
have a high conductivity at low voltages,13 since it possesses 1) a high degree
of rr-electron conjugation, 2) a very simple structure, 3) a planar rt-system, and
4) is expected to have a very low ionization potential associated with a very
large bandwidth. Besides a high conductivity, polyacene should have a higher

thermostability and mechanical strength due to its ladder structure.

100

5

Figure 4. Six methyl-substituted polyacenes studied by Wang and Hou

Band gap largely determines whether the polymer is intrinsically a
metal, a semiconductor or insulating material. Besides band gap, there are two
other important properties that related to the electronic structure: the ionization
potential (IF) and electron affinity (EA). For intrinsically semi-conducting
polymers, IP and EA are measures of the ease that the polymers can be
oxidized and reduced, respectively.14 The theoretical value of the IP of
polyacene is about 4.71 eV,11 which is rather low, suggesting that polyacene
can easily be doped with electron acceptors such as l2 or AsF5. The EA of
polyacene is predicted to be about 4.21 eV, signifying that polyacene can be
readily doped with electron donors such as alkali metals.“16 As a whole,

although polyacene is particularly stable, it is still very interesting as a

101

conductive material either by itself or after doping with electron acceptors or

donors.

1.2.2. Electronic structures of alkyl substituted polyacene

The preparation of polyacenes is difficult, partially due to the low
solubility of linear acenes. We are interested in alkylated polyacenes and how
the alkyl group will affect the electronic structure and conductivity of polyacene.
There are few studies on this subject. Wang and Hou17 studied the electronic
structure of three of the possible isomers of six different methyl substituted
polyacenes (Figures 4, 5). Their calculation, using a one-dimensional tight-
binding SCF—CO method at the CNDO/2 level, predicted that the most stable
configuration of all six methyl-substituted polyacenes is either structures b or c
of Figure 2, or in another words, like the unsubstituted polyacene, the quinoid
structure is a more stable structure than the non-altemant form. Some
electronic properties from their calculations are listed in Table 1. The band
gaps of structures 2, 3, and 6 are the largest because of the unsymmetric
substitution pattern of the methyl groups in these three derivatives. From the
values of ionization potential and electron affinity, both n-type and p-type

doping are expected to be effective for these three methyl derivatives. The

 

Figure 5. Three possible isomers of structure 5.

102

band gap of structure 1 is the smallest (highest symmetry) and therefore its
conductivity is expected to be the highest. The band gaps of 4 and 5 are small,
and therefore they too should be conductive materials, especially when n or p-
doped. Thus, to achieve both structural stability and good electrical
conductivity, synthetic efforts should be aimed at molecules with structures 4 or

5.

Table 1. Electronic properties of structures 1-6

 

 

 

 

 

1 2 3 4 5 6
E9 (eV) 1.77 4.32 2.80 2.45 2.48 3.04
IP“ (eV) 4.75 6.11 5.61 5.23 5.23 5.41
EA“ (eV) 2.80 1.79 2.81 2.79 2.76 2.37

 

 

 

 

 

 

 

a: ionization potential; b: electron affinity.

103

 

1.2.3. Properties of polyacenes

The most interesting feature of polyacene is its narrow band gap,
estimated at ~0.4 eV from theory and by experimental approximation. In terms
of their electronic properties, materials are usually classified as insulators
(band gap > 5 eV), semiconductors (band gap of 3 - 5 eV), and conductors
(band gap = 0) (Figure 6). The most conductive materials are metals (no gap
between the valence band and conducting band), where there are many free
unbound electrons that make metal conductive.18 Most conductive polymers
can be considered to be semiconductors. For example, the E9 of polypyrrole is
3 eV,19 that of polythiophene is 2 eV,19 and polyacetylene, one of the first

conducting polymers, has an E9 of 1.2 eV.19

1.3. Applications

1.3.1 . Electronically conductive materials

Conducting polymers have attracted considerable attention since the
1960s due to the wide range of possible applications for these materials. In
principle, conducting polymers offer two potential advantages over the
traditional inorganic materials used as conductors. First, the processing of
conducting polymers by molding and other plastics processing techniques is
easy compared to the metallurgical processes used for inorganic materials.
Second, the low densities of polymeric materials would make certain types of
applications more practical and economical. Polyacenes are promising

conductors due to their low band gap energy and are predicted to be highly

104

_NuOE

 

 

 

8.2205 0233283 83. 3mm 6 05a:

LouozocoEEom

20:35.3

3928....on
c

H>ov61r >oo.N >00.» >omv mm

a coo a: on

 

m: 9.. 0.2528 _

 

 

586—32.:

>omxom

 

 

105

have not yet been prepared,

although low molecular weight oligomers can be obtained by thermal or photo

induced polymerization.20

Acene-like materials, often termed PAS (polyacenic semiconductive
material), are prepared from pyrolytic treatment of phenol-formaldehyde
resin,21 styrene-divinyl benzene copolymer22 and polyphenylene.23 Most
studies of the applications of acenes use PAS since it is easy to prepare from
low cost starting materials. PAS has some of the attributes of graphite. It is air
stable, and its electrical conductivity ranges from 10'"-10° S cm", depending
on the degree of graphitization. The properties of PAS can be controlled by
the pyrolysis temperature.24 According to Tanaka, as the pyrolysis temperature
increases, the H/C molar ratio decreases and at the same time the conductivity
increases. Similar to most other conductive polymers, PAS can be readily
doped and its conductivity increases by up to 11 orders of magnitude.“25 The

highest reported conductivity after doping is around 103 S/cm.24

“Doping” is an important method used to improve the conductivity of a
material. Doping can be accomplished through chemical and electrochemical
processes. Chemical doping can be considered as a redox reaction in which

neutral PAS is oxidized or reduced to polycations or polyanions.

@
(CH)x + 0x1 == (CH)x + Red, (1)

e
(CH)x + mag: (CH)x + 0x2 (2)

106

 

where 0x1 and redz denote the oxidizing and reducing agents,
respectively, while red1 and 0x2 are their conjugates formed as a result of the

charge transfer reaction.

In electrochemical doping, the doping process corresponds to the
removal or addition of electrons at electrodes. Positively charged polyacene
molecules are said to be p-doped, and negatively charged polyacenes are n-
doped. The oxidation or reduction of polyacene requires the inclusion of an
oppositely charged counter ion in the polyacene. These so-called dopant ions
may be formed during the electron transfer reaction or they may be an original
component of the system. In the case of electrochemical doping, the dopant
ions correspond to the ions of the supporting electrolyte.

The obvious result of successful doping is an increase in the electrical
conductivity. ln conjugated materials, there is a finite energy that separates the

HOMO (valence band) and LUMO levels (conducting band). This energy gap,

(CH)x - e (CH)xEB (3)

(CH)? (4)

l l

(CH)x + e
E9, corresponds to the energy that must be applied to the system to observe
conductivity. Since E9 is usually >>kT, the gap must be decreased for there to
be appreciable conductivity at ambient temperatures. As shown in Figure 7,
oxidation of polyacene (p-doping) removes an electron from the HOMO, and

thus the energy of that level increases and EQ decreases. Similarly, adding an

107

 

 

 

 

 

 

 

 

 

 

1 LUMO A LUMOf
LUMO ‘ +e -e 159”
." """ Eg """ ‘T 1
59
T L T j HOMO
HOMO HOMO
n-doped p-doped

Figure 7. Molecular orbital energy diagram of polyacene and doped polyacene

electron to the LUMO (n-doping) lowers its energy and decreases E9. Thus,
both p and n-doping decrease E9 and increase the conductivity.

PAS can be readily doped because both structural and electrical
transport studies have shown that PAS has a porous, amorphous structure.26
This enables PAS to be heavily doped with high stability. Porous PAS samples
have macropores and micropores.27 The dimensions of the macropore is
around 104 A, suggest that it may play a role in promoting the transport and
diffusion of electrolyte during electrochemical doping, while the dimensions of
the micropores is about 10 A, roughly the same as that of the dopant
molecules and hence is considered to be closely related to the chemical
doping and undoping processes. The pores can be reduced to around 5 A by
controlling the specific pyrolysis temperature.27 The high porosity of PAS opens
the possibility of using PAS in rechargeable batteries. A detailed examination
of the electrochemical doping-undoping processes in PAS samples showedthat

p and n-doping were reversible, a requirement for rechargeable batteries.28

108

1.3.2. Optical limiters

Optical limiters are devices that strongly attenuate intense optical
beams while exhibiting high transmittance for ambient low-intensity light. These
nonlinear optical devices are of significant interest for the protection of human
eyes and optical sensors from intense laser pulses, which can pose
considerable hazards in laboratory and field environments. The simplest
limiters use materials whose optical transmittance decreases at high light
levels. Molecules with weak ground-state absorptions that form strongly

absorbing exited states can be used in optical limiters.29

At low fluences, the absorption of the light passing through materials

follows the Beer-Lambert law: log (lo/l). = ecl, where lo is the input fluence, | is

 

Output influence

 

 

 

Input fluence /Jcm'2

Figure 8. Relationship between output fluence and input fluence of a 70%
transmitting solution of a polyacene-based oligomer in toluene

109

the output fluence, and log (lo/l) is the absorption. This law implies that the
transmittance (mg) is independent of the input fluence. Figure 8 shows the
input-output response of a solution of a polyacene-based oligomer in toluene.
At low input fluence, the transmittance obeys the Beer-Lambert law, but at high
input fluence, the transmittance decreases with input fluence, an effect termed
“optical limiting”. Optical limiting was first observed by Giuliano and Hess for
indanthrone compounds.30 Since then, many compounds were shown to
exhibit optical limiting, especially conjugated rt-electron systems such as
metallophthalocyaines?‘'3‘2 diphenypolyenes,33 C50,34 070,34 and polyacene-
based materials.35 Mechanisms proposed for optical limiting include reverse
saturable absorption36 (non-linear absorption such as sequential single photon
absorption), nonlinear scattering,37 thermal blooming,38 and multi-photon

absorption.35

 

 

 

 

 

 

 

2 “a 4
. .21”

Triplet states
Singlet states

. o: Absoprtion cross section
0 TI Relaxation time

Figure 9. Five-level energy diagram of polyacene

110

The reverse saturable absorption mechanism, shown in Figure 9, yields
a reasonable explanation for optical limiting of conjugated n-electron systems
such as C50 at low input fluences (1J/cm2). As polyacenes also are conjugated
n-electron systems, the optical limiting behavior of polyacenes might also be
described using the five-level energy diagram shown in Figure 9. The initial
absorption is from a ground state (level 1), typically a singlet, to an excited
singlet state (level 2). This exited state can relax to its original ground state, be
excited to a higher singlet state (level 3), or transfer to the lower level of a
triplet state (level 4) by intersystem crossing. Triplet state absorption involves
excitation from level 4 to level 5. To obtain optical limiting, the upper singlet
and lower triplet must have larger absorption cross-sections (css and 0.) than
the ground state (09). The relaxation times between levels 3 and levels 2, 132,
and between level 5 and level 4, 154, are extremely fast. The upper singlet level
and lower triplet level should have large relaxation times (121 and 1:...)

compared with relaxation times 132 and 154.

1.3.3. Electrodes for lithium ion batteries

In 1982, a rechargeable (secondary) polymer battery was proposed in
which the charge-discharge cycle corresponds to the electrochemical doping-
undoping process of an electrically conductive polymer.11 Since then,
investigations of polymers as electrodes have shown that a serious limitation of
polymer batteries stems from the chemical instability of most conductive

polymers.39

lll

Perhaps the most important application of acene-like materials is their
use as anodes in lithium ion batteries. Porous PAS can be electrochemically
and reversibly doped with both electron acceptor and donors, and can be used
to store electrical energy.40 Since porous PAS can serve as both the cathode

and the anode in a rechargeable battery, two types of PAS-containing batteries

—I LOAD I
e'T O Q 19'
P'Li+ p
-e'l \» Li+ —~\l +e'

P P'Li+

 

 

(a) PAS/ PAS

 

 

 

 

 

Anode Cathode
PAS PAS

I LOAD I

e-lle_'——‘—elle-

P'Li+ MOx
-e'l \» Li” «\1 +e‘

P MOx'Li+

 

(b) PAS/ PAS

 

 

 

 

 

Anode Metal Oxide
PAS PAS

Figure 10. Cell structure and electrochemistry of electrodes in PAS batteries

112

were developed, PAS/PAS“42 and PAS/metal oxide cells.43'Both types of
batteries cycle well. The cell structure and electrochemistry of electrodes of
these two types of batteries are shown in Figure 10. In the PAS/metal oxide
battery, the charged cell corresponds to lithium intercalated in the carbon, and
the discharge reaction is the transfer of lithium ions from the PAS anode and
intercalation of the ions in the metal oxide structure. Rechargeable batteries
function by repeatedly cycling Li+ between the metal oxide and the PAS
anode. Fully charged batteries have potentials of 3.5-4.5 volts, depending on
the chemical potential of the host materials.

