5 PLACE N RETURN BOX to roman this checkout from your record.
5 TO AVOID FINES Mum on or More data duo.

DATE DUE DATE DUE DATE DUE

___Jl
l-___:
m fl
__J|__L__J
ICJI ICE
l|___J:]
Jl I I

MSU I. An Afflrmatlvo Action/Equal Opportunity Institution
m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HIS-9.1

 

 

COMPOSITE POLYMER ELECTROLYTES USING FUNCTIONALIZED FUMED
SILICA AND LOW MOLECULAR WEIGHT PEO: SYNTHESIS AND
CHARACTERIZATION
By

Jun Hou

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department Of Chemistry

1997

ABSTRACT

COMPOSITE POLYMER ELECTROLYTES USING FUNCTIONALIZED
FUMED SILICA AND LOW MOLECULAR WEIGHT PEO:
SYNTHESIS AND CHARACTERIZATION

By
Jun Hou

A new class of highly conductive and processable composite polymer electrolytes
for use in rechargeable lithium batteries have been synthesized. The principal characteristic
of these systems is the use of surface functionalized firmed silica fillers to control the
mechanical properties of the composites. When a low molecular weight polyether solution
is immobilized in fumed silica, an open three-dimensional network forms due to van der
Waals attractions Or hydrogen bond linkages between silica particles. The open network
structure is conducive to high ionic mobility and responsible for the reinforcement of the

composite.

The most promising electrolytes are cross-linkable composite polymer electrolytes
that incorporate a silica that bears both hydrophobic and cross-linkable fimctional groups.
Hydrophobic groups contribute to the formation of open networks, while cross-linking
groups provides mechanical strength and dimensional stability through the formation of

covalent bonds between silica particles.

Cross-linkable composite electrolytes were prepared from PEGDM-SOO, LiClO4,
firmed silica, and 10 wt% of methyl, butyl, or octyl methacrylate. Before cross-linking, the
electrolytes were thixotropic and had ionic conductivities >10"4 S/cm. Afier UV-induced
cross-linking, the electrolytes were rubbery and dimensionally stable. DSC results showed
that the addition of silica and monomer to the polyether-Li salt matrix did not affect the
glass transition of polyether. In addition, the conductivities were unchanged according to
impedance spectroscopy. After solvent extraction of the cross-linked composites to
remove the polyether, lithium salt, and homopolymer, the solid phase was analyzed. The
FTIR spectrum showed qualitatively the grafiing of the guest polymer, and TGA analyses
demonstrated that nearly all the polymer had been chemically bound to the silica surface.
These data all support a model where the added methacrylate monomer and initiator phase
' separate from polyether-Li salt matrix and form a thin layer surrounding the silica
network. Upon photoinitiation, the added monomer copolymerizes with cross-linkable
groups on silica surface to yield a silica/polymer phase and little or no polymer in the
PEGDM-SOO phase. Thus, the mechanical properties of the composite electrolyte and its

ionic conductivity are de-coupled, and can be optimized independently.

To Hong and Annie.

iv

ACKNOWLEDGMENTS

I wish to express my deep appreciation to Professor Gregory L. Baker for his
guidance, patience and constant encouragement throughout the course of this research. I
would also like to thank Professors Chris Chang, Robert Cukier, Mercouri Kanatzidis, and
Robert Hollingsworth for serving in my committee. I would also like to thank all members
in Baker group for making my life in the lab enjoyable.

Financial support from Department of Chemistry and the U. S. Department of
Energy is also acknowledged.

Finally, I want to thank my wife Hong for her support and bringing to this world

our daughter Annie with whom life had never been the same.

