A Luna...‘
:13?) .2! I

k,“

4 2... .. a
. Izmir? n
l .51. xi)!

 

IiAIvab .qul. A

. .. ...
3.

 

. . . , 13...... .mmhmxm... 1.5%....»
§Z.§ém...».%.;» 5...». . . . . _ » .».. -

1..

T;- 3‘51”? LIARBR

‘5 '5 llllllllllllllllllllllllll Illil lillllsllllllllll

“7“)? 31293018209076

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

SYNTHESIS AND CHARACTERIZATICN 0F RIGID PQYWER
SYSTEMS: SUBSTITUTED POLY(D-Pl-ENYLENE)S, POLYACENIC
NETWORKS, IWD PG.Y(TRIlVE'll-|YLSILYLPROPYNE) COPOLYIVERS

presented by

Cory J. Ruud

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Chemistry

 

 

a? éé a jar professor <

Date 8/11/99

 

MSU is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

i

LIBRARY

Michigan State
University

 

 

PLACE iN RETURN Box to remove this checkout from your record.
To AVOID FINE return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

it 3/"

\J

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11” WW“

SYNTHESIS AND CHARACTERIZATION OF RIGID POLYMER SYSTEMS:
SUBSTITUTED POLY(p-PHENYLENE)S, POLYACENIC NETWORKS, AND
POLY(TRIMETHYLSILYLPROPYNE) COPOLYMERS
By

Cory J. Ruud

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

1999

ABSTRACT
SYNTHESIS AND CHARACTERIZATION OF RIGID POLYMER SYSTEMS:
SUBSTITUTED POLY(p—PHENYLENE)S, POLYACENIC NETWORKS, AND
POLY(TRIMETHYLSILYLPROPYNE) COPOLYMERS '
By

Cory J. Ruud

The Simplest mathematical model of a polymer chain in space is the freely
jointed chain. It is defined as a chain having n links, each of length L, joined in a
linear sequence with no restrictions on the angles between successive bonds. The
model is equivalent to a random walk by a particle where n is the number of jumps
by the particle and L is the length of each jump, and can be characterized by the net
displacement or its end-to-end distance of the random walk. To characterize a
polymer’s behavior in solution, it is common to define its characteristic ratio, the
ratio of the end-to-end distance of the real polymer to that of a freely jointed
polymer of the same molecular weight. Flexible polymers like polyethylene or
poly(ethylene oxide) usually have small characteristic ratios because of low barriers
of rotation about the polymer backbone, and high degrees of polymerization. Rigid
polymers, however, typically have large characteristic ratios. Rigidity in polymers
occurs when the backbone is locked into a particular conformation by chemical
bonding or by large rotational barriers. Usually, flexible polymers are associated
with coiled backbones while rigid polymers in solution have rod-like geometries.

This is true in the cases of poly(p-phenylene) (PPP) and polyacene. The structure of

PPP consists of a series of aromatic rings connected in a linear 1,4-substitution
pattern. Since the carbon atoms in PPP are Sp2 hybridized, bond motion is limited to
rotation about the central axis and PPP takes a rigid rod-like conformation.
Polyacenes are the long chain members of the family of edge-fused aromatic
compounds that includes naphthalene, anthracene, and higher homologs. Bond
rotation within the acene backbone is impossible since all of the sp2 carbons are
conjugated forming a fused planar structure. The correspondence between chain
rigidity and rod-like conformation apparently breaks down for substituted
polyacetylenes like poly[l-trimethylsilyl-l-propyne] (PTMSP). The reported
solution properties suggest a stiff polymer chain, yet PTMSP is highly soluble and
amorphous with no report of liquid crystalline behavior. Large barriers of rotation
about the backbone are caused by the steric interaction of the bulky trimethylsilyl
and methyl groups, effectively preventing the polymer from changing its
conformation. It iS deduced that PTMSP has a rigid but disordered backbone
contour, i.e., a rigid random coil conformation. The characteristic ratio of PTMSP is
expected to be lower than that of rigid rod-like polymers, therefore the geometries of
rigid polymers are not restricted to a particular shape.

The three polymer systems investigated: PPP, polyacene, and PTMSP each
possess a set of unique electronic, optical and mechanical properties. However rigid
polymers often exhibit poor solubility, which hinders efficient processing. Herein
the physical properties and potential applications of these rigid polymer systems will

be discussed.

ACKNOWLEDGMENTS

I would like to thank my advisor, Professor Greg Baker, for guiding me
through this learning process. His friendship, patience and understanding have
meant a great deal to me. I would also like to thank my committee members, Gary
Blanchard and Mitch Smith, for their willingness to give “free advice” at a price.

I would like to thank the Baker group members, past and present, for their
assistance and support throughout my journey. I would like to extend Special thanks
to Chun who has become a better chemist by learning from my mistakes. I wish
them all happiness and success in the future.

Of course I could not continue without mentioning the excellent staff: Russ
and Sam in the machine ShOp, the glass blowers: Manfred, Scott, and Keki, Ron and
Scott in the electronics shop who kept our equipment running, Tom in the
stockroom, and Bob who kept our lab livable. I would also like to thank Cindy,
Carmen, Ann, Lisa, Cathie. and Diane who were always willing to go out of their
way to help me.

Most importantly, I would like to thank my Mom and Sally for their constant

support and encouragement.

TABLE OF CONTENTS

List of Tables .................................................................................. viii

List of Figures .................................................................................. ix

Chapter 1. Introduction

1.1. Structure/Property Relationships of Rigid Polymer Systems. ................1

1.1.1. Poly(p-phenylene)s .................................................... 6
1.1.2. Poly(acene) ........................................................... 10
1.1.3. Poly(l-trimethylsilyl-l-propyne) ................................... 12
1.2. Syntheses of Rigid Polymer Systems ......................................... 16
1.2.1. Poly(p-phenylene)s .................................................. 16
1.2.2. Poly(acene) ........................................................... 25
1.2.3. Poly(acetylene)s — Poly(l -trimethylsilyl-l-propyne) ............ 29
1.3. References ....................................................................... 32

Chapter 2. Synthesis and Characterization of Ethylene Oxide Substituted
Poly(p-phenylene)s.

2.1. Introduction ..................................................................... 39
2.2. Experimental Section .......................................................... 42
2.3. Results ........................................................................... 48
2.4. Discussion ....................................................................... 69
2.5. Conclusions ..................................................................... 77

2.6.

References ....................................................................... 78

Chapter 3. Synthesis and Characterization of Polyacenic Networks.
3.1. Introduction ...................................................................... 81
3.2. Experimental Section .......................................................... 83
3.3. Results ............................................................................ 89
3.4. Discussion ...................................................................... 103
3.5. Conclusions .................................................................... 107
3.6. References ...................................................................... 108
Chapter 4. Synthesis and Characterization of Poly[(l-trimethylsilyl-l-
propyne)-co-(l-(4-azidobutyldimethylsilyl)-1-propyne)] Copolymers.
4.1. Introduction ................................................................... 110
4.2. Experimental Section ......................................................... 113
4.3. Results ......................................................................... 119
4.4. Discussion ..................................................................... 142
4.5. Conclusions ................................................................... 150
4.6. References ..................................................................... 151
Chapter 5. PTMSP as a Permeable Solid Support in Oxygen Sensor
Applications.
5.1. Introduction ................................................................... 154
5.2. Experimental Section ......................................................... 159
5.3. Results ......................................................................... 163

vi

5.4. Discussion ...................................................................... 178
5.5. Conclusions .................................................................... 181

5.6. References ...................................................................... 182

vii

Table

Table 1.1

Table 2.1

Table 4.1

Table 4.2

Table 4.3

LIST OF TABLES

Solubility of phenylene Oligomers in toluene at room

temperature .......................................................

Page

.......... 7

Representative molecular weights of polymers 5a-5h .............. 53

Elemental analyses of poly[( 1 -trimethylsilyl-l -
propyne)-co-(l -(4-azidobutyldimethylsilyl)-1-

propyne)] copolymers ...........................................

Permeability coefficients of oxygen and the
Oz/Nz separation factor (O) at 23i1 0C of
thermally cross-linked poly[( l -trimethylsilyl-

1 -propyne)-c0-(1 -(4-azidobutyl-dimethylsilyl)-

1-propyne)] copolymer membranes ..........................

Permeability coefficients of oxygen and the
Oz/Nz separation factor (OI) at 23i1 °C of
photo cross-linked poly[( l -trimethylsilyl-
l-propyne)-c0-( 1-(4-azidobutyl-dimethylSilyl)-

1-propyne)] copolymer membranes ..........................

viii

...... 125

....... 146

....... 147

Figure
Figure 1.1

Figure 1.2

Figure 1.3
Figure 1.4
Figure 1.5

Figure 1.6

Figure 1.7

Figure 1.8
Figure 1.9
Figure 1.10
Figure 1.11
Figure 1.12

Figure 1.13

Figure 1.14

Figure 1.15

LIST OF FIGURES

A random walk of a freely jointed chain ..........................

Relationships between the characteristic ratio

and rigidity in polymers .............................................
Structures of a) poly(p-phenylene) and b) polyacene ............
Structure of poly(] -trimethylsilyl-1 -propyne) ....................

Structure of a ladder poly(p-phenylene) ...........................

Possible electronic configurations of polyacene:

(a) aromatic, (b) cis-quinoid, (c) trans-quinoid ..................

Hypothetical rotational barriers for disubstituted
polyacetylenes where 0 degrees corresponds to

an eclipsed conformation in which two double bonds
are coplanar. (a). trans-polyacetylene, 6 Kcal/mol

7t barrier; (b) 20 Kcal/mol rotational barrier

that simulates Significant 1-3 steric interaction;

(c) 100 Kcal/mol barrier that simulates polyenes

with bulky substituents ..............................................
Polymerization of 1-bromo-4-lithiobenzene ......................
Polymerization of diiodoarenes .....................................
Catalytic cycle of palladium mediated aryl coupling ............
Polymerization of an enediyne .....................................

Diels-Alder polymerization of biscyclopentadienones ..........

Synthesis of PPP via soluble poly(cyclohexadiene)

Intermediates ..........................................................
Polymerization of poly(ethynyl acetylene) .......................

Polymerization of Langmuir-Blodgett films of diacetylenes.... . .

Page

....... 1

....... 2

....... 4

....... 5

....... 9

...... 11

...... l3

...... 19

..... 20

...... 22

...... 23

..... 24

...... 25

...... 26

27

Figure 1.16
Figure 1.17

Figure 2.1

Figure 2.2
Figure 2.3
Figure 2.4
Figure 2.5
Figure 2.6

Figure 2.7

Figure 2.8

Figure 2.9

Figure 2.10
Figure 2.11

Figure 2.12

Figure 2.13
Figure 2.14
Figure 2.15

Figure 2.16

Figure 2.17

Polymerization of cycloproparenes ...................................... 28
Initiation and propagation of acetylene polymerization ............... 31

Structure/property relationships of ethylene oxide

substituted PPP’S ........................................................... 41
Synthetic route to monomers 4a-4h ..................................... 48
Melting points of monomers 4a-4h ...................................... 49
Nickel catalyzed polymerization ......................................... 50
Catalytic cycle of nickel mediated aryl coupling ....................... 51
Degrees of polymerization of polymers Sa-Sh ......................... 54

Fluorescence spectra of a solution of polymer 5d
in methanol excited at 317 nm ............................................. 56

Comparison of the orientation of repeat units in

2,5-disubstituted and 2-substituted PPP’S ............................... 57
Therrnogravimetric analysis of a) 5b, b) 5c, and

c) 5d in nitrogen heated at 10 °C/min ................................... 59
DSC results for polymers Sa-Sh ......................................... 60
Glass transition temperatures of polymers Sd-Sh ..................... 61

Schematic representation of rigid rod PPP backbone

substituted with flexible ethylene oxide Side chains ................... 62
Melting points of polymers 5g—5h ....................................... 63
Effect of Li+ salt on the Tg of polymer 5d .............................. 66
Effect of Li+ salt on the Tg of polymer 5e .............................. 67

Schematic phase diagram of ethylene oxide
substituted PPP’S 5a-5h ................................................... 70

Synthetic route to exact length ethylene glycol
monomethyl etherS ......................................................... 71

Figure 2.18

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6
Figure 3.7
Figure 3.8

Figure 3.9

Figure 3.10

Figure 3.11
Figure 3.12
Figure 3.13

Figure 4.1

Figure 4.2

Synthetic route to polymers 5a-5h with uniform
backbone length ............................................................. 74

Polymerization of methoxymethylbenzene ............................. 89

Thermogravimetric analysis of the networks formed
from monomers 1-3 heated in nitrogen at 10 °C/min .................. 91

Thermogravimetric analysis of the networks
formed from a) monomer 1, and b) monomer 3
in a stagnant air environment heated at 10 °C/min .................... 92

Idealized structures resulting from the
polymerizations of the biS(methoxymethyl)benzenes 1-3 ............ 94

Infrared absorption spectra of the polybenzyl
network formed from monomer 1 heated in nitrogen
at 10 °C/min to a) 30 °C, b) 380 °C, c) 500 °C,

d) 600 °C, and e) 800 °C .................................................. 95
Synthesis of l,4-bis(methoxymethyl)-2,5-dibromobenzene. . . ... . . ..97
Synthesis of substituted poly(p-phenylene) 6 .......................... 98
Cross-linking of substituted PPP 6 ....................................... 99

Thermogravimetric analysis of a) substituted
polymer 6, and b) cross-linked network 7 in
nitrogen heated at 10 °C/min ............................................ 100

Infrared Spectra of a) polymer 6, b) network 7,
and 0) carbon char of network 7 after thermal

treatment at 700 °C ....................................................... 102
Alkyl substituted polyacene ............................................. 104
Synthesis of a soluble polyacene precursor ............................ 105
Chemical dehydrogenation of polyacene precursor .................. 106

Synthesis of azide-containing PTMSP via
allylic bromination ....................................................... 1 19

Synthetic route to 1-(4-bromobutyldimethylsilyl)-
1 -propyne .................................................................. 1 22

xi

Figure 4.3
Figure 4.4

Figure 4.5

Figure 4.6

Figure 4.7

Figure 4.8

Figure 4.9

Figure 4.10

Figure 4.11

Figure 4.12

Figure 4.13

Figure 4.14

Figure 4.15

Figure 5.1
Figure 5.2

Figure 5.3

Synthetic route to copolymers 6a-6g .................................. 123
FTIR spectra of PTMSP and copolymers 6a-6g ...................... 126
Integrated azide absorbance intensity at 2100 cm'1

versus the azide content in copolymers 6a-6g as
calculated from the comonomer feed ratios ........................... 127

TGA of a) PTMSP under nitrogen; b) copolymer
6g under nitrogen; and c) copolymer 6g under air ................... 129
DSC results for copolymers 6a-6g ...................................... 131

Heat evolved during thermal decomposition of
azide copolymers 6a-6g measured by DSC ........................... 132

FTIR Spectra Showing the thermal decomposition
of copolymer 6f at 225 0C under vacuum ............................. 134

FTIR Spectra Showing the thermal decomposition
of copolymer 6f at 250 °C under vacuum ............................. 135

FTIR spectra showing the thermal decomposition
of copolymer 6f at 275 °C under vacuum ............................. 136

LOSS of [N3] in copolymer 6f versus time
at the indicated temperatures ........................................... 137

Arrhenius plot of the rates of thermal decomposition
of copolymer 6f .......................................................... 138

Swelling of thermally cross-linked films of copolymers
6a and 6f in solvents with different solubility

parameters, 6 ([J/m3]”2x10'3) ........................................... 140
Swelling of thermally cross-linked films of copolymers

6a-6g in toluene ........................................................... 141
Structure of (M06C112)n clusters ........................................ 155
Stem-Vollmer plot of emission intensity versus [02] ................ 157

The UV/V is absorption spectra of a) PTMSP and
b) MO6C112, and the fluorescence Spectra of c) MO6C112. . . . . . . . 1 64

xii

Figure 5.4 Synthesis ofpoly[(1-trimethylsilyl-1-propyne)-

co-(l -(4-cyanobutyldimethylsilyl)-1-propyne)] ...................... 1 66
Figure 5.5 Luminescence spectra from a M06C11 g/nitrile-

PTMSP copolymer composite film in laboratory

air (bottom curve) and under Ar (top curve) .......................... 168
Figure 5.6 Observed emission from Fiber 1 in varying

oxygen/nitrogen environments .......................................... 171
Figure 5.7 Peroxide induced chain degradation of PTMSP ...................... 172
Figure 5.8 Observed emission intensity from Fiber 2 in nitrogen

and a) 2.1%, b) 4.1%, c) 6.2%, d) 12.4%, and

c) 20.6% oxygen containing environments ............................ 174
Figure 5.9 Observed emission intensity from Fiber 2 in nitrogen

and a) 0.1%, b) 0.2%, c) 0.4%, d) 0.6%, e) 0.8% and

t) 1.0% oxygen containing environments ............................. 175

Figure 5.10 Stem-Vollmer plot of emission data from Fiber 2 .................... 176

xiii

Chapter 1

Introduction

1.1. Structure/Property Relationships of Rigid Polymer Systems

The simplest mathematical model ofa polymer chain in Space is the freely
jointed chain. It is defined as a chain having n links, each Of length L, joined in a
linear sequence with no restrictions on the angles between successive bonds. The
model is equivalent to a random walk by a particle where n is the number of jumps
by the particle and L is the length of each jump. A random walk such as that shown
in Figure 1.1 for a random walk of 30 steps can be characterized by the net

displacement or its end-tO-end distance, < r2 >, ofthe random walk.

 

Figure 1.1. A random walk of a freely jointed chain.

 

rod
(PPP, polyacene)

/ /\J

’ extended chain
(polyimide)

 

 

flexible chain

(PE,PEO) \\\\\\\$
9

random coil
0 (PTMSP)

characteristic ratio (C)

 

 

 

0 rigidity —> 1

Figure 1.2. Relationships between the characteristic ratio and rigidity in polymers.

TO characterize a polymer’s behavior in solution, it is common to define its
characteristic ratio, C. This ratio (Equation 1) is the ratio of the end-to-end distance
ofthe real polymer to that ofa freelyjointed polymer ofthe same molecular weight.1

As Shown schematically in Figure 1.2, flexible polymers like polyethylene (PE) or

c =<F>/(n- L2) (1)

poly(ethylene oxide) (PEO) usually have small characteristic ratios because of low
barriers of rotation about the polymer backbone, and high degrees of polymerization.
Rigid polymers, however, typically have large characteristic ratios. Rigidity in
polymers occurs when the backbone is locked into a particular conformation by

chemical bonding or by large rotational barriers.

Usually, flexible polymers are associated with coiled backbones while rigid
polymers in solution have rod-like geometries. This is true in the cases of poly(p-
phenylene) (PPP) and polyacene. The structure of PPP consists of a series of aromatic
rings connected in a linear 1,4-substitution pattern (Figure 1.3 a). Since the carbon
atoms in PPP are sz hybridized, bond motion iS limited to rotation about the central
axis and PPP takes a rigid rod—like conformation. Polyacenes, shown in Figure 1.3
b, are the long chain members of the family of edge-fused aromatic compounds that
includes naphthalene, anthracene, and higher homologs. Bond rotation within the
acene backbone is impossible since all of the Sp2 carbons are conjugated forming a

fused planar structure.

b)

 

Figure 1.3. Structures of a) poly(p-phenylene) and b) polyacene.

The correspondence between chain rigidity and rod-like conformation
apparently breaks down for substituted polyacetylenes like poly(l-trimethylsilyl-l-
propyne) (PTMSP), shown in Figure 1.4. The reported solution properties suggest a
stiff polymer chain, yet PTMSP is highly soluble and amorphous with no report of
liquid crystalline behavior. Large barriers to rotation about the backbone are caused
by the steric interaction of the bulky trimethylsilyl and methyl groups, effectively
preventing the polymer from changing its conformation. It is deduced that PTMSP
has a rigid but disordered backbone contour, i.e., a rigid random coil conformation.
The characteristic ratio of PTMSP is expected to be lower than that of rigid rod-like
polymers, therefore the geometries of rigid polymers are not restricted to a particular

Shape.

CH3

\

n

31...
I” H
CH3/\ C 3
CH3

Figure 1.4. Structure ofpoly(1-trimethylsilyl-l-propyne).

The three polymer systems investigated: PPP, polyacene, and PTMSP each
possess a set of unique electronic, optical and mechanical properties. However rigid
polymers often exhibit poor solubility, which hinders efficient processing. Herein
we describe the physical properties and investigate potential applications of these

rigid polymer systems.

1.1.1. Poly(p-phenylene)s

PPP’s have excellent thermal, oxidative, and electrical stability, and are
among the most thermally stable and intractable polymers known. PPP undergoes
no glass transitions and decomposes before it melts. Since there are no weak linkages
in the polymer backbone, PPP loses only 7% of its mass when heated to 900 °C in an
inert atmosphere. Although PPP takes a rigid rod geometry, the polymer backbone is
not planar; adjacent rings twist ~23° to relieve steric interactions between adjacent

2

ortho hydrogens. The twisting about the central axis limits the extent of conjugation

between the rings. PPP’S are attractive for applications as electrical conductors,
light emitting diodes (LED’S), dielectric layers, and luminescent materials.3'6

The major drawback of PPP is its poor solubility in all organic solvents. AS
the number of rings in the backbone is increased, the solubility dramatically decreases
due to rt-stacking of the backbone. As Shown in Table 1.1, the solubility of Short
chain poly(p-phenylene) oligomers approaches zero even for low degrees of

polymerization.7

Table 1.1. Solubility of phenylene oligomers in toluene at room temperature.

Compound Solubility (g/L)

8.0

0.2

0.01

©©© 3°

O 0 ©
Ewi

© © O

O ©

©

Recent work has shown that adding side chain substituents8

to polyphenylene
or connecting the aromatic rings through ortho and meta linkages increases solubility
and processibilityg'11 Both strategies prevent efficient n-stacking of the backbone,
hindering crystallization and therefore increasing solubility. Adding ortho and/or
meta linkages to the PPP backbone however decreases the desired physical properties
of the polymer. Ortho and meta linkages place kinks in the polymer backbone, and
the polymer can no longer exist in a rigid rod conformation. Although the solubility
of these systems is increased since the polymers take a coiled conformation,
polyphenylenes with mixed linkages are not as thermally stable as PPP and tend to
be easily oxidized.

There are numerous examples of adding side chains to the PPP backbone to
increase itS solubility and processibility. 12‘] 6 It has been reported that when groups
are added in the 2.5-positions, the twist angle between adjacent rings increases to
~45° to alleviate steric interactions of the adjacent groups. The solubility dramatically
increases even for the addition of small side chains. For example, as shown in Table
1.1, terphenyl has a solubility of 8.0 g/L in toluene, while hexamethylterphenyl has a
solubility of 30 g/L. It is not quite clear how the added methyl groups improve the
solubility of terphenyl. One argument suggests that substitution increases the ring
twist and simply creates more disorder in the backbone and hinders crystallization.
17,18

This argument can not hold true in the case of ladder polyphenylenes however.

As shown in Figure 1.5, the aromatic rings in ladder polyphenylenes are fused together

through benzyl linkages, forcing the backbone to become nearly planar. Soluble
ladder polyphenylenes have been synthesized by adding Side chains to the benzyl
positions. In this case, the solubility of these polymers must be due solely to the Side
chains themselves Since ring twisting is impossible. It is proposed that the side
chains can act as spacers to disrupt the rt-stacking of the polymer backbone, hindering
crystallization. The ladder polymers also have a structural source for solubility since
the R group can either be above or below the plane of the ladder. This is analagous

to tacticity seen in polyolefins.

 

 

 

Figure 1.5. Structure of a ladder poly(p-phenylene).

Substituted PPP’S exhibit interesting properties due to their unique structures.
With flexible Side chains radiating out from the rigid rod core, soluble substituted
PPP’S take on a “hairy rod-like” conformationlg’20 Liquid crystallinity is often
observed in these systems.21’22 In Chapter 2, the properties of a substituted PPP
will be described. It was observed that the properties of these polymers varied as the

volume fraction ofthe Side chain was increased.

1.1.2. Poly(acene)

Polyacene can be viewed as a one-dimensional analog of graphite, a ribbon
taken from a single graphene Sheet. The structure can also be viewed as a ladder
polymer, where the rungs of the ladder connect two trans-polyene chains. Since the
two polyene chains are locked into a planar conformation, they should possess a
very high degree of conjugation. Theory predicts that polyacenes Should have very
small band gap energies and be promising electrically conductive materials. However,
due to the high degree of conjugation, polyacenes are highly reactive and susceptible
to oxidation.