An important parameter of these lithium cells is their capacity, usually
expressed as Ah/g, which is limited by the moles of lithium that can be inserted
reversibly into the carbon. For highly crystalline graphite, the interlayer
distance of graphite is 3.35 A and it can be doped with lithium only up to the

CeLi stage44 with expansion of the interlayer distance to 3.70 A as indicated in

@: $0
00‘0“

Figure 11. The CsLi configuration observed in the first
stage Li-graphite intercalation compound (Li-GIC)

113

Figure 11.25 The strong electrostatic repulsions prevent occupation at nearest
neighbor sites, therefore in this state, Li atoms are located over every three
benzene rings. The theoretical capacity of the fully intercalated graphite is 372
mANg

I. 45 I. 46

Recent work by Dahn et a and Sawada eta using carbon heat-
treated at low temperatures show that the CeLi ratio can be exceeded. The key
to the increase in capacity seems to be the high degree of disorder found in
carbon pyrolyzed at low temperature. PAS materials are one some of the most
promising amorphous carbons. PAS materials have a larger interlayer distance
((10 = 3.7-4.0 A than that of graphite (d6 = 3.5 A), and the results of electron
spin resonance (ESR),47 X-ray diffraction (XFiD),48 and 13C-NMR49 have been

rationalized to provide the illustrative model of PAS shown in Figure 12.

Owing to such a loose structure, PAS materials have been intensively
examined as anodes for Li-rechargeable batteries.5°'51 Yata reported that PAS
electrodes have capacities as high as 850 mAh/g with good stability and

52 Moreover, an extraordinary high capacity of 1100 mAh/g,

reversibility.
corresponding to the molar ratio [Li]/[C] = 1/2, was reported in 1993.44 The CzLI
stage is obviously much more highly doped than the CsLi stage of graphite,

and has an energy density on a volume basis almost equal to Li metal.

The Li storage mechanism of PAS materials has been studied
extensively, but is still controversial, clue to their disordered structures. A model
with Li atoms located over every benzene ring (C2Li) was proposed on the

basis of Li NMR measurement by Sato et al.53 proposed at the excess

114

 

capacity corresponds to cluster-like lithium doping that fills the void space in
the disordered carbon,54 while Dahn et al. argued that lithium can also be
located at the edges of the graphite sheets.55 The latter view is supported by
experiments that show that the increased capacity is correlated with the
hydrogen content of the electrode,45 as well as by recent calculations.56 Yata et
al. studied the electronic and geometrical structures of Li-doped PAS using a
semiempirical molecular orbital (MO) method and studied the Li storage
mechanism and capacity in detail. According to their calculation, Li can edge
intercalate between carbon layers of PAS, as well as absorb on the surface
andplanes of carbon substrates. Since the maximum ratio of intercalating Li is
only CaLi, it is the Li absorbed on the surface and peripheral edges that push

the capacity to as great as C2Li.

Since the charge-storing capacity is highest when the residual hydrogen
content is between 0.15 and 0.30 H atoms per C atom, it is important to be
able to prepare acene-like structures with a substantial fraction of edges.
Current routes to carbons use petroleum pitches or phenolic resins as the base
material, and when pyrolyzed, these materials give partially graphitized
structures. However, the high temperatures needed for the pyrolysis allows
little control of the acene-like structural elements that are the precursors to
graphite. In contrast, if elements of the acene structure can be designed into
the starting material, it should be possible to obtain partially graphitized

structures at low temperatures that retain some memory of the chemical

116

structure before pyrolysis. In principle, PAS could be prepared that requires no

thermal treatment for it to be useful electrode materials.

However, one disadvantage of PAS materials is that there often is a
large capacity loss in the first charge-discharge cycle in rechargeable Li ion
batteries, and there is a large hysteresis in charge-discharge potentials.50 PAS
materials with higher hydrogen contents usually have higher capacities but
show larger hysteresis. This suggests that the hysteresis is caused by Li
binding at the hydrogen terminated edges of graphene fragments. Bonding
changes in the host have been previously shown to cause hysteresis in
electrochemical measurements.57 If the inserted Li binds to an edge carbon
atom, then the carbon-carbon bond changes from sp2 to spa, which would in

turn, cause the hysteresis.

Besides hysteresis, PAS materials usually exhibit capacity loss on the
first charging cycle. The irreversible capacity observed in the first discharge
step is defined as the difference in capacity between the first discharge and the
subsequent charge. The irreversible capacity is usually ascribed to the
formation of a solid electrolyte interface (SEI) layer on the carbon surface due
to the decomposition of electrolyte. Lithium can also irreversibly react with
surface functional groups such as hydroxyl and carboxyl groups. Since PAS
materials are highly disordered, the samples likely have many defect sites
where surface oxides may form. Thus PAS or polyacene-like materials offer

the highest capacities found thus far in anode materials, but more work is

117

needed to verify the reported values of specific energy, improve the discharge

curves, and reduce the irreversible capacity lost on first charging.

1.3.3. Capacitors

The high capacities and deep-discharge cycle of PAS makes it an
attractive material for capacitors.”59 Yata et al. studied PAS-based capacitors
and found that their capacities were much higher than that of conventional
electric double-layer capacitors (EDLCs) which uses activated carbon as the
electrode.60 A PAS capacitor can maintain more than 70% of the initial capacity
even after 100,000 cycles. Moreover, this capacitor can be charged and
discharged in a few minutes as well as at lower rates. PAS capacitors are
produce in coin-type and cyclindrical geometries, the coin-type capacitors are
currently used for memory back-up in electrical and communications
equipment, while the cyclindrical geometry provides higher power densities

and is used as back-up power for starting the drive parts of electric equipment.

1.4. Synthetic Approaches to Polyacenes

Efficient methods for the preparation of well-defined polyacenes are not
yet known. Problems with stability and solubility have stymied the synthesis of
long chain acenes, although oligomers containing several rings have been
synthesized. Efforts to synthesize polyacenes have been reported,“62
however insolubility, low purity, and crystallinity prevent these materials from

being directly compared to the theoretical work on idealized infinite polyacene

chains.

118

Unsubstituted linear polyacenes up to heptacene have been
synthesized.63 As the length of the polyacene chain increases, the chemical
reactivity also increases. Materials become so sensitive toward photooxidation
that polyacenes higher than pentacene and hexacene must be handled under
nitrogen. Heptacene has never been obtained in a pure state. The high
reactivity of acenes shows up in acene oligomers such as anthracene, where
oxidation and halogenation at the 9,10-positions is facile. Similarly, an infinite
polyacene becomes unstable toward formation of two sp3 carbons flanked by
two polyacene chains. Not surprisingly, the higher acenes have proved too
reactive to isolate under ordinary conditions, and only derivatives of

octacene,“65 nonacene,66 and undecacene67 have been prepared.

There are two general synthetic routes to polyacenes: “topochemical

polymerization”, and “Diels-Alder polyadduction.

119

1.4.1. The Diels-Alder approach

Diels-Alder reactions have been employed in the stepwise construction
of compounds with molecular architectures that resemble ladders, such as
polyacenes.”74 An example of this strategy that of Clardy,72 shown in Scheme
1, is the reaction of a dienophile with an electron rich diene. Changing the
nature of the olefin changes the end groups of the acene oligomer (Figure 13).
Although the reported yields are very high, these reactions cap the acene, and
do not allow for further extension of the polyacene chain. Thus, only
anthracene derivatives were prepared. As shown in Scheme 2, Mullen et al.
used benzyne (generated with 1,2-dibromobenzene and n-butyllithium at -40
°C) as the dienophile to synthesize higher acenes.75 The key to the synthesis is
to mask the center ring using a group that can be later eliminated via a retro

Diels-Alder reaction to give the fully conjugated acene. Due to the driving force
Ph Ph Ph
Ph Ph A Ph
+ 00 —- 00000
ph Ph Ph
Ph Ph Ph

COPh

PhOC
PhOC COPh KOH/MeOH PhOC COPh
—> 000
PhOC COPh PhOC COPh
Scheme 1. Polyacene oligomer synthesized by via Diels-Alder adduction

120

00000

92%

00©©©00

90%

Figure 13. Other acene derivatives made by Clardy’s group
of forming benzene derivatives, this reaction takes place at low temperatures
(70 °C) to give the pentacene product in 90 % yield. However, the same
approach failed when applied to nonacene. Because the anthracene unit in the
nonacene precursor itself can act as a diene, the reaction of the nonacene
precursor and tetrachlorothiophenedioxide and its bromo homolog turned out
to be very complicated.

Diels-Alder reactions have also played an important role in the synthesis

7376 and spiro acenes.68

of various acene-related species, such as cyclacene
Thus, we see that while the DieIs-Alder reaction, in principle, allows for the
stepwise growth of polyacenes, its use has been limited to heptacene and

shorter acenes. This approach is probably not a practical method for

synthesizing long chain linear polyacenes.

121

. :(I
0 : 00000

Br toluene, -40°C

 

toluene, reflux

00000

X

X
/
3:...
X

X

l chloranil

X= Cl, Br

 

X
: :X
eliminate
X
X

(pentacene)

Scheme 2. Polyacene oligomer synthesized via Diels-Alder addition and retro
Diels-Alder elimination.

122

1.4.2. Topochemical polymerizations

Besides sequential Diels-Alder reactions, the other common approach
to polyacene synthesis is through topochemical polymerization of butadiyne”
79 or 1-buten-3-yne70'78'79 derivatives followed by pyrolysis. As shown in
Scheme 3, metathesis polymerization of the diyne is selective for the 1,2
pathway. Triple bonds are more reactive than the double bonds in coordination
polymerizationeo, and thus the acetylene is the major reaction site during
metathesis polymerization of 1-buten-3-yne derivatives. The pyrolysis step is a
high-temperature polycondensation that yields graphite-like pyropolymers. The
pendant substituents on the pre-polymers not only suppress the gasification of
the polymer itself, they also control the morphology of the final pyropolymers.
Bulky substituents, such as phenyl group, are thought to hinder the
development of polymer order during pyrolysis, and therefore cause the final
graphite-like polymer to contain defect sites.

Compared to the Diels-Alder approach, topochemical polymerization
provides much higher molecular weight polyacenes, but with poor control over
the configuration and morphology of the final product. More important, most of

the pyrolysis products resemble two-dimensional graphite rather than the one-

dimensional linear polyacenes.

123

1.4.3. Acene-like materials

Although polyacene haves promising properties, the synthesis of perfect
linear polyacenes is very difficult. Despite various attempts to synthesize
polyacenes, this intriguing, fully unsaturated ribbon polymer is still unknown.
One solution to this problem is the preparation of acene-like products from
well-commercialized starting materials. These materials are termed “polyacenic

semiconductors” or PAS, and are the materials that are studied in most

 

NbCl5 or TaCl5 NbCl5 or TaC|5
chlorobenzene chlorobenzene
80°C, Ar 80°C, Ar
Ph

 

A ' in vacuo

defect sites and edges

 

Scheme 3. Topochemical polymerization of butadiyne and
1-buten-3-yne to prepare polyacene

124

application-based research.

Among the acene-like materials, the most well-studied and broadly-
utilized is the pyrolyzed product from phenol-formaldehyde resin, which is
named “PAS-PF”. As shown in Scheme 4, porous PAS-PF is prepared by the
pyrolysis of phenol-formaldehyde resin molded with 211012.24 The ZnClz serves
as the pore-forming agent, resulting in open pores that grow into a three-
dimensional network structure in the cured material. This leads to

aloosestructure that can be easily

OH OH OH

9 e CHZOH
(j + HCHO $333+ + (hr

0 D O O

OH OH OH OH OH OH OH

O O O O

phenol-formaldehyde resin

-H

   
 

       
               

Scheme 4. Preparation acene-like material from phenol-formaldehyde resin.