TABLE OF CONTENTS

Page
List of Abbreviations .................................................................................................. xi
List of Schemes ........................................................................................................... xii
List of Tables ............................................................................................................... xiv
List of Figures ............................................................................................................. xv
INTRODUCTION ....................................................................................................... l
I. Electrolyte Systems .............................................................................. 1
1. Primary and Secondary Batteries .......................................... l
2. High-Energy-Density Battery ............................................... 2
3. Electrolytes for Lithium Batteries ......................................... 4
4. New Polymer Electrolyte Systems ........................................ 11
II. Photoinitiaticn of Polymerization .......................................................... 29
1. Theory ................................................................................. 29
2. Polymerization Initiators ...................................................... 32
3. UV Induced Polymerization ................................................. 38
4. Inhibition and Retardation .................................................... 40
RESULTS ................................................................................................................... 44
I. Silylation Conditions ............................................................................. 46
1. General ................................................................................ 46
2. Amounts of Amine Catalyst .................................................. 47
3. Reaction Time ...................................................................... 47
4. Amount and Type of Silane .................................................. 51
H. Hydrophobic Silica ............................................................................... 54
III Hydrophilic Silica ................................................................................. 67

IV. Polymerizable Silica .............................................................................. 74

1. Silylation Using Single Silanes .............................................. 74

2. Silylation Using Mixed Silanes ............................................. 82

V. Composite Polymer Electrolytes ........................................................... 92

1. Compatible Composite Polymer Electrolytes ........................ 92

2. Cross-linked Composite Polymer Electrolytes ...................... 94
DISCUSSION ............................................................................................................. 121
I Current System. .................................................................................... 12 1
1. Composite Reinforcing mechanism: phase separation ............ 121

2. Cross-Linking and Emulsion Polymerization: A

Comparison ............................................................................... 132
11. Future Work and Suggestions ............................................................... 138
1. Optimization of Current System ........................................... 138
2. Core-active Initiators ............................................................ 138
3. Two-Component Core-Shell Composites ............................. 139

4. Highly Dispersible & Cross-Linkable Silicas for

Emulsion Polymerization ........................................................... 141

III. Summary .............................................................................................. 143
EXPERIMENTAL ...................................................................................................... 145
1. Materials .............................................................................................. 145

11. Analytical Methods ............................................................................... 146

III. Hydrophobic Silicas ............................................................................... 149
FS-TEMS-l ................................................................................. 149

FS-TEMS-Z ................................................................................. 149

FS-TEMS-3 ................................................................................. 149

vii

F S-TEMS-O ................................................................................. 149

FS-TEMS-ZH .............................................................................. 149
FS-TEMS-13H ............................................................................ 149
FS-TMCS-l ................................................................................. 1 50
FS-TMCS-Z ................................................................................. 150
TMCE-3 ...................................................................................... 150
TMES-l ...................................................................................... l 50
F S-C2 .......................................................................................... 150
FS-C8 .......................................................................................... 151
FS-C 1 8 ........................................................................................ 1 5 1
FS-DM-CS-l ............................................................................... 1 5 1
FS-DM-C8-2 ............................................................................... 1 5 1
F S-DM-C8-3 ............................................................................... 151
FS-DM-H ............................. . ...................................................... 152
FS-H ........................................................................................... 1 52
FS-OEt-H .................................................................................... l 53
FS-OH-H ..................................................................................... l 53
F S-TMS-H .................................................................................. 153
IV. PEG-Modified Silicas ........................................................................... 155
Ethylene Glycol Allyl Methyl Ether (EGAME) ............................. 155
Diethylene Glycol Allyl Methyl Ether (DEGAME) ....................... 155
Tiethylene Glycol Allyl Methyl Ether (TEGAME) ........................ 156

Poly(ethylene glycol-350) Allyl Methyl Ether (PEGAME-
350) ............................................................................................. 156
Poly(ethylene glycol-550) Allyl Methyl Ether (PEGAME-
550) ............................................................................................. 156

viii

Poly(ethylene glycol-750) Allyl Methyl Ether (PEGAME-

750) ............................................................................................. 156
Chlorodimethyl(4,7-dioxaoctyl)silane ........................................... 1 57
Chlorodimethyl(4,7,lO—trioxaundecyl)silane ................................. 157
Chlorodimethyl(4,7, 10,13-tetraoxatetradecyl)silane ...................... 1 57