Polyacene has long been the Object of theoretical interest because it can
have several different electronic structures (Figure 1.6), each having a different
bandgap energy. Predicted band gap energies range from 0 to 0.45 eV.23'25 The
trend of decreasing band gap energy can be seen in the electronic absorption spectra
of lower homologs of polyacene. The leading edge of the absorption Spectra of
anthracene, pentacene, and heptacene decrease in energy as the extent of conj ugation
is increased. A practical application proposed for polyacenes is their use as optical
limiters.26’27

Polyacenicmaterials. polymer networks that contain a large fraction of acene

units, are also materials of great interest. The acene units imbedded in these materials,

10

usually formed during the high temperature pyrolysis of a suitable precursor, give
the materials interesting chemical and electronic properties. Potential applications
of polyacenic materials include their use as electrical conductors and electrodes for

lithium ion batteries.

.
I 303330
.. OOOOOO

Figure 1.6. Possible electronic configurations of polyacene:
(a) aromatic, (b) cis-quinoid, (c) trans-quinoid.

11

1.1.3. Poly(l-trimethylsilyl-l-propyne)

PTMSP is an example ofa rigid random coil. For PTMSP, simple molecular
mechanics calculations suggest that steric interactions between the substituents on
adjacent carbon atoms can cause Significant deviations from planarity.28 Because
of the shorter distances between carbon atoms in the Sp2 skeleton, the magnitude of
the interactions are expected to be much larger than for their saturated analogs and
there Should be a large barrier to rotation. Plotted schematically in Figure 1.7 is a
potential energy diagram for rotation about single bonds in substituted trans-
polyacetylenes. The rotational barrier can be modeled using two energy terms; one
with a c0520 dependence that accounts for the 7t contribution to the rotational barrier,
and a second term with a cos 20 dependence that corresponds to steric barriers to

29 the It barrier is set at 6

rotation. Based on experimental results for butadiene,
Kcal/mol. With small substituents, the combination of a small It energy term and
limited 1,3 steric interactions leads to a broad Shallow potential well with the most
stable conformations being those that are nearly planar. As the magnitude of the
steric interactions is increased, a double well potential becomes well defined and the
It contribution becomes a minor perturbation to the total energy. For large substituents
(i.e. t-butyl, trimethylsilyl), the barrier to rotation at room temperature may be large

enough to make planar conformations so unfavorable that the polymer is driven into

one of two sharply defined potential wells. With bulky substituents, the barrier to

12

50

 

 

 

 

 

 

 

 

. C
I
40 j
‘6‘
g 30 .
a
o
E .
>. I
9 20 ' b
on
c
a)
10 .,
i a
0 .I____. . . . _ A .1» _ I . '.__ __.-_.E ....J
-180 - 120 -60 0 60 120 180

dihedral angle (°)

Figure 1.7. Hypothetical rotational barriers for disubstituted polyacetylenes
where 0 degrees corresponds to an eclipsed conformation in which two double
bonds are coplanar. (a). trans-polyacetylene, 6 Kcal/mol 1t barrier; (b) 20
Kcal/mol rotational barrier that simulates Significant 1-3 steric interaction; (c)
100 Kcal/mol barrier that simulates polyenes with bulky substituents.

13

rotation may become large enough to limit access to a Single well. Thus, the polymer
conformation is fixed, and overall changes in the shape of the polymer are limited to
oscillations within a Single well.

Thermal annealing experiments provide an illustration Of PTMSP’S inability
to change conformations.30 Depending on the catalyst used to prepare the polymer,
different amounts of cis and trans double bonds are found in the polymer. For
polyacetylene in the solid state, isomerization of cis double bonds to trans double
bonds occurs readily at temperatures above 100 °C. At 180 °C for example, the
isomerization of semicrystalline films of cis-polyacetylene to the trans isomer is
completed in one minute.30 Similar experiments on PTMSP Show no signs of
isomerization.31 For annealing at temperatures above 200 °C, 13C NMR spectra
Show that the isomer content is unchanged after 24 hours, and instead random chain
scission leads to lower molecular weight polymers with the cis/trans ratios preserved.
Remarkably, the barrier to isomerization apparently exceeds the energy required to
break the polymer backbone. The barrier calculated from molecular dynamics
Simulations (40 Kcal/mol) is in accord with such behavior. Thus, PTMSP fits well
the model of a rigid chain with a highly disordered backbone contour.

Due to its rigid backbone, PTMSP is one of the most unusual polyenes
synthesized. In contrast to polyacetylene which is insoluble, appears blue in

transmitted light, and can be doped to form highly conducting materials, PTMSP

14

readily dissolves in a wide range of non-polar solvents, is colorless, and is unreactive
toward common chemical dopants. The twisting caused by the steric repulsion of
the trimethylsilyl and methyl groups misaligns adjacent 1t bonds, leading to PTMSP’S
lack of color. Its lack of absorption above 280 nm in the UV-vis spectrum implies a
degree of conjugation as extensive as butadiene. PTMSP is amorphous according to
X-ray diffraction and thermal analysis.32

Amorphous polymers exhibit different physical and mechanical behavior
depending on the temperature and the polymer structure. At low temperatures,
amorphous polymers are glassy, hard, and brittle. As the temperature is raised, they
go through a glass transition (Tg) and enter a rubbery state. However, PTMSP shows
no evidence of a Tg to the upper limit of its chemical stability, usually 300 °C.

32

Mechanical measurements place the modulus of PTMSP between that of typical

glassy and rubbery materials. Due to these unique properties, PTMSP has found

33,34

practical applications as a material for deep-UV photolithography, and as a

35-38 39-42

highly permeable membrane material for gas and liquid separations.

With its rigid backbone structure and the presence of the large bulky
substituents, inter-segmental packing is hindered and thus the polymer has a large
free volume. Therefore films of PTMSP have many molecular-scale holes just after
its preparation which results in a high gas permeability.43 The permeability decays
with time however, dropping to 10 % of its original value after 100 days.44 One

model for the decrease is the Slow interdiffusion of polymer coils in the solid state.

From solution, the polymer is deposited under kinetic control as an ensemble of

15

spheres. With time, the chains entangle causing an increase in the density and decrease
in free-volume and the permeability. Stabilization of PTMSP in its porous form is
needed for PTMSP to be useful.

Cross-linking the polymer Should lead to stabilized PTMSP networks in its
highly permeable form. Chapter 4 describes a method to stabilize PTMSP by cross-
linking. An application for these mechanically stable, highly permeable materials is
described in Chapter 5. Cross-linked PTMSP was used as a permeable solid support

in a novel oxygen-sensing scheme.

1.2. Syntheses of Rigid Polymer Systems

In order to obtain these rigid polymer systems, controlled synthetic methods
are necessary to produce ideal structures. The methods employed must be
regiospecific and lead to stable products. Any defects or side reactions can lead to a
dramatic decrease in the desired properties. For example, a single flexible link in a
rigid rod backbone dramatically reduces the characteristic ratio. A decrease in
solubility with increases in the degree of polymerization often must be overcome to

lead to high molecular weight products.

l.2.l.Poly(p-phenylene)s
PPP has traditionally been a very difficult polymer to synthesize. The struggle
to find efficient methods for producing PPP with both a regioregular linear structure

and high molecular weight, together with the inherent lack of processibility of the

16

material, have retarded the development of PPP as a useful material. A successful
synthetic approach to poly(p-phenylene)s must employ methods that lead to
exclusively para aromatic linkages and high molecular weights. However, PPP’S
decreasing solubility as the extent of polymerization increases limits the molecular
weight of the product. Degrees of polymerization approaching 50 must be obtained
in order for the materials to be classified as polymeric.

The electrochemical coupling of aromatic monomers to give polyphenylenes
on inert electrodes (usually Pt) has been reported.45’46 Electrochemical methods,
namely the oxidative coupling of unsubstituted aromatic monomers and the reductive
dehalogenation of halogenated aromatic monomers, only yield small quantities of
polymer. Most electrochemical syntheses are designed to produce thin conductive
films on electrodes. not bulk amounts. The structures of the deposited polymers are
irregular and contain substantial amounts halogens and a mixture of ortho, meta,
and para aromatic linkages between the rings, indicating that electrochemical coupling
produces lead to materials of uncontrolled structure.

Chemical oxidative coupling of aromatic rings is possible if the catalyst
system used can act as a good Friedel-Crafts catalyst and as a good oxidizing agent.47
The polymerization of benzene using AlCl3 and CuCl2 proceeds under mild conditions

to yield a low molecular weight brownish powder, which almost exclusively consists

ofpara-linked phenylene rings. Other reagents, such as PbO, MnOz, and chloranil,

17

have been used as co-catalysts with AlCl3 but result in products with even lower
molecular weights and more structural irregularity (branching and high halide
content). The polymerization of substituted benzenes resulted in products with higher
molecular weights. The increase, however, is not due to an increase in the solubility
from the side chains, but instead is a consequence of the increase in the number of
mixed linkages (ortho and meta) that connect the rings.

Some of the earliest reported attempts to synthesize poly(p-phenylene)
involved the reductive coupling of substituted aromatic compounds. Using the Fittig
reaction, dihalo-benzenes were refluxed in diethyl ether with sodium or potassium
metal to give brownish oligomers that are soluble in benzene. The relatively low
reported melting points (<3 00 OC) and structural irregularities, such as chain branching
and residual halide content, indicate that the reaction led to low molecular weight
products with a less than ideal PPP character.

More recently, an analogous process was developed that involves the
polymerization of a 1-bromo-4-lithiobenzene intermediate (Figure 1.8). It was found
that the intermediate, generated by the mono lithium-halogen exchange of 1,4-
dibromobenzene and t-butyllithium, polymerized nearly instantaneously at —78 0C
upon addition of hexamethylphosphoramide (HMPA) or after 2 hours in
tetrahydrofuran at room temperatureg’10 The structure of the resulting polymer

however contains many non-para linkages and has a high bromide content.

18

Br Li )\

1 . HMPA, ~78°C ,/

/ , t-BuLi A or f\ l
1/ I
V J

 

Br Br

Figure 1.8. Polymerization of 1-bromo-4-lithiobenzene.

Slightly better results are Obtained from the Ullmann homocoupling of 1,4-
diiodobenzene using activated copper powder.48'50 Starting with the para diiodo-
substituted benzene ring ensured that the resulting polymer has the correct para
substitution pattern. As shown in Figure 1.9. exact length poly(p-phenylene)
oligomers were also synthesized from the coupling of monoiodo biphenyls and
terphenyls. The major drawback of Ullmann coupling is the extreme conditions
required to promote the reaction. While the Ullmann coupling method has been used
in the polymerization of nitro-substituted diiOdO-biaryls, the scope of applications
for this method is limited Since most functional groups can not tolerate the prolonged
reaction temperatures of greater than 200 °C.

The use Of transition metal catalysis allows for milder reaction conditions
and greater chemoselectivity leading entirely to para-linked polyaromatics when

starting with para-substituted aromatic monomers.8’51'56 High yield palladium

l9

 

Figure 1.9. Polymerization of diiodoarenes.

20

and nickel catalyzed aryl coupling reactions were found to be ideal for
polymerizations.5 7 Transition metal catalyzed polymerizations are, in general, more
tolerant of a wide variety of functional groups. This allows for the synthesis of a
gamut of substituted polyphenylenes previously unattainable by earlier methods. The
catalytic cycle that most transition metal catalyzed polymerizations follow involves
three major steps. A typical palladium catalytic cycle is shown in Figure 1.10. First,
a zero valent metal complex, such as tetrakis(triphenylphosphine)palladium,
undergoes an oxidative addition with a bromo- or iOdO-arene. Transmetallation with
an electrophilic aromatic species. such as an aryl boronic acid or aryl tin complex,
forms a diary] metal complex. It is proposed that this intermediate rapidly undergoes
reductive elimination to yield the coupled product and regenerate the zero valent
metal complex. These polymerizations can be run at very low catalyst loadings, i.e.
below 5 mol%, due to the efficiency of each step in the cycle. The reactions are very
sensitive to impurities however. Since highly reactive zero valent metals are generated
during the cycle, great care must be taken to remove all oxygen from the
polymerization vessel. Oxidized catalyst usually Shows up as an insoluble black
powder that precipitates from the polymerization. The specific nickel catalyzed
polymerizations used to prepare substituted poly(p-phenylene)s will be discussed in

more detail in Chapters 2 and 3.

21

L : triphenyl phosphine
Pd°L4

ArX

2L 2L

X = Br, I, OMS, OTr

(Ar)2Pd"L2 "
ArPd XL2
x--E(OH)2 Ar—B(OH)2
or 01'
X-SnR3 Ar-SnR3

Figure 1.10. Catalytic cycle of palladium mediated aryl coupling.

22

Synthetic methods other than aromatic coupling have been reported to yield
poly(p-phenylene)s as well. One example involves the use of Bergman cyclization
to thermally generate a benzene 1.4-diradical from an enediyne which then
polymerizes (Figure 1.1 1).58 This route is attractive for several reasons. It requires
no chemical catalysts or reagents for the polymerization and heteroatomic coupling
sites such as halogens are not necessary. The monomers can also be readily prepared

with a variety of substitution patterns leading to soluble poly(p-phenylene) derivatives.

 

Figure 1.11. Polymerization of an enediyne.

Soluble substituted poly(p-phenylene)s have also been prepared via Diels-
Alder routes by reacting phenyl-substituted arylene biscyclopentadienones (or arylene
bispyrones) with m- or p-diethynylbenzene (Figure 1.12). The retro DielS-Alder
reactions are prevented due to the evolution of CO or C02. The polymerizations
were conducted in concentrated toluene solutions under pressure at elevated
temperatures (200-300 °C). Nearly quantitative yields and satisfactory elemental

analyses were Obtained. High molecular weights were obtained from the

23

 

0 Ph Ph O H—CECQCEC—H Ph Ph Ph Ph
Ph Ph _ 2 CO E I I I I 3['1
Ph

Ph Ph Ph
Figure 1.12. DielS-Alder polymerization of biscyclopentadienones.

polymerization using biscyclopenatadiones (Mn>10,000), but only low viscosities
were obtained for the polymers derived from the bispyrone monomers.
Another novel strategy used to prepare PPP proceeds through soluble

POIY(CyClohexadiene) intermediates. 14’59'61

In the first precursor route Studied,
poly(1,3-cyclohexadiene) was treated with Brz, followed by thermal
dehydrobromination to form the aromatic polymer. Unfortunately, this method yielded
a poorly processible material that underwent only partial conversion to polyphenylene
and contained large percentages of ortho and meta linkages. Slightly improved
results were obtained using soluble precursor polymers derived from the radical
polymerization ofcis-5,6-dihydroxy-1,3-cyclohexadiene. These precursors were then
converted to PPP by the acid catalyzed elimination of the pendant groups. Although

offering high processibility, this route suffered from the low selectivity of the radical

polymerization. This resulted in precursor polymers with random tacticity and

24

regiochemistry, which contributed to thermal depolymerization during the final
conversion to PPP. As Shown in Figure 1.13, the use of an organonickel catalyst that
could stereospecifically polymerize protected hexadiene monomers with the desired
tacticity and 1,4-regiochemistry led to structurally regular PPP with high molecular

weight after acid catalyzed elimination.

 

 

1) allyl-Ni cat. A
n r; ———>
2) TBAF. Ac20 acid cat. n
- 2n ROH
Me3SiO OSIME3
R = — Cl:_ CH3

Figure 1.13. Synthesis of PPP via soluble poly(cyclohexadiene) intermediates.

1.2.2. Poly(acene)

To date, well-defined high molecular weight polyacenes have not been
successfully prepared and characterized due to their sensitivity towards oxidation
and increasing insolubility with increasing chain length. The longest straight chain
acene analog that has been prepared is heptacene. Some related structures, however,
have been prepared. For example, various planarized polyphenylenes have been
synthesized that can be thought of as end-linked fluorene and phenanthrene

structures.18’62'64 Soluble polyacenequinone structures were synthesized by

25

repetitive Diels-Alder reactions.65 Diels-Alder reactions have also been used to
prepare cyclic belt or collar acene-like structures.66’69 Although related structurally,
the electronic properties of these materials do not match those expected for the doubly
conjugated ladder structure of polyacene.

The synthesis of linear polyacene via the step-wise polymerization of a
monosilylated diacetylene has been reported.70 The first polyacetylene strand was
polymerized using a WCl6-Ph4Sn mixture. After desilylation of the resulting
poly(trimethylsilyl ethynyl acetylene), poly(ethynyl acetylene) was thermally
polymerized to form the second polyacetylene strand and yield polyacene (Figure
1.14). The products from this reaction, however, contained substantial amounts of
structural defects. The catalyzed polymerization leading to poly(trimethylsilyl ethynyl
acetylene) did not result in the required pure trans isomer. Bond rotation to relieve
steric hindrance in the original polyacetylene strand would also lead to cross-linking

between adjacent chains upon thermal treatment.

W W,
HI —» I .0 I“:
CCC CC / I

ll
1 l | (I: |
SIR, SIR, SIR, H

|
H H

Figure 1.14. Polymerization of poly(ethynyl acetylene).

26

To overcome this problem, Langmuir-Blodgett (LB) films ofthe diacetylene
monomer were prepared by attaching the monomers to a calcium fluoride surface
through Short tethers (Figure 1.15).71 The first polyacetylene strand was
photopolymerized by UV irradiation, followed by thermal treatment to form the
second. Characterization of the resulting films indicated that only small acenic units
were obtained, most likely due to the fact that the LB molecular packing density of

the monomer was lower than expected.

H H H \$%%\%%\as
II II II II || ||
hv ‘ A
——b ——->
or
e-beam

 

 

 

 

 

 

 

 

 

 

 

/ /////// /////// / // //////

Figure 1.15. Polymerization of Langmuir-Blodgett films of diacetylenes.

Other routes to linear acene materials utilize iterative chain growth reactions.
One example, shown in Figure 1.16. exploits the chemistry used to produce 9,10-
dihydroacenaphthalene from benzocyclopropene.72 Bicycloproparenes were reacted

with silver ions in chloroform to yield acene precursors, which upon aromatization

27

400* —-* @0009

DDQ

benzene
polyacene +—— . I

Figure 1.16. Polymerization of cycloproparenes.

using DDQ or other suitable oxidizing agents, affords polyacene. The insolubility of
the acene precursors is the major drawback to this method.

The synthesis ofpolyacenic materials. which are defined as three-dimensional
networks that contain some amount of acene units, usually involve the pyrolysis of a
suitable precursor. The precursor polymers most commonly used are phenol-
formaldehyde and cellulOse resins.73’74 It is proposed that upon high temperature
treatment, the precursors eliminate water and hydrogen to form aromatic materials.
A more accurate description of the “curing” process is graphitization since the
statistical probability of producing linear acenes is low. The characterization data
from the charred materials could also be interpreted as the presence of graphite or

three-dimensional aromatic networks.

28

Using an analogous reaction, we found that biS(methoxymethyl)benzenes
undergo acid-catalyzed polymerization to yield network structures. The idealized
chemical structure for the polymerization of bis(methoxymethyl)benzenes is a linear
polyacene precursor, that after elimination of H2, Should yield polyacenes with an
extended p electron system. To obtain the ideal linear acene, the synthetic strategy
employed must impart some degree of control over the outcome of the resulting
structure. Defects and side reactions quickly lead to less than ideal structures. In
Chapter 3, the methods we used to control the structure of the network and the
characterization of these materials are described. We also report the preparation of a
methoxymethyl-substituted poly(p-phenylene) that appears to convert to an acene

cross-linked polyphenylene when reacted with concentrated sulfuric acid.

1.2.3. Polyacetylene — Polyll-trimethylsilyl-l-propyne]

Syntheses of polyacetylene. The most popular synthetic method for
preparing linear polyacetylenes with high crystallinity is the polymerization of
acetylene with a Ziegler-Natta catalyst.75 High quality polyacetylene films are
prepared from homogeneous catalyst mixtures consisting of Ti(OR)4 and R’3Al.
Titanium halides and other metal complexes, such as vanadium-based organometallic
compounds, have also been used. Polyacetylene can also be prepared via ring-opening
polymerization of cyclooctatetraene.76 The monomer is polymerized neat using a

Grubbs-type catalyst to give free-standing films. This approach has been extended

29

to the polymerization of trimethylsilyl substituted cyclooctatetraene to afford soluble
polyacetylenes which have been used to construct solar cells and layered structures.

Polyacetylene can be Obtained through indirect methods as well. Some of
the earliest attempts at preparing polyacetylene involved the thermal elimination Of
small molecules from soluble precursors. Poly(vinyl alcohol) and poly(vinyl chloride)
were heated, evolving water and HCI, respectively, to form polyacetylene. However,
the thermal elimination was Often inefficient and led to irregular structures. The
isomerization of poly(benzvalene), prepared through ring-opening polymerization,
is an alternative precursor approach to polyacetylene.77 The precursor rearranges by
heat or light to polyacetylenic materials.

Syntheses of substituted polyacetylenes. Substituted acetylene monomers
can be polymerized using suitable transition-metal catalysts.7S Metathesis
polymerization using Group 5 and 6 transition metal complexes (namely Nb, Ta,
Mo, and W) is believed to proceed via metal carbenes. According to the proposed
mechanism (Shown in Figure 1.17), propagation proceeds through the reaction of a
metal carbene with an acetylenic monomer to give a metallocyclobutene, followed
by ring opening to regenerate the metal carbene. This mechanism resembles that of
the ring-opening metathesis polymerization (ROMP) of cycloolefins. Some of the
catalysts that have been effective include: 1) the chlorides of Nb, Ta, Mo, and W; 2)

1:1 mixtures of those metal chlorides with suitable organometallic cocatalysts; and

30

C:C C—C
1. ——-.__ t. ——~
a... 1:1. I I
WCZM MC—‘M Wfi llvll
: l 1 ~———*
ego 0:0 C—C

Figure 1.17. Initiation and propagation of acetylene polymerization.

3) catalysts obtained by UV irradiation of CC I 4 solutions of MO and W hexacarbonyls.
In general, the Group 6 metal complexes (Mo and W) polymerize monosubstituted
and less sterically hindered disubstituted acetylenes, while Group 5 metal complexes
(Nb and Ta) are used to polymerize sterically hindered disubstituted acetylenes. With
proper choice of catalyst and solvent systems, polymers with molecular weights
exceeding 106 can be obtained from substituted acetylene monomers.

The polymerization of l-trimethylSinI-l-propyne, a sterically hindered
disubstituted acetylene, was accomplished using a co-catalyst system consisting of
tantalum pentachloride and triphenyl bismuth.78‘80 Copolymerization of 1-
trimethylsilyl-l-propyne with functional acetylenic monomers was also catalyzed
by the same co-catalyst system to yield polymers with high molecular weights (Chapter

4).

31

1.3. References

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

2)Elsenbaumer, R. L.; Jen, K. Y.; Oboodi, R. Synth. Met. 1986, 15, 169-174.
3)Edwards, A.; Blumstengel, S.; Sokolik, 1.; Yun, H.; Okamoto, Y.; Dorsinville, R.
Synth. Met. 1997, 84, 639-640.

4)Segura, J. L. Acta Polym. 1998, 49, 319-344.

5)Leising, G.; Tasch, S.; Meghdadi, F.; Athouel, L.; Froyer, G.; Scherf, U. Synth.
Met. 1996, 81, 185—189.

6)Kim, H. K.; Ober, C. K. J. Macromol. Sci-Pure Appl. Chem. 1993, A30, 877-
897.

7)Kern, W.; Wirth, H. O. Makromol. Chem. 1959, 31, 154-164.

8)Rehahn, M.; Schluter, A. D.; Wegner, G.; Feast, W. J. Polymer 1989, 30, 1060-
1062.

9)Tour, J. M.; Stephens, E. B. J. Am. Chem. Soc. 1991, 113, 2309-2311.

10)Tour, J. M.; Stephens, E. B.; Davis, J. F. Macromolecules 1992, 25, 499-500.
11)Tour, J. M. Adv. Mater. 1994, 6, 190-198.

12)Huber, J.; Scherf, U. Macronml. Rapid Commun. 1994, 15, 897-902.
13)Vahlenkamp, T.; Wegner, G. Macromol. Chem. Phys. 1994, 195, 1933-1952.

14)Gin, D. L.; Conticello, v. P. Trends Polym. Sci. 1996, 4, 217-223.

32

15)McCarthy, T. F.; Witteler, H.; Pakula, T.; Wegner, G. Macromolecules 1995, 28,
8350—8362.

16)Rehahn, M.; Schluter, A. D.; Wegner, G. Makromol. Chem, Macromol. Chem.
..Phys 1990,191,1991-2003.

17)Scherf, U.; Mullen, K. The synthesis ofladder polymers, 1995; Vol. 123, pp 1-
40.