125

doped with bulky dopants. To prepare PAS-PF, a mixture of resole-type
phenolic resin, 211012 and water was heated to 100 °C for 1 hr. The cured resin
was further heat-treated at 500-800 °C in a non-oxidative atmosphere. The
reaction is essentially a carbonization process consisting of dehydration and
dehydrogenation, the degree of which can be controlled by the temperature
applied. ZnCl2 was removed by washing with water at the last step. As shown
in Scheme 5, similar acene-like materials can be produced by pyrolysis of

styrene-divinylbenzene copolymers (PAS-SD) and poly (p-phenylene) (PAS-

.ggg 0 0 0 0
/ 0 0

pyrolysis ; .. PAS-SD

 

 

O O O _. ..
" COO

Scheme 5. Acene-like material synthesized from styrene-
divinylbenzene copolymers and poly(p-phenylene)

126

PP).

1.5.4. Soluble polyacenes

Polyacenes and PAS are promising and useful organic conductive
materials, however, the established methods for the synthesis of polyacenes
are inadequate. Most synthetic efforts fail because of the poor solubility of
polyacenes. To date, heptacene has been the longest acene prepared.81
Having soluble polyacenes would enable full characterization of their structure

and properties, and improve the processability of conductive polyacene

Pr Pr
LiAlH
Pr COOMe —“>

PBra
2m1Lii:: R
Pr COOMe 7

 

Pr Pr

Cp22r802

 

MeOOC-E—COOMe

 

 

Pr COOMe

000.0
Pr COOMe

Pr COOMe

Pr 3“ /ECOOMe

Scheme 6. Synthesis of highly alkyl-substituted pentacene
derivatives by homologation.

127

materials.

One common approach to solve the solubility problem is by use of
soluble precursors of acenes, which can be converted to the desired acenes
after processing.75'82 Scheme 2 shows an example of this approach, the
synthesis of pentacene. A retro Diels-Alder reaction is an appropriate step for
the solid state transformation of a precursor to the product because a small,
volatile molecule (benzene in this case) is eliminated and can be easily

removed from the reaction medium.

It is well known that solubility problems can be solved by introducing
alkyl substituents to rc-conjugated aromatic compounds,75 and a similar
strategy can be applied to improve the solubility of polyacene. However, there
is no general method for the synthesis of highly substituted alkyl-acenesm“36
Nakajima et al.71 recently reported a general method for the synthesis of alkyl-

substituted pentacenes and naphthacenes by the homologation scheme shown

in Scheme 6.

I 1.6. Characterization of Polyacenes

The intractability of polyacenes severely limits the ability to accurately
determine structures of polyacenes. UVNIS, near IR, elemental analysis, X-
ray diffraction, Raman, and other spectroscopic methods have been used to

analyze the polyacene network.

128

1.6.1. X-ray diffraction

There are several X-ray diffraction studies of polyacenes, with most of
the analyses performed on polyacenic semiconductive materials derived from
PAS-PF, a pyrolyzed phenol-formaldehyde resin. Sugawara et al. studied the
graphitization of these materials86 and Ban et al. reported X-ray studies of
alkali metal intercalation in amorphous carbon.87 Tanaka et al. studied porous
PAS-PF,”52 and the X-Ray results from pristine samples pyrolyzed from 25 to
935 °C show two broad peaks in all samples, indicating that the PAS-PF
samples are essentially amorphous. As the I pyrolysis temperature (T p)
increases, peaks around 20 = 20-24° shift to larger angles and appear to
evolve toward the (002) peak of graphite. Thus the samples have a layered
structure that becomes increasingly graphitized as Tp increases. The interlayer
distances, dc, and the crystallite size, LC, were derived from the (002) band.
The range of (10 values for PAS (4.42 A to 3.76 A) is considerably larger than
that of graphite (3.35 A), and Lc remains small (8-9 A) for all the pristine

samples, indicating that PAS lacks long range order.

1.6.2. 13c NMR

Solid state NMR has also been used to characterize undoped and
doped PAS-PF derived from pyrolyzed phenol-formaldehyde resin. The NMR
spectra clearly indicate that heat-treatment converts the resin from a phenolic
resin to a polyacenic material by the decrease in peaks at 150 ppm, assigned
to aromatic carbons bonded to the hydroxy group, and at 40 ppm which

corresponds to the methylene carbons. Ando et al.88 obtained dipolar-

129

dispersion (DDPh)/MAS NMR spectra of PAS-PF. From an analysis of the data
and decomposition of overlapping peaks by computer-aided curve fitting, they
assigned each peak to different carbons of the polymer. These experiments
are hampered by the complexity of the pyrolyzed structure, making it difficult to

make more than basic structural assignments.

1.6.3. ESR (Electron Spin Resonance)

The ESR studies of PAS-PF mainly focus on the doping and undoping
p,rocesses.89'9° The ESR peak seen in pristine samples is believed to result
from carbon-centered radicals associated with structural defects. When acenes
are oxidized or reduced, the ESR spectral lines broaden, probably due to the
formation of charged sites on the polyacene structure.90 Ando et al.89 studied
Li-doped PAS-PF which was used as an anode of the Li rechargeable
batteries. initially, the line width increased, but a narrow spike showed up in the
ESR spectrum of heavily doped PAS (Li doping amount x > 50%, for CzLi, x =
50%). This spike is due to the formation of Li° clusters, which is consistent with

021.1 being the highest doped stage for PAS.

1.6.4. UV and fluorescence

Among all the analytical methods, UVNIS and fluorescence studies
should provide the most straightforward characterization of the electronic
structure of polyacenes. The presence of polyacene units can be detected by
their strong absorbance through the visible region and into the near IR. The
absorbance of a perfectly linear polyacene is expected to tail into the IR region.

For soluble acene materials, the distribution of the length of acene segments

130

can be estimated by comparing the UV and fluorescence spectra91 with those
of known linear oligomers. For PAS materials, UV and fluorescence are not

applicable due to the poor solubility of these materials.

1.6.5. Raman spectroscopy

Raman spectroscopy is an important technique for the characterization
of polyacene networks. Although no Raman spectra of authentic linear acenes
have been reported, Raman spectroscopy is widely used to characterize
polyacene-like material such as PAS-PF.92 Two peaks in the Raman spectrum
are used for diagnostic purposes, one at 1575 cm'1 assigned to the E29
vibration mode of the crystalline graphite, and a second at 1355 cm‘1 from the
A19 vibration from amorphous portions of graphite. The intensity of the latter
band is inversely proportional to the crystallite size and also allows an

estimation of the crystallite size in the surface layer of any carbon sample.

131

EXPERIMENTAL
2.1. General

Tetrahydrofuran (T HF) and toluene were first distilled from powdered
calcium hydride under nitrogen, then from sodium benzophenone ketyl under
nitrogen. Unless otherwise specified, ACS reagent grade starting materials
and solvents were used as received from commercial suppliers without further
purification. 1H and 13C NMR spectra were measured using a Varian Gemini-
300 spectrometer at 300 MHz. All samples were run at room temperature in
CDCl3. The chemical shifts were calibrated using residual CHCI3 and reported
in ppm (6) relative to tetramethylsilane. Infrared spectra were obtained from
KBr pellets under nitrogen at room temperature on a Nicolet Magna-IR 550
Fourier Transform lR spectrometer. A Hitachi U-4001 UV-visible spectrometer
was used to obtain the UV-visible spectra of cyclohexane solutions of the
polymers. The spectra of Raman was taken by exciting the sample at 633nm
on a Holoprobe Raman Spectrometer. Polymer molecular weights were
determined by gel permeation chromatography (GPC) using a PLgel 2011
Mixed A column with THF as the eluting solvent at a flow rate of 1 mL/min, a
Waters R401 Differential Refractometer detector at room temperature, and a
Varian 2050 Variable Wavelength detector set at 254 nm. Monodisperse
polystyrene standards were used to calibrate the molecular weights. The
concentration of the polymer solutions used for GPC measurements was 1
mg/mL. Therrnogravimetric analyses (TGA) were performed in nitrogen and air
atmospheres at a heating rate of 10 9C/min on a Perkin-Elmer TGA 7

instrument.

132

2.2. Synthesis

bis(methoxymethyl)benzenes. A mixture of hexanes/MeOH (1:1, 50
mL) was added drop-wise to a slurry of sodium hydride (NaH, 1.4 g, 57 mmol)
that had been rendered oil-free by washing with hexanes. The solution was
stirred at room temperature until all of the NaH reacted. The sodium
methoxide solution was transferred via cannula to a solution of the
bis(bromomethyl)benzene (6.0 g, 23 mmol) in hexanes (100 mL). The mixture
was gently refluxed for 3 hr, during which the color of the solution turned
slightly tan. After cooling, the solution was washed with water (3 x 50 mL).
The organic layer was passed through a short pad of Celite, and dried with
MgSO4. The solvent was removed under reduced pressure and the products
were purified by vacuum distillation to yield clear oils.

1,2-bis(methoxymethyl)benzene. Prepared in 91% yield (3.45 9); bp
98 °C/25 mm Hg (111.93 4045 °C/2 mm Hg). 1H NMR (000).) 6 3.35 (s, 6H), 4.5
(s, 4H), 7.2-7.35 (m, 4H).

1,3-bis(methoxymethyl)benzene. Prepared in 85% yield (3.21 9); bp
110 °C/25 mm Hg (111.93 25-130 °C/2 mm Hg). 1H NMR (000).) 6 3.35 (s, 6H),
4.4 (s, 4H), 7.25-7.4 (m, 4H).

1,4-bis(methoxymethyl)benzene. Prepared in 89% yield (3.36 9); bp
105 °C/25 mm Hg ((11.93 55-60 °C/2 mm Hg). 1H NMR (000(3) 6 3.35 (s, 6H),
4.4 (s, 4H), 7.4 (s, 4H).

Polymerization of the bis(methoxymethyl)benzenes. A 100 mL

round bottomed flask was charged with the bis(methoxymethyl)benzene (1.0 g)

133

and cooled in an ice bath. Concentrated sulfuric acid (3 mL) was added, and
the mixture was stirred at 0 °C for 36 hr. The resulting suspension was
washed with 20 % (w/w) sodium hydroxide solution (3 x 25 mL), water (3 x 25
mL), and rinsed with ether (3 x 50 mL). The collected solid was dried overnight
at 120 °C under vacuum to yield 0.44 g (72%) of a finely powdered, tan solid.
Chemical dehydrogenation of polyacenic networks. The tan powder
(0.270 g) from the acid catalyzed polymerization of the bis(methoxymethyl)
benzene, 2,3-dichloro-5, 6-dicyano-1, 4-benzoquinone (DDQ, 1.20 g, 5.3
mmol), and xylenes (60 mL) were added to a 100 mL round bottomed flask.
The yellow DDO immediately tumed dark violet. The flask was fitted with a
water condenser and purged with nitrogen. After refluxing for 48 hr, the
suspension was collected via filtration and the solid was repeatedly washed
with benzene, ethanol, water, and acetone until the filtrate was colorless. The
solid was dried at 120 °C under vacuum for 3 hr to give 0.25 g of a tan powder.
1,4-bis(bromomethyl)-2,5-dibromobenzene. Into a 500 mL round-
bottom flask was added 2,5-dibromo-p-xylene (20.0 g, 0.076 mol), N-
bromosuccinimide (28.4 g, 0.160 mol), a catalytic amount of benzoyl peroxide
(10 mg), and carbon tetrachloride (CCI4, 250 mL). The flask was fitted with a
water condenser and the system was flushed with argon. The reaction was
initiated by a 250 W sunlamp. The solution was gently refluxed until the
succinimide floated atop the mixture (1-2 hours). Upon cooling, the
succinimide was removed by filtration and rinsed thoroughly with CCl4. The

combined organic layers were washed with water (3 x 50 mL) and dried with

134

MgSO4. Filtration and evaporation of the solvent in vacuo resulted in a slightly
yellow solid. Repeated recrystallization from ethanol yielded 10.6 g (33%) of
1,4—bis(bromomethyl)-2,5-dibromobenzene as a fine white powder. mp 156-
157 ‘2C (lit.94 161-162 9C). 1H NMR (CDCI3) 5 7.65 (s, 2H), 4.50 (s, 4H).
1,4-bis(methoxymethyI)-2,5-dibromobenzene. Over a period of one
hour, anhydrous methanol (5 mL) and hexanes (10 mL) were added drop-wise
via a pressure equalizing addition funnel to an argon filled 100 mL round
bottomed flask containing a stirred solution of oil-free sodium hydride (2.1 eq,
0.051 mol) in 25 mL hexane. After the addition was complete, the flask was
fitted with a reflux condenser and the solution was gently refluxed for an
additional hour until all of the sodium hydride reacted. Upon cooling, the
solution was transferred via cannula to argon filled 250 mL round bottomed
flask containing 1,4-bis(bromomethyl)-2,5-dibromobenzene (10.30 g, 0.024
mol) dissolved in warm hexanes (150 mL). The mixture was gently refluxed for
3-4 hours. The cooled reaction mixture was poured into a separatory funnel
containing 100 mL dilute HCI (1N) and extracted with dichloromethane
(CHZCIZ). The combined organic layers were dried with magnesium sulfate,
filtered, and the solvent evaporated under reduced pressure to yield the crude
product as a yellow solid. Recrystallization from ethanol yielded 6.64 g (84%)
of 1,4-bis-(methoxymethyl)-2,5-dibromobenzene as a pale yellow crystalline
solid. mp 68.5-70 °C (lit.94 71-72 °C). 1H NMR (CDCla) 5 7.61 (s, 2H), 4.44 (s,

4H), 3.45 (s, 6H).