Poly(ethylene glycol-3 50) methyl 3-
(chlorodimethylsily)propyl ether (PEGMPE-350) ......................... 158
Poly(ethylene glycol-550) methyl 3-
(chlorodimethylsily)propyl ethers ................................................. 158
Poly(ethylene glycol-750) methyl 3-

(chlorodimethylsily)propyl ethers ................................................. 158
F S-DM-EOI ................................................................................ 158
FS-DM-EOZ ................................................................................ 159
FS-DM-EO3 ................................................................................ 159
F S-DM-E08 ................................................................................ 159
FS-DM-EOIZ .............................................................................. 159
FS-DM-EO] 7 .............................................................................. 159
FS-EOI ....................................................................................... 159
FS-EO3 ....................................................................................... 160
FS-EO8 ....................................................................................... 160
FS-EOIZ ..................................................................................... 160
FS-DM-S—EOI ............................................................................ 160
FS-DM-S-EO8 ............................................................................ 160
V. Polymerizable Silicas ............................................................................ 161
5-(Tetrahydropyran-2-yloxy)pent-1-ene ....................................... 161
2—(8-Chlorooctyloxy)tetrahydropyran ........................................... 161
2-(7-Octenyloxy)tetrahydropyran ................................................. 162

ix

FS-C8-P ........................ ' .............................................................. 162

FS-C8-OH ................................................................................... 162

FS-C8-M-l .................................................................................. 163
7-Octen-1-ol ................................................................................ 163

Acrylic acid pent-4-enyl ester. ...................................................... 163

Methacrylic acid pent-4-enyl ester ................................................ 164

7-Octenyl methacrylate ................................................................ 164
8-(Trichlorosilyl)octyl methacrylate .............................................. 165
FSC8M2165

FS-C3-M ..................................................................................... 166

FS-OTMS .................................................................................... 166

FS-TOM-l ................................................................................... 166

F S-TOM-2 ................................................................................... 167

FS-TOM-3 ................................................................................... 167

FS-TOM-4 ................................................................................... 167

F S-TOM—S ............................................ . ...................................... 167

VI. Preparation of composites ..................................................................... 168
1. Compatible composites ........................................................... 168

2. Cross-linkable composites ....................................................... 168
BIBLIOGRAPHY ....................................... . ............................................................... 169

AIBN
BMA
BPO
CMC
CP/MAS
DEA
DEAP
DEGAME
DMAEMA
DMPA

dd

ddt

dt

DSC
ETCS

OMA
ODMCS
ODTCS
OTCS
OTMS
PEGAME
PEGEMA
PPO

SDS
TEGAME
TEM
TGA
TMCS
TMES

LIST OF ABBREVIATIONS

2,2’-Azobisiso(butylronitrile)

Butyl methacrylate

Benzoyl peroxide

Critical micelle concentration

Cross -Polarization Magic Angle Spinning
Diethyl amine
2,2’-Diethoxyacetophenone
Di(ethylene glycol) allyl methyl ether
Dimethylaminoethylmethacrylate
2,2’-Dimethoxy-2-phenyl-acetophenone
Doublet of doublet

Doublet of doublet of triplet

Doublet of triplet

Differential scanning calorimetry
Ethyltrichlorosilane

Methyl methacrylate

Multiplet

n—Vinyl-pyrrolidone

Octyl methacrylate
Octyldimethylchlorosilane
Octadecyltrichlorosilane
Octyltrichlorosilane
Octyltrimethoxysilane

Poly(ethylene glycol) allyl methyl ether
Poly(ethylene glycol) ethyl methacrylate
1 -Phenyl- l ,2-propanedione—2-O-benzoyloxime
Singlet

Sodium dodecyl sulfate

Tri(ethylene glycol) allyl methyl ether
Transmission electron microscopy
Thermogravimetric analyses
Trimethylchlorosilane.
Trimethylethoxysilane

Triplet

Scheme

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

Scheme 9.

Scheme 10.

Scheme 11

Scheme 12.
Scheme 13.
Scheme 14.
Scheme 15.
Scheme 16.
Scheme 17.
Scheme 18.
Scheme 19.
Scheme 20.
Scheme 21.
Scheme 22.