18)Kertesz, M.; Hughbanks, T. R. Synth. Met. 1995, 69, 699-700.

19)Petekidis, G.; Vlassopoulos. D.; Fytas, G.; Rulkens, R.; Wegner, G.
Macromolecules 1998, 31, 6129-6138.

20)Park, K. C.; Dodd, L. R.; Levon, K.; Kwei, T. K. Macromolecules 1996, 29,
7149-7154.

21)Witteler, H.; Lieser, G.; Wegner, G.; Schulze, M. Makromol. Chem, Rapid
Commun. 1993, 14, 471-480.

22)Steuer, M.; Ballauff, M. J. Polym. Sci. Pol. Chem. 1993, 31, 1609-1619.
23)Salem, L.; Longuett-Higgins, H. C. Proc. R. Soc. London, Ser. A 1960, 255,
435.

24)Tanaka, K.; Yata, S.; Yamabe. T. Synth. Met. 1995, 71, 2147-2149.
25)Whangbho, M. H.; Hoffmann, R.; Woodward, R. B. Proc. R. Soc. London, Ser.
.41979,366,23.

26)Kojima, Y.; Matsuoka, T.; Sato, N.; Takahashi, H. Macromolecules 1995, 28,

2893-2896.

33

27)Kojima, Y.; Matsuoka, T.; Takahashi, H. J. Appl. Polym. Sci. 1998, 68, 129-
135.

28)Clough, S. B.; Sun, X. F.; Tripathy, S. K.; Baker, G. L. Macromolecules 1991,
24, 4264-4269.

29)Bock, C. W.; George, P.; Trachtman, M.; Zanger, M. J. Chem. Soc, Perkin
Trans. 1979, 1, 26.

30)Rolland, M.; Bernier, P.; Lefrant, S.; Aldissi, M. Polymer 1980, 21, 1111.
31)Baker, G. L.; Shelburne III, J. A.; Bates, F. S. J. Am. Chem. Soc. 1986, 108,
7377.

32)Masuda, T.; Tang, B. Z.; Tanaka, A.; Higashimura, T. Macromolecules 1986,
19, 1459.

33)Baker, G. L.; Klausner, C. F.; Gozdz. A. S.; Shelburne, J. A.; Bowmer, T. N.
Adv. Chem. Ser. 1990, 663-678.

34)Gozdz, A. S.; Baker, G. L.; Klausner, C.; Bowden, M. J. Proc. SPIE 1987, 771,
18.

35)Nagai, K.; Higuchi, A.; Nakagawa, T. J. Appl. Polym. Sci. 1994, 54, 1207-
1217.

36)Morisato. A.; Shen, H. C.; Sankar. S. S.; Freeman, B. D.; Pinnau, 1.; Casillas,
C. G. J. Polym. Sci. Pt. B-Polym. Phys. 1996, 34. 2209-2222.

37)Koros, W. 1.; Chem, R. T. Separation ofGaseous Mixtures Using Polymer

Membranes; Rousseau, R. W., Ed; Wiley Interscience: New York, 1987.

34

38)Savoca, A. C.; Surnamer, A. D.; Tien, C. F. Macromolecules 1993, 26, 6211-
6216.

39)Huang, R. Y. M.; Feng, X. Sep. Sci. Technol. 1992, 27, 1583.

40)Schmidt, S. L.; Myers, M. D.; Kelley, S. S.; McMillan, J. D.; Padukone, N.
Appl. Biochem. Biotechnol. 1997, 63-5, 469-482.

41)Robeson, L. M.; Langsam, M. Sep. Sci. Technol. 1992, 27, 1245-1258.
42)Masuda, T.; Tang, B.; Higashimura, T. Polym. J. 1986, 18, 565.
43)Shimomura, H.; Nakanishi, K.; Odani, H.; Kurata, M.; Masuda, T.;
Higashimura, T. Kobunshi Ronbunshu 1986, 43, 747.

44)Masuda, T.; Higashimura, T. Silicon Based Polymer Science; Ziegler, J. M. and
Fearon, F. W. G., Ed.; American Chemical Society: Washington, DC, 1990; Vol.
224, pp 641.

45)Arnautov, S. A.; Gavrilenko, V. V. Synth. Met. 1997, 84, 283-284.

46)Shimura, T.; Funaki, H.; Nishihara, H.; Aramaki, K.; Ohsawa, T.; Yoshino, K.
Chem. Lett. 1992, 457-460.

47)Kovacic, P.; Jones, M. B. Chem. Rev. 1987, 87, 357-379.

48)Bringmann, G.; Walter, R.; Weirich, R. Angew. Chem-Int. Edit. Eng]. 1990, 29,
977-991.

49)Fanta, P. E. Synthesis 1974, 9-21.

50)Wirth, H. 0.; Herrmann, F. U.; Hern, W. Makromol. Chem. 1964, 80, 120-140.

35

51)Percec, V.; Hill, D. H. Step-polymerization reactions via nickel- and palladium-
catalyzed carbon-carbon bond formation, 1996; Vol. 624, pp 2-56.

52)Percec, V.; Pugh, C.; Cramer, E.; Okita, S.; Weiss, R. Makromol. Chem,
Macromol. Sym. 1992, 5—1-5. 113-150.

53)Novak, B. M.; Wallow, T. I.; Goodson, F.; Loos, K. Abstr. Pap. Am. Chem. Soc.
1995, 209, 215-POLY.

54)Durairaj, K. Curr. Sci. 1994, 66, 833-838.

55)Saa, J. M.; Martorell, G. J. Org. Chem. 1993, 58, 1963-1966.

56)Kwiatkowski, G. T.; Colon, 1.; Elhibri, M. J.; Matzner, M. Makromol. Chem,
Macromol. Sym. 1992, 54-5, 199-224.

57)Suzuki, M. J. Synth. Org. Chem. Jpn. 1993, 5], 141-156.

58)Tour, J. M.; John, J. A. Abstr. Pap. Am. Chem. Soc. 1993, 206, 43-POLY.
59)Gin, D. L.; Conticello, V. P.; Grubbs, R. H. J. Am. Chem. Soc. 1994, 116,
10934-10947.

60)Gin, D. L.; Conticello, V. P.; Grubbs, R. H. J. Am. Chem. Soc. 1994, 116,
10507-10519.

61)Gin, D. L.; Avlyanov, J. K.; Macdiarmid, A. G. Synth. Met. 1994, 66, 169-175.
62)Kertesz, M.; Frapper, 0; Hong. S. Y.; Lee, Y. S.; Kim, 0. K. Synth. Met. 1993,
57, 4344-4349.

63)Kertesz, M.; Ashertehrani, A. Macromolecules 1996, 29, 940-945.

36

64)Grem, G.; Paar, C.; Stampfl, J.; Leising, G.; Huber, J.; Scherf, U. Chem. Mat.
1995, 7, 2-4.

65)Miller, L. L.; Jozefiak, T. H. Synth. Met. 1988, 27, 8431-8437.

66)Ashton, P. R.; Brown, G. R.; Isaacs, N. S.; Giuffrida, D.; Kohnke, F. H.;
Mathias, J. P.; Slawin, A. M. Z.; Smith, D. R.; Stoddart, J. F.; Williams, D. J. J.
Am. Chem. Soc. 1992, 114, 6330-6353.

67)Ashton, P. R.; Brown, G. R.; Smith, D. R.; Stoddart, J. F.; Williams, D. J.
Tetrahedron Lett. 1993, 34, 8337—8340.

68)Mayer, A.; Meier, H. J. Prakt. Chem-Chem. Ztg. 1997, 339, 679-681.
69)Graham, R. J.; Paquette, L. A. J. Org. Chem. 1995, 60, 5770-5777.
70)Kojima, Y.; Tsuji, M.; Matsuoka, T.; Takahashi, H. J. Polym. Sci. Po]. Chem.
1994, 32,1371-1376.

71)Minakata, T.; Nagoya, 1.; Ozaki, Ml. Synth. Met. 1991, 42, 1501-1508.
72)Billups, W. E.; Luo, W. M.; McCord, D.; Wagner, R. Pure Appl. Chem. 1996,
68, 275-280.

73)Zheng, T.; Zhong, Q.; Dahn, J. R. J. Electrochem. Soc. 1995, 142, L211-L214.
74)Wang, T.; Kang, E. T.; Neoh, K. G.; Tan, K. L. Synth. Met. 1997, 84, 405-406.
75)Shirakawa, H. Synth. Met. 1995, 69, 3-8.

76)Novak, B. M.; Risse, W.; Grubbs, R. H. Adv. Polym. Sci. 1992, 102, 47-72.

37

77)Gibson, V. C. Adv. Mater. 1994, 6, 37-42.

78)Nagai, K.; Higuchi, A.; Nakagawa, T. J. Polym. Sci. Pt. B-Polym. Phys. 1995,
33, 289-298.

79)Nagai, K.; Nakagawa, T. J. Membr. Sci. 1995, 105, 261-272.

80)Masuda, T.; lsobe, E.; Higashimura, T. Macromolecules 1985, 18, 841.

38

Chapter 2

Synthesis and Characterization of Ethylene Oxide
Substituted Poly(p-phenylene)s.

2.1. Introduction

Addition of flexible side chains to rigid polymer backbones has proven to
be an extremely effective procedure for Obtaining tractable and fusible materials.
Not only does this modification enable the utilization of conventional polymer
processing methods, side chain substituted polymers often exhibit properties that do
not exist in the unsubstituted analogs. Some of these properties include pronounced

l

thermochromism and solvatochromism, thermotropic and lyotropic liquid

2-6 7-9

crystallinity, and structural self-assembly.

Most Of the Side chain containing rigid polymers synthesized to date employ
alkyl, alkoxy or phenylalkyl side chains of varying lengths that are chemically
anchored at sites along the polymer backbone.6’10'16 From a structural standpoint,
the most interesting characteristic of these substituted rigid polymers is the presence
oftwo chemically dissimilar building blocks; the non-polar flexible Side chains and
the stiff rt-conjugated main chains. Given the opportunity, these polymers will adopt
structures that produce phase separation of the two fundamental units restricted by

the chemical linkage that binds them together and any additional steric packing

constraints. The net result is that solvent cast films or precipitated powders commonly

39

form layered, semicrystalline structures whereby the hydrophobic side chains appear
as spacers between nested stacks of the main chains.l7'19

Interesting molecular architectures may also be obtained by attaching
hydrophilic Side chains to the polymer backbonezo'23 Short poly(ethylene oxide)
(PEO) Side chains have been utilized to increase the solubility of poly(phenylene)
and poly(thiophene) systemsl’24 An advantage Of using polar side chains is that
the solubility of the substituted polymers is increased in a wider variety of solvents.
The ethylene oxide Side chains also increase the functionality of the polymer systems
by introducing the ability to coordinate with metal ions. Coordination of Li+ occurs
with the lone pair of electrons on oxygen in the ethylene oxide segments. When
lithium salts are incorporated into the flexible Side chains, the polymer system may
possess a high ionic conductivity.24

We have prepared a series of soluble poly(p-phenylene)s substituted with
ethylene oxide side chains using a transition metal catalyzed aryl coupling reaction.
These substituted rigid rod polymers most likely form comb-like or hairy rod-like
structures. As shown schematically in Figure 2.1, we predict that the physical
properties of these polymers will vary as the volume fraction of the side chains is
increased. Polymers with short side chains Should possess properties similar to the
parent PPP, while the properties of the polymers with longer Side chains should begin

to resemble those of poly(ethylene oxide). We also investigated how the addition of

lithium salts into the side chains of these polymers affects the physical properties.

40

 

 

 

  
    

 

 

 

A isotropic
e i
B :
‘3 5
ID 5 9'
o. , an
g s g
"" E ,g o- e- 5
I (x. (I) i
s a 1‘9 .“s’ '
I a U 2 = .
i 'O '0 .53 '
g E 3 '5 ..Z.‘ 3 : 8
0. i 8. g 2 o CL
0 volume fractionside chains 1

Figure 2.1. Structure/property relationships of ethylene oxide substituted PPP’S.

41

2.2. Experimental Section

Materials. Reagent grade nickel chloride hydrate was heated (250 °C) under
vacuum to give a fine. orange powder with constant weight. Triphenylphosphine
was purified by recrystallization from hexanes. Powdered zinc was purified by stirring
in acetic acid, filtered, rinsed thoroughly with diethyl ether, and dried under vacuum.
It was then crushed with a mortar and pestle to increase its surface area. The
polymerization solvent, N,N-dimethylformamide (DMF), was stirred over calcium
hydride, distilled under vacuum, and stored over 3A molecular sieves. Pyridine was
stored over 3A molecular sieves. Unless otherwise specified, all other ACS reagent
grade starting materials and solvents were used as received from commercial suppliers
without further purification.

Characterization. Proton nuclear resonance ('H NMR) spectra were
measured using a Varian Gemini-300 Spectrometer at 300 MHz. All samples were
run at room temperature in CDCI3. Chemical shifts were calibrated using residual
CHCl3 and are reported in ppm (5) relative to tetramethylsilane. Infrared spectra of
polymers were obtained under nitrogen at room temperature on a Nicolet Magna-IR
550 Fourier Transform IR spectrometer. A Hitachi U-4001 UV—visible Spectrometer
was used to Obtain the UV-visible spectra of polymer solutions. Molecular weights
of polymers were determined by gel permeation chromatography (GPC) using a PLgel
20m Mixed A column and a photodiode array detector at room temperature with

THE as eluting solvent at a flow rate 1 mL/min. Monodisperse polystyrene standards

42

were used to calibrate the molecular weights. The concentration of the polymer
solutions used for GPC measurements was 1 mg/mL. Differential scanning
calorimetry (DSC) data were obtained under a He environment using a Perkin-Elmer
DSC 7 instrument at a heating rate of 10 °C/min. The DSC 7 temperature was
calibrated with an indium Standard. DSC results of the polymer samples are the first
heating scan, taken after flash cooling the sample from an isotropic state.
Thermogravimetric analyses (TGA) were performed under nitrogen and air
atmospheres at a heating rate of 10 oC/min on a Perkin-Elmer TGA 7 instrument.
Elemental analyses were performed using a Perkin-Elmer 2400 Series II Analyzer.

2,5-Dichlorobenzoyl chloride (2). A stirred suspension of 2,5-
dichlorobenzoic acid (1) (19.02 g, 99.01 mmol) in thionyl chloride (SOCIZ, 100 mL)
was refluxed overnight. The excess SOCl2 was removed by distillation at atmospheric
pressure. The crude product was purified by vacuum distillation at 70-74 °C/40
mtorr to yield 18.94 g (90%) Of2.5-dichlorobenzoyl chloride (lit.25 bp 81 °C/2 torr).
1H NMR 8: 7.4 (d. 2H), 8.0 (t, 1H).

Monomer synthesis: General Procedure. The alcohol (3a-3h, 37 mmol)
and acid chloride 2 (5.16 g, 24.6 mmol) were dissolved in pyridine (50 mL) and
gently heated overnight. Upon cooling, methylene chloride (100 mL) was added and
the solution was washed three times with 50 mL of 1N HCl (aq) and once with 50
mL of H20. The organic layer was collected, dried with magnesium sulfate, filtered,

and the solvent removed with a rotary evaporator. The compounds were purified by

43

recrystallization (4a), vacuum distillation (4b-4d), or column chromatography (4e-
4h). The monomers were passed through a Short pad of activated alumina inside a
drybox prior to use.

Methyl-2,5-dichlorobenzoate (4a). Compound 43 was purified by
recrystallization from methanol to yield 4.74 g (94%). mp 35 °C (DSC) (lit.26 mp
37-40 °C). 1H NMR 8: 3.9(S, 3H), 7.4(d, 2H), 7.8(t, 1H).

3-Oxybutyl-Z,5-dichlorobenzoate (4b). Compound 4b was purified by
vacuum distillation at 105 °C/80 mtorr to yield 5.63 g (92%). 1H NMR 8: 3.4(s, 3H),
3.7(t, 2H), 4.45(t, 2H), 7.4(d, 2H), 7.8(t, 111). Anal. Calcd. for C,0H,0CI,O,: C,
48.22; H, 4.05; CI, 28.47; O, 19.27. Found: C, 48.26; H, 4.03.

3,6-DioxyheptyI-2,5-dichlorobenzoate (4c). Compound 4c was purified by
vacuum distillation at 170 00130 mtorr to yield 5.98 g (83%). 'H NMR 8: 3.4(s,
3H), 3.55(t, 2H), 3.7(t, 2H), 3.85(t, 2H), 4.5(t, 2H), 7.4(d, 2H), 7.8(t, 1H). Anal.
Calcd. for C12H14C1204: C, 49.17; H, 4.81; CI, 24.19; O, 21.83. Found: C, 49.01; H,
4.50.

3,6,9-Trioxydecyl-2,5-dichlorobenzoate (4d). Compound 4d was purified
by vacuum distillation at 195 °C/200 mtorr to yield 6.55 g (79%). 1H NMR 6: 3.35(s,
3H), 3.5(t, 2H), 3.65(m, 6H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H), 7.8(t, 1H). Anal.
Calcd. for CMHISCIZOS: C, 49.87; H, 5.38; CI, 21.03; O, 23.72. Found: C, 49.75; H,

5.39.

44

Monomer 4e. Compound 4e was purified by column chromatography (silica
gel) with hexane/ethyl acetate (2:1) as the eluting solvent to yield 11.06 g (86%). lH
NMR 8: 3.35(s, 3H), 3.5(t, 2H), 3.65(b, 22H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H),
7.8(t, 1H). Anal. Calcd. for C22H34C1209: C, 51.47; H, 6.67; CI, 13.81; 0, 28.05.
Found: C, 51.08; H, 6.68.

Monomer 4f. Compound 4f was purified by column chromatography (silica
gel) with hexane/ethyl acetate (2:1) as the eluting solvent to yield 15.12 g (85%). 1H
NMR 6: 3.35(s, 3H), 3.5(t, 2H), 3.65(b, 42H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H),
7.8(t, 1H). Anal. Calcd. for C32H54CIZOM: C, 52.39; H, 7.42; CI, 9.66; O, 30.53.
Found: C, 51.15; H, 7.17.

Monomer 4g. Compound 4g was purified by column chromatography (Silica
gel) with hexane/ethyl acetate (2:1) as the eluting solvent to yield 18.39 g (81%). 1H
NMR 5: 3.35(s, 3H). 3.5(t, 2H), 3.65(b, 58H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H),
7.8(t, 1H). Anal. Calcd. for C40H70C12018: C, 52.80; H, 7.75; CI, 7.79; O, 31.65.
Found: C, 51.04; H, 7.15.

Monomer 4h. Compound 4h was purified by column chromatography (silica
gel) with hexane/ethyl acetate (2:1) as the eluting solvent to yield 32.07 g (60%). 1H
NMR 8: 3.35(s, 3H), 3.5(t, 2H), 3.65(b, 174H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H),
7.8(t, 1H). Anal. Calcd. for C98H186C12047: C, 53.81; H, 8.57; CI, 3.24; O, 34.38.

Found: C, 53.80; H, 8.73.

45

Nickel Catalyzed Polymerization: General Procedure. Inside a helium
atmosphere drybox, the monomer (1.33 mmol), NiCl2 (8.6 mg, 0.067 mmol),
bipyridine (10.4 mg, 0.067 mmol), triphenyl phosphine (34.9 mg, 0.134 mmol), zinc
powder (0.27 g, 4.12 mmol), and DMF (1 mL) were added into a 100 mL round
bottomed flask. The flask was then fitted with a reflux condenser capped by a rubber
septum. After removal from the drybox, the system was connected to an argon bubbler
through the rubber septum and placed in an oil bath. The solution was stirred
magnetically and the mixture was heated to 80 oC. The green solution became brown
after 10 minutes, eventually turning deep red-brown after 30 minutes. The mixture
was stirred overnight. Upon cooling, the solution was transferred into a flask
containing 200 mL Of2N HCl/MeOH (1 :1) and stirred until the excess zinc dissolved.
The polymer was extracted from the HCl/MeOH mixture with methylene chloride (3
x 50 mL). The combined organic layers were dried with MgSO4, filtered, and the
solvent removed under reduced pressure to yield a pale yellow film.

Preparation of polymer/lithium salt composites. Polymer/Li+ composites
were prepared from polymer and lithium perchlorate (LiClO4). Acetonitrile (6-8
mL) was added to a vial containing the polymer sample (0.5 g) and the appropriate
amount of LiClO4. The mixture was stirred until a homogeneous solution was
obtained. The solution was concentrated by solvent evaporation at 70 °C, followed

by drying under vacuum for 24 h. The composites were further dried under high

46

vacuum at 120 °C for 2h. Polymer composites from 5d and 5e were prepared with
O:Li ratios of 64:1, 32:1, 16:1, 8:1, 4:1, and 2:1 for each polymer. The polymer

composites were stored in a dry, inert atmosphere.

47

2.3. Results

Monomer synthesis. The monomers are easily synthesized from the readily
available 2.5-dichlorobenzoic acid in ~85% yields (Figure 2.2). This route offers
the flexibility of preparing a series of monomers by simply interchanging various
ethylene glycol monomethyl ethers. All of the alcohols used are commercially
available, however the oligomeric alcohols 3e-3h used to prepare monomers with
long Side chains are not monodisperse, which leads to sets of monomers with Slightly
differing side chain lengths. The series of monomers was purified by different methods
depending upon their physical properties. While monomer 4a was recrystallized
from methanol, short side chain monomers 4b-4d with lower boiling points were
vacuum distilled to yield clear oils. Since the monomers with longer Side chains

thermally decomposed before distilling, monomers 4e-4h were purified by column

chromatography.
0' O C' O CH,(OCHZCH2)mOH
OH 60012 C, 3(a-h) © O(CH20H20)mCH,
—» ———>

pyridine

Cl Cl C'
a: m=0

1 2 b: m=1 4(a-h)
c: m=2
d: m=3
e: m=7 (avg.)
f: m=12 (avg)
g: m=16 (avg)
h: m=45(avg.)

Figure 2.2. Synthetic route to monomers 4a-4h.

48

The melting points of the series of monomers as measured by DSC are Shown
in Figure 2.3. Short side chain monomers (4a-4c) exhibited higher melting points,
presumably dominated by the p-stacking of the aromatic moiety. For monomer 4d,
no melting point was observed above -—150 °C, the low temperature limit ofthe DSC

apparatus. AS the lengths of the side chains are increased, the observed melting

60

 

40

30

10

Temperature (°C)

 

 

 

0 1 0 20 30 40 50

Side chain ethylene oxide units (m)

Figure 2.3. Melting points of monomers 4a-4h.

49

points increase to where the melting point of monomer 4h nearly approaches that of
PEO. The trend in the melting points of monomers 4a-4h is a strong indication that
increasing the Side chain lengths attached to the substituted polyphenylenes Should
affect the physical properties of the polymers.

Polymerization. The monomers were polymerized utilizing a nickel-
catalyzed aryl coupling reaction to yield substituted poly(p-phenylene)s 5a-5h (Figure
2.4). This coupling reaction was originally proposed as an inexpensive aryl coupling
reaction to produce substituted biphenyls.27 Due to the reported high yields, lack of

terminating Side reactions, and tolerance to a variety of functional groups, it was

 

O O
OICHZCHZOMCHCI NiClz, Zn, PPh, 0(CH2CH20)mCH3
bipyridine, DMF
CI 0 Cl ¢
100 °C, 8 hr n
4(a-h) 5(a-h)

Figure 2.4. Nickel catalyzed polymerization.

also found to be an efficient method for the polymerization of dichloro aromatic
compoundszg'3 1 The desired all para ring linkages in the resulting polymers are
ensured by starting with para dichloro-substituted aryl monomers. The proposed

catalytic cycle is shown in Figure 2.5. The active catalytic species (N10) is generated

50

NiIICI2 L : triphenyl phosphine

Zn,L
Zn012 ch'z ArCl
\ N10L3 /
Zn L
ArNIIICIL2
NIICIL,
C3 /.,.
L (Ar)2Ni"|CIL2 ArNi'La ZnCIZ
ArCl

Figure 2.5. Catalytic cycle of nickel mediated aryl coupling.