135

1,4-bis(methoxymethyI)-2,5-dialkylated-benzenes. To a round bottom
flask were added an alkene (5.5 mmol), dry THF (5 mL) and then a solution of
9-BBN (0.5 M solution in THF, 5.5 mmol). The solution was warmed slowly to
room temperature and then stirred for 4-6 h to give a solution of the alkyl-9-
BBN derivative. To the above solution were added PdCI2 (dppf) (0.15 mmol),
the 1,4-bis(methoxymethyl)—2,5-dialkylated-benzene (5 mmol), an additional 12
mL THF and aqueous NaOH (5 mL of 3 M solution) at room temperature. The
mixture was refluxed overnight. After the reaction was complete, the reaction
mixture was diluted with dichloromethane (20 mL), and the residual borane
was oxidized by addition of H202 (30%, 2 mL) at room temperature. The
product was extracted, washed with brine, and dried over magnesium sulfate.
The crude product was purified by recrystallization from ethyl alcohol.

1,4-bis(methoxymethyl)-2,5-dioctylated-benzene. Prepared with a
yield of 64%. ‘H NMR (CDCI3) 5 7.46 (s, 2H), 4.34 (s, 4H), 3.42 (s, 6H), 2.84(t,
4H), 1.94(m, 4H), 0.82-1.40 (m, 26H).

1,4-bis(methoxymethyl)-2,5-dioctylated-benzene. Prepared with a
yield of 57%. 1H NMR (coma) 5 7.42 (s, 2H), 4.32 (s, 4H), 3.452(s, 6H),
2.86(t,4H), 1.95 (m,4H), 0.80-1.48 (m, 34H).

Polymerization of bis(methoxymethyl)benzenes. To a 25 mL
Schlenk flask containing 10 mL of CHzclz was added the 1,4-
bis(methoxymethyl)-2,5-dialkylated-benzene (0.26 mmol). Following 3 freeze-
pump-thaw cycles, BF3.H3PO4 (0.8 mL, 9 mmol) was added slowly to the

solution. The mixture was purged by argon for four days under room

136

temperature, quenched with H20, and extracted with CHzclz. Evaporation of
the solvent gave the polyacene precursor as a brown solid.

Dehydrogenation. To 10 mg of the above polyacene precursors was
added a deoxygenated solution of D00 (20 mg, 0.088 mmol) in 10mL of
toluene. The reaction was refluxed under argon overnight, and then the D00
was removed by an Alea column.

2-Benzoylbenzoic acid. A mixture of 5 g of phthalic anhydride, 9 g of
AICI3 and 80 mL of benzene were shaken for 30 min. at room temperature. The
temperature was then raised to the reflux temperature, and after heating for 2
hr, the reaction mixture was hydrolyzed with 1N HCI. A grayish-white solid was
removed by filtration, and the filtrate was extracted three times with diethyl
ether. The combined ethereal extracts were washed with 5% aqueous sodium
bicarbonate and then were acidified with cone. HCI. A milky suspension
formed, and after storing the product overnight in a refrigerator, white crystals
were obtained. A 98% yield of the product was obtained after collecting the
crystals by filtration. m.p. 96°C.95 1H NMR (CDCla) 5 11.6 (s, 1H), 8.00-7.27
(m, 9H).

2-Benzylbenzoic acid. 2-Benzoylbenzoic acid (7.6 g) was reduced at
approximately 65 °C in 20 mL of acetic acid and 1.2 g of 10% Pd on carbon.
After 2 hr, the catalyst was remover by filtration and the solution was poured
into 100 mL of water. After the mixture was chilled for 12 hr, the product was

collected by filtration and dried under in vacuum to yield 4.5 g (65%) of

137

colorless 2-benzylbenzoic acid. mp. 111-113°C.96 1H NMR (CDCI3) 6 11.6 (s,
1H), 7.33-8.00 (m, 9H), 4.46 (s, 2H).

2-Benzylbenzyl alcohol. To a stirred suspension of 1.1 g of LiAlH4 in
50 mL of anhydrous ether was added a solution of 4 g of 2-benzylbenzoic acid
in 20 mL of anhydrous ether. The resulting mixture was stirred and refluxed for
one hour, and the excess hydride was destroyed by addition of 2 mL of ethyl
acetate. The ether solution, after washing with 10 % sulfuric acid and 10%
sodium bicarbonate, was dried over magnesium sulfate and the solvent was
later removed. The oily residue was crystallized from 30-60 °C petroleum ether
to give 3.2 g of 2-benzylbenzyl alcohol, in 53% yield. mp. 40-41 °C.97 1H
NMR, (CDCI3) 5 7.40-7.07 (m, 9H), 4.58 (s, 2H), 4.04 (s, 2H), 1.80 (br s, 1H).

2-Benzylbenzyl methyl ether. To a round bottom flask containing a
stirred solution of NaH in anhydrous THF under argon, was added drop-wise a
solution of methyl iodide in THF. After the addition was comlete, the mixture
was refluxed gentlly for 1hr. Upon cooling, a solution of 2-benzylbenzyl alcohol
in dry THF was added drop wise, and the mixture was refluxed overnight. The
mixture was cooled and the suspension was filtered. The collected solid was
rinsed with dichloromethane, and the the combined organic layers were
washed with water, dried over magnesium sulfate, filtered and evaporated to
dryness. The product was purified by column chromatography to give a pale
yellow liquid in 68% yield. 1H NMR (CDCI3) 5 3.42 (s, 3H), 4.15 (s, 2H), 4.49

(s, 2H), 7.0-7.6 (m, 9H).98

138

Cyclization of model compound in Cflzclz. A solution of 25 mg of 2-
benzylbenzyl methyl ether in 40 mL of CHzclz was degassed via 3 freeze-
pump-thaw cycles, and then 0.5 mL of BF3/PA was added drop-wise. The
reaction was stirred for 12 hr at rrom temperature and was quenched by
adding 30 mL of water. The organic layer was washed with water until it was
no longer acidic, dried over magnesium sulfate and the solvent removed. A
white solid was obtained in 95% yield. 1H NMR (CDCI3) 8 3.9 (s, 2H), 7.14 (dd,
4H), 7.36 (dd, 4H).

Nickel catalyzed polymerization of 1,4-bis (methoxymethyl)-2,5-
dibromobenzene. Inside a helium atmosphere dry box, NlC'z (8.6 mg, 0.067
mmol), bipyridine (10.4 mg, 0.067 mmol), and zinc powder (0.27 g, 4.12 mmol)
were added into a 100 mL round bottomed flask fitted with a reflux condenser
capped by a rubber septum. After removal from the dry box, the system was
connected to an argon bubbler through the rubber septum and placed in an oil
bath. DMF (1 mL) was added to the flask and the solution was stirred
magnetically. The mixture was then heated to 80 °C. The green solution
became brown after 10 minutes, eventually turning deep red-brown after 30
minutes. The 1,4-bis(methoxymethyl)-2,5-dibromobenzene (0.43 g, 1.33
mmol) was dissolved in DMF (4 mL) in an argon filled 50 mL round bottomed
flask and transferred via cannula to the catalyst mixture. The solution returned
to its original green color upon addition of the monomer but turned deep red-
brown after 15 minutes. The mixture was refluxed overnight. Upon cooling,

the solution was poured into a flask containing 200 mL of 2N HCVMeOH (1 :1)

139

and stirred until the excess zinc dissolved. The yellow precipitate was
collected by filtration and was repeatedly washed with distilled water and
methanol. The polymer was dried under vacuum at 110 °C overnight to yield
0.194 g (88%) of a fine, pale yellow powder that was insoluble in common
organic solvents. Anal. Calcd. for (C10H1202)n: C, 73.17; H, 7.32; O, 19.51.
Found: C, 69.85; H, 7.19; 0, not determined.

Cross-linking of methoxymethyl substituted PPP. The bis
(methoxymethyl) substituted PPP 6 (0.035 g) and concentrated sulfuric acid (3
mL) were mixed in a scintillation vial and cooled in an ice bath. Stirring at 0 °C
was continued for 36 hr. The suspension was washed with 20 %(w/w) sodium
hydroxide solution (3 x 25 mL), water (3 x 25 mL), and rinsed with ether (3 x 50
mL). The collected solid was dried overnight at 120 °C under vacuum to give

0.009 g (30%) of the cross-linked polymer as a tacky, black solid.

140

RESULTS AND DISCUSSION

Realization of the many applications of polyacenes is limited by their
instability (facile oxidation in air) and poor solubility.99 Thus the longest acene
synthesized to date is heptacene.100 The primary goal of this work was to
devise and execute the synthesis of well-defined polyacenes and acene-like
materials. We investigated two types of polyacene syntheses based on acid-
catalyzed electrophilic aromatic substitution reactions. The first targets simple
routes to acene-like materials that contain a substantial fraction of acene
subunits. Materials resulting from this approach are relevant to applications
such as electrical conductors, anti-static agents, and carbon electrodes, where
it is important to be able to supply large amounts of material at relatively low
cost. The second target is to develop a synthetic scheme for the preparation of
soluble well-defined polyacenes that would enable characterization of the
structure-property relationships of polyacenes and comparisons of the

experimental data with theoretical predictionsf”101

3.1. Polyacenic materials.

3.1.1. Simple routes to acene-like materials

It is well-know that methoxymethyl benzenes condense via acid-
catalyzed electrophilic aromatic substitution reactions to give polybenzyl
networks (Scheme 7).44 While most reported syntheses of polybenzyls are
polycondensations utilizing benzyl halides and various Lewis acids (AICI3,
TiCl4, and SnCl4), we found that we could obtain a product with a similar

structure by adding concentrated sulfuric acid to neat methoxymethyl benzene.

141

 

w 0 0 0
. / >

n

Scheme 7. Acid catalyzed polymerization of methoxymethylbenzene.

The polymerization yielded a poorly soluble tan solid with a glass transition at
180 °C. When cooled from above that temperature, the polybenzyl product
formed a glassy resin. Thermal gravimetric analysis runs at 10°C/min (Figure
14) indicate that the polymer is stable in nitrogen to 400 °C, but then the benzyl
linkages that connect the aromatic rings degrade and the polymer converts to
volatile products and evaporates. The material is completely decomposed
before it reaches 600 °C.

If we apply similar chemistry to bis(methoxymethyl)benzenes, the
resulting networks should contain a large fraction of fused ring aromatic
compounds that upon elimination of hydrogen would yield materials with a
large proportion of acene-like structures. To test this idea, we prepared 1,2-
bis(methoxymethyl) benzene, 1,3-bis(methoxymethyl benzene), and 1,4-
bis(methoxymethyl)benzene and used sulfuric acid to catalyze their
condensation (Scheme 8). Compared to the condensation of methoxymethyl
benzene, instead of forming two benzyl bridges per ring to give a linear
polymer, having two methoxymethyl groups per ring should result in formation
of four benzyl bridges per ring. These starting materials have the potential to

form polymeric edge-fused ring compounds where alternate rings are aromatic.