LIST OF SCHEMES

Page
Step-wise Silylation mechanism ............................................................. 14
Chlorination/reduction of silica surface ................................................. 15
Silanization/hydrosilation of silica surface ............................................. 17
Basic functions of SCAs ....................................................................... 22
TPM modified of silica .......................................................................... 24
Composite ofcolloidal crystals ofsilica in PMMA 24
Functionalization offumed silica 25
Preparation of coreactive initiators ............................................. 27
Pathways to generate initiating radicals ................................................. 33
Fragmentation ofketoxime . ..............36
Benzoquinone as an inhibitor ................................................................ 41
Nitrobenzene as an inhibitor: attack on ring .......................................... 41
Nitrobenzene as an inhibitor: attack on nitro group ............................... 42
Tertiary amines as anti-oxidants ................................................. 43

Silylations of A200 using modifiers of different lengths. .. .... ...
Silylations of A200 with varying amounts of ODMCS ......................
Silylations ofA200 with dimethylchlorosilane...
Silylations of A200 with triethoxysilane ...............................................
Preparation of PEO tethered firmed Silicas: first approach. . . . .. .....
Preparation of PEO tethered fumed Silicas: Second approach...
Preparation of cross-linkable firmed Silicas: Method 1... . . . . . . . . . .

Synthesis of 7-octen- 1 -ol .........................................................

xii

54
57
6O
61
68

...72

75
76

Scheme 23. Preparation ofcross-linkable silicas: Method 2 78
Scheme 24. Preparation of cross-linked composite polymer electrolytes. .. . . . .. 95
Scheme 25. Solvent extraction experiments to analyzed surface-grafted

polymer. ............................................................................................... 109

Table

Table 1.
Table 2.
Table 3.

Table 4.
Table 5.
Table 6.
Table 7.
Table 8.
Table 9.

Table 10.
Table 11.
Table 12.
Table 13.
Table 14.
Table 15.
Table 16.
Table 17.
Table 18.
Table 19.

LIST OF TABLES

Useful IR absorptions for characterizing silica surfaces. . . . . . . .

Results ofsilylation offumed silica with XSIMC3......

Results of silylations of A200 with

octyldimethylchlorosilane (ODMCS). .........................................
Modified silica fillers and their properties ............................
Mixed Silylation of A200 with OTMS and TPM .....................
Conductivity data of electrolytes with compatible fillers. . . . . . . . . . .
IRdata forUV initiated polymerization.......................................

Thermal properties of PEGDME-LiClO4 system ...................... . ...... 103

Page

13
52

.......... 57
......... 69
........ 84

93
.. 100

Thermal transition data of composites before and after cross-linking ..... 108

Conductivity data (x104) for Figure 43 ................................
Conductivity data for Figures 47 and 48 ..............................
Formulations ofour and Bourgeat-Lami’s systems. . . . ..
Reaction conditions for FS-TEMS series........

TMS modified srlrcas

Reaction conditions for FS-Cn series............. . . .... .....

Reaction conditions for FS-DM-C8-n series..............
Preparation ofPEO attached silicas: Method 1
Preparation ofPEO attached silicas: Method 2

Products ofmixed Silylation reactions...........................

xiv

.......... 118
......... 126

135
173
174
I75
176
177
177
177

Figure
Figure 1.

Figure 2.

Figure 3.
Figure 4.

Figure 5.

Figure 6.
Figure 7

Figure 8.

Figure 9.

Figure 10.
Figure 11.
Figure 12.
Figure 13.
Figure 14.

Figure 15.

LIST OF FIGURES

Schematic of a battery cell showing its main components

and electron flow during discharge ...................................................

Hydrogen bonding between AEROSIL particles and with water

molecules .......................................................................

Thixotropy: decrease in the viscosity during shearing ...... . . .

Schematic representation of the interaction between AEROSIL

particles in liquids ........................................................................

Characteristics of photo-induced electronic transitions in carbonyl

mompounds .. ..