51

by the reduction of nickel(ll) chloride using powdered zinc. The Nio species then
undergoes an oxidative addition with an aryl chloride followed by transmetallation
with another monomer unit to form a diaryl nickel(II) complex. In order for reductive
elimination to occur and form the coupled product, the aryl ligands must take a cis
conformation on the square planar nickel complex. The bipyridine acts as a bidentate
ligand that forces the aryl groups into a cis orientation to facilitate reductive
elimination. The catalyst is then regenerated by zinc in the highly reducing
environment.27

The condition of the catalyst can be monitored as the reaction proceeds due
to a built-in indicator. The reaction takes on a blood red color shortly after heating
due to the formation of N10. Contamination of the polymerization by oxygen and
proton sources must be prevented. The active N10 species is consumed by oxidation
forming black NiO, and proton sources reduce the growing polymer, terminating the
chain. Either contaminant leads to the loss of the deep red reaction color. Potential
proton sources in this reaction are moisture absorbed in the solvent (DMF) and
monomers, and high boiling alcohols contaminating the monomer after esterification.
The solvent was dried by passing it through a pad ofoven-dried silica gel inside the
dry box and storing it over 3A molecular sieves. Any possible contamination of
residual alcohol in the monomers was removed by running monomer solutions through
a Short pad of oven dried silica gel using toluene as solvent inside the dry box. The

toluene was then removed under reduced pressure and the monomers were stored

inside the dry box.

52

Polymer characterization. The physical properties of the resulting polymers
ranged from powders to tacky solids depending upon the side chain length and degree
of polymerization. The molecular weights of the polymers as measured by GPC

relative to polystyrene standards are shown in Table 2.1. The obtained degrees of

Table 2.1. Representative molecular weights of polymers 5a-5h.

 

 

poflmer M MW 391 DP in)

5a 4700 8900 1 .9 35
5b 17500 46600 2.6 98
SC 15000 41000 2.7 68
5d 14600 63000 4.3 55
5e 18200 59500 3.3 40
St 14000 26000 1 .9 22
59 18000 35000 1 .9 21
5h 9400 23300 2.5 5

 

 

 

 

 

 

 

polymerization for the series of polymers were not consistent however (Figure 2.6).
The molecular weights of the short side chain polymers are limited by insolubility at
higher degrees of polymerization, while the molecular weight ofthe polymers with
longer side chains are limited by either steric hindrance, or a reduction in monomer
concentration as the aryl ring becomes a smaller part of the monomer. GPC traces
also indicate an extremely high molecular weight fraction for polymers 5e-5h. It is
believed that these polymers are lightly cross-linked aggregates. The long chain
alcohols used to prepare the monomers most likely contained some diol that when
reacted with 2 equivalents of acid chloride, formed a cross-linkable monomer. It is

difficult to analyze for the presence Of this contaminant since its amount is small and

53

Degree of polymerization (n)

120

 

100 f.
80
60 -

401 o

20 9.

 

 

 

0 1 0 20 3O 40 50

Side chain ethylene oxide units (m)

Figure 2.6. Degrees of polymerization of polymers 5a-5h.

54

it can not be readily distinguished from the monomer by 1H NMR integration. The
amount of solvent used in the polymerizations also affected the resulting molecular
weight of the product. Since dilute polymerizations yielded lower molecular weight
polymers than concentrated polymerizations, the minimum amount of solvent was
used to dissolve the monomer and catalytic reagents.

The conformation of the aromatic rings in the polymer backbone can be
determined from the UV/Vis absorption spectra. Twisting of the rings as a result of
steric interactions of the side chains limits the n-conjugation of the phenylene
backbone. This is evident by the Amax observed for the polymers. Regardless of
molecular weight or length of side chain, the polymers’ maximum intensity wavelength
absorption was at 317 nm, corresponding to the benzoyl chromophore. With twisting
of the rings, the Side chains randomly extend outward from the rotated backbone
creating comb-like or hairy rod structures. As shown in Figure 2.7, polymer solutions
excited at 317 nm strongly fluoresce from 350 to 450 nm. The iridescent blue emission
can also be seen when polymer films are irradiated with a black lamp.

There is a multitude of effects directly related to the side chain placement
and chemical architecture that have a tremendous impact on the final structural
ordering and phase behavior of these polymers (Figure 2.8). For example, it has
been observed that varying the degree of regioregularity of the monomeric repeat
units can alter many physical properties. In the case of 2,5-disubstituted poly(p-

phenylene)s, there are considerable steric interactions that frustrate the formation of

55

Emission

 

 

 

 

300 350 400 450 500

wavelength (nm)

Figure 2.7. Fluorescence spectra of a solution of
polymer 5d in methanol excited at 317 nm.

56

 

Figure 2.8. Comparison of the orientation of repeat units
in 2,5-disubstituted and 2-Substituted PPP’S.

a planar main chain conformation. Due to the twisting of adjacent aromatic rings to
relieve steric interactions between substituents, these polymers are not extensively
rt-conjugated.32 Since the monomer unit is symmetric, these polymers can organize
by stacking and tend to form ordered structures. However, the repeat units in
monosubstituted polyphenylenes. such as Sa-Sh, can be coupled in three distinct
orientations: head-to-tail, head-to-head, and tail-to-tail. Although the orientations
of the repeat units in the resulting polymers may be random or highly regular,
considerably less steric interaction exists between adjacent rings and the degree of
twisting is less than that in 2,5-disubstituted PPP’S. We believe that polymers 5a-5h
form more disordered structures due to randomness of the ring twisting and the

positions of the Side chains attached to the polymer backbone.

57

Thermogravimetric analyses of the series of polymers under nitrogen Show
the onset ofa weight loss around 300 °C due to the decomposition ofthe Side chains.
Depending upon the volume fraction of polyphenylene backbone in the original
sample, degradation results in a 10-50 % char yield at 600 °C. As shown in Figure
2.9, the temperature at which volatile by-products begin to evolve is dependent upon
the length ofthe side chain. Since the polymers with short side chain substituents do
not melt before decomposing, it is also possible that these solid samples tend to
degrade at slightly higher temperatures.

The phase transitions ofthe series of polymers were measured by DSC. The
thermograms shown in Figure 2.10 are the first heating scans taken after flash cooling
the sample from 250 °C. DSC measurements on polymers 5a-5c Show no thermal
transitions below the decomposition temperature (< 300 °C). Polymers Sd-Sg with
longer Side chains exhibit glass transitions (Tg). Plotted in Figure 2.1 l , the measured
glass transition temperatures approach the Tg of PEO, ~ -60 °C, indicating that the
properties of the substituted polymers are approaching those of PEO. By definition,
a glass transition occurs when molecular motion within a polymer is no longer
restricted and results in a second order transition, softening the sample. AS shown
schematically in Figure 2.12. the Tg’s ofthese polymers occurs when the side chains
become flexible enough to allow the hairy rods to pass back and forth across each

other. This is analogous to the feel ofa wire bristled brush versus a soft paint brush.

58

Weight %

 

 

 

 

 

100 ~

80 .

60 ~
a

40 ” b
C

20 .

0

0 200 400 600 _ 800

Temperature (°C)

Figure 2.9. Thermogravimetric analysis of a) 5b, b) 5c,
and c) 5d in nitrogen heated at 10 °C/min.

59

Heat flow ( endo -> )

 

PEC)

 

5h

 

59

 

5f

 

5e

5d

 

 

 

50

 

 

 

 

-100

Figure 2.10. DSC results for polymers 5c-5h, and PEO.

-75

-25 0

Temperature (°C)

60

25

50

75

4o

 

30 O

20

10«

Temperature (°C)
é-s

O

 

 

 

0 5 10 15 20

Side chain ethylene oxide units (m)

Figure 2.11. Glass transition temperatures of polymers 5d-Sh.

61

It.

Figure 2.12. Schematic representation of rigid rod PPP backbone
substituted with flexible ethylene oxide side chains.

62

70

 

 

 

 

60 ~ 0
PEG
8 50 - .
Ow
5h
2 4o
3
H
t
E
o
I- 20
o
10 ” 59
o
o 50 100 150 200 250

Side chain ethylene oxide units (m)

Figure 2.13. Melting points of polymers Sg-Sh, and PEO.

63

Crystallization of the side chains is observed in polymers 5g and 5h. It is
likely that the non-uniform side chain lengths for the polymers with m>7 hinder
crystallization. Plotted in Figure 2.13, the observed melting points appear to approach
that of PEO. The crystallinity was also detected by observing birefringence using
polarized optical microscopy.

Liquid crystalline behavior was Observed for polyphenylenes 5b-5f. The
term liquid crystal refers to the phase that lies between the rigidly ordered solid
phase where mobility of individual molecules is restricted and the isotropic phase
where molecular mobility and a general lack of molecular order exists. When the
rod-like polymer molecules are aligned either through mechanical or other means,
the sample will polarize light in the ordered domains. Rod-shaped organic molecules
typically align in a nematic phase. The molecules within a nematic mesophase are
organized in one dimension, with the chains lying parallel to each other at equilibrium.
Polymers 5b and 5c exhibited liquid crystalline behavior when the samples were
mechanically sheared at elevated temperatures. These polymers however began to
thermally degrade before becoming isotropic. Polymers 5d and 5e exhibited liquid
crystallinity at room temperature when mechanically sheared, but became isotropic
when heated slightly. Polyphenylene 5f showed signs of temporary liquid crystallinity
immediately after mechanical shearing at room temperature, but returned to an
isotropic state within a few seconds indicating that the interactions aligning the rod-

like molecules are not as strong as those for short side chain polymers. When the

64

crystalline polymers 5g and 5h were melted, they showed no signs of liquid crystalline
behavior when mechanically sheared. Because the low degrees of polymerization
for polymers 5f—5h may not be high enough to think of the polymer molecules as
being truly rod-like, it is likely that the polymers’ ability to align in a nematic domain
is greatly hindered.

The effect of lithium salts on the physical properties of ethylene oxide side
chain substituted PPP was studied by adding various amounts of lithium perchlorate
to polymers 5d and 5e. In PEO, Li+ coordination with the oxygen atoms in adjacent
polymer chains acts as transient cross-linking, increasing the Tg of the polymer as
the amount of added lithium salt is increased. The coordination stabilizes the
networks, therefore more energy is needed to overcome the cross-linking. The
opposite appears to happening in the case of the composites prepared with polymer
5d. As the amount of lithium perchlorate added into the polymer is increased, the
measured Tg’s decrease as shown in Figure 2.14. Since the Short m=3 side chains
are relatively rigid without any added lithium salts, we believe that the lithium
perchlorate essentially dilates the network lowering chain-chain interactions.
Interestingly, another trend is seen in the Tg’s of the composites prepared from polymer
5e (Figure 2.15). At low Li+ loadings, the observed Tg’s initially increase, indicating
that the added lithium perchlorate presumably organizes the chains through
coordination and transient cross-linking. At higher Li+ loadings however the same

trend aS seen in Figure 2.14 appears. At these high loadings the increased amount of

65

 

 

 

 

40
30 l
A I
g 20 ~-
q’ I
'5 10 ~
1“
‘3
a: 0
O. I
E
B -10 + .
-2o '
-30 . . 4 .
0 0.1 0.2 0.3 0.4 0.5 0.6

Li:0

Figure 2.14. Effect of Li+ salt on the Tg of polymer 5d.

66

 

 

 

20 *
I

15 I
A I
e 5 ~
3
‘5 0
h I
8 -5 f
E, -10
a

-15 I-

-20

-25

 

0 0.1 0.2 0.3 0.4 0.5
Li:O

Figure 2.15. Effect of Li+ salt on the Tg of polymer 5e.

67

lithium salt again induces disorder, lowering the Tg. In this sense, high loadings of
lithium perchlorate can be thought Of as a plasticizer in the polymer sample. If a set
of composites were prepared from a high molecular weight sample of polymer 5h,
one would expect the observed Tg’s to increase Similarly to PEO since the properties

of the polymer 5h begin to mirror those of PEO.

68

2.4. Discussion

Structure/property relationships. As predicted, the physical properties of
polymers 5a-5h exhibit a strong dependence upon the length of the oxyethylene side
chains attached to the polymer backbone. The structure/property relationships are
summarized in the schematic phase diagram Shown in Figure 2.16. The side chains
in polymers 5a-5h can be thought of as ropes attached to a wall. The chemical
connectivity ofthe chain to the PPP backbone limits the mobility of the side chains.
It was observed that the short chains could not aggregate, and only the longest side
chains showed Signs of order. Presumably the long chains are less affected by the
limited mobility.

The random orientation of the repeat units in the rigid polymer backbone
also prevents organization. Whereas stereoregular polymers, which usually have
symmetric repeat units, exhibit reasonably well-ordered structures, stereoirregular
polymers typically exhibit relatively poor degrees of order. The position ofthe side
chain relative to that of another on an adjacent ring and the degree of twisting along
the polymer backbone resulting from the steric repulsion of the ring substituents
creates a disordered “hairy rod-like” structures. Therefore, ordering of the side chains
could only be expected for longer side chains which are not affected by the chemical

connectivity.

69

 

n Iaia$am

 

 

 

    
  
   
 

7

o 0(CH20H20)mCH3
[ i i n

 

 

 

isotropic

temperature —>

PPP
amorphous
powder

 

PEO

 

 

0 1 2 3 7 12 16 45
side chain length (m)

Figure 2.16. Schematic phase diagram of ethylene oxide substituted PPP’S 5a-5h.

70

 

A drawback to the assessment of the role of Side chain length on the overall
properties of the series of substituted polymers is the inconsistency of side chain
length for polymers 5e-5h. The long chain alcohols used in the synthesis of monomers
4e-4h are not of uniform length. Phase transitions such as crystallization may be
hindered due to variance in the side chain lengths. A proposed synthesis of exact
length long chain ethylene glycol monomethyl ethers is suggested in Figure 2.17.
This route involves the iterative coupling of short chain ethylene glycol monomethyl
ethers and deprotonated tetraethylene glycol, protected with a tetrahydropyran (THP)
group. The first step involves tosylation of the commercially available, exact length
ethylene glycol monomethyl ethers 3a-3d. Tetraethylene glycol can then be mono-
protected by reaction with dihydropyran and deprotonated by reaction with sodium

hydride. A Williamson-type ether synthesis gives a coupled product that after

TsCl
CH3(OCH2CH2)XOH —+ CH3(OCH2CH2)XOTS
X = 0-3 CH3(OCH2CH2)mOH
NaH m = 4-7

Figure 2.17. Synthetic route to exact length ethylene glycol monomethyl ethers.

71

deprotection, yields the exact length alcohols with m = 4-7. To obtain chain lengths
greater than seven ethylene oxide repeat units. the products from the first coupling
can be recycled back into the reaction sequence to yield alcohols with 8 to 11 ethylene
oxide units. It may be challenging to purify the alcohols however. Since the boiling
points of the products are greater than 300 °C, it is likely that the alcohols will
thermally degrade before distilling. The differences in polarity of the alcohols and
diols could be useful in the separation and purification by column chromatography
or HPLC.

Another major problem hampering the evaluation of the effects of side chain
length on the properties of polymers 5a-5h is the inconsistency of the degrees of
polymerization between samples. Not only are the Side chain lengths being increased,
but the backbone lengths are generally decreasing in size. We believe that the degrees
of polymerization vary for several reasons. First, the addition of short side chains to
the polyphenylene backbone initially results in higher degrees of polymerization due
to the increased solubility of the substituted polymer. However, as the side chain
length is increased, the molecular weights steadily decline. Initially we believed
this to be due to electronic effects caused by the side chains coordinating to the
catalytic sites. Addition of diglyme to the polymerization mixture did not result in
lower molecular weights however. Another possibility is the steric blocking of the
catalytic Site by the side chains. The results from this effect Should eventually become

constant; beyond some Side chain length, the total length of the chain becomes

72

irrelevant. It appears that the decrease in the measured molecular weights is actually
due to a dilution of the reactive aryl chloride Sites in the monomers as the length of
the Side chain is increased. Although the monomer/catalyst ratio was held constant,
the amount of aryl chloride Sites in the polymerization decreased as the total volume
remained constant. Even polymerizations of monomer 4h attempted using a minimum
amount of added solvent resulted in low molecular weight products.

A potential route to a series of polymers with uniform backbone lengths
involves the large scale synthesis of a pre-polymer. As shown in Figure 2.18, a
butoxy-substituted monomer capable Of yielding high molecular weight polymer could
be polymerized in high yield. After hydrolysis of this polymer, a carboxylic acid
substituted pre-polymer could be obtained. Portions of this polymer could then be
re-esterified by two possible routes. The first route involves the reaction of the
carboxylated polymer with thionyl chloride, followed by esterification with the desired
alcohol (3a-3h). The other route involves the direct esterification of the pre-polymer
with the desired alcohol. These routes do present several challenges however. The
insolubility of the pre-polymer could hinder high reaction yields. Presumably as the
amount of side chains grafted unto the polymer backbone is increased, the solubility
should also increase. Steric hindrance caused by the attached Side chains may limit
the conversion of esterification. It may also be difficult to determine the extent of

reaction by standard means.

73

 

 

03“ NiC|2,Zn,PPh, OBU 0“
bipyridine, DMF 1) NaOH (aq)
.. O .. - ————~
100 °C, 8 hr n 2)HCI (aq) n
SOCI2
0
CI
ROH,H*
n
ROH "
0
OR

Figure 2.18. Synthetic route to polymers Sa-Sh with uniform backbone lengths.

74

Applications. Mixtures of PEO with lithium salts have been researched in
hopes of developing solid polymer electrolytes (SPE) for rechargeable lithium
batteries. In such a device, the SPE has to work as an ion conducting medium as
well as a separator preventing direct contact between both electrodes. To do this, the
polymer electrolyte requires considerable mechanical strength. One way to achieve
this is to form PEO networks. Another possibility is to build molecular composites
from rigid rod polymers with ethylene oxide side chains attached to the backbone.
The reinforcing effect of the main chains on the disordered ethylene oxide Side chains
should give materials with distinctly improved mechanical properties compared to
pure amorphous PEO systems.24

The UV-visible absorption Spectrum of neutral conjugated polymers can be
modified by varying their conformational structure. In most cases, these optical
changes have been attributed to planar/non-planar conformational transitions of the
conjugated backbone. The nature of the side chains and the environment surrounding
the polymer affect the electronic properties of the polyphenylene backbone. The
strong dependence upon the position and the bulkiness of the substituents seems to
indicate that side chain disorder could be the driving force of these Optical effects.
Non-covalent interactions between alkali metal ions and oxyethylene side chains
could also modify the conformation of the Side chains and induce a twisting of the

main chain. These features are particularly promising for the development of novel

affinitychromic and photochromic sensors3 3 based on the series of polymers 5a-5h.

75

Due to the electronic nature of the rigid backbone and the metal ion
coordinating ability of the side chains, these polymers may also have applications as
energy transfer materials. Luminescent inorganic salts, such as terbium or europium
halides, possess a strong absorption band within the range of the observed fluorescence
of polymers Sa-Sh. The emission from the polymer backbone could excite the
inorganic salts coordinated to the Side chains, resulting in a second lower energy
emission (~ 600 nm). The materials could be tuned to emit specific wavelengths by
utilizing any number of luminescent inorganic complexes, making it an interesting

material for LED devices.

76

2.5. Conclusions

We have prepared a series of substituted polyphenylenes from ethylene oxide
substituted 2,5-dichlorobenzoates utilizing nickel-catalyzed coupling reactions. The
number average molecular weights of the polymers ranged from 3000 to 30,000 g/
mol. Polymers with short side chains are pale yellow powders, but high molecular
weight samples with longer side chains are tacky solids. Thermal analyses and
polarized optical microscopy show that the polymer side chains crystallize, and in
some cases form liquid crystalline phases. The results of the characterization
performed on the title series of substituted poly(p-phenylene)s indicate that the
physical, thermal, and optical properties observed Show a strong dependence on the
length of the pendant side chain and on the length of the rigid backbone. The thermal
stability of the polymers is limited to the stability of the side chains. An intriguing
property of the polymers is the liquid crystalline nature observed after mechanical
shearing at elevated temperatures. Due to the physical properties of these polymers,
they may have practical applications as LED’S, materials for sensing devices, and

rigid building blocks for molecular architectures.

77

2.6. References

1)Levesque, 1.; Leclerc, M. J. Chem. Soc, Chem. Commun. 1995, 2293-2294.
2)Chang, 8.; Han, C. D. Macromolecules 1997, 30, 1670-1684.

3)Steuer, M.; Ballauff, M. J. Polym. Sci. P0]. Chem. 1993, 31, 1609-1619.
4)Majnusz, J.; Lenz, R. W. Eur. Polym. J. 1989, 25, 847-855.

5)Majnusz, J.; Catala, J. M.; Lenz, R. W. Eur. Polym. J. 1983, 19, 1043-1046.
6)Stern, R.; Ballauff, M.; Lieser, G.; Wegner, G. Polymer 1991, 32, 2096-2105.
7)Steiner, U. B.; Caseri, W. R.; Suter, U. W.; Rehahn, M.; Rau, I. U. Langmuir
1994, 10, 1164-1170.

8)Delnoye, D. A. P.; Sijbesma, R. P.; Vekemans, J. A. J. M.; Meijer, E. W. J. Am.
Chem. Soc. 1996, 118, 8717-8718.

9)Kato, T.; Kihara, H.; Uryu, T.; F ujishima, A.; F rechet, J. M. J. Macromolecules
1992, 25, 6836-6841.

10)Rodriguez-Parada, J. M.; Duran, R.; Wegner, G. Macromolecules 1989, 22,
2507-2516.

ll)Schrauwen, C.; Pakula, T.; Wegner, G. Makromol. Chem, Macromol. Chem.
Phys. 1992, 193, 11-30.

12)Chen, T. A.; Obrien, R. A.; Rieke, R. D. Macromolecules 1993, 26, 3462-3463.
13)Chen, S.-A.; Ni, J.-M. Macromolecules 1992, 25, 6081-6089.
l4)Rughooputh, S. D. D. V.; Nowak, M.; Hotta, S.; Heeger, A. J.; Wudl, F. Synth.

Met. 1987, 21, 41-50.

78

15)Corish, J.; Morton-Blake, D. A.; Beniere, F.; Lantoine, M. J. Chem. Soc.-
Faraday Trans. 1996, 92, 671-677.

16)Segura, J. L. Acta Polym. 1998, 49, 319-344.

17)Moore, J. S. Curr. Opin. Solid State Mat. Sci. 1996, 1, 777-788.

18)Fahlman, M.; Rasmusson, J.; Kaeriyama, K.; Clark, D. T.; Beamson, G.;
Salaneck, W. R. Synth. Met. 1994, 66. 123-127.

19)Duran, R.; Ballauff, M.; Wenzel, M.; Wegner, G. Macromolecules 1988, 21,
2897-2899.

20)Lauter, U.; Meyer, W. H.; Wegner, G. Macromolecules 1997, 30, 2092-2101.
21)Allcock, H. R.; O’connor, S. J. M.; Olmeijer, D. L.; Napierala, M. E.;
Cameron, C. G. Macromolecules 1996, 29, 7544-7552.

22)Pei, Q.; Yang, Y. J. Am. Chem. Soc. 1996, 118, 7416-7417.

23)Aliotta, E; Di Marco, G.; Fontanella, M. E.; Lanza, M. J. Phys. Condes. Matter
1998, 10. 545-556.

24)Lauter, U.; Meyer, W. H.; Enkelmann, V.; Wegner, G. Macromol. Chem. Phys.
1998, 199, 2129-2140.

25)Cambie, R. C.; Coulson, S. A.; Mackay, L. G.; Janssen, S. J.; Rutledge, P. S.;
Woodgate, P. D. J. Organomet. Chem. 1991, 409, 385-409.

26)Chaturvedi, V.; Tanaka, S.; Kaeriyama, K. Macromolecules 1993, 26, 2607-
2611.

27)Colon, I.; Kwiatkowski, G. T. J. Org. Chem. 1986, 5]. 2627.

79

28)Colon, 1.; Kwiatkowski, G. T. J. Polym. Sci. Po]. Chem. 1990, 28, 367-383.
29)Kwiatkowski, G. T.; Colon, 1. Contemporary Topics in Polymer Science 1992,
7, 57-74.

30)Kwiatkowski, G. T.; Colon, 1.; Elhibri, M. J.; Matzner, M. Makromol. Chem,
Macromol. Sym. 1992, 54-5, 199-224.

31)Kwiatkowski, G. T.; Colon, 1.Abstr. Pap. Am. Chem. Soc. 1991, 201, 167—
POLY.

32)Pasco, S. T.; Baker, G. L. Synth. Met. 1997, 84, 275-276.

33)Levesque, 1.; Leclerc, M. Chem. Mai. 1996, 8, 2843-2849.

80

Chapter 3

Synthesis and Characterization of Polyacenic Networks

3.1. Introduction

Perhaps the most important application of acene-like materials is their use
as anodes in lithium ion batteries. 1'5 These rechargeable batteries consist of a metal
oxide cathode (typically C002, MnO2 or V205) and a graphitized carbon anode, both
being able to reversibly intercalate lithium ions into their structure. The anode
materials are derived from petroleum pitches or phenolic resins pyrolyzed at high

temperatures to give partially graphitized structures.6

A charged cell corresponds to
high lithium content in the carbon anode, with the discharge reaction corresponding
to the transfer of lithium ions from the anode to the metal oxide cathode.