142

 

100i

 

sample weight (%)

20—

 

 

 

 

o l l l L

0 200 400 600 800 1000
temperature (° C)

Figure 14. Therrnogravimetric analysis results for
poly(methoxymethylbenzene). Heating rate: 10 °C/min.

CC:
WXJMw—»:eeeeor
3006600

Scheme 8. Acid-catalyzed polymerization of bis(methoxymethyl)benzenes.

 

143

Dehydrogenation of these “polyacene precursors” should yield polyacenes. In
all three cases, we observed rapid formation and precipitation of brown solids.
Characterization of these materials is difficult. Not only are they insoluble, but
also the structural differences between polybenzyl and the polyacene
precursor are subtle and not easily detected by IR, NMR, and other standard
spectroscopic techniques.

Thermal gravimetric analyses of the products showed that the networks

formed from the acid-catalyzed polymerization from each of the methoxymethyl

 

 

 

 

 

 

100 l
90 l-
80 ~
5": 70 .
fit 60 ~
'6
3 50 -
3
g 40 - methoxymethyl
a benzene
‘0 30 F
20 .
10 ) 7
0
0 200 400 600 800 1000
temperature(°C)

Figure 15 Therrnogravimetric analysis results for
poly(methoxybenzene) and poly(bis(methoxymethyl)benzene) when 02
is rigorously excluded. Heating rate: 10 °C/min.

144

benzenes exhibit surprising thermal stability in nitrogen. As shown in Figure
15, after losing a small weight fraction near 400 °C, presumably due to the loss
of a low molecular weight fraction or residual unreacted methoxymethyl
groups, the polymers derived from the di-substituted monomers were thermally

stable to >700 °C and resulted in 70% char yields.

We made an interesting observation from a series of thermal
gravimetric analysis runs where the nitrogen atmosphere was not continuously
purged, and thus was contaminated with a slight amount of oxygen (Figure
16). Under these conditions, the sample prepared form the ortho
bis(methoxymethyl) benzene retained high thermal stability with a 60% char

yield at 700 °C. In contrast, the para and meta derivatives degraded more

rapidly, with the para derivative completely degrading before reaching 500 °C.

These results imply that the three materials are structurally different.
As shown in Scheme 9, there are two limiting structures that can be obtained
from bis(methoxymethyl) benzenes. The desired polyacene precursor (top)
forms when the polymerization is exclusively intramolecular, i.e. both
methoxymethyl groups react to give linear chains. The other extreme is a
hyperbranched polybezyl structure (bottom), highly cross-linked but connected
via thermally labile benzyl bridges. The acene structure is particularly stable
thermally because of is ladder-like architecture. Unlike a single stranded
polybenzyl, two chains must be severed to form volatile products. Based on

the thermal gravimetric analysis runs, the ortho derivative presumably contains

145

 

a higher fraction of thermally stable rings, while the thermal behavior of the

 

 

 

 

 

 

100 7
L
A 80 ortho
if» 60 r
0
3
2
O.
E, 40 3 meta
para
20 r
O I I k=—~1
O 200 400 600 800 ’ 1000
temperature (°C)

Figure 16. Therrnogravimetric analysis results for poly(methoxymethylbenzene)
and po|y(bis(methoxymethyl)benzene run under nitrogen contaminated with
10% oxygen. Heating rate: 10 oC/min.

least stable

146

product, the para derivative, resembles that of a linear polybenzyl. Thus
it appears that ortho substitution favors cyclization, para substitution leads to a

polybenzyl structure, and meta substitution represents an intermediate case.

A plausible argument for explaining the differences in the structures
obtained from the three monomers can be developed from considering steric
factors during the substitution reactions. As shown in Scheme 10, the growing
polyacene chain ends for the ortho, para, and meta methoxymethyl benzenes
are structurally different. The ortho case is unique in that in forming the first
connection to the growing chain, it can add to two sites without having to form
a bond ortho to either a ring juncture or a methoxymethyl group on the chain
end. Both the meta and para cases require at least one such interaction.

We note that in a rigorously inert atmosphere, the ortho, para, and meta
methoxymethyl benzene products have similar thermal stabilities. Linear
polyacene precursors eliminate hydrogen to give acenes, while polybenzyl
networks rearrange and graphitize as commonly seen in phenol-formaldehyde
resins. Elimination of oxygen is critical for limiting oxidation of the benzyl

linkages and formation of volatile hydrocarbon products.

147

inc-at,

ideal case: only linear structure is formed

OCH3
OCH3

 

H300 OCH3 H+
> —-

OCH3

 

intermediate case: acene precursors in a polybenzyl network

J 00
000
00

limiting case: only polybenzyl network is formed

 

06H3

Scheme 9. Possible outcomes from the polymerization of
bis(methoxymethyl)benzenes.

148

\ :
OCH3 .
(Io-H. ,ss: ortho

 

00H3 t
me a
06H3
so
' para

00113 H300 0O 1

Scheme 10. The substitution pattern on the benzene
ring leads to different condensation products.

149

3.1.2. Chemical dehydrogenation

The dihydroanthracene units of polyacene precursors dehydrogenate
thermally, but the high temperatures involved probably cause rearrangement of
the carbon skeleton. An alternative method is to use a chemical agent such as
2,3-dichloro-5,6-dicyano-1,4-benzoquinone (D00). D00 was refluxed with
suspensions of the polyacenic materials in refluxing xylenes for 2 days. The
product was collected by filtration, washed and dried. Neither IR spectroscopy
nor TGA runs showed changes in the materials (Figures 17 and 18), not

surprising given the insolubility of the materials.

3.1.3. Polymerization in solution

The synthesis of a perfectly linear polyacene is difficult due to the lack of
synthetic control over the electrophilic aromatic substitution reaction used to
prepare the polyacenic materials. Instead of a bulk polymerization, which
tends to favor the intermolecular reaction, we briefly considered the use of
different solvents, including ethanol, ethyl ether, and ethyl acetate. 10 wt% of
the 1,2-bis(methoxymethyl)benzene was added to a 1:1 mixture of sulfuric acid
and the solvent. The usual brown precipitate formed, was washed and dried
under vacuum. The char yields from these polymerizations (Figure 19) were
less than the yields from bulk polymerizations. It is likely that the solvent

reacted under these conditions.

150

 

sample weight (%)

100

 

90

  
  
 

80 Without BBQ

70 " with 000 <—

60—

40r

30-

10L

 

 

 

O l l l l
O 200 400 600 800 1 000

temperature (° C)

Figure 17. Thennogravimetric analysis results for poly (bis 1,2-
(methoxymethyl)benzene) in nitrogen. Heating rate: 10 °C/min.

151

sample weight (%)

100

 

90

80

  
   

70 with BBC treatment

60~

without DDQ treatment

50~

40~

30~

20L

10~

 

 

 

O 200 400 600 800 1000

temperature (°C)

Figure 18. Therrnogravimetric analysis results for poly(bis—1,4-
(methoxymethyl)benzene) in nitrogen. Heating rate: 10 °C/min.

152

sample weight (%)

 

   

 

 

 

 

100
90
30 inbulk
70 - .
inethylacetate
m D
ine ether
50 - thyl
40 l-
inethanol
30 l-
20 -
10 -
O l l 4 l
0 200 400 600 800 1000
temperature (°C)

Figure 19. Therrnogravimetric analysis results for poly(1,2-bis
(methoxymethyl)benzene) prepared via solution and bulk reactions.
Heating rate: 10 °C/min.

153

 

 

3.1.4. Characterization of acene-like materials

3.1 .4.1. Infrared

To better understand the structural changes that occur in the polymer
derived from the ortho substituted monomer, we heated the products under
nitrogen to various temperature in a TGA apparatus and the residues were
analyze using infra red spectroscopy. The samples were made as KBr pellets,
and for meaningful comparisons, the spectra were normalized for sample
weight (mg sample/100 mg KBr) and the thickness of the pellet (related to the
absolute concentration of sample in the pellet). An internal standard added to
the sample (CBr4) and the residual water in the KBr also were used to
normalize the IR spectra. As shown in Figure 20, the IR spectra clearly show
progressive chemical structure evolution from that of the “as-obtained” polymer
with a high proportion of C-H bonds to an intermediate structure with
predominantly aromatic C-H bonds. Assuming the ortho precursor contains a
large fraction of edge-fused rings, the sample treated at 500 °C should contain
an appreciable fraction of polyacenes.

A signature of polyacenes would be the appearance of materials with a
relatively low band gap, but with the retention of aromatic C-H bonds. From
the IR spectra, we find that the 500 and 650° samples have a strong electronic
absorption that tails into the infrared region, indicating extensive n-conjugation.
In addition, the IR bands corresponding to C-H stretching are retained. The
bands due to aliphatic C-H bands (2900-3000 cm") decrease in intensity while

those due to aromatic C-H bonds increase, as expected for the formation of

154

acenes. Further treatment of the samples at 800 °C resulted in the near
complete loss of C-H bonds and presumably the formation of a graphitic
structure. The broad absorption at 3500 cm'1 was due to residual moisture in

the KBr.

3.1.4.2. Raman

We also studied the structure of these acene-like materials using
Raman spectroscopy. The polymer samples characterized were derived from
the acid-catalyzed condensation of the ortho, meta, and para
methoxymethylbenzenes that had been heat treated at 800 °C in an oxygen-
free environment. The Raman spectra (Figure 21) are very similar to those of
graphitic materials such as activated charcoal and carbon black.102 Two
characteristic peaks were observed for all samples, a peak at 1355 cm'1
assigned to an A19 vibration mode of amorphous carbon, and a peak at 1600
cm'1 from the E29 mode of graphite. Consistent with thermal gravimetric
analyses of the precursors, the ortho derivative had the strongest 1600 cm“
band and thus the most ordered graphite structure. The same peak in the
spectrum of the meta and para derivatives is weaker, consistent with their

precursors being more polybenzyl-like compared to the ortho sample.

3.1.4.3. Elemental Analysis

Elemental analyses of the condensation products of the
methoxymethylbenzenes provide only limited structural information since the
polybenzyls and polyacenes have identical chemical formulae. Elemental

analysis can however provide some information about the purity of the

155

 

materials. The data obtained show that the samples are impure. Analysis for
C, H, and S gave 76% C, 5.4% H and 2.2% S; the remainder is presumably
oxygen. This gives a C/H ratio of 14 instead of the 16 that would be expected
if the polymers were linear acene precursors (C8H5 for each repeat unit). As
expected, heat treatment increased the C/H ratio (to 71), and the elemental
analysis results were 83% C, 1.2%H and 1.1% S. The presence of sulfur is a
problem as it may substantially affect the properties of the polyacene product.
How it is incorporated into the materials is unknown, but we suspect
sulfonation of the polymer or occlusion of sulfates in the insoluble polymer. The
lower than expected C/H ratio for the acene precursors is likely linked to
incomplete condensation of the methoxymethyl groups, while the higher than
expected C/H ratio after heat treatment is consistent with substantial

graphitization.

156

 

relative absorbance

 

800

650

500

380

w

 

 

 

lit.
M M

1 1 1 1

 

4500 3500 2500 1 500 500

wavenumbers (cm")

Figure 20. Infrared absorption spectra of the polybenzyl
network formed from monomer 1 heated in nitrogen at 10
°C/min to 30 °C, 380 °C, 500 °C, 650 °C, and 800 °C.

157

 

 

 

 

 

600
500 ~
3‘ 400 -
'5
5 300 ’ pomox 900
.‘é
200 pprnox 900
100 ~
prrmox900
0 MM
500 1000 1500 2000
wavenumber(cm-1)

Figure 21. Raman spectra of ortho,para and meta-
(methoxymethyl) benzene after heated to 800 °C.

158

3.2. Well-defined Soluble Polyacene

As seen in the previous sections, the poor solubility of polyacenes
stymies characterization of the polymers. A new synthetic approach to soluble
polyacenes is needed for there to be substantial progress toward their
characterization. To improve the solubility of polyacene, we adopted a well-
known strategy for solubilizing rigid-rod polymers like polyphenylenes
(PPPs),‘°3“°6 attaching flexible alkyl chains to the polymer backbone. Soluble
PPP derivatives were obtained by adding n-hexyl chains to the polymer repeat
units. As shown in Scheme 11, alkyl chains attached to the polyacene

backbone will yield ”hairy rods” that like PPPs, should be soluble.