Schematic ofUVinduced photo-polymerization

IR spectra of amine-catalyzed silylation of A200 with excess methyl

triethoxysilane

IR spectra showing the effects of reaction time on silylation of

A200 with excess MeSi(OEt)3 .........................................................

IR spectra of the silylation products of A200 and Me3SiCl

at three silane/Si-OH ratios
IR spectra of silylation products of A200 with RSiCl3. . . . . . . .. . . . ..
TGA curves of silylation products of A200 with RSiCl; ...................... 56
IR spectra of silylation products of A200 and CgI-Il7(Me)2$iCl ........

TGA curves for silylation products of A200 and Cer-I(Me)2$iCl .....

IR spectra of hydride modified A200 as defined by Schemes l6 and

17 ....................................................................................................
TGA curves ofsamples FS-H and FS-DM-H

Page

..... 2

..... 18

.19

..... 20

...31
39

49

..... 50

53
55

..... 58
..... 59

..... 64

65

Figure 16.
Figure 17.
Figure 18.
Figure 19.
Figure 20.
Figure 21.
Figure 22.

Figure 23.

Figure 24.

Figure 25.

Figure 26.

Figure 27.

Figure 28.

Figure 29.
Figure 30.

Figure 31.

Figure 32.

Figure 33.
Figure 34.

DSC traces ofsamples FS-H and FS-DM-H
TGAcurves ofsamplesFS-DM-EOn......................
IR spectra of samples F S-DM-EOn .........................................
IRspectra of(a) FS-DM-S-EOr;(b) FS-DM-S-EOs.............
IRspectra ofsamples FS-CS-M-l and FS-C8-M—1..............
IR spectra of polymerizable silicas: F S-C3-M and FS-C8-M-2. .........

Schematic representations of A200 (top left),

FS-C3-M (top right), and FS-C8-M-2 (bottom)..............

A portion of the 1H NMR spectrum of the residue from sample

Schematic representations of R805 (top) and F S-TOM—4

(bottom)................ ......................................................
TPM fractions on silica surface and in residue......................

IR spectra ofrepresentative samples from Table 5.........

29Si CP/MAS spectra of samples TOM4 (top spectrum) and

FS-C8-M (bottom spectrum)................

l3C CP/MAS spectra of samples TOM4 (top) and FS-C8-M

(bottom) ........................................................................................
Schematic of PEO modified Silicas ...........................................
Designed models for cross-linkable composite polymer electrolyte:

before (top) and afier (bottom) cross-linking .....................................

IR spectra of composites before (top) and after (bottom) cross-

IR-generated cross-linking kinetics.Spectra are a series, and

correspond to the original spectrum-the spectrum taken at time t. .....

Conversion ofpolymerization vs irradiation time...

DSC second heating scans for PEGDM-SOO/LiClO4 electrolytes.

66
7O
71
73
77
80

81

84

86

...87

88

90

....91
....92

....96

99

101

102

Figure 35.

Figure 36.

Figure 37.

Figure 38.

Figure 39.
Figure 40.
Figure 41.
Figure 42.
Figure 43.
Figure 44.

Figure 45.

Figure 46.
Figure 47.

Figure 48.
Figure 49.

Conditions: run under He at a heating rate of 10 °C/mrn

DSC second heating scans for PEGDM-SOO/fumed silica/LiClO4

electrolytes

DSC second heating scans for cross-linked PEGDM-SOO/fumed

silica/LiClO. electrolytes .....................................................

IR spectra of surface modified silicas. Top scan: Aerosil A-200
firmed silica; middle scan: A-200 modified with surface-bound

methacrylates; bottom scan: cross-linked composite electrolyte

after exhaustive extraction with acetone..........................

TGA scans of surface-modified firmed silicas. T op scan: Aerosil
A-200 fumed silica; middle scan: A-200 modified with surface-

bound methacrylates; bottom scan: cross-linked composite

electrolyte after exhaustive extraction with acetone. ... ....................
TEM micrographs ofA200 ..
TEM micrographs ofcross-linkable srlrca
TEM micrograph of extraction residue of cross-linked composite ........
Freeze-Fracture TEM images of composites before cross-linking .........