An important parameter of lithium ion cells is their capacity, which is limited
by the amount of lithium that can be inserted reversibly into the carbon. Until recently,
it was thought that the limiting stoichiometry for lithium insertion into graphitized
carbon was C 6Li, the value obtained with highly ordered graphite. Recent work7'10
using carbons heat treated at low temperatures has shown that the C6Li ratio can be
exceeded. It is unclear if the excess capacity corresponds to cluster-like lithium

11 or ifthe excess lithium is

doping that fills the void Space in the disordered carbon,
located at the edge of the graphene sheets. 12‘” Recent experimental evidence and

calculations indicate that the increased capacity is correlated with the hydrogen content

81

of the electrode. 1 4 Therefore, it is important to be able to prepare acene-like structures
with a substantial fraction of edges in large quantities.

The high temperatures needed for the pyrolysis allows little control of the
acene-like structural elements that are the precursors to graphite. 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 structure before pyrolysis. In principle, polyacenes could be prepared
that require no thermal treatment for them to be useful electrode materials.

The primary goal of this work is 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 work are relevant
to applications like 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 route targets the synthesis of a methoxymethyl substituted poly(p-
phenylene) that when chemically cross-linked, Should form a two-dimensional
network closely resembling a graphite Sheet. It is believed that this precursor can be
converted to highly ordered graphite at much lower temperatures than previously

obtained.

82

3.2. Experimental Section

Materials. Tetrahydrofuran (THF) and toluene were distilled from powdered
calcium hydride under nitrogen, and then from sodium benzophenone ketyl under
nitrogen. N,N-dimethylformamide (DMF) was stirred over powdered calcium
hydride, distilled under reduced pressure and stored over 3A molecular sieves.
Reagent grade NiCl2 was treated with SOCl2 and then dried at 220°C under vacuum.
Triphenylphosphine (PPh3) was purified by recrystallization from hexanes. Granular
zinc was stirred with glacial acetic acid, filtered, rinsed thoroughly with diethyl ether,
dried under vacuum, and crushed with a mortar and pestle to increase surface area.
All other chemicals were purchased from commercial sources and were used without
further purification except as noted.

Characterization. ‘H and 13C NMR spectra were measured using a Varian
Gemini-3 00 Spectrometer at 300 MHz. All samples were run at room temperature in
CDCI3. The chemical shifts were calibrated using residual CHCl3 and reported in
ppm (8) relative to tetramethylsilane. Infrared spectra were obtained from KBr pellets
under nitrogen at room temperature on a Nicolet Magna-IR 550 Fourier Transform
IR Spectrometer. A Hitachi U-4001 UV-visible spectrometer was used to obtain the
UV-visible spectra of cyclohexane solutions of the polymers. Polymer molecular
weights were determined by gel permeation chromatography (GPC) using a PLgel
20p 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

83

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. Thermogravimetric
analyses (TGA) were performed in nitrogen and air atmospheres at a heating rate of
10 °C/min on a Perkin-Elmer TGA 7 instrument. Elemental anlyses were performed
using a Perkin-Elmer 2400 Series II Analyzer.

Synthesis of bis(methoxymethyl)benzenes. A mixture of hexanes/MeOH
(1:1, 50 mL) was added dropwise 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 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 (1). Compound 1 was prepared in 91%
yield (3.45 g); bp 98 °C/25 mm Hg (111.15 40-45 °C/2 mm Hg). 1H NMR: 8 3.35 (s,

6H), 4.5 (s, 4H), 7.2-7.35 (m, 4H).

84

1,3-Bis(methoxymethyl)benzene (2). Compound 2 was prepared in 85%
yield (3.21 g); bp 110 °C/25 mm Hg (Iit.15 125-130 °C/2 mm Hg). lH NMR: 6 3.35
(s, 6H), 4.4 (s, 4H), 7.25-7.4 (m, 4H).

1,4-Bis(methoxymethyl)benzene (3). Compound 3 was prepared in 89%
yield (3.36 g); bp 105 °C/25 mm Hg (III.15 55-60 °C/2 mm Hg). 1H NMR: 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) 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 0C under vacuum to yield
0.44 g (72%) ofa 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 DDQ immediately
turned dark violet. The flask was fitted with a water condenser and purged with
nitrogen. After refluxing for 48 hrs, 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 hrs to give

0.25 g of a tan powder.

85

1,4-Bis(bromomethyl)-2,5-dibromobenzene (4). Into a 500 mL round-
bottom flask was added 2.5-dibromo-p-xylene (20.0 g, 0.076 mol), N-
bromosuccinimide (28.41 g. 0.160 mol), a catalytic amount ofbenzoyl peroxide (10
mg), and carbon tetrachloride (CC14, 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 CC14. The combined organic layers were washed with
water (3 x 50 mL) and dried with MgSO4. Filtration and evaporation of the solvent
in vacuo resulted in a Slightly yellow solid. Repeated recrystallization from ethanol
yielded 10.57 g (33%) of1,4-bis(bromomethyl)-2,5-dibromobenzene as a fine white
powder. mp 156-157 °C (lit.16 161-162 °C). 'HNMR: 6 7.65 (s, 2H), 4.50 (s, 4H).

1,4-Bis(methoxymethyl)-2,5-dibromobenzene (5). Over a period of one
hour, anhydrous methanol (5 mL) and hexanes (10 mL) were added dropwise via a
pressure equalizing addition funnel to an argon filled 100 mL round bottomed flask
containing a stirred solution ofoil-free sodium hydride (2.1 eq, 0.05] 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 an argon
filled 250 mL round bottomed flask containing 4 (10.30 g, 0.024 mol) dissolved in

warm hexanes (150 mL). The mixture was gently refluxed for 3-4 hours. The cooled

86

reaction mixture was poured into a separatory funnel containing 100 mL dilute HCl
(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.17 71-72 °C). 1H NMR: 6 7.61 (s, 2H), 4.44 (s,
4H), 3.45 (s, 6H).

Nickel catalyzed polymerization of 1,4-bis(methoxymethyl)-2,5-
dibromobenzene (6). Inside a helium atmosphere drybox, NiCl2 (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 drybox, 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 l,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 HCl/MeOH

87

(1 :1) 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 oC overnight to yield 0.194 g (88%) of a
fine, pale yellow powder that was insoluble in common organic solvents. Anal.
Calcd. for (CIOH1202)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 (7). 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.

88

 

3.3. Results

Polyacenic materials. A simple route to acene-like materials is suggested
by the acid catalyzed condensation of alkoxymethyl substituted benzenes. AS shown
in Figure 3.1, methoxymethylbenzene undergoes electrophilic aromatic substitution
to give polybenzyl. While most reported syntheses of polybenzyl are
polycondensations utilizing benzyl halides and various Lewis acids (AlCl3, TiCl4,

and SnCl4), similar results are Obtained using benzyl methyl ether and concentrated

C HZOC H3 £31],

 

 

Figure 3.1. Polymerization of methoxymethylbenzene.

sulfuric acid. The polymerization yields a slightly soluble, tan solid that exhibits a
glass transition at 180 °C to form a glassy resin. Thermal analysis indicates the
polymer is thermally stable to 400 °C, at which temperature the benzyl linkages
connecting the aromatic rings are broken and volatile by-products are evolved.

If this chemistry were applied to biS(methoxymethyl)benzenes, the resulting
networks should contain large amounts of fused aromatics, that upon elimination of

hydrogen would yield materials with high acenic content. To test this idea, we

89

prepared the ortho-, meta-, and para- isomers Of bis(methoxymethyl)benzene (1-3)
and polymerized them using acid catalysis. In all cases, the addition of sulfuric acid
to the monomer led to the rapid formation of yellow or brown solids, which
precipitated from solution. Structural characterization of these materials is difficult.
Not only are they insoluble, but the structural differences between polybenzyl and
the acene precursor are subtle and not readily detected by NMR, IR, and other standard
spectroscopic techniques. However, thermal analysis studies provide circumstantial
evidence that the ortho isomer tends to form acene-like structures, while the para
and meta isomers yield predominantly polybenzyl architectures.

The networks formed from the acid catalyzed polymerization of the
corresponding bis(methoxymethyl)benzenes exhibit surprisingly high thermal stability
in a nitrogen atmosphere. As shown in Figure 3 .2, 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 disubstituted
monomers are thermally stable past 700 °C resulting in ~70% char yields. When run
in air, all the polymers were easily oxidized and completely degraded by 500 °C.

An interesting observation was made however during a thermal analysis
experiment in which the environment around the samples was not continuously
purged. Figure 3.3 shows the TGA scans of the polymers taken in a stagnant
environment containing a small percentage of oxygen. The initial weight loss with

an onset near 400 °C was again observed for all the samples. However, for the para

90

 

 

 

 

100
80
S" 60 I
H
.C
.9
Q)
g 40
20
0
0 200 400 600 800 1000

Temperature (°C)

Figure 3.2. Thermogravimetric analysis of the networks formed from
monomers 1-3 heated in nitrogen at 10 °C/min.

91

 

 

 

 

100
a
80
°\o 60 ~
v0.0
.c
2"
Q)
3
40
20
b
0 %

 

0 100 200 300 400 500 600 700 800

Temperature (°C)

Figure 3.3. Thermogravimetric analysis of the networks formed from a) monomer
1, and b) monomer 3 in a stagnant air environment heated at 10 °C/min.

92

and meta derivatives, low degrees of oxidation led to chain scission and the complete
degradation of the samples. In contrast, the ortho sample shows the same remarkable
stability as it did in a nitrogen environment, a property we attribute to the formation
of a substantial proportion of fused rings during polymerization. It appears that the
substitution pattern on the benzene ring leads to different structures, with ortho
substitution favoring cyclization while the meta and para derivatives yield a
predominately polybenzyl network as Shown in Figure 3.4. For materials with ladder-
like structures, the scission of a Single chain does not necessarily lead to the severing
of the chain, but chain scission in a single strand polymer like polybenzyl, breaks the
polymer chain into smaller pieces. The prediction is that the polymers derived from
the ortho substituted monomer contain polyacene-like precursors that are thermally
stable and are able to dehydrogenate at temperatures above 600 °C.

To better understand the structural changes occurring in the polymer derived
from the ortho substituted monomer, polymer samples were heated in N2 to various
temperatures in a TGA apparatus and the residues were analyzed using infrared
spectroscopy. The spectra, shown in Figure 3.5, Show the progressive chemical
structure evolution from that of a standard polymer with a high proportion of aliphatic
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 samples
taken at 500 °C Should contain an appreciable fraction ofpolyacenes. One signature

of polyacenes would be the appearance of materials with a relatively low bandgap,

93

 

CHZOCH3 H.
.. :CHZOCH3
1

6.1.66.4,
H+
\

CHZOCH3 L O O O
O O C

Figure 3.4. Idealized structures resulting from the polymerizations of
the bis(methoxymethyl)benzenes 1-3.

94

 

relative absorbance
O D. (‘D
If //

 

 

 

4500 3500 2500 1500 500
wavenumber (cm'1)

Figure 3.5. Infrared absorption spectra of the polybenzyl network formed
from monomer 1 heated in nitrogen at 10 °C/min to a) 30 °C, b) 380 °C, c)
500 °C, (1) 600 °C, and e) 800 °C.

but with the retention of aromatic C-H bonds. From the IR Spectra, we find that
samples taken at 500 and 650 °C have a strong electronic absorption that tails into
the infrared region, indicating extensive Tt conjugation. In addition, we note that 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
stretching (3000-3150 cm") increase in intensity indicating the Shift from aliphatic
to aromatic C-H bonds as expected for the formation of acenes. Further treatment of
the samples at 800 °C results in the near complete loss of C-H bonds and presumably
the formation of a graphite structure. The broad absorption at 3500 cm'1 is due to
residual moisture in the KBr and the amount was assumed to be constant for all
samples.

An alternative method to dehydrogenate dihydroanthracenes is by using
chemical dehydrogenation reagents, such as 2,3-dichloro-5,6-dicyano-l,4-
benzoquinone (DDQ). The DDQ was refluxed with the polybenzyls in xylenes for 2
days. The product was collected by filtration, washed, and dried. Analysis by IR
Spectroscopy and TGA show little evidence for the success of the chemical treatment.
It is quite possible that the extent of dehydrogenation was limited due to the
insolubility of the polybenzyl networks.

Acene/poly(p-phenylene) hybrids. If a polymer containing the reactive
methoxymethyl functionalities was synthesized, it is logical that it should exhibit

‘ the same type of chemistry as the bis(methoxymethyl) benzenes in the presence of

96

strong acids such as concentrated sulfuric acid. If the same type of reaction occurs
when the polymer is exposed to the strong acid, it should result in a network in
which the polymer backbones are cross-linked via benzyl bridges. Furthermore, if
the polymer backbone exhibited extremely limited bond rotation, it seems reasonable
that the most probable cross-linking of adjacent reactive functionalities would result
in cyclized products. If this network propagates, it would result in a structure
resembling a sheet of graphite, and should have high thermal stability.

The monomer was synthesized from readily available starting materials as
Shown in Figure 3.6. The free-radical bromination of 2,5-dibromo-p-xylene using
N-bromosuccinimide led to l,4-bis(bromomethyl)-2,5-dibromobenzene (4). The
overall yield of this reaction was only ~30% due to the difficulty of separating the
product from impurities. Although two equivalents of the brominating agent were
used, the bromination exhibited no selectivity and the reaction resulted in a statistical

distribution ofnon-, mono-, and di-brominated methyl groups. Due to the symmetry

 

Br 1 Br
: NaOCH
,CH3 NBS. BPO CH Br 3 /l CH OCH
l//‘\\/ 001,, reflux ,1 " \/ 2 hexanes W 2 3
U ,, I 9; ——- CI
CH3/\ BrCHz/ Y CH,OCH2/\'/
Br Br Br
4 5

Figure 3.6. Synthesis of 1,4-bis(methoxymethyl)-2,5-dibromobenzene.

97

of the desired product 4, it could be recovered by repeated recrystallization.
Compound 4 was then reacted with 2 equivalents of sodium methoxide to form the
bis(methoxymethyl)benzene 5. When run in tetrahydrofuran, this reaction leads to 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.

Methoxymethyl substituted poly(p-phenylene) was synthesized from the
corresponding dibromobenzene monomer 5 using a nickel catalyzed coupling reaction
(Figure 3.7). 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.%) indicated a degree of

polymerization less than 10 (relative to polystyrene standards).

 

O/ O/
NiClz, Zn
PPh3, bipy
B =
Br I DMF, 80 °C n
O O
/ /
5 6

Figure 3.7. Synthesis of substituted poly(p-phenylene) 6.

98

Polymer 7 was reacted with sulfuric acid following the same procedure as
for the polymerization of bis(methoxymethyl)benzenes 1-3. Due to the rigid nature
of the PPP backbone, intramolecular cyclization leading to the formation of acene
precursors should be favored (Figure 3.8). Figure 3.9 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 0C 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

from the cross-linking reaction was neutralized with base and rinsed repeatedly with

 

Figure 3.8. Cross-linking of substituted PPP 6.

99

 

 

 

 

b
80
,,\° 60
E
5"
'5
3 a
40 »
20
0
0 200 400 600 800

Tem pe ra ture (°C)

Figure 3.9. Thermogravimetric analysis of a) substituted polymer 6, and b)
cross-linked network 7 in nitrogen heated at 10 °C/min.

100

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.

These materials were also analyzed by IR spectroscopy, shown in Figure
3.10. 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 network formed from
bis(methoxymethyl)benzene 1, high temperature treatment of network 7 at 700 °C
also results in the near complete loss of C-H bonds and the tailing of the electronic
absorption into the IR. The results from IR and thermal analysis indicate that thermal
treatment of the cross-linked network structure also leads to materials with extensive

rt-conjugation and low H content.

101

 

Absorbance

 

 

 

4

3500 2500 1 500 500

 

Wavenumber (cm'1)

Figure 3.10. Infrared Spectra of a) polymer 6, b) network 7, and c) carbon
char of network 7 after thermal treatment at 700 °C.

102

3.4. Discussion

The synthesis of a perfectly linear polyacene is difficult due to the lack of
synthetic control of the electrophilic aromatic substitution reaction used to prepare
the polyacenic materials. All of the unsubstituted aromatic sites have the same
reactivity. Selectivity could be obtained by polymerizing monomers or cross-linking
polymers in a pre-ordered state. Highly ordered graphite sheets could be prepared
from the cross-linking of assembled alkoxymethyl substituted PPP’S. If a soluble
PPP polymer with longer side chains were synthesized in high molecular weight, the
polymer would likely form a nematic liquid crystalline phase at some temperature,
as described in Chapter 2. Formation of the network could be initiated by the
photogeneration of a strong Lewis acid that cross-links the substituted PPP polymer
in an organized liquid crystalline state.

The intrinsic insolubility of a linear polyacene also hinders its synthesis.
Both acenes and poly(p-phenylene)s (PPP) are rod-like polymers with extended TE
conjugation and both become insoluble at low degrees of polymerization. The problem
of poor solubility for PPP has been overcome by attaching side chains to the polymer
backbone. Soluble, high molecular weight PPP can be obtained by polymerizing a
monomer substituted with n-hexyl side chains. 1 8 A similar approach can be used to
solubilize polyacenes. As shown in Figure 3.11, alkyl chains attached to the polyacene
backbone to give “hairy rods” that like the polyphenylenes should also have enhanced

solubility. These materials have the potential of forming very interesting

103

L n

Figure 3.11. Alkyl substituted polyacene.

macromolecular architectures. If the linear, alkyl substituted polyacene could be
synthesized in high molecular weight free of structural defects, this board-like polymer
Should exhibit sanidic liquid crystallinity.19’20 Unlike other sanidic polymers, this
polymers should remain soluble due to the long alkyl side chains that act as Spacers
to disrupt the p-stacking of the polymer backbones.

To obtain a substituted polyacene, a monomer containing both
methoxymethyl functionalities and unsubstituted aromatic Sites capable of undergoing
electrophilic aromatic substitution must be designed. Unlike monomers 1-3 however,
these monomers would also have to possess solubilizing side chains. Strategies

Shown in Figure 3.12 could be easily used to synthesize the monomers from available

104

R

: CHZOC H3
CHZOC H3

 

R _ R R
LeWIs acid
catalyst
., = I000.
n
R R R
: CHzoCHg
CH3OCH2
R

Figure 3.12. Synthesis of a soluble polyacene precursor.

starting materials, such as compound 5. The long alkyl side chains (R = n-hexyl or
n-octyl) may help the monomers to align into lamellar phases, as seen with other
similar compounds. Designing monomers capable of organizing into an ordered
phase may increase the probability of intramolecular cyclization.

In order to obtain linear polymer, it is necessary to control the chemistry
such that intramolecular cyclization reactions are favored over intermolecular
reactions that lead to branching and cross-linking. Examples of similar cyclization
reactions involving polymeric structures have appeared in the literature.
Polyphenylenes with pendant secondary alcohols undergo an acid-catalyzed

intramolecular cyclization to yield soluble planarized polyphenylenes.21 A similar

105

cyclization that leads to an S-substituted ladder polymer has also been reported.21
Certainly one factor favoring these reactions’ success was carrying out the reaction
in dilute solution.

Once polymerized, dehydrogenation of the polyacene precursor (Figure 3.13)
could be performed using common dehydrogenation reagents, such as DDQ or Pd/C.
Although an advantage of using Pd/C it that it is easy to remove from the reaction by
filtration, longer reaction times may be necessary to reach high conversions since
the mechanism of dehydrogenation requires the polymeric precursor to come in contact

with the catalyst surface.

IOOOO? :IOOOO

Figure 3.13. Chemical dehydrogenation of polyacene precursor.

Dehydrogenation of the precursor would alter the physical conformation of
the polymer into a completely planar backbone. The extent of conjugation in the
soluble polymer could then be characterized by UV/Vis spectroscopy. The predicted
electronic absorption of the polymer backbone will extend into the far IR and possibly
IR if the extent of conjugation is large. The properties of this soluble polyacene

could then be compared to those predicted by theory.22'24

106

3.5. 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 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 p electron system. The second
route targeted the synthesis of a methoxymethyl substituted poly(p-phenylene) 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
proposed. Once synthesized, the electronic properties of this material could be

compared to those predicted by theory.

107

3.6. References

1)Megahed, S.; Scrosati, B. J. Power Sources 1994, 51, 79—104.

2)Salomon, M.; Scrosati, B. Gazz. Chim. Ital. 1996, 126, 415-427.

3)Brandt, K. Solid State Ion. 1994, 69. 173-183.

4)Zhang, S. S.; Liu, Q. G.; Yang, L. L. J. Electrochem. Soc. 1993, 140, LlO7-L108.
5)Yata, S.; Hato, Y.; Kinoshita, H.; Ando, N.; Anekawa, A.; Hashimoto, T.;
Yamaguchi, M.; Tanaka, K.; Yamabe, T. Synth. Met. 1995, 73, 273-277.

6)Wang, T.; Kang, E. T.; Neoh, K. 0.; Tan, K. L. Synth. Met. 1997, 84, 405-406.
7)Ago. H.; Nagata. K.; Yoshizawa. K.; Tanaka, K.; Yamabe, T. Bull. Chem. Soc.
Jpn. 1997, 70, 1717-1726.

8)Tanaka, K.; Yata, S.; Yamabe, T. Synth. Met. 1995, 71, 2147-2149.

9)Yata, S.; Kinoshita, H.; Komori, M.; Ando, N.; Kashiwamura, T.; Harada, T.;
Tanaka, K.; Yamabe, T. Synth. Met. 1994, 62, 153-158.

10)Yazami, R.; Guerard, D. J. Power Sources 1993, 43, 39-46.

ll)Yamabe, T.; Tanaka, K.; Ago, H.; Yoshizawa, K.; Yata, S. Synth. Met. 1997, 86,
2411-2414.

l2)Xing. W. B.; Xue, J. 8.; Zheng, T.; Gibaud, A.; Dahn, J. R. J. Electrochem. Soc.
1996, 143, 3482-3491.

l3)Zheng, T.; McKinnon, W. R.; Dahn, J. R. J. Electrochem. Soc. 1996, 143,

2137-2145.

108

l4)Buiel, E.; George, A. E.; Dahn, J. R. J. Electrochem. Soc. 1998, 145, 2252-
2257.

15)Ortiz, B.; Walls, F.; Yuste, F.; Barrios, H.; Sanchez-Obregon, R.; Pinelo, L.
Synth. Commun. 1993, 23, 749-756.

l6)Otsubo, T.; Kohda, T.; Misumi, S. Bull. Chem.Soc.Jpn. 1980, 53, 512-517.
17)Bedard, T. C.; Moore, J. S. J. Amer. Chem. Soc. 1995, 117, 10662-10671.
18)Rehahn, M.; Schluter, A. D.; Wegner, G.; Feast, W. J. Polymer 1989, 30, 1060-
1062.

l9)Voigt-Martin, 1. C1; Simon, P.; Bauer, 8.; Ringsdorf, H. Macromolecules 1995,
28, 236-242.

20)Voigt-Martin, I. G.; Simon, P.; Yan, D.; Yakimansky, A.; Bauer, 8.; Ringsdorf,
H. Macromolecules 1995, 28. 243-254.

21)Scherf, U.; Mullen. K. Synthesis 1992, 23-38.

22)Whangbho, M. H.; Hoffmann, R.; Woodward, R. B. Proc. R. Soc. London, Ser.
A 1979, 366, 23.

23)Salem, L.; Longuett-Higgins, H. C. Proc. R. Soc. London, Ser. A 1960, 255,
435.

24)Li, Z. J.; Xu, H. B.; Yao, K. L. Mod. Phys. Lett. B 1997, 11, 477-483.

109

Chapter 4

Synthesis and Characterization of Cross-linkable
Poly[(1-trimethylsi1yI-l-propyne)-c0-(l-(4-azidobutyldimethylsilyl)-
l-propyne)] Copolymers

4.1. Introduction

Since it was first synthesized,1 poly(l-trimethylsilyl-1-propyne) (PTMSP)
has been a material of particular interest for gas separations. Its oxygen permeability
is roughly ten times that of poly(dimethylsiloxane), the polymer typically used as a

2

benchmark for high permeability. Rubbery polymers, such as poly(dimethylsiloxane)

around room temperature, usually have a high gas permeability attributed to its flexible

backbone and resultant high free volume.3

On the other hand, typical glassy polymers
have low gas permeabilities. The reason for PTMSP’s unusually high permeability
can be traced directly to its rigid random coil backbone. With its rigid coiled backbone
structure and the presence of the large bulky substituents, inter-segmental packing is
hindered and thus the polymer has a large free volume.