3.2.1 Monomer design and synthesis
To obtain a substituted polyacene, a monomer must be designed that

contains two methoxymethyl groups, unsubstituted sites capable of

n

Scheme 11. Alkyl substituted polyacene.

159

electrophilic aromatic substitution, and alkyl groups to solubilize the polymer.
Scheme 12 shows the route chosen for the synthesis. The free-radical
bromination of 2,5-dibromo-p-xylene using N-bromosuccinimide led to 1,4-
bis(bromomethyl)-2,5-dibromobenzene (4) in 30% yield. The poor yield is due
to the difficulty of separating 4 from impurities. Two equivalents of N-
bromosuccinimide were used in the reaction, and the bromination yielded a
statistical mixture of mono-, di-, and unbrominated products. Due to the
symmetry of 4, it could be recovered from the reaction mixture by repeated
crystallization. Compound 4 was treated with 2 equivalents of sodium
methoxide to give bis(methoxymethyl)-2,5-dibromobenzene 5. When run in
THF, this reaction forms a substantial amount of low molecular weight
poly(phenylene vinylene). This side reaction was avoided by running the
reaction in non-polar, non-coordinating solvents such as hexanes. Introduction
of the alkyl group by a Pd-catalyzed coupling reaction via the Grignard reagent
was not very successful. Clean disubstitution in high yield is needed, but TLC
revealed at least 5 products with the disubstituted product present in low yield.
While the coupling reaction works well on substrates like dibromobenzenes,
the larger steric hindrance of our substrate presumably hinders the coupling
reaction. In an alternative Pd-catalyzed coupling scheme (Scheme 13), we
replaced the Grignard reagent with the 9BBN derivatives of 1-octene and 1-

7

decene.10 This reaction was successful, and we obtained a 56% of the

coupled product.

160

 

 

Bl’ NBS, C32
CH3 uv light
H3C
Br
R
OCH3
H3CO
R PdCl2(dPPF)2
+ <
THF
Br
RMgBr, R=CBH17
OCH3
H3CO

+ other products

Br

CH28T
BrCH2
Br
CH30H, NaH
THF
Br
OCH3
H3CO
Br

Scheme 12. First approach to the synthesis of 1,4-bis—(methoxymethyl)-2,5—dialkyl

benzenes

161

Br NBS, 082 Br

 

 

CH3 UV light CHzBr
H3C BrCH2
Br Br
CH3OH, NaH

THF

R Br
OCH3 A PdCI2(dPPF)2 OCH3
H300 ‘ THF H300
R RBBN, R=CBH17. CtoHZI Br
9 BEN + W —_’ C8H1yBBN +

Scheme 13. Modified synthesis of 1,4-bis-(methoxymethyl)-2,5-dialkyl benzenes

162

3.2.2 Polymerization condition development

A model reaction was used to develop appropriate conditions for the
condensation reactions of the monomer. Octylated
bis(methoxymethyl)benzene was treated with different catalysts in an excess of
benzene (Scheme 14). The reactions were assayed for the yield of 2,5-
dibenzyl-1,4-dioctylbenzene and its purity. The reaction with concentrated
H2804 formed insoluble products. BFa-etherate did not catalyze the desired
reaction, but instead 1H NMR indicated formation of an ethylated aromatic
product. This was confirmed when the reaction of BFa-etherate and benzene
alone yielded hexaethyl benzene (Scheme 15). However, a combination of
BF3 and phosphoric acid resulted in clean formation of the dibenzylated

product in 85% yield. When applied to bis(methoxymethyl)benzene, 1H NMR

3.9 ppm
08H” CaH17 /
@330 i 0 0 0

H3C‘O
C8H17 excess T C8H17
3.9 ppm

Scheme 14. Control experiment used to develop polymerization conditions

 

© BF3 etherate

Scheme 15. Control experiment to prove BF3 caused side reactions.

163

showed the loss of the methoxymethyl group and the appearance of the benzyl
linkages. The product was soluble, and based on GPC measurements, had a

molecular weight of 3000 g/mol.

3.2.3. Model cyclization study

To optimize the cyclization conditions, we first synthesized the model
compound shown in Scheme 16. The ring closing reaction of the benzyl
substituted methoxymethyl benzene (Scheme 17) should provide a guide to
the conditions that would best yield polyacenes. The model compound

resembles the product of the first condensation step, which then can react

©+E>§ofl+$

Wolff-Kishner
NH2NH2/KOH
ethylene glycol
O O LiAlH4
EtZO
OOH
OH
CH3I, NaH

OCl-l3

Scheme 16. Synthesis of a model compound for the cyclization study

164

 

 

intramolecularly to give the poyacene precursor, or interrnolecularly to give
hyperbranched polybenzyl products (Scheme 19). We tested a series of
reaction conditions such as monomer concentration, monomer to catalyst ratio,
temperature, reaction time, etc. to optimize the conditions that favor the
intramolecular reaction. In each trial, we always observed some insoluble
material, which must correspond to the branched polybenzyl product. After
removing the insoluble fraction, we dried the soluble fraction, and determined
the crude yield of 9,10-dihydroanthracene, the expected product. The crude
product was analyzed by HPLC and the yield was corrected for purity. The
results are tabulated in Table 2-8. From these series of experiments, we
adopted standard conditions (monomer concentration = 2mM,
catalyst/monomer ratio = 33, 12 hr reaction time

in CH2CI2 at room temperature) for all polymerization reactions. When applied

to the polymerization of dioctyl bis(methoxymethyl)benzene, we obtained an

0.0

intram olecular cyclization

(soluble)

interm olecular bridge to rrnation

(soluble + insoluble)

Scheme 17. Model cyclization condition study

165

85% yield of soluble product with a molecular weight of 7000. This

corresponds to an average of 22 monomer units per chain.

With the presumed acene precursor in hand, we examined conditions
for dehydrogenation by Pd/C‘oa'109 and DDQW'111 to form the acene structure.
An advantage of Pd/C is that it can easily be removed from reaction mixtures
via filtration, however we were concerned that the polyacene product might
strongly bind to the catalyst. In addition, the mechanism of the hydrogenation
reaction requires that the substrate come in contact with the catalyst, a step
that might be hindered by the alkyl chains. DDQ, being soluble, should not
have these problems. Nevertheless, one disadvantage of DDQ is that it may

react with the acene and form a Diels-Alder adduct.

We carried out model dehydrogenations as shown in Scheme 18 and
found that the DOC) reaction worked well (no detected side reactions) as long
as the temperature was held to no higher than the boiling point of toluene, and
the reaction time was held to 12 hours. When this scheme was applied to the
acene precursor polymer (Scheme 20), we did not see the characteristic
absorption in the visible that one would expect from the n-conjugated system of
a polyacene. This implies that the length of the acene segments must be
small, and that the intermolecular condensation reaction dominated the desired

lntramolecular reaction needed to form rings.

0.0 COO
———-—>
12hr

Scheme 18. Model dehydrogenation test

166

30°va 0.30.0683 Eco. 30 m_ 0.30.0QE03 8:000. .0303 05. ..

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

men 0.3. Ron Nmm ”F «-3 ..NN A058 _E $8035.: 93
. . . . . . .8060
m E v 5 m m mm N on ._. no? ..N N EEEovSmE ohm
.00 .5258
as. 20> 3... So: es. _e_o_> e. ..woos. 3.5.5 53580
__m.0>0 E2. >250 003.0 :95 030m 204. 5303:0800
>u3m cosz=o>o .08... E 203 .w> 0.30.0203 cozom0m .n 030...
.2 Mb? ..NN w_ coach—.0250 3009 05....

08 s a- 2 .. 2 _EEESEEE 3K
5mm 0.5 Emu 0mm H_. v 0.. .. we _EtootmE who
mém v.5 mdm Nam ”F cor ..NN _E {_EovtmE ohm
v.50 o .8 ohm Nov ”F m-o_. Ld 3. {_EonE osm

.00 SEER.
as. o_o_> €00 So: as. 20> o. woos. 3:95
__m.0>O Eo: 3:30 003.0 eoEv ozm. 29o. 5303:0080 cozficoo

 

 

 

 

 

 

 

.630 53056.6 000:. E 203 .m> 5303:0980 .N 030...

167

:090 :_ 2 533300 :03000. .0300 0:3 ..

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. . . . . . ._0\.:N EoomN
m 5 v _.m 3 mm N mm .w wow ..N N _EEEovtmEmo
. . . . . . . .3.: NtoomN
N mm 3 Nm 0 3 0 mm .M ".3 ..v N _ESEow 23E on
_00 3o
32.. 22> 22.. 03a... 3... 22> 2 five: 58 3.5.5 5.3.250
__0.0>O E03 3333 003.0 coEv 030. 0_0< :0303:00:00 . .
>030 :030N__0>0 _000E :_ 0_0_> .0> 093a :09< .m 030k
.3. NF 0_ 0E3 :03000. .0300 0E. ..
. . . . . . ._0\.33N\0omN
. . . . . . ._0\.:N toemN
m 5 v E 0 mm 0 mm .m so? ..N 0 52503558
.00 :30 E00
$0. 22> 303 03a: 3... 22> 2 fine: 3:95 5.3.250
__0.0>O E03 3330 003.0 coEv 030. 0_0< :0303:00:00 . .

 

 

 

 

 

 

 

>030 :030~__0>0 _000E :_ 20$ .0> 0E3 :03000m .v 030...

168

 

052:. 2 2 2cm 23 2cm 2 2 5228 .253 ofi ..

 

0525 8 2 22... 0228

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Now 0.00 0.00 mdm ”F cor . mmN .33... NtoomN
52:53.05 2
SEE or :_ 000 050000
Ndm 3N0 0.5 mdm ”P n- or .. omN .33.: NtoemN
5:53:05 2
a
$05 22> 35 03a: 05 22> 252mg“. 58 3.5.5 52950
__0.0>O E03 >3_.3n_ 003.0 coEv 030. 0_0< :0303:00:00
>030 :030~__0>0 _000E :_ 0_0_> .0> 000 03 35000 30 030. 05. K 030...
0mm; H0_ 000 03 0:303E00 _000E 30 030. 3000 03.. ..
. . . . . .<\.:N 2.3..
0 mm m mm 0 oo oo: cor .. 00 N >Em>Eo¢EmEooé>
. . . . . . .3.: NE .3..
3 mm 3 Nm 0 mm mm m: 03 .. mm N _EmN.o\_EovtmE©.o>
. . . . . . .3.: N20 N
N mm P Na 0 3 m mm .3 wow .0 N _E_\_Eov:m:e. on
_00 :300E00
05 22> 35 03a: 05 22> 2 moves. 3.5.5 82250
__0.0>0 E03 3333 003.0 :0303:00:00 . .

 

 

 

:95 2.0. 2o<

 

 

 

 

>030 :030N__0>0 _000E :_ 0_0_> .0> 030. 0_0< .0 030...

169

 

 

 

0:0..30E0.0_.._0_0 0_ E0200 .0300 03... ..

 

 

.__o 525 2 5.595 .0525 E 2 2o0 0280
>0on 2 2:8. 2.. .3 .2058. 00: 0.00 2 a- or .. 00.0 2...; 020000
25858 502.. 05 =0 0320 «=22 _Eiaozfov _Eoezme E
052:. E 2 28 02260
0.00 :00 0.00 0.00 3 a- 3 .. 00.0 .3... 020000

5:204:05 _Eovzme E

 

 

_00 :300E00
$052.0; 20503:: 352.3; 2 ..wuos. 3295 52250
__0.0>O E03 3330 003.0 coEv 030. 0.0.4 :030 3:00:00 . .

 

 

 

 

 

 

 

>030 :030N__0>0 _000E :_ 0_0_> .0> E0200 3:00.35 .m 0.00...

0
7

M.\AW

 

 

VWVMW

 

Scheme 19. lntramolecular cyclization vs. intermolecular bridge
formation

171

 

n R (+3 n
A
o/ BF3 / H3PO4 CH2 CH2
/
R n

 

—’ i
OOCOOQ.

Scheme 20. Synthesis of alkylated polyacenes.