Freeze-Fracture TEM images of composites after cross-linking .......

Conductivity of composite electrolytes at selected

temperatures before and after cross-linking ...............................

Arrhenius plots of the conductivity of cross-linked composite

electrolytes

Scenario 1 (top) and 2 (bottom) asdefined in the text............... .. . ..

Conductivity vs. solubility parameters difference between

PEG and PMA .................................................................
Conductivity vs. carbonto oxygen ratioinPMA..................
TGAweightlosses for samplesaslabeled

.105

106

107

110

111
112
113
114
115
116

119

.120
.124

127
129
130

Figure 50.

Figure 51.

Figure 52.
Figure 53.

Dynamic moduli (G’ and G”) before and after cross-linking for

composites containing 10% cross-linkable silica and 10% butyl

methacrylate in a PEGDME-LiClO4 matrix............. 136
Conductivity of cross-linkable composite polymer electrolytes

before and after cross-linking. Solution (SOL) in legend is Li

imide + PEGDME. Proportions are given in weight percentage ............. 137
A proposed route to a core-shell composite polymer electrolyte . .......... 140
Schematic ofhydrophilic firmed silicas 142

INTRODUCTION

1. Electrolyte Systems

1. Primary and Secondary Battery

A battery is defined as a power generating device which transforms chemical
energy into electrical energy. Batteries can be classified into two different types: primary
and secondary. A primary system is one where the stored energy is inherently present in
the chemical substances used and whose useful life is ended once these substances have
been consumed by the discharge processes. In contrast, a secondary battery, also known
as a rechargeable battery, is capable of being charged or discharged when its chemical
substances have been used up. The chemical reaction inside the cell can be reversed by
passing current through the cell in the opposite direction to that of cell discharge. A

secondary battery can therefore be considered as an electrochemical energy storage unit.

The major components of a simple battery cell consist of anode, cathode, electrode
separator and electrolyte as illustrated schematically in Figure 1 where the two electrodes
are connected electronically together and placed in an ionically conducting electrolyte. The
electrolyte can be different materials with diverse characteristics and will be the center
topic of the following discussion. Based on the physical state, an electrolyte can be defined »

as liquid, molten, solid, or gel electrolyte.

 

 

 

 

 

 

load
Anode Cathode
N Separator (
Electiolyte

 

 

 

Figure 1. Schematic of a battery cell showing its main components

and electron flow during discharge.

2. High-Energy-Density Battery

It is always desirable to develop new batteries incorporating high energy density
and long operating life. In designing such a battery system, one has to consider factors
associated with energy density. In the above example, if zinc is used as anode and copper

as cathode, the half-cell reactions during charging are shown in equations 1 and 2:

Atanode Zn 2 Zn2++ 2e' (1)

At cathode Cu2‘+ 2e- :- Cu (2)

At equilibrium, the overall cell reaction is given by equation 3:

Cu‘+211.—:Cu+znz* (3)

The maximum electrical work available from this reaction can be expressed by the Gibbs

free energy change of the reaction (equation 4):

AG=~nFE (4)

where n is the number of mol of electrons passed, F is the Faraday constant (96487

C/mol), and E is the electromotive force (e.m.f.) and can be calculated from equation 5:

E = Emma: - 13...... (5)

The energy density of a battery (ED) is directly related to the free energy
variation (AG) of the cell reaction and to the weight of all chemical reactants (Zwt,)

(equation 6):

ED. = (-AG) / 2 wt, = n F E / Swt. (6)

Based on this equation, light electropositive metals such as alkaline and alkaline earth
metals are therefore the elements of choice for a high energy-energy-density battery, since
a more negative AG (large E) and a small £wt, will be possible. Among these elements,
lithium is particularly attractive due to its high standard potential and low equivalent

weight. The following sections will focus on Li-based electrolytes.

3. Electrolytes for Lithium Batteries

a. Liquid Electrolytes

As lithium reacts with water and other protic solvents, only aprotic organic
solvents are feasible as the basic materials for lithium cell electrolytes. Polar organic
compounds are often used. The most commonly