Unfortunately a major drawback of PTMSP is that the permeability declines
with time,4 an effect that has been attributed to aging or densification. We propose
that if the polymer chains were cross-linked, the interdiffusion of chains would be
prevented and the decline in the desired physical properties could be avoided.

To accomplish this, we chose organic azides as cross-linking moieties.

5-ll

Organic azides are widely used for cross-linking polymers and have three

110

advantages as cross-linking agents. They are easily synthesized, most commonly
via 8N2 substitution chemistry using sodium azide and a phase transfer agent, but
preparation from a large number of other functional groups has also been reported. 12
Azides are latent cross-linking agents and can be activated by thermal, photochemical,
or chemical means when cross-linking is desired, increasing the flexibility of material
processing. Finally. azide functional groups decompose to nitrogen gas and reactive
nitrenes that undergo insertion reactions.13 Thus, no low molecular weight by-
products are formed that could contaminate the material.

Azide-based cross-linking is particularly attractive for the stabilization of
membranes used for gas and liquid separations. In a recent publication, the preparation
of cross-linked PTMSP membranes using bis(aryl azide)s was explored.11 Cross-
linking PTMSP membranes with aryl azides stabilized their permeability and rendered
the membranes insoluble. However. one limitation ofthe bis(aryl azide) cross-linking
approach is that the azides phase separate from PTMSP at relatively low
concentrations (z 5 wt%), and thus the cross-link density is limited by the solubility
ofthe cross-linking agents in the polymer matrix. Increased cross-link densities are
needed for better control of the pore size in PTMSP and to minimize polymer swelling
in liquid separation schemes.

Higher cross-link densities could be obtained by preparing bis(azide)s that
are more compatible with PTMSP, or by directly attaching azides to the polymer

backbone. A disadvantage of the former approach is that the added azides occupy

lll

much of the free volume in the polymer, leading to a large drop in permeability.11

Direct attachment of azides to the polymer eliminates the phase separation problem
and minimizes the size of the cross-linking agent. The structure of PTMSP suggests
three strategies for incorporating azides into PTMSP: reactions involving the main
chain double bonds, substitution reactions at allylic methyl sites, or copolymerization
with an acetylenic comonomer having either an azide group or a suitable azide
precursor. In this chapter, strategies for the direct attachment of azides to the PTMSP
framework and characterization of the cross-linking behavior of the resulting polymers

are explored.

112

4.2. Experimental Section

Materials. Tetrahydrofuran (THF) and toluene were purified by distillation
from calcium hydride followed by a second distillation from sodium/benzophenone
ketyl. Unless otherwise specified. all other ACS reagent grade starting materials
and solvents were used as received from commercial suppliers without further
purification.

Characterization. Proton nuclear resonance (‘H NMR) spectra were
measured using a Varian Gemini-300 spectrometer at 300 MHz. All samples were
run at room temperature in CDC13. Chemical shifts were calibrated using residual
CHCl3 and are reported in ppm (6) relative to tetramethylsilane. Infrared spectra of
polymers were obtained under nitrogen at room temperature on a Nicolet Magna-IR
550 Fourier Transform IR spectrometer. IR measurements of samples before and
after thermal treatment were carried out on films cast on NaCl disks. A Hitachi U-
4001 UV-visible spectrometer was used to obtain the UV-visible spectra of
cyclohexane solutions of the polymers. Molecular weights of the polymers were
determined by gel permeation chromatography (GPC) using a PLgel 20}; Mixed A
column and a Waters R401 Differential Refractometer detector at room temperature
with THF as eluting solvent at a flow rate 1 mL/min. Monodisperse polystyrene
standards were used to calibrate the molecular weights. The concentration of the
polymer solutions used for GPC measurements was 1 mg/mL. Differential scanning

calorimetry (DSC) data were obtained under a He environment using a Perkin-Elmer

ll3

DSC 7 instrument at a heating rate of 10 °C/min. The DSC 7 temperature was
calibrated with an indium standard. Thermogravimetric analyses (TGA) were
performed under nitrogen and air atmospheres at a heating rate of 10 °C/min on a
Perkin-Elmer TGA 7 instrument. Elemental analyses were performed at the
Microanalysis Laboratory at the University of Illinois, Urbana-Champaign, using a
CE44O Carbon, Hydrogen, Nitrogen Analyzer (Exeter Analytical, Inc) and
conventional Br analysis (titration).

Poly[l-trimethylsilyl-l-propyne]. PTMSP was prepared according to the
literature procedure. 1 4'1 6 A solution of 1-trimethylsilyl-l -propyne (0.64 g, 5.7 mmol)
in dry toluene (10 mL) was degassed by three freeze/pump/thaw cycles. The monomer
solution was then added via cannula to a mixture of tantalum pentachloride (120 mg,
0.34 mmol) and triphenylbismuth (150 mg, 0.34 mmol). The reaction mixture was
heated to 80 °C and stirred for 16 hours. The resulting gel-like product was dissolved
by adding more toluene, centrifuged to remove catalyst residues and precipitated by
slow addition into an excess of methanol with rapid stirring. Redissolving in toluene
and reprecipitating into methanol yielded 0.51 g (80%) of poly(l-trimethylsilyl-l-
propyne). GPC: Mn =750.000. Mw/Mn =2.3. 1H NMR: 6 0.25 (br, 9H), 1.75 (br,
3H).

Allylic bromination of PTMSP. Partially brominated PTMSP (50%) was
prepared following the literature procedure17 by reacting PTMSP (1.0 g) with N-

bromosuccinimide (0.79 g, 4.5 mmol) in refluxing CCl4 (100 mL) in the presence of

114

benzoyl peroxide (100 mg). Conversion to the brominated product was determined
by monitoring the formation ofa signal at 2.2 ppm in the 1H NMR and an absorption
band at 1219 cm‘1 in the IR spectra due to the introduction of C-Br stretching. The
yield was nearly quantitative.

Azidization of brominated PTMSP. Brominated PTMSP (0.50 g, 2.2 mmol
Br) was dissolved in either 25 mL benzene or THF under a nitrogen atmosphere.
The phase transfer reagent (18-crown-6 ether, 29 mg, 0.11 mmol), and NaN3 (0.58 g,
8.9 mmol) or LiN3 (0.44 g, 8.9 mmol) were added at the desired temperature (50 °C
for THF and 60 °C for benzene). The reaction mixture was sampled at different
times, precipitated into methanol, and the conversion determined by lH NMR and IR
spectroscopy. For reactions using ZnClz, the same procedure was used except that
ZnCl2 (0.18 g, 0.73 mmol) and pyridine (0.18 g, 2.2 mmol) were used in place ofthe
phase transfer reagent.

4-Bromo-l-butene (2). 4-Bromo-l-butene was prepared according to the
literature procedurelg in 37% yield (21 g); bp 98-100 °c (lit.18 98-100 °C). ‘H
NMR: 8 2.7 (m, 2H), 3.3 (t, 2H), 5.1 (m, 1H), 5.8 (m, 2H).

4-BromobutyIdimethylchlorosilane (3). Compound 3 was prepared
following the procedure of Kunzler and Percec19 in 85% yield (17.9 g); bp 90-95
oC/23 mm Hg (lit.19 92-95 °C/5 mmHg). IH NMR: 6 0.4 (s, 6H), 0.8 (t, 2H), 1.5 (m,

2H), 1.9 (m, 2H), 3.4 (t, 211).

115

l-(4-Bromobutyldimethylsilyl)-1-propync (4). A solution of
propynyllithium was prepared by the drop-wise addition of n-BuLi (2.5 M in hexanes,
12 mL) to a cold propyne-ether solution obtained by bubbling propyne gas through
cold diethyl ether (-78 °C). To the solution of propynyllithium was added 4-
bromobutyldimethylchlorosilane (6.7 g, 29 mmol) dropwise at 0 °C. The reaction
mixture then was stirred overnight at room temperature and then poured into 400 mL
of ice water. The ether layer was collected and washed with distilled water, dried
over MgSO4, filtered and the solvent removed by rotoevaporation. The remaining
oil was purified by vacuum distillation at 59-65 oC/3 mmHg to yield 4.6 g (68%) of
1-(4-bromobutyldimethylsilyl)-1-propyne (lit.19 70-75 °C/15 mmHg). lH NMR: 5
0.15 (s, 6H), 0.6 (t, 2H), 1.5 (m, 2H), 1.85 (s, 3H), 1.9 (m, 2H), 3.4 (t, 2H).

Poly[(l-trimethylsilyI-1-propyne)-co-(l-(4-bromobutyldimethylsilyl)-1-
propyne)] (5a-5g). The following procedure. illustrated for the synthesis of the
copolymer containing 20 mol% of the l-(4-bromobutyldimethylsilyl)-1-propyne co-
monomer, was used to prepare all of the copolymers. A mixture of l-trimethylsilyl-
l-propyne (5.23 g. 0.047 mol) and 1-(4-bromobuty1dimethylsilyl)-l-propyne (2.72
g, 0.012 mol) was dissolved in dry toluene (100 mL) and degassed by three freeze/
pump/thaw cycles. The monomer solution was then added via cannula to a mixture
oftantalum pentachloride (0.835 g, 2.33 mmol) and triphenylbismuth (1.025 g, 2.33
mmol). The reaction mixture was heated to 80 OC and stirred for 16 hours. The

resulting gel-like product was diluted with toluene, centrifuged to remove residual

116

catalyst, and precipitated by slow addition into an excess of methanol with rapid
stirring. Redissolving in toluene and reprecipitating into methanol yielded 6.54 g
(83%) of poly[(l-trimethylsilyl-l -propyne)-co-(l -(4-bromobutyl-dimethylsilyl)-1-
propyne)]. GPC: Mn = 432,000. Mw/Mn = 2.4. 1H NMR: 5 0.2 (br), 1.6 (br), 1.8
(br), 3.4 (br).
Poly[(l-trimethylsilyI-l-propyne)-co-(l-(4-azidobutyldimethylsilyl)-l-
propyne)] (6a-6g). Sodium azide (0.65 g, 10 mmol) and a catalytic amount of
tetrabutylammonium hydrogen sulfate (TBAH, 100 mg) were added to a stirred
solution of 1.0 g of poly[(l-trimethylsi1yl-1-propyne)-co-(1-(4-bromobutyl-
dimethylsilyl)-1-propyne)] (5g) in 100 mL of THF. The reaction mixture was stirred
for 120 hours at 60 0C. The resulting polymer was dissolved in additional THF and
centrifuged to remove most of the catalyst and other undissolved solids. The cloudy
polymer solution was washed with water to remove salts not removed by
centrifugation. The resulting clear solution was precipitated into methanol, filtered,
redissolved in THF and again precipitated in methanol. The collected polymer was
vacuum dried at 50 OC overnight to obtain 0.90 g (90%) of poly[(]-trimethylsilyl-l-
propyne)—co-(l-(4-azidobutyldimethylsilyl)-1-propyne)]. GPC: Mn = 410,000, Mw/

Mn=2.6. ‘H NMR16 0.2 (br), 1.6 (br), 1.8 (br), 3.25 (br).

117

Film casting and cross-linking. Viscous casting solutions were obtained
by dissolving the copolymer (200 mg) in toluene (10 mL). Polymer films were
prepared by pouring the casting solutions onto a flat glass plate and drying for 48
hours. The films were then completely dried in a vacuum oven at 50 °C overnight.
Thermal cross-linking of the resulting free standing copolymer films was achieved
by heating the samples in a vacuum oven at 250 °C for 3 hours.

Swelling experiments. Cross-linked copolymer films were cut into 2 cm x
2 cm pieces, weighed, and immersed in solvents with solubility parameters (5) ranging
from 14-20 x 103 (J/m3)“2. After 24 hours, the films were removed from the solvent,
blotted dry, weighed, and placed back into the solvent. This procedure was repeated
until the samples reached a constant weight. The reported swelling values are the

average of three individual film measurements.

118

4.3. Results

Despite numerous attempts and the use of various reaction conditions, it
has been shown that the double bonds of the PTMSP backbone are unreactive.19
Low reactivity for these double bonds is not unexpected because of the severe steric
effects caused by the methyl and trimethylsilyl side groups.

Incorporation of azides at allylic sites. Direct attachment of azides to the
allylic methyl groups of PTMSP is an attractive strategy for accessing PTMSP
polymers with high azide contents and high degrees of cross-linking because the
small size of the azide group should result in minimal structural changes compared

the parent polymer. As shown in Figure 4.1, brominated PTMSP was prepared using

 

CH3 CH3 CH3 CH3 CH3 CHzBI‘
NBS

n CCI4 n
TMS TMS TMS TMS TMS TMS
transfer agent

CH3 CH3 CH2N3

cross-linked :1) cast film

films 2) ho or heat n

TMS TMS TMS

Figure 4.1. Synthesis of azide-containing PTMSP via allylic bromination.

119

17,20

standard free-radical conditions, and various methods were examined for

nucleophilic displacement of the bromide by azide. Allyl azides are easily prepared

21

from allyl bromides by reaction with NaN3 or trimethylsilyl azide, and similar

methods have been used to partially convert halogen-containing polymers such as

2234 and poly(epichlorohydrin)25 to polymeric azides. However,

poly(vinyl chloride)
we found that substitution chemistry at the allylic methyl group in PTMSP is highly
hindered. Both IR and 1H NMR spectroscopy showed no evidence for incorporation
of azide into the polymer.

Since SN2 chemistry failed, we followed the procedure of Ravindranath et
al.26 and used Lewis acids to form allylic carbocations that could be trapped by the
azide anions. The product from the azidization reaction of 50% brominated PTMSP
showed a small peak at 2100 cm‘1 in the IR spectra indicating the presence of the N3
group, but a strong BrCHZ- vibrational band at 1219 cm" was still evident. Proton
NMR also indicated only partial introduction of the azide. Our best results show
only limited success, with less than 5% conversion to the azido derivatives after five
days.

Incomplete conversion to the azide is a serious problem since brominated
PTMSP degrades rapidly in the presence of UV light and oxygen.20 Before these
polymers can be cross-linked into insoluble films, the remaining bromide must be

removed while retaining the azides. We tried to replace the remaining bromine with

hydrogen by treating the azido derivatives with Bu3SnH/AIBN. After reacting in

120

toluene at 75 °C for 2 hours, IR spectra showed the disappearance of both the -N3
and the BrCHZ- bands, indicating that the reaction was not selective and that hydride
replaced the bromide and reduced the azide group.

We also examined anionic methods for substitution at the methyl group.
For example, brominated PTMSP treated with either BuLi or Li metal generated the
allylic anion, but gelation was observed in these reactions and the resulting polymers
became insoluble and unprocessable. Presumably, slow formation of the anion
allowed the coupling of anions with residual allyl bromides to give a cross-linked
network.

Incorporation of azides via copolymerization. Because steric hindrance
at the allylic methyl group resulted in only partial conversion of the allylic bromide
to the azide, we decided to increase the reactivity of the halide by moving it away
from the polymer backbone. While the tether could be connected to either the allylic
methyl or trimethylsilyl group of the monomer, attachment through the trimethylsilyl
group should have the least effect on the polymerization rate since the site of the
substitution is farther from the triple bond. Following the approach of Kunzler and

19 we prepared copolymers of 1-trimethylsilyl-l-propyne and 1-(4-

Percec,
bromobutyldimethylsilyl)-1-propyne. Azide groups can be introduced into these
copolymers by simple nucleophilic displacement reactions. Because monomers

containing azides are unlikely to survive the polymerization conditions, we did not

explore the direct polymerization of an azide-containing monomer.

121

O

HMPA
BrCH2CH2CH2CHzBI’ 2—003 BrCH2CH2CH=CH2

 

 

1 2
(EH3 H2PtC|5/2-propanol (EH3
2 + Cl-ISI-H toluene 7- CI—ISI—(CH2)4Br
CH3 CH3
3
CH3CECLi EH3
> CH3C—EC— I—(CH2)4BI'
ether, -78°C (5,43
4

Figure 4.2. Synthetic route to l-(4-bromobutyldimethylsilyl)-1-propyne.

The comonomer, prepared by the route outlined in Figure 4.2, was
copolymerized with 1-trimethylsilyl-1-propyne to give a series of copolymers that
contain 2-20 mol% of 1-(4-bromobutyldimethylsilyl)-1-propyne (Figure 4.3). GPC
measurements showed that the molecular weights of the copolymers were similar to
those of PTMSP, ranging from 4-5x10S g/mol with a polydispersity of2.5. Treatment
of the copolymer with sodium azide and a phase transfer agent smoothly converted
the bromobutyl side chain to the corresponding azide. The reaction was sampled at
different times and analyzed by lH NMR and FTIR spectroscopy to determine the

conversion. In the IR spectra, the intensity ofthe band at 1219 cm'l decreased (CH2

122

CH3
x CH3C—EC—Si—CH3 + y 4
(EH3 toluene, 80°C

CH3 CH3
1 1
CH3— si—CH3 CH3—Si—CH3

CH3 OHzCl-lzCHzCHzBr
5(a-9)

TaCldPh3Bi

 

CH3 C

l l”"
NaN3/TBAH
o urc- H0 r
THF, soc (I: x \cl: y
CH3—si—CH3 CH3—si—CH3

CH3 CH2CH2CH2CH2N3
6(a-9)

 

Figure 4.3. Synthetic route to copolymers 6a-6g.

123

wagging, CH2Br group), and a new band grew in at 2100 cm'l that corresponds to

.le3 group27 in poly[(l-trimethylsilyl-1-propyne)-co-(1-(4-

the
azidobutyldimethylsilyl)-1-propyne)]. Corresponding changes were seen in the lH
NMR data. The peak at 3.4 ppm from the -CHZBr segment in poly[(l-trimethylsilyl-
1-propyne)-co-(1-(4-bromobutyldimethylsilyl)-1-propyne)] decreased as a new
resonance at 3.25 ppm for -CH2N3 in poly[(l-trimethylsilyl-1-propyne)-co-(1-(4-
azidobutyldimethylsilyl)-1-propyne)] increased. The peak at 3.4 ppm in the 1H NMR
spectra and the absorption band at 1219 cm'I in the IR spectra both decreased to
baseline, demonstrating that complete displacement of the bromide was achieved
after 5 days.

Polymer characterization. Elemental analyses (Table 4.1) of the
copolymers showed no bromine detectable by titration, confirming the high degrees
of bromide to azide conversion indicated in the 1H NMR and IR data. The absolute
azide content in the copolymers can also be calculated from the experimental CfN
ratios. We found that the azide contents of copolymers calculated from elemental
analyses data result in z20% lower values than expected. The deviations from the
expected values may be due to PTMSP’s tendency to form refractory oxides in
oxidizing environments.28

Shown in Figure 4.4 are the IR spectra of the series of copolymers with

increasing azide contents. The spectra are normalized relative to the intensity of the

C=C absorption (1600 cm“) of the polymer backbone. The most prominent feature

124

Table 4.1. Elemental analyses ofpoly[(1-trimethylsilyl-l-propyne)-co-(1-(4-azido-
butyldimethylsilyl)-1-propyne)] copolymers.

 

 

 

 

 

3:33:31 s w 20
theo. exp. theo. exp. theo. exp.

C% 63.46 63.26 62.76 62.80 61.52 60.61
H% 10.61 10.02 10.45 10.25 10.17 7.68
N% 1.80 1.36 3.49 2.71 6.52 5.45
Si% 24.13 b 23.30 27.79 21.79 24.95

Experimental 7.8

N3 (mol%)°’d

 

 

 

 

co-(l -(4-bromobutyldimethylsilyl)-1 -propyne)]

b. not analyzed
calculated from elemental analysis results

.0

(:1. Br was not detected using titration method

125

calculated by assuming 100% conversion from poly[(l-trimethylsilyl-l-propyne)-

 

 

20°

15%

12%
10%l K )1 JULUU‘

7%

Absorbance

5% . l

 

PTMSP

‘-
_

 

 

l l l

 

3500 3000 2500 2000 1500 1000

Wavenumber (cm'1)

Figure 4.4. FTIR spectra of PTMSP and copolymers 6a-6g.

126

 

Integrated Intensity

 

 

 

I l l I

O 5 10 15 20 25

mol% azide (y)

Figure 4.5. Integrated azide absorbance intensity at 2100 cm'1 versus the azide
content in copolymers 6a-6g as calculated from the comonomer feed ratios.

127

in the spectra is the increase in the intensity of the absorption band for the alkyl
azide at 2100 cm". This increase is mirrored by a growth in the intensity of the
aliphatic C-H stretching bands (2800-3000 cm") caused by the increase in the
concentration of alkyl side chains in the higher azide-content copolymers. When the
integrated intensities of the azide absorption bands are plotted versus the theoretical
azide calculated from the comonomer feed ratios (Figure 4.5), we obtain a linear
relationship indicating that the overall composition of the samples are comparable
to the initial monomer ratios.

UV—visible spectra of cyclohexane solutions of poly[(l-trimethylsilyl-1-
propyne)-co-(l-(4-azidobutyldimethylsilyl)-1-propyne)] show no absorption above

300 nm. Absorption bands of alkyl azides27

typically occur at kmax = 287 nm (e =
25) and Kmax = 216 nm (e = 500), but the strong C=C absorption of the polymer
backbone masks the azide absorption in these samples. Like PTMSP, the azide-
containing copolymers are snow-white solids after precipitation into methanol and
form clear, colorless films when cast from solvents such as THF or toluene.
Thermogravimetric analyses (Figure 4.6) of poly[(l-trimethylsilyl-1-
propyne)-co-(l-(4-azidobutyldimethylsilyl)-l-propyne)] under nitrogen show the
onset ofa weight loss at 225 °C due to the decomposition ofthe azide groups and the
release of nitrogen from the polymer. Depending upon the density of the original

sample, the subsequent cross-linking reactions result in a 10-20 % char yield at 600

0C. In contrast, PTMSP is thermally stable to 350 °C and then rapidly decomposes

128

 

100--

 

80-

 

60-

Weight %
2’3

40-

30-

 

20-

10- I
O n l 1 1 J

O 100 200 300 400 500 600

 

 

 

 

Temperature (° C)

Figure 4.6. TGA of a) PTMSP in nitrogen; b) copolymer 6g in
nitrogen; and c) copolymer 6g in air.

129

to monomer, yielding no residue. When run in air, the copolymers again evolve
nitrogen at 225 °C, but the resulting products are less stable and decompose more
rapidly. The weight loss due to evolution of nitrogen could, in principle, be used to
quantitate the N3 contents in the copolymers. However, the weight loss from nitrogen
evolution partially overlaps with a weight loss associated with a slow degradation of
the polymer, and consistent values could not be obtained.

DSC measurements on poly[(l-trimethylsilyl- l -propyne)-co-(1-(4-
azidobutyldimethylsilyl)-1-propyne)] copolymers show no thermal transitions below
the azide decomposition temperature (< 225 OC) indicating that like PTMSP,29 the
copolymers do not exhibit a glass transition. We interpret the similarity of the PTMSP
and copolymer DSC scans as evidence that the added side chains did not drastically
alter the physical properties of the PTMSP backbone. The exothermic transition for
azide decomposition is centered at 270 0C in the DSC thermograms (Figure 4.7) and
can be quantitatively related to the azide content in the copolymers. Figure 4.8 shows
that the enthalpy change associated with the exotherm is linearly related to the azide
content in the copolymers calculated from the original co-monomer feed ratios. The
plot yields a positive y-intercept since the DSC exotherm also includes a contribution

from the change in heat capacity (ACp) of the cross-linked networks.

130

Heat flow (endo —’)

 

2W—
1W

 

 

 

 

 

 

 

12% \f—
10%
“V”—
7% J
5W
2%
f
125 175 225 275 325

Temperature (°C)

Figure 4.7. DSC results for copolymers 6a-6g.

131

Heat evolved (Jlg polymer)

1 000

 

900

800

700

600

500

400

300

200

100

 

 

 

0 1 1 1 1
0 5 10 15 20 25

 

mol % azide (y)

Figure 4.8. Heat evolved during thermal decomposition of
azide copolymers 6a-6g measured by DSC.

132

Thermal cross-linking of copolymer films. We studied the thermal
decomposition ofthe azide-containing copolymers from 225-275 °C, the temperature
range that corresponds to the exotherm measured in the DSC curves (Figure 4.7) and
the nitrogen weight loss seen in the TGA thermograms (Figure 4.6). The rate ofthe
reaction was determined by monitoring the decrease in the intensity of the N3
absorption at 2100 cm" in the IR (Figures 4.9 — 4.11). We assumed that the azide
concentrations in the films are proportional to the integrated absorption intensities

in the normalized IR spectra. Using Equation 1, where [N3]o and [N3]t are the azide

[N3]:
-ln — = kdt (1)
[N310

concentrations in a film before cross-linking and a film at time t, respectively, the
azide degradation rate constant (k d) were obtained for each temperature studied (Figure
4.12). An Arrhenius plot (Figure 4.13) ofthe resulting kd values gave the activation
energy for azide decomposition (Ea) in the copolymer. For copolymer 6f, the activation

energy was calculated to be 130 kJ/mol.