172

3.2.4. Characterization

Since the attachment of the alkyl chains improved the solubility of the
acene precursor, we were able to better characterize the acene precursor
polymer and the dehydrogenated product. UV and fluorescence spectroscopy
(Figure 22) are the most direct techniques that can be used to assay the
conjugation length of the polyacene product. When the soluble polymer was
excited at 420 and 570 nm, the lowest energy emissions detected were near
480 and 610 nm, the longest wavelength absorption bands of tetracene and
pentacene, respectively. Thus, the longest acene segments in the polymers
contained 4 to and 5 fused rings. The bulk of the emission envelope appeared

at shorter wavelengths, and we did not try to quantify the distribution of chain

 

 

 

 

lengths.
. Ex=420nm

g EX=570 I:
'5 .2

0
3 2
9 E
< m

400 500 600 700 h ‘ 800

 

wavelength (nm)

Figure 22. UV and fluorescence spectra of soluble polyacene products.

173

3.2.5. Discussion

Why were the lengths of the polyacene segments so short? We relied
on the optimized reaction conditions of model compounds to favor cyclization
over polybenzylation, but the data imply that the polymers had a higher degree
of polybenzyl content than would be predicted from the model study. In the
polymer system, the competition between intra and intermolecular cyclization
apparently is biased toward the latter. It is possible that a different set of
reaction conditions, such as a more dilute monomer concentration, would be
more appropriate for the polymerization. There also may be undetected side
reactions that are more prevalent in the polymerization. It is important to keep
in mind that the polymerizations are probably best described as step growth
systems, where monomers form dimers and dimers need to find each other to
extend the chain. High concentrations favor more rapid growth, and possibly

polybenzyl formation.

To solve this dilemma, the condensation and cyclization steps must be
separated, i.e. a linear polymer must first be synthesized at high monomer
concentrations, and after purification, the polymer can then be cyclized in dilute
solution to give the acene precursor. Based on this idea, we propose the
route shown in Scheme 21. By preparing a dialkylated monomer with the
methoxymethyl and a second functional group, labeled X, the acid catalyzed
reaction can be initiated to prepared a linear polybenzyl. After purification of

the polymer, the second functional group can be accessed to close the ring.

174

3.3. Acenelpoly (p-phenylene) hybrids.

The lack of control over the electrophilic aromatic substitution reactions
used to prepare polyacenic materials resulted in our inability to prepare
perfectly linear polyacenes. An approach that should yield higher selectivity is
to align the functional groups in a pro-ordered state. Designing monomers
capable of organizing into an ordered phase may increase the probability of
intramolecular cyclization. If a polyphenylene containing the reactive
methoxymethyl functionalities is synthesized, then it should exhibit the same
type of chemistry as bis(methoxymethyl)benzenes in the presence of strong
acids and yield a network structure where the polymer backbones are cross-
linked by benzyl bridges. Furthermore, if the polymer backbone has limited
bond rotation, then the most probable cross-linking reaction is the formation of
cyclized products. The over all structure of the product would resemble a

sheet of graphite and should have high themal stability.

To realize this idea, we synthesized 1,4-dibromo-2,5-
bis(methoxymethyl)benzene from readily available starting materials as shown in
Scheme 12. Methoxymethyl substituted poly(p-phenylene) was synthesized
from the dibromobenzene monomer using a nickel catalyzed coupling reaction
(Scheme 22). The resulting polymer was insoluble in DMF and precipitated from
solution during the course of the polymerization. The polymer was collected by
filtration and washed repeatedly with water and ethanol to remove any
polymerization reagents. After drying, the polymer was obtained as a finely

powdered, pale yellow solid. GPC analysis of the THF soluble fraction (5 wt%)

175

indicated a degree of polymerization less than 10 (relative to polystyrene
standards).

The polymer was reacted with sulfuric acid following the same procedure
as the polymerization of bis(methoxymethyl)benzenes. Since the carbon atoms
in PPP are sp2 hybridized, bond motion is limited to rotation about the central
axis and the geometry of polyphenylenes is rigid rod; therefore the PPP chain
can easily packed together. Once the PPP rods align with each other,
intramolecular cyclization should be favored, leading to the formation of acene

precursors (Scheme 23).

 

/ /
O O
NiClz, Zn
Br 0 Br PPh3,blpy= O
DMF, 80°C n
O O
/ /
PPPMOX

Scheme 22. Synthesis of a methoxymethyl-substituted polyphenylene

Figure 23 shows the thermal stabilities of the polymer and that of the
network resulting from the acid catalyzed cross-linking. The polymer is thermally
stable to 350 °C at which point the side chains are evolved as volatile products
leaving behind the polyphenylene backbone, which is stable to 700 °C. The
weight loss observed in the TGA scans of the cross-linked network at lower
temperatures is most likely due to the evolution of methanol resulting from the

reaction of residual acid and methoxymethyl groups. Even though the product

176

 

from the cross-linking reaction was neutralized with base and rinsed repeatedly
with water, some sulfuric acid did remain in the tacky solid. Despite the early
weight loss, the network exhibits excellent thermal stability, producing a 70%
char yield at 700 °C. There are two possible products from the heat
dehydrogenation: one is the idealized 20 polyacene structure, the other is the

more likely formed graphitized product. Both exhibit high thermal stability.

 

 

 

 

100 ~
PPPMOX
after xlink
80 -
°\0 60 -
E PPPMOX
4O ’ before xlink
2O -
o 1 l
0 200 400 600 800

temperature (°C)

Figure 23. Therrnogravimetric analysis of 2D polyacene

177

  
   

PPPMOX
PPPMOX after Xlink

 

/ ..
Idealized 2D stmctures more likely graphitized structures

Scheme 23. Synthesis of sheet structures from methoxymethyl-
substituted polyphenylenes.

These materials were also analyzed by IR spectroscopy, shown in
Figure 24. The most noticeable evidence of the structural changes brought on
by the acid catalyzed cross-linking of the polymer can be seen by the
decreases in the C-H and C-0 stretching intensities at 2900-3000 and 1100
cm", respectively, due to the loss of the methoxy groups. As seen in the
pyrolysis of the networks formed from bis(methoxymethyl)benzenes, high
temperature treatment of the polyphenylene-derived network at 700 °C also
results in the near complete loss of C-H bonds and the tailing of the electronic

absorption into the IR.

178

Raman spectrum also indicate a graphite-like material was formed after
high temperature treatment (800 °C) of cross-linked PPPMOX (Figure 25).
Among the two peaks observed, the one at 1355 cm’1 belong to the A19
vibration mode, attributed to the amorphous portion of graphite; the peak at
1600 crn'1 is from the E29 vibration mode, which is assigned to the lattice
graphite. All the results from Raman, IR and thermal analysis indicate that
thermal treatment of the cross-linked network structure also leads to materials

with extensive n—conjugation and low H content.

 

30

    
 

PPNDX after xlink heated to 900°C
25 .

20*

PPPMOX after xlink

.....J‘k -

  

15—

 

10-

relative absorption

PFPMVDX before xlink

 

 

 

 

3000 2400 1 800 1 200 600

wave number (cm")

Figure 24. Infrared of cross-linked and heat treated samples of a
methoxymethyl-substituted polyphenylene.

179

3.4. Conclusion

The high temperatures needed for the pyrolysis of phenolic and
cellulose resins allows little control of the acene-like structural elements that
are the precursors to graphite. If elements of the acene structure are designed
into the starting material, it is possible to obtain partially graphitized structures
at much lower temperatures that still retain some memory of the chemical
structure before pyrolysis. Two types of polyacene syntheses based on a
Lewis acid catalyzed electrophilic aromatic substitution reaction were

investigated. The first targeted simple routes to acene-like materials that

 

60

 

   

 

 

500 1000 1500 2000 2500

wave number (cm")

Figure 25. Raman spectrum of a cross-linked and heat treated sample of a
methoxymethyl-substituted polyphenylene.

180

contain a substantial fraction of acene subunits. We found that
bis(methoxymethyl)benzenes undergo acid-catalyzed polymerization to yield
network structures. The idealized chemical structure for the polymerization of
1,2-bis(methoxymethyl)benzene is a linear polyacene precursor that after
elimination of H2, yields polyacenic materials with an extended electron
system. The second route targeted the synthesis of a methoxymethyl
substituted poly(p-phenyl) that when chemically cross-linked, ideally formed a
two-dimensional network closely resembling a graphite sheet. The IR and
thermal analysis measurements indicate that much lower temperatures are
required to pyrolyze these materials into partially graphitized structures. A
synthetic route to a novel, soluble linear polyacene based on the same acid
catalyzed electrophilic aromatic substitution reaction was also investigated.
We obtained soluble polymer with molecular weights as high as 7000.
However, fluorescence data show that the maximum length of the acene

segments is ~5 units.

181

BIBLIOGRAPHY

#ODN

10

11

12

13

14

15

16

17
18

Kertesz, M. Macromolecules 1995, 28, 1475-1480.
Darosa, A. L. S.; Demelo, C. P. Phys. Rev. 81988, 38, 5430-5437.
Mizes, H.; Conwell, E. Phys. Rev. 81991, 44, 3963-3969.

Su, W. P.; Schrieffer, J. R.; Heeger, A. J. Phys. Rev. Lett. 1979, 42,
1698-1701.

Su, W. P.; Schrieffer, J. R.; Heeger, A. J. Phys. Rev. B 1980, 22, 2099-
2111.

Wilboum, K.; Murray, R. W. J. Phys. Chem. 1988, 92, 3642-3648.

Pietrass, P. Proceedings of the 13th Annual International Symposium on
Electronic Structure of Metals and Alloys; Johnsbach: Dresden, 1983.

Boon, M. R. Theoret. Chim. Acta 1971, 23, 109-110.

Whangbo, M. H.; Hoffmann, R.; Woodward, R. B. Proc. R. Soc. London
Ser. A-Math. Phys. Eng. Sci. 1979, 366, 23-46.

Cioslowski, J. J. Chem. Phys. 1993, 98, 473-477.

Tanaka, K.; theki, K.; Nankai, S.; Yamabe, T.; Shirakawa, H. J. Phys.
Chem. Solids 1983, 44, 1069-1075.

Kertesz, M.; Lee, Y. S.; Stewart, J. J. P. Int. J. Quantum Chem. 1989,
35, 305-313.

Bredas, J. L. Handbook of Conducting Polymers, Dekker, New York:,
1986; Vol. 2.

Pomerantz, M.; Cardona, R.; Rooney, P. Macromolecules 1989, 22,
304-308.

Dewar, M. J. S.; Dougherty, R. C. The PMO Theory of Organic
Chemistry, Plenum: New, York, 1975.

Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry Part A; 2nd
ed.; Plenum: New York, 1984.

Wang, R. S.; Hou, W. L. Synth. Met. 1991, 42, 1637-1640.

Albery, W. J.; Doblhofer, K.; Lyons, M. E. G.; Mount, A. R.; Vorotyntsev,
M. Electroactive Polymer Electrochemistry, Plenum: New York, 1994.

182

19

20

21

22

23
24

25

26

27

28

29

30
31
32
33

34
35

36
37

Feast, W. J. Handbook of Conducting Polymer; Marcel Dekker:, 1986;
Vol. 1.

lyer, V. S.; Wehmeier, M.; Brand, J. D.; Keegstra, M. A.; Mullen, K.
Angew. Chem-Int. Edit. Engl. 1997, 36, 1604-1607.

Trick, K. A.; Saliba, T. E. Carbon 1995, 33, 1509-1515.

Dang, T. D.; Wang, C. 8.; Click, W. E.; Chuah, H. H.; Tsai, T. T.;
Husband, D. M.; Arnold, F. E. Polymer1997, 38, 621-629.

Dhami, T. L.; Bahl, O. P.; Jain, P. K. Carbon 1995, 33, 1517-1524.

Tanaka, K.; theki, K.; Yamabe, T.; Yata, S. Synth. Met. 1984, 9, 41-
52.

Yamabe, T.; Tanaka, K.; theki, K.; Yata, S. J. Phys. (Paris) Col/eq.,
Suppl. 1983, 44. 03-645.

Yata, S.; Hato, Y.; Osaki, T.; Sakurai, K. In US Patent No. 4 753 717
1988.

Yata, S.; Hato, Y.; Sakurai, K.; Satake, H.; Mukai, K.; Tanaka, K.;
Yamabe, T. Synth. Met. 1990, 38, 169-175.

Yata, S.; Osaki, T.; Hato, Y.; Takehara, N.; Kinoshita, H.; Tanaka, K.;
Yamabe, T. Synth. Met. 1990, 38, 177-184.

Kojima, Y.; Matsuoka, T.; Sato, N.; Takahashi, H. Macromolecules
1995, 28, 2893-2896.

Giuliano, C. R.; Hess, L. R. IEEE J. Quantum Electron 1979, 15, D71.
Coulter, D. R.; Perry, J. W.; Wei, T. H. Proc. SPIE1989, 1105, 42.
Perry, J. W.; Khundkar, L. R. NATO ASI Ser. E1991, 194, 369.

Bunning, T. J.; Natarajan, L. V.; Schmitt, M. G.; Epling, B. L.; Crane, R.
L. Appl. Optics 1991, 30, 4341-4349.

Tutt, L. W.; Kost, A. Nature 1992, 356, 225-226.

Wood, G. L.; Clark, W. W., Ill; Miller, M. J.; Salamo, G. J.; Sharp, E.
Proc. SPIE1989, 1105, 154.

Tutt, L. W.; McCahon, S. W. Opt. Lett. 1990, 15, 700-702.
Mansour, K.; Soileau, M. J.; Vanstryland, E. W. J. Opt. Soc. Am. B-Opt.
Phys. 1992, 9, 1100-1109.

183

38

39

40

41

42

43
44

45
46

47

48

49

50

51

52

53

54

Branelik, D. M.; McLean, D. G.; Schmitt, M. G.; Epling, B. L.; Colclasure,
C.; Tondiglia, V.; Pachter, R.; Oberrneirer, K.; Crane, R. L. In Electrical
Optical and Magnetic Properties of Organic Solid State Materials;
Materials Research Society: Pittsburgh, 1992; Vol. 247, p 361.

Pochan, J. M.; Pochan, D. F.; Rommelmann, H.; Gibson, H. W.
Macromolecules 1981, 14, 110-114.

Shacklette, L. W.; Chance, R. R.; Ivory, D. M.; Miller, G. G.; Baughman,
R. H. Synth. Met. 1980, 1, 307-320.

Yata, S.; Hato, Y.; Sakurai, K.; Osaki, T.; Tanaka, K.; Yamabe, T. Synth.
Met. 1987, 18, 645-648.

Jobst, K.; Sawtschenko, L.; Wuckel, L.; Doge, H. G.; Schwarzenberg, M.
In DD Patent 240 285 1986.

Fauteux, D.; Koksbang, R. J. Appl. Electrochem. 1993, 23, 1-10.

Yata, S.; Kinoshita, H.; Komori, M.; Ando, N.; Kashiwamura, T.; Harada,
T.; Tanaka, K.; Yamabe, T. Synth. Met. 1994, 62, 153-158.

Zheng, T.; Xue, J. S.; Dahn, J. R. Chem. Mat. 1996, 8, 389-393.

Tatsumi, K.; Akai, T.; lmamura, T.; Zaghib, K.; lwashita, N.; Higuchi, S.;
Sawada, Y. J. Electrochem. Soc. 1996, 143, 1923-1930.

Tanaka, K.; Koike, T.; Yamabe, T.; Yamauchi, J.; Deguchi, Y.; Yata, S.
Phys. Rev. B 1987, 35, 8368-8373.

Tanaka, K.; Ueda, M.; Koike, T.; Yamabe, T.; Yata, S. Synth. Met. 1988,
25, 265-275.

Yazami, R.; Nakajima, T. Synth. Met. 1991, 40, 387-392.

Mohri, M.; Yanagisawa, N.; Tajima, Y.; Tanaka, H.; Mitate, T.; Nakajima,
S.; Yoshida, M.; Yoshimoto, Y.; Suzuki, T.; Wada, H. J. Power Sources
1989, 26, 545-551.

Yata, S. In US Patent No. 4 601 849788 1986.

Tanaka, K.; Koike, T.; Yamabe, T.; Yata, S. Synth. Met. 1989, 29, E489-
E494.

Sato, K.; Noguchi, M.; Demachi, A.; Oki, N.; Endo, M. Science 1994,
264, 556-558.

lshikawa, M.; Sonobe, N. Synth. Met. 1991, 40, 231.

184

55

56

57

58

59

60

61
62
63

64
65
66
67
68

69
70

71

72

73

Zheng, T.; McKinnon, W. R.; Dahn, J. R. J. Electrochem. Soc. 1996,
143, 2137-2145.

Tokumitsu, K.; Mabuchi, A.; Fujimoto, H.; Kasuh, T. J. Electrochem.
Soc. 1996, 143, 2235-2239.

Selwyn, L. S.; McKinnon, W. R.; Vonsacken, U.; Jones, C. A. Solid
State Ion. 1987, 22, 337-344.

Nogami, T.; Mikawa, H.; Nawa, M. J. J. Phys. Colloq. (France) 1983,
C3, 573.

Yata, S.; Sakurai, K.; Osaki, T.; Inoue, Y.; Yamaguchi, K.; Tanaka, K.;
Yamabe, T. Synth. Met. 1990, 38, 185-193.

Yata, S.; Okamoto, E.; Satake, H.; Kubota, H.; Fujii, M.; Taguchi, T.;
Kinoshita, H. J. Power Sources 1996, 60, 207-212.

Sidman, J. W. J. Chem. Phys. 1956, 25, 122.
Kause, N. J.; Small, G. J. J. Chem. Phys. 1972, 56, 2985.

Clar, E. Polycyclic Hydrocarbons, Academic Press: London, 1964; Vol.
1,2.

Marschalk, C. Bull. Soc. Chim. Fr. 1938, 5, 306.

Marschalk, C. Bull. Soc. Chim. Fr. 1941, 8, 354.

Marschalk, C. Bull. Soc. Chim. Fr. 1950, 311.

Notario, R.; Abboud, J. L. M. J. Phys. Chem. A 1998, 102, 5290-5297.

Horn, T.; Baumgarten, M.; Gerghel, L.; Enkelmann, V.; Mullen, K.
Tetrahedron Lett. 1993, 34, 5889-5892.

Graham, R. J.; Paquette, L. A. J. Org. Chem. 1995, 60, 5770-5777.
Suh, M. C.; Shim, S. C. Macromolecules 1995, 28, 8707-8712.

Takahashi, T.; Kitamura, M.; Shen, B. J.; Nakajima, K. J. Am. Chem.
Soc. 2000, 122, 12876-12877.

Molinski, T. F.; Fahy, E.; Faulkner, D. J.; Van Duyne, G. D.; Clardy, J. J.
Org. Chem. 1988, 53, 1341-1343.

Mathias, J. P.; Stoddart, J. F. Angew. Chem. Int. Ed. Engl. 1987, 26,
805.

185

 

74

75

76

77
78
79
80
81

82

83
84

85

86

87

88
89

90

91

92

Alder, R. W.; Allen, P. R.; Edwards, L. S.; Fray, G. l.; Fuller, K. E.; Gore,
P. M.; Hext, N. M.; Perry, M. H.; Thomas, A. R.; Turner, K. S. J. Chem.
Soc-Perkin Trans. 1 1994, 3071-3077.

Herwig, P. T.; Mullen, K. Adv. Mater. 1999, 11, 480.

Girreser, U.; Giuffrida, D.; Kohnke, F. H.; Mathias, J. P.; Philp, D.;
Stoddart, J. F. Pure Appl. Chem. 1993, 65, 119-125.

Ozaki, M.; lkeda, Y.; Nagoya, I. Synth. Met. 1987, 18, 485-490.
Kim, D. S.; Suh, M. C.; Shim, S. C. Synth. Met. 1996, 80, 291-296.
Suh, M. C.; Shim, S. C. Chem. Mat. 1997, 9, 192-200.

Kertesz, M.; Ashertehrani, A. Macromolecules 1996, 29, 940-945.

Allison, G.; Bushby, R. J.; Jesudason, M. V.; Pailaud, J. L.; Taylor, N. J.
J. Chem. Soc. Perkin Trans. 1997, 2, 147.

Vogel, T.; Blatter, K.; Schluter, A. D. Makromol. Chem, Rapid Commun.
1989, 10, 427-430.

Luo, J. H., H. J. J. Org. Chem. 1987, 52, 4833.

Netka, J.; Crump, S. L.; Rickbom, B. J. Org. Chem. 1986, 51, 1189-
1199.

Goodings, E. P.; Mitchard, D. A.; Owen, G. J. J. Chem. Soc. Perkin
Trans. 1972, 1310.

Kobayashi, K.; Sugawara, S.; Toyoda, 8.; Honda, H. Carbon 1968, 6,
359.

Jenkins, G. M.; Kawamura, K.; Ban, L. L. Proc. Royal Soc. (London)
1972, A327, 501.

Kurosu, H.; Nakajima, Y.; Ando, I. J. Mol. Struct. 1995, 355, 27-32.

Tanaka, K.; Ago, H.; Matsuura, Y.; Kuga, T.; Yamabe, T.; Yata, S.;
Hato, Y.; Ando, N. Synth. Met. 1997, 89, 133-139.

Tanaka, K.; Koike, T.; Nishino, H.; Yamabe, T.; Yamauchi, J.; Deguchi,
Y.; Yata, S. Synth. Met. 1987, 18, 521-526.

Nijegorodov, N.; Ramachandran, V.; Winkoun, D. P. Spectroc. Acta Pt.
A-Mo/ec. Biomolec. Spectr. 1997, 53, 1813-1824.

Tuinstra, F.; Koenig, J. L. J. Phys. Chem. 1970, 53, 1126.

186

93

94

95

96

97

98

99

100
101

102
103

104

105
106

107

108

109

110

Ortiz, 8.; Walls, F.; Yuste, F.; Barrios, H.; Sanchez-Obregon, R.; Pinelo,
L. Synth. Commun. 1993, 23, 749-756.

Otsubo, T.; Kohda, T.; Misumi, S. BulI.Chem.Soc.Jpn. 1980, 53, 512-
517.

Bedard, T. C.; Moore, J. S. J. Amer. Chem. Soc. 1995, 117, 10662-
10671.

Zhang, S. S.; Liu, Q. G.; Yang, L. L. J. Electrochem. Soc. 1993, 140,
L107-L108.

Yata, S.; Hato, Y.; Kinoshita, H.; Ando, N.; Anekawa, A.; Hashimoto, T.;
Yamaguchi, M.; Tanaka, K.; Yamabe, T. Synth. Met. 1995, 73, 273-277.

Wang, T.; Kang, E. T.; Neoh, K. G.; Tan, K. L. Synth. Met. 1997, 84,
405-406.

Ago, H.; Nagata, K.; Yoshizawa, K.; Tanaka, K.; Yamabe, T. Bull.
Chem. Soc. Jpn. 1997, 70, 1717-1726.

Tanaka, K.; Yata, S.; Yamabe, T. Synth. Met. 1995, 71, 2147-2149.

Salem, L.; Longuett-Higgins, H. C. Proc. R. Soc. London, Ser. A 1960,
255, 435.

Yazami, R.; Guerard, D. J. Power Sources 1993, 43, 39-46.

Voigtmartin, I. G.; Simon, P.; Bauer, 8.; Ringsdorf, H. Macromolecules
1995, 28, 236-242.

Voigtmartin, l. G.; Simon, P.; Yan, D.; Yakimansky, A.; Bauer, 8.;
Ringsdorf, H. Macromolecules 1995, 28, 243-254.

Scherf, U.; Mullen, K. Synthesis 1992, 23-38.

Rehahn, M.; Schluter, A. D.; Wegner, G.; Feast, W. J. Polymer1989,
30, 1060-1062.

Miyaura, N.; lshiyama, T.; Sasaki, H.; lshikawa, M.; Satoh, M.; Suzuki,
A. J. Am. Chem. Soc. 1989, 111, 314-321.

Buiel, E.; George, A. E.; Dahn, J. R. J. Electrochem. Soc. 1998, 145,
2252-2257.

Megahed, S.; Scrosati, B. J. Power Sources 1994, 51, 79-104.

Brandt, K. Solid State Ion. 1994, 69, 173-183.

187

l l l Salomon, M.; Scrosati, B. Gazz. Chim. Ital. 1996, 126, 415—427.

188

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

llllllllLIIIIIIIIIIIIIWIIIIIIII