133

 

7hr

6hr
d)
g 4hr
m
.0
h
8 3hr
.0
<
2hr
1hr
0hr

 

 

 

3500 3000 2500 2000 1 500 1 000 500

Wavenumber (cm'1)

Figure 4.9. FT IR spectra showing the thermal decomposition
of copolymer 6f at 225 °C under vacuum.

134

 

240rni

-- Mi
:2: » ill}

 

     
 

d,
U
C
a T
.a l k
h
:3 45th") It I
.o .I
< l
30 min IW‘
15 min ‘ «W*

Ornhi

l l l l l

3500 3000 2500 2000 1 500 1 000 500

 

 

 

Wavenumber (cm'1)

Figure 4.10. FTIR spectra showing the thermal decomposition
of copolymer 6f at 250 °C under vacuum.

135

 

75 min _ AM

60 min

45 min

$19

 

30 min

20 min

9

_J

Absorbance

10 min

5 min A

0 min

L

 

 

l 1 l l l

3500 3000 2500 2000 1500 1000 500

 

Wavenumber (cm'1)

Figure 4.11. FTIR spectra showing the thermal decomposition
of copolymer 6f at 275 °C under vacuum.

136

-ln([N3]1l[N3]o)

3.0

 

275° C
2. 5 ~
+ 250° C

 

 

 

 

 

 

400

time (minutes)

Figure 4.12. Loss of [N3] in copolymer 6f versus time
at the indicated temperatures.

137

 

-ln kcl

 

 

 

 

O 1 1
1.8 1.9 2.0 2.1

1000/r (K'1)

Figure 4.13. Arrhenius plot of the rates of thermal
decomposition of copolymer 6f.

138

Swelling measurements. Polymer films used for swelling experiments were
cast from toluene solutions and were dried under vacuum at 50 °C. The conditions
selected for cross-linking the copolymer films (250 °C, 3 hours) were chosen to
maximize the extent of cross-linking while minimizing thermal degradation. Thermal
decomposition shows up as a visible yellowing ofthe samples and the appearance of
new bands in the IR spectra at ~1700 cm". Swelling measurements were performed
on cross-linked copolymers that contained 2 mol% (6a) and 15 mol% azide (6f)
before cross-linking. Pieces of the cross-linked copolymer films were cut and
immersed in a series of poor H-bonding solvents that are good solvents for PTMSP.
After 24 hours, the films were removed from the solvent, blotted dry, weighed, and
returned to the solvent. After an additional 24 hours, the procedure was repeated to
ensure that the films had reached equilibrium. Figure 4.14 shows the results for
these two cross-linked networks. Cross-linked copolymer 6a swells significantly in
octane, cyclohexane, and toluene, while cross-linked copolymer 6f exhibits minimal
swelling over the entire solvent range, with only a slight preference for octane,
cyclohexane, and toluene. Cross-linked films of the entire series of azide copolymers

3)”2 x 103) and the degree of swelling

(6a-6g) were placed into toluene (5 = 18.2 (J/m
measured. Figure 4.15 shows a dramatic decrease in swelling as the cross-linking

density increases, with copolymers with azide contents greater than 5 mol% showing

negligible swelling in toluene.

139

 

Q (ml/g)

2%

 

 

 

14 15 16 17 18 19 20

5 ( [J/mfi1’2x10'3)

Figure 4.14. Swelling of thermally cross-linked films of copolymers 6a
and 61' in solvents with different solubility parameters, 6 ([J/m3]“2x10'3).

140

Q (ml/g)

4.5

 

4.0 -

3.5 -

2.5 e

2.0 -

1.0 -

0.5 -

 

 

 

00 1 1 1 1
0 5 10 15 20 25

mol% azide (y)

Figure 4.15. Swelling of thermally cross-linked films of
copolymers 6a-6g in toluene.

141

4.4. Discussion

Synthetic strategy. In a previous report, we explored the use of aryl azides
as polymer cross-linking agents. 1 1 Cross-linkable PTMSP films were easily prepared
by casting from polymer solutions containing added bis(aryl azide), but a major
limitation of this scheme is that only low levels of aryl azide can be incorporated
into most polymers before the onset of phase separation. In principle, the compatibility
of aryl azides with polymer hosts could be improved through chemical modifications
of the aryl azide structure. In similar fashion, alkyl azides could also be used in the
form of an additive. The higher decomposition temperature of alkyl azides means
that the processing window for polymers containing alkyl azides is wider than for
those with aryl azides. One potential drawback, however, is that the higher
temperatures needed for cross-linking might lead to some loss of the cross-linking
agent from the polymer due to the volatility of the low molecular weight alkyl azide.

Our goal of reaching higher azide contents (and higher degrees of cross-
linking) was achieved by directly attaching the azide group to the polymer backbone.
In most polymers this would be an easy task since azides are good nucleophiles and
can be introduced using simple SN2 chemistry. However, the steric hindrance
associated with the PTMSP backbone blocked our attempts to introduce the azide
group. Copolymerization with a functional acetylenic comonomer followed by its
conversion to the azide using a nucleophilic displacement reaction was necessary to
incorporate larger amounts of azide functionality into PTMSP.

The addition of the tethered side chains does not appear to drastically alter

142

the physical properties of the PTMSP backbone. The UV-vis absorption spectra of
the copolymers are nearly identical to that of PTMSP. Neither PTMSP nor the
c0polymers strongly absorb above 280 nm, indicating that the C=C bonds are not
conjugated and the rigid polymer backbone is retained. This conclusion is also
supported by DSC data that indicate that the copolymers, like PTMSP, do not exhibit
a glass transition below its thermal decomposition temperature.

The copolymerization route enabled us to overcome the problem of phase
of separation at low azide contents and reach high degrees of cross-linking. While
the highest azide content we explored in this work was 20 mol%, much higher levels
are accessible by changing the monomer feed ratios during the synthesis of PTMSP
copolymers.

C ross-Iinking. Both aryl and alkyl azides can be activated using thermal or
photochemical processes. The UV-vis spectra of PTMSP films containing bis(aryl
azide)s have two absorption maxima, one from PTMSP and a second from the bis(aryl
azide) (e. g. 300 nm for 4,4’-diazidobenzophenone).11 Because their absorption
maximum occurs at lower energies, aryl azides can be photochemically decomposed
selectively in optically clear substrates like PTMSP. Thus, the irradiation of PTMSP
containing bis(aryl azide)s at 300 nm led to cross-linked films with no discernable
damage to the host polymer. For alkyl azides, the absorption maximum is shifted to
shorter wavelengths and coincides with the absorption band for PTMSP. Irradiation

of the alkyl azide groups in poly[(l-trimethylsilyl-l-propyne)-c0-(1-(4-

143

azidobutyldimethylsilyl)-1-propyne)] may lead to photochemistry that alters the
structure of the polymer backbone. In addition, thick samples or those with high
azide contents may be difficult to cure by UV.

DSC data show that polymers containing either alkyl or aryl azides thermally
decompose, leading to cross-linked products. The limited compositional range
available for aryl azide additives prevented us from examining the relationship
between thermal decomposition and azide content in detail, but the data reported in
this paper for alkyl azides indicate that the thermal properties show a linear
dependence on composition over the entire range studied. Although the conditions
required to cross-link alkyl azide substituted PTMSP are more severe than those
required for the activation of bis(aryl azide)s, mechanically stable optically clear
cross-linked networks with low levels of degradation were obtained. We also observed
that copolymer films cast on glass exhibit excellent adhesive properties after cross-
linking, either through chemical or physical interactions with the surface. Only after
prolonged immersion in good solvents and sonication did a cross-linked film of
copolymer 6f fragment and dislodge from the glass surface.

Cross-linking the rigid polymer chains before they have a chance to
interdiffuse results in stabilized low-density open structures with high amounts of
free volume. Swelling measurements made on the series of cross-linked copolymer
films (6a-6g) show that a low level of cross-linking yields materials impervious to

solvent. Flexible polymers, such as polysiloxanes, typically require higher degrees

144

of cross-linking to yield non-swelling films,3 1 but the low levels of swelling observed
in lightly cross-linked PTMSP networks could be anticipated from the known rigidity
ofthe PTMSP chains. The slight swelling of PTMSP networks with greater than 10
mol% cross-linking, as measured by weight gain, is presumably due to solvent
molecules occupying the free volume of the polymer.

Permeability and temporal stability of cross-linked membranes. The
oxygen and nitrogen permeability coefficients of freshly prepared and cross-linked
membranes were measured.32 As shown in Tables 4.2 and 4.3, the permeability
coefficient of uncross-linked films decreases slightly as the amount of azide
incorporated in the copolymer increases, presumably due to the side chains occupying
the free volume of the membranes.

Cross—linked membranes exhibit vastly different properties depending upon
the mode of cross-linking employed. The different behavior of thermal and photo
cross-linked films may be due to more free volume in the thermally cross-linked
membranes. During the thermal process, thermal expansion of the membranes initially
creates a dilated network with high free volume. Cross-linking then interlocks the
chains, maintaining the high free volume (Table 4.2). Unlike flexible polymers,
which could easily relax at these relatively low cross-linking densities, the rigidity
of the PTMSP backbone prevents bond rotations that would enable interdiffusion of
the polymer chains leading to decreased free volume, and therefore decreased

permeability.

145

Table 4.2. Permeability coefficients of oxygen, P(Oz), and the oxygen/nitrogen
separation factor (or) at 23i1 0C of thermally cross-linked poly[(]-trimethylsilyl-l-
propyne)-co-(l -(4-azidobutyldimethylsilyl)-1-propyne)] copolymer membranes.
P(Oz) is in units of 10’7 cm3(STP) ' cm/cm2 ° s * cm Hg.

 

 

 

mol% N3 in before cross-linking after cross-linkifl
”polymer P(02) 01(02/N2) P(02) 01(02/N2)
PTMSP 7.4 1.7 6.0 1.6
2 5.4 1.7 3.5 1.7
3 5.2 1.7 3.0 1.9
5 3.5 1.7 2.1 2.1
7 3.3 1.5 2.1 2.1
10 3.7 1.6 2.2 2.1

 

 

 

 

 

 

 

146

Table 4.3. Permeability coefficients of oxygen, P(Oz), and the oxygen/nitrogen
separation factor (or) at 23¢] 0C of photo cross-linked poly[(l-trimethylsilyl-l-
propyne)-co-( 1 -(4-azidobutyldimethylsilyl)-1 -propyne)] copolymer membranes.
P(Oz) is in units of 10’7 cm3(STP) ° cm/cm2 ° 3 ° cm Hg.

 

 

 

mol% N3 in before cross-linking after cross-linking
copolymer
P(02) O“Oz/1‘12) P(02) 0«Oz/1‘12)
PTMSP 7.4 1.7 0.70 3.2
2 5.6 1.8 0.46 3.8
5 4.2 1.5 0.37 3.7
7 3.4 1.5 0.30 3.9

 

 

 

 

 

 

 

147

In contrast, membranes cross-linked photochemically are not dilated and
cross-linking leads to densification of the network (Table 4.3). The high energy
required to activate the alkyl azides also leads to photo-oxidative cross-linking of
the irradiated surface due to residual oxygen in the irradiation chamber. Consistent
with this interpretation are the measured pyncometric densities of the copolymer
membranes. Thermal cross-linking results in only a small change in density, while
photo cross-linking results in significant densification.

Ideally, a greater content of azide groups in the copolymer should also
produce higher degrees of cross-linking. However, higher azide contents did not
lead to higher separation factors. We believe that for azide contents >5%, the cross-
linked network is too rigid to allow further densification and losses in free volume.
This observation shows that there is a limit of modification beyond which the
selectivity is not affected.

The permeabilities of cross-linked membranes, unlike PTMSP membranes,
did not decline when stored under vacuum.32 If we assume that interchain diffusion
leads to densification and causes the decreases in the free volume and permeability,
this suggests that cross-linking of PTMSP inhibits the interchain diffusion and
stabilizes the gas permeability.

Comparison to energetic materials. Finally, we need to address the
potential application of azide-containing PTMSP as an energetic material. For

copolymer 6f. an activation energy of 130 kJ/mol was calculated from IR data using

148

the Arrhenius equation. This value compares well to the reported activation energy
of another azide containing polymer, GAP, which has an E8 = 120 kJ/mol.33 The
copolymers are stable under ambient conditions and are not nearly as “shock sensitive”
as trinitrotoluene (TNT) or nitroglycerine, which have activation energies of
approximately 30 kJ/mol and 2 kJ/mol respectively.34

The heat released during the thermal decomposition of azide in the
copolymers is comparable to that of other azide-based materials. The enthalpy of
azide decomposition for the copolymer with the highest azide content in the series,
6g, is -1.2 kJ/ g, as measured by DSC. Two other azide containing materials, GAP3 5
and poly(vinyl nitrate) partially substituted with azide (PVAZ),36 have reported
enthalpies of -1.8 and -2.7 kJ/g, respectively. The higher enthalpies for these two
polymers are due to the higher azide content incorporated in the structures. PTMSP
with increased azide contents (>20 mol%) could easily be prepared by increasing the
co-monomer feed ratio or by homopolymerization of the functional co-monomer.
The molar enthalpies for thermal decomposition of copolymers 6a-6g were also
calculated from the DSC results. All copolymers in this series evolved approximately
-800 kJ/mol N3. This value is comparable to the measured molar decomposition
enthalpy of a sample of PTMSP containing a known amount of non-volatile dodecyl

azide. Thus, the data indicate that the azide groups decompose independently and

the thermochemistry of the copolymers is additive.

149

4.5. Conclusions

Copolymers of l -(4-azidobutyldimethylsilyl)-l-propyne and l-trimethylsilyl-
1-propyne were prepared by functionalizing the bromobutyl side chain of poly[(]-
tri-methylsilyl-l-propyne)-co-(1-(4-bromobutyldimethylsi1yl)-1-propyne)]
copolymers. The amount of N3 in the copolymers was determined quantitatively by
elemental analysis, DSC, and IR spectroscopy. Compared to the cross-linking of
PTMSP by the physical addition of bis(aryl azide)s, copolymerization allowed
incorporation of a higher concentration of cross-linking sites without phase separation.
The measured activation energy and enthalpies of thermal decomposition of azide
contained in the series of copolymers are comparable to other reported azide
containing materials. Thermally induced cross-linking of the copolymer films resulted
in rigid, non-swelling materials at relatively low cross-link densities. Mechanically
stable membranes were prepared by thermal and photochemical cross-linking of azide-
functionalized PTMSP copolymers. The separation properties of the cross-linked
membranes are dependent upon the method of cross-linking and can be ascribed to
the amount of free volume in the cross-linked membranes. Measurements taken after
one month of storage under vacuum indicate that all cross-linked membranes exhibit

no decline in permeability.

150

4.6. References

1)Masuda, T.; Isobe, E; Higashimura. T.; Takada, K. J. Am. Chem. Soc. 1983, 105,
7473.

2)Shimomura, H.; Nakanishi, K.; Odani, H.; Kurata, M.; Masuda, T.; Higashimura,
T. Kobunshi Ronbunshu 1986, 43, 747.

3)Takada, K. M., H.; Masuda, T.; Higashimura, T. 1985, 30, 1605.

4)Masuda, T.; Higashimura, T. Silicon Based Polymer Science; Ziegler, J. M. and
Fearon, F. W. G., Ed; American Chemical Society: Washington, DC, 1990; Vol.
224, pp 641.

5)Morita, H.; Mori, S.; Uchino. N.; Yokoyama, S. Z. Phys. Chemie-lnt. J. Res.
Phys. Chem. Chem. Phys. 1993, 182, 209-216.

6)Murray, K. A.; Holmes, A. B.; Moratti, S. C.; Friend, R. H. Synth. Met. 1996, 76,
161-163.

7)Gonzalez, L.; Rodriguez, A.; deBenito, J. L.; MarcosFernandez, A. J. Appl.
Polym. Sci. 1997, 63, 1353-1359.

8)Yan, M. D.; Cai, S. X.; Wybourne, M. N.; Keana, J. F. W. J. Mater. Chem. 1996,
6,1249-1252.

9)Naito, 1.; Morihara, H. J. Imaging Sci. Technol. 1996, 40, 50-55.

10)Yamamoto, J.; Uchino, S.; Hattori, T.; Yoshimura, T.; Murai, F. Jpn. J. Appl.

Phys. Part 1 - Regul. Pap. Short Notes Rev. Pap. 1996, 35, 6511-6516.

151

11)Jia, J. R; Baker, G. L. J. Polym. Sci. Pt. B-Polym. Phys. 1998, 36, 959-968.
l2)Scriven, E. F. V.; Turnbull, K. Chem. Rev. 1988, 88, 297-368.
13)Abramovitch, R. A.; Kyba. E. P. The Chemistry ofthe Azido Group; Patai, 3.,
Ed.; Interscience Publishers: London, 1971, pp 221-330.

l4)Nagai, K.; Higuchi, A.; Nakagawa, T. J. Polym. Sci. Pt. B-Polym. Phys. 1995,
33, 289-298.

15)Nagai, K.; Nakagawa, T. J. Membr. Sci. 1995, 105, 261-272.

l6)Masuda, T.; Isobe, 13.; Higashimura, T. Macromolecules 1985, 18, 841.
17)Baker, G. L.; Klausner, C. F.; Gozdz, A. S.; Shelburne, J. A.; Bowmer, T. N.
Adv. Chem. Ser. 1990, 663-678.

18)Kraus, G.; Landgrebe, K. Synthesis 1984, 10, 885.

19)Kunzler, J .; Percec, V. New Polym. Mater. 1990, 1, 271.

20)Bowmer, T. N.; Baker, G. L. Polym. Prepr, ACS Div. Polym. Chem. 1986, 2 7,
218.

21)Nishiyama, K.; Karigomi, H. Chem. Lett. 1982, 9, 1477.

22)Gilbert, E. E. J. Polym. Sci. ,' Polym. Chem. Ed. 1984, 22, 3603.

23)Hidalgo, M.; Gonzalez, L.; Mijangos, C. J. Appl. Polym. Sci. 1996, 61, 1251-
1257.

24)Jayakrishnan, A.; Sunny, M. C.; Rajan, M. N. J. Appl. Polym. Sci. 1995, 56,
1187-1195.

25)Cohen, H. L. J. Polym. Sci. Poly. Chem. Ed. 1981, 19, 1337-1347.

152

26)Ravindranath, B.; Srinivas. P. 1nd. J. Chem. 1985, 248, 1178.

27)Gurst, J. E. The Chemistry of the Azido Group; Patai, S., Ed.; Interscience
Publishers: London, 1971, pp 191-202.

28)Gozdz, A. S.; Baker, G. L.; Klausner, C.; Bowden, M. J. Proc. SPIE 1987, 771,
18.

29)Tang, B. Z.; Masuda, T.; Higashimura, T.; Yamaoka, H. J. Polym. Sci. Pol.
Chem. 1989, 27, 1197-1209.

30)Burrell, H. Polymer Handbook; 2nd ed.; Brandrup, J. and Immergut, E. H., Ed.;
John Wiley & Sons: New York, 1975, pp IV-337.

31)Pleshakov, D. V.; Lotmentsev, Y. M. Vysokomol. Soedin. 1994, 36, 2026-2030.
32)Jia, J. Cross-linked Poly[1-trimethylsilyl-1-propyne]; Michigan State
University: East Lansing, Michigan, 1997.

33)Eroglu, M. S. Polym. Bull. 1998, 41, 69-76.

34)Johansson, C. H.; Persson, P. A. Detonics oingh Explosives; lst ed.;
Academic Press Inc.: New York, 1970.

35)Arisawa, 11.; Brill, T. B. Combust. Flame 1998, 112, 533-544.

36)Mishra, S. C.; Pant, J.; Dutta, P. K.; Durgapal, U. C. Def Sci. J. 1997, 47, 131-

137.

153

Chapter 5

PTMSP as a Permeable Solid Support in Oxygen Sensor Applications

5.1. Introduction
Quantitative detection of oxygen is important for medical applications,
control of industrial processes, and combustion. Electrochemical cells are commonly

used to detect oxygen at high temperatures,1'3

while ambient temperature sensors
are often based on the quenching of luminescence from organic and organometallic
chromophores.4’5 Optical schemes are particularly attractive since they can be
adapted for use in fiber optic geometries that are robust and insensitive to electronic
noise. However. a drawback of these schemes is that the chromophores often
decompose and have limited temperature ranges.

In contrast, the physical properties and photophysics of M06Cl12 and related
metal clusters are well suited for oxygen sensing schemes. The photosensitive
MoéCl12 cluster, shown in Figure 5.1, is composed on an octahedral core of
molybdenum atoms with eight face bridging chlorides and four axial chlorides, which
are shared among neighboring cluster subunits to form a (M06C112)n polymeric

6 Pumping the clusters with 300 to 400 nm radiation generates a strong red

structure.
luminescence that is efficiently quenched by ground state 302. The clusters show no

signs of degradation and are thermally stable to 600 °C. The observed emission

intensity (I) from the clusters is related to the concentration of quencher, [Oz], by the

154

 

 

 

 

 

 

 

 

 

Figure 5.1. Structure of (Mo(,Cl.2)n clusters.

155

Stern-Vollmer equation (Equation 1), where K q is the product of the fluorescence

7

lifetime and the quenching rate constant of the cluster. Shown in Figure 5.2 is the

relationship between signal intensity and quencher concentration for the molybdenum

lo

‘1 - 1 + qu021 (1)

clusters in solution.8

The large changes in signal intensity with small changes in the
oxygen concentration indicate that these clusters would be ideal probes for sensing
oxygen in the range of 10'4 to 10'1 M.

To adapt Mo6Cl12 to fiber optic sensing, the clusters need to be immobilized
at the end ofa silica optical fiber. Polymers and silica have been used to immobilize
Mo clusters for oxygen detection and as sensitizers for 1O2 generation,8'10 but for
gas sensing applications, the polymer matrix needs to have a high permeability to
ensure fast response times.

Because of its attractive physical properties and high oxygen permeability,
poly(l-trimethylsilyl-l-propyne) (PTMSP) should be an excellent support for fiber
optic based sensing applications. PTMSP is an amorphous polymer with a low density
and a large free volume caused by the inefficient packing of its rigid random coil

backbone. Its permeability coefficient for oxygen is ten times greater than

poly(dimethylsiloxane), the polymer typically used as the benchmark for high

156

 

100

80

60~

Intensity (%)

40~

20

 

 

 

 

0 4 W em
LO v to N "‘ o
C? c? 9 9 c? S?
LU Lu LIJ uJ UJ w
‘— v- 1- "— ‘— ‘_

[02] M

Figure 5.2. Stern-Vollmer plot of emission intensity versus [02].

157

permeability. 11 PTMSP is tough and mechanically stable, with a modulus between
that of typical amorphous materials and rubbery polymers. PTMSP exhibits no glass
transition below its degradation temperature (350 0C), is relatively chemically inert,
and is optically clear with no absorption above 280 nm. Transparent films can be
produced by spin casting or solvent evaporation

Combining the optical properties of MoéCl12 clusters with the physical
properties of PTMSP should lead to a composite ideal for fiber optic based oxygen
sensing applications. This chapter describes the effectiveness and stability of PTMSP
as a solid support, and the procedure used to prepare and bind the PTMSP/Mo‘sCl12
composite to the end of a large core silica optical fiber. This work led to the
development of a reflection mode fiber optic capable of the real time monitoring of

oxygen in 0 - 21% gaseous environments.

158

5.2. Experimental Section

Materials. The preparation of PTMSP, poly[(l-trimethylsilyl-1-propyne)-
co-(l-(4-bromobutyldimethylsilyl)-1-propyne)] and poly[(l-trimethylsilyl-1-
propyne)-co-(l -(4-azidobutyldimethylsilyl)-1-propyne)] copolymers are described in
Chapter 4. MoéCl12 was prepared according to literature procedures.2’6’12
Tetrahydrofuran (THF) and toluene were purified by distillation from calcium hydride
followed by a second distillation from sodium/benzophenone ketyl. Unless otherwise
noted, all other reagent grade materials were purchased from commercial suppliers
and used without further purification.

Characterization. Proton nuclear resonance (‘H NMR) spectra were
measured using a Varian Gemini-300 spectrometer at 300 MHz. All samples were
run at room temperature in CDC13. Chemical shifts were calibrated using residual
CHCl3 and are reported in ppm (6) relative to tetramethylsilane. Infrared spectra
were obtained under nitrogen at room temperature on a Nicolet Magna-IR 550 Fourier
Transform IR spectrometer. A Hitachi U-4001 UV—visible spectrometer was used to
obtain the UV—visible spectra of cyclohexane solutions of the polymers and acetonitrile
solutions of MobCllz. Fluorescence spectra of the molybdenum clusters dissolved in
acetonitrile were obtained using a Hitachi F-4500 fluorescence spectrophotometer.
Molecular weights of the polymers were determined by gel permeation

chromatography (GPC) using a PLgel 2011 Mixed A column and a Waters R401

Differential Refractometer detector at room temperature with THF as eluting solvent

159

at a flow rate 1 mL/min. Monodisperse polystyrene standards were used to calibrate
the molecular weights. The concentration of the polymer solutions used for GPC
measurements was 1 mg/mL. Thermogravimetric analyses (TGA) were performed
under nitrogen and air atmospheres at a heating rate of 10 °C/min on a Perkin-Elmer
TGA 7 instrument.
Poly[(l-trimethylsilyl-l-propyne)-co-(1-(4-cyanobutyldimethylsilyl)-l-
propyne)]. Poly[( 1 -trimethylsi1yl- 1 -propyne)-co-(1 -(4-bromobutyldimethylsilyl)-1 -
propyne)] (80 mg, 0.6 mmol Br) was dissolved in toluene (5 mL). Sodium cyanide
(NaCN, 0.294g, 6 mmol) and a catalytic amount of tetra-n-butylammonium iodide
(n-Bu4NI, 0.022g, 0.06 mmol) were dissolved in 1 mL distilled water and added to
the polymer solution. The flask was fitted with a water condenser and placed into an
oil bath heated to 105 °C. The mixture was stirred vigorously overnight. Upon
cooling, water (50 mL) was added and the polymer was extracted with methylene
chloride (3 x 50 mL). The combined organic layers were dried over magnesium
sulfate, filtered, and the solvent removed under reduced pressure. The polymer was
redissolved in toluene and precipitated dropwise into an excess of methanol. The
polymer was collected by filtration to yield 48 mg (60%). GPC: Mn = 400,000. Mw/
Mn = 2.3. lH NMR: 5 0.2 (br), 1.6 (br), 1.8 (br), 2.3 (br). IR: CzN, 2250 cm".
Anal. Calcd. for (C6.8H13N0.28i)n: C, 64.99; H, 10.43; N, 2.23; Si, 22.35. Found: C,

63.23; H, 10.67; N, 1.91; Si, not determined.

160

Casting/dipping solutions. Solutions used to coat fibers and cast films
were prepared by dissolving MoéCll2 (300 mg, 0.3 mmol) in 300 mL of spectroscopic
grade acetonitrile. The solution was concentrated to give a yellow oil, presumably
the corresponding acetonitrile adduct of the molybdenum complex. A solution of
PTMSP in tetrahydrofuran was added to the yellow oil. After filtration through a 2
pm filter, the solution was concentrated to the desired viscosity (typically 1-2 wt %)
with a steady stream of dry nitrogen and gentle heating.

Film casting. Films used for measuring the optical properties of the
composites were spin cast from the previously described solutions onto SuprasilTM
high-purity quartz slides (1x2 cmz) inside a He atmosphere drybox. Spinning rates
ranged from 1000 to 2000 rpm depending upon the viscosity of the sample. Film
thickness was measured using a Dektak SFM instrument.

Fiber coating. The sensor was fabricated from a commercially available
multimode fiber, FT-1.5-UMT from 3M Specialty Optical Fibers, which is designed
for UV to visible transmission. The large core diameter (1 500 um) and high numerical
aperture (0.39) of the fiber allows for efficient excitation and collection of the
luminescence from clusters immobilized at the back face of the fiber. The exposed
surfaces of the cleaved 1.5-meter length fibers were cleaned in a UV-ozone
photoreactor for 30 min, and then immediately transferred into a He-atmosphere
drybox. A thin primer layer of poly[(l-trimethylsilyl-l-propyne)-co-(1-(4-

azidobutyldimethylsilyl)-1-propyne)] (azide content 15%) was applied by dip coating

161

the fiber tip into a 1 wt% solution in toluene. After drying, the primer layer was
cross-linked at 250 0C for 3 hours. A transparent, uniform bead (~ 100 pm) of
Mo6C112/polymer composite was then applied by dipping the primed fiber tip into a
2 wt% casting solution. The composite bead was allowed to dry overnight and then
was dried under vacuum at ambient temperature for 1 hour.

Detection apparatus. The coated fiber tip was secured in a glass cell by a
rubber septum that was fed through from the front end of the fiber. The “flow-
through” cell was equipped with a gas inlet and outlet in order to control the
environment around the sensor. Radiation from a 325 nm He/Cd laser was coupled
via a microscope objective into the front end of the fiber to provide the incident
pump beam. A maximum of 2.5 mW could be coupled into the fiber. The reflected
luminescence was collected at the front end of the fiber with either a photodiode or
photomultiplier tube. Separation of the signal and pump beams was accomplished

using a 45° long wave pass (LWP) dichroic beamsplitter and a 630nm LWP filter.

162

5.3. Results

To use PTMSP as a permeable support for MoéCl12 clusters, it is necessary
that the polymer does not interfere with the optical properties of the clusters. As
shown in Figure 5.3, PTMSP does not absorb above 280 nm.13 Therefore, the support
matrix will not absorb the excitation energy of the molybdenum clusters at 325 nm.
Separation of the emission radiation from the initial excitation energy is also simplified
due to the large red shift of the cluster emission.

PTMSP and Mo6Cl12 must be miscible in order to form homogeneous
composites that possess good optical properties. However, PTMSP is soluble in
non-polar solvents while MoéCll2 clusters are only soluble in polar solvents such as
methanol and acetonitrile. Only THF was found to efficiently dissolve both. Our
initial scheme involved physically dispersing and suspending M06C112 in the polymer
by mixing solutions of the polymer and salt in THF and concentrating the solutions
until viscous. The casting solutions were prepared to yield composites with ~10%
of the volume occupied by the clusters. Unfortunately, during film formation, M06C l12
clusters precipitated in the polymer matrix creating heterogeneous films with poor
optical properties. In THF, two of the bridging chlorides are replaced by solvent,
and the (M06C112)n polymer clusters dissolve as discrete clusters. But in the absence
of suitable ligands, the clusters reassemble through the bridging chlorides as the
solvent evaporates. Attempts to kinetically control the crystallization by speeding
up the evaporation rate (i.e. increasing the spinning rate) resulted in films which

were too thin with salt concentrations below limits of detection.

163

 

I nte ns ity

 

 

 

L

200 300 400 500 600 700 800 900
wavelength (nm)

 

Figure 5.3. The UV/V is absorption spectra of a) PTMSP and b) Mo6C112, and
the fluorescence spectra of c) M06C112.

164

Similar metal clusters have been dissolved in various polymers by utilizing

the axial coordination sites to ligate the cluster core to the polymeric backbones"

10’14’15 Dissolving the clusters in PTMSP could be accomplished by using a PTMSP
copolymer with side chains that contain polar ligands capable of solvating the
molybdenum salt. Since the photophysics of interest is due to electronic transitions
in the core of the cluster, attachment of the cluster to a polymer matrix does not add
non-radiative pathways to the lumophore. Of the ligand-containing copolymers that
were investigated, a nitrile containing copolymer proved to be the most chemically
stable and easiest to synthesize.

The nitrile-containing copolymer was synthesized from a brominated PTMSP
copolymer, poly[(l -trimethylsilyl-1 -propyne)-co-(1 -(4-bromobutyl-dimethylsilyl)-1-
propyne)], with 20% of the repeat units having bromobutyl side chains. 16 As shown
in Figure 5.4, treatment of the copolymer with sodium cyanide and a phase transfer
agent smoothly converted the bromobutyl side chain to the corresponding nitrile.
The reaction was analyzed by 1H NMR and FTIR spectroscopy to determine the
conversion. In the IR spectra, the intensity of the band at 1219 cm" decreased (CHZBr
group), and a new band grew in at 2250 cm'1 corresponding to the -CEN group in
poly[(l-trimethylsi1yl-1-propyne)-co-(l-(4-cyanobutyldimethylsilyl)-1-propyne)].
Corresponding changes were seen in the 1H NMR data. The peak at 3.4 ppm from
the -CH2Br segment in poly[(l-trimethylsilyl-l-propyne)-co-(1-(4-

bromobutyldimethylsilyl)-1-propyne)] decreased as a new resonance at 2.3 ppm for

165

CH3 CH3

W
CH3‘ISI'CH3 CH3‘ I I‘CH3

CH3 CH2CH2CH20HzBr

NaCN, n-Bu4NI
toluene, H20

CH3 CH3

W m
CH3-Si-CH3 CH3- Ii—CH3
CH3 CHZCHZCHZCHZCN

Figure 5.4. Synthesis ofpoly[(1-trimethylsilyl-1-propyne)-co-(1-
(4-c yanobutyldimethylsilyl)-1 -propyne)].

166

-CH2CN in poly[(l -trimethylsilyl-1-propyne)-co-(1-(4-cyanobutyldimethylsilyl)-1-
propyne)] increased. The peak at 3.4 ppm in the 1H NMR spectra and the absorption
band at 1219 cm“l in the IR spectra both decreased to baseline, demonstrating that
complete displacement of the bromide was achieved.

The copolymer was redissolved in toluene and stirred over powdered
molybdenum clusters for several days. The solution was then filtered to remove the
undissolved clusters and concentrated to a viscous solution. The luminescent
properties of the MoéCln/nitrile copolymer composites were characterized by
fluorescence measurements of films spin cast onto quartz slides. The amount of
cluster dissolved in the copolymers was too low to be detected, indicating that either
the nitrile ligands were not strong enough to displace the bridging chloride ligands
or the concentration of the pendant nitriles was too low to ligate a useful quantity of
cluster. IR spectra of the composite films indicated no shift in the frequency of the
nitrile absorption band due to coordination with the cluster.

A casting solution of the nitrile copolymer containing a higher concentration
of clusters was prepared in THF as was previously done for PTMSP/Mo6Cll2
composites. Although films cast from this solution still exhibited some phase
separation, modulation of the fluorescence from the clusters was observed in air and
argon environments (Figure 5.5). Quenching of the fluorescence by oxygen in air

resulted in a 5x decrease in the luminescence intensity.

167

 

Reelative intensity

 

 

 

 

600 650 700 750 800 850

Wavelength (nm)

Figure 5.5. Luminescence spectra from a MoéCllz/nitrile-PTMSP copolymer
composite film in laboratory air (bottom curve) and under Ar (top curve).

168

The problems of low cluster concentration and phase separation in the
composites were overcome by preparing a soluble derivative of (M06C112)n that can
co-dissolve with PTMSP in high concentration. Although the clusters are soluble in
both THF and acetonitrile, the clusters precipitate from THF after evaporation
reforming a yellow powder. However after evaporation of acetonitrile, a soluble
yellow oil formed. The ideal structure of the adduct is M06C112(CH3CN)2, but it is
unlikely that acetonitrile efficiently forms single solvated clusters. The only reported
examples of isolated single clusters were synthesized by replacing the bridging
chlorides with covalently bound ligandslo’l7 We assume that acetonitrile merely
acts as a temporary ligand, solvating small domains of clusters that once dispersed
in the PTMSP framework are unable to phase separate. Films prepared on quartz
slides using casting solutions containing the acetonitrile solvated clusters exhibited
high intensity emission due to the high cluster loading and minimal loss of signal
due to scattering caused by phase separation.

The composites showed poor adhesion to freshly cleaved fiber tips, and
peeled off the fiber upon drying. To achieve better adhesion to the fiber, we first
applied a primer layer to the fiber tip to enable better wetting by the composite. The
primer layer consisted of a cross-linkable PTMSP derivative, poly[( 1 -trimethylsilyl-
1-propyne)-co-(1-(4-azidobutyldimethylsilyl)-1-propyne)], which after heating,
adhered to the fiber surface. Uniform beads of the PTMSP/M06Cl12 composite were

then applied to the fiber tip by repeated clipping of the tip into the viscous solutions.

169

After drying, the beads of MoéCl I2/PTMSP composite were mechanically stable and
adhered well to the fiber end.

The measured emission intensity from the coated fiber (Fiber 1) in varying
oxygen concentrations is shown in Figure 5.6. Although steps in the observed intensity
are seen as the atmosphere is changed from pure oxygen to nitrogen, re-introduction
of nitrogen into the cell resulted in a positive spike in the signal that decayed to the
previous intensity level after a short period of time. We propose that this is due to
the presence of additional quenching species. As shown in Figure 5.7, the PTMSP
backbone is susceptible to radical chain oxidation, and the degradation by-products
may act as efficient quenchers of the excited clusters. The degradation pathway can
be initiated by a radical source generated either chemically or photochemically. Upon
irradiation of the composite in nitrogen, a steady state amount of photooxidation
products in the polymer is generated, causing a decay of the observed emission
intensity. Introduction of oxygen accelerates the chain degradation process, forming
additional quenching species that further lower the emission intensity. Although
reintroduction of nitrogen into the cell results in the maximum emission intensity,
continued chain degradation reforms the steady state amount of quenchers, decreasing
the observed intensity.

Investigation of the casting solutions revealed that the tetrahydrofuran used
to dissolve the composite tested positive for peroxides. After evaporation of the

solvent, a ~4% non-volatile weight residue remained consisting of peroxidized and

170

 

800
02 02 02 02

7°07 1111
600— M\\

500
400 .

300 -

Intensity (counts)

200 *

100~T T

N2 N2 N2 N2

 

 

 

 

 

 

 

_.1
_,

 

 

 

time (min)

Figure 5.6. Observed emission from Fiber 1 in varying
oxygen/nitrogen environments.

171

TMS TMS TMS
M N
——->
CH3 'CH2 00—O/CH2
_ (H' ) PTMSP
+ (H°)
TMS TMS
HO . N ._ (N/
CH CH
-o’ 2 Ho—o’ 2

Figure 5.7. Peroxide induced chain degradation of PTMSP.

172

polymerized THF. When used for the preparation of the composites, this non-volatile
material remained in the composite after drying, aiding in the degradation. To avoid
this contamination, freshly distilled THF was passed through an activated alumina
column immediately before use.

New composite coated fibers prepared using peroxide-free solvent were
tested in varying oxygen concentration environments (Fiber 2). The intensity of the
observed emission decreased in a stepwise manner as the concentration of oxygen
surrounding the probe was systematically increased over time (Figures 5.8 and 5.9).
The emission intensity stepped up to its original level upon reintroduction of pure
nitrogen into the cell. The data indicated no spiking due to peroxide formation and
polymer degradation, as was seen in the previous samples. When the raw data is
fitted to the Stern-Vollmer equation, the value of K q extracted from the experimental
data from the solid state sensor probe is comparable to the quenching constants

7 indicating that the clusters behave similarly in PTMSP

calculated from solution data,
(Figure 5.10).

To determine the sensor’s operating temperature range, thermogravimetric
analysis was performed on the components of the fiber sensors. Pieces of the fiber
jacket and inner cladding were removed from the fiber core. The outer jacket is
thermally stable to 325 °C under nitrogen, at which point it begins to yellow and

becomes brittle. The inner cladding is slightly less thermally stable, rapidly degrading

at 275 0C under nitrogen. The PTMSP/MoéCl12 composites are thermally stable to

173

 

700

600

500 .

A
O
O

300

 

Intensity (counts)

200 - a

 

100 . C

 

 

 

0 10 20 30 40
time (min)

Figure 5.8. Observed emission intensity from Fiber 2 in nitrogen and a) 2.1%,
b) 4.1%. c) 6.2%, d) 12.4%, and e) 20.6% oxygen containing environments.

174

800

 

700 - N2

600 »

01
O
O

400 ~

 

 

Intensity (counts)

300 ~

200 .

100 r

 

 

 

0 10 20 30 40

time (min)

Figure 5.9. Observed emission intensity from Fiber 2 in nitrogen and a) 0.1%, b)
0.2%, c) 0.4%, d) 0.6%, e) 0.8% and f) 1.0% oxygen containing environments.

175

Normalized intensity (1/10)

0.8 <

0.6

0.4

0.2

 

 

 

 

O
O
O
O
O
O
O
O

“of 3 8 8' ‘5
LU. Lb. Lb. u'J. u'J.

02 [M]

Figure 5.10. Stem—Vollmer plot of emission data from Fiber 2.

176

~350 °C under nitrogen, but begin to degrade about 10 degrees lower in air.
Interestingly, the composites form significant amounts of residue (15-20 wt%) that
are thermally stable past 400 °C. Under nitrogen, PTMSP alone completely degrades
by 400 oC, leaving no residue. In air however, PTMSP exhibits only a 10 wt%
residue at 400 °C, presumably due to the formation of cross-linked material. These
data indicate that the added MoéCl12 clusters provide an alternative degradation
pathway leading to thermally cross-linked material.

The stability of the composites at elevated temperatures was also investigated
over an extended period of time. PTMSP and the PTMSP/M06Cl12 composites were
isothermally heated at 200 °C under nitrogen and air for 4 hours. Under inert
conditions PTMSP undergoes a minor weight loss (0.7 wt%), but in air the weight
loss increases (2.5 wt%) and the sample shows signs of yellowing. The composites
lose 1.4 wt% under nitrogen, of which a small portion of this weight loss could be
attributed to the loss of acetonitrile ligands. In air, the composite loses 2.6 wt% and
begins to brown. From the thermal analysis results, the upper temperature limit of

the sensor is 200 °C, currently limited by the stability of the composite matrix.

177

5.4. Discussion

Phase separation. A major problem encountered during the design of the
fiber optic sensor was the phase separation of the molybdenum clusters from the
polymer matrix. The molybdenum clusters favor reforming its polymeric structure
after removal of solvent. Incorporation of the clusters into a PTMSP copolymer
containing nitrile ligands proved ineffective, but preparation of a soluble adduct that
could be physically dispersed in PTMSP allowed the formation of transparent samples.
By replacing the bridging chlorides with acetonitrile ligands, the clusters did not
phase separate from the matrix after removal of solvent. Transmission electron
microscopy (TEM) was performed on the samples to determine the size of the clusters
imbedded in the composite. However, the concentration of clusters was too high to
yield useful data.

Permeability of composite. PTMSP was chosen as the polymer support
due to its mechanical properties and high gas permeability. The addition of
molybdenum clusters will no doubt occupy some of the free volume of the matrix,
and as a result, lower the permeability. The oxygen permeability of the composite
was not measured because of the difficulty of preparing large, uniform membranes
free of defects. An estimate of the permeability can be made by comparing the
composite to the PTMSP membranes containing added bis(aryl azide).13 Since the
clusters and the bis(aryl azide)s are roughly the same size and the wt% amounts

added are similar, the permeability through the composite should be comparable.

178

Degradation of composite matrix. The radical chain degradation process
of PTMSP is a serious limitation to the lifetime ofthe composite based sensor. Small
traces of oxidation in the polymer can lead to significant decreases in the molecular
weight after a short period of time. After approximately 2 hours of testing, a sample
of polymer was extracted off a sensor fiber tip. The molecular weight of the sample,
as measured by GPC, dropped from 1,000,000 g/mol to 50,000 g/mol after testing.
The sample still possessed good mechanical properties, but extended exposure would
likely decrease the molecular weight below a usable level. More importantly, this
degradation pathway introduces additional quenching species that interfere with the
optical properties of the clusters. Sensor aging could be seen after prolonged cycling
between oxygen and nitrogen. The signal intensity weakened and the depth of
modulation needed to detect small changes in oxygen concentration decreased. The
use of peroxide-free solvents to prepare the samples decreased the rate of sensor
aging, but residual peroxides can lead to significant degradation over time.

Cross-linking. To expand the number of potential applications of these
sensors, schemes to make the composite impervious to solvents would be desirable.
This could be accomplished by using a cross-linkable polymer support. Fibers were
prepared using composite solutions of poly[(l-trimethylsilyl-1-propyne)-co-(l-(4-
azidobutyldimethylsilyl)-1-propyne)] and Mob-C112. We were able to adhere
transparent beads of composite to the fiber tips, but during the cross-linking procedure,

the samples severely darkened. The emission from these composite coated fibers

179

lacked the depth of modulation previously observed in the MoéCl] 2/PTMSP composite
coated fibers.

Although we are unable to cross-link the composite using azide chemistry,
the sensors made from the PTMSP/MoéCl12 composites may also operate in aqueous
conditions. Since PTMSP does not swell in water, the large molybdenum clusters
are unlikely to leach out of polymer matrix. This may expand the potential uses of
the sensor to biological and biomedical applications, such as monitoring oxygen

levels in blood.

180

 

5.5. Conclusions

A fiber optic oxygen sensor based on oxygen quenching of the red emission
from molybdenum halide clusters was developed. PTMSP served as a mechanically
stable, permeable support for the luminescent probes. Phase separation of the
molybdenum clusters from the polymer matrix was overcome by forming a soluble
acetonitrile adduct that dissolved in the polymer. Distinguishable steps in the emission
intensity from the MoéCllz/PTMSP composite coated fiber probes were observed in
varying oxygen concentration environments ranging from 0.1 to 20%. Fast response
times due to the highly permeable matrix were observed. The sensors do exhibit
limited operating lifetimes however due to the photooxidation of the polymer matrix.

Sensors capable of detecting oxygen at lower concentrations could be
fabricated using other molybdenum clusters (i.e. MoéClM'z) that have different
quenching constants (Kq). Theoretically, a family of sensors capable of monitoring
a wide gamut of gaseous oxygen concentrations could be produced by this method.

The present probe is thermally stable to at least 200 °C, limited by the polymer
matrix. Since the Mob—Cl12 clusters are synthesized at 1000 °C, immobilization of
the clusters in an inert, thermally stable sol-gel matrix is an attractive alternative. 1 8’ 1 9

This would enable oxygen sensing in gaseous environments at elevated temperatures

as well as liquid environments with varying salinity and pH.

181

5.6. References

1)Yamamoto, K.; Kasuga, T.; Nogami, M. Appl. Phys. Lett. 1998, 73, 3297-3299.
2)Mussell, R. D.; Nocera, D. G. J. Phys. Chem. 1991, 95, 6919-6924.

3)Azad, A. M.; Akbar, S. A.; Mhaisalkar, S. G.; Birkefeld, L. D.; Goto, K. S. J.
Electrochem. Soc. 1992, 139, 3690-3704.

4)McNamara, K. R; Li, X. P.; Stull, A. D.; Rosenzweig, Z. Anal. Chim. Acta 1998,
361, 73-83.

5)McNamara, K. P.; Rosenzweig, Z. Anal. Chem. 1998, 70, 4853-4859.

6)Schafer, V. H.; Schnering, H. G.; Tillack, J.; Kuhnen, F.; Wohrle, H.; Baumann,
H. Z. Anorg. Chem. 1967, 353, 281.

7)Jackson, J. A.; Turro, C.; Newsham, M. D.; Nocera, D. G. J. Phys. Chem. 1990,
94, 4500-4507.

8)Jackson, J. A.; Newsham, M. D.; Worsham, C.; Nocera, D. G. Chem. Mat. 1996,
8, 558-564.

9)Robinson, L. M.; Shriver, D. F. J. Coord. Chem. 1996, 37, 119-129.
10)Robinson, L. M.; Lu, H.; Hupp, J. T.; Shriver, D. F. Chem. Mat. 1995, 7, 43-49.
ll)Shimomura, H.; Nakanishi, K.; Odani, H.; Kurata, M.; Masuda, T.;
Higashimura, T. Kobunshi Ronbunshu 1986, 43, 747.

12)Mussell. R. D._; Nocera, D. G. Inorg. Chem. 1990, 29, 3711-3717.

l3)Jia, J. R; Baker, G. L. J. Polym. Sci. Pt. B-Polym. Phys. 1998, 36, 959-968.

182

l4)Sergeev, G. B.; Petrukhina. M. A. Prog. Solid State Chem. 1996, 24, 183-211.
15)Golden, J. H.; Deng, H. B.; Disalvo, F. J.; Frechet, J. M. J.; Thompson, P. M.
Science 1995, 268, 1463-1466.

16)Ruud, C. J.; Jia, J.; Baker, G. L. submitted to Macromolecules 1999.
l7)Ehrlich, G. M.; Warren, C. J.; Haushalter, R. C.; Disalvo, F. J. Inorg. Chem.
1995, 34, 4284-4286.

l8)McDonagh, C.; MacCraith, B. D.; McEvoy, A. K. Anal. Chem. 1998, 70, 45-50.

l9)Lin, J.; Brown, C. W. True-Trends Anal. Chem. 1997, 16, 200-211.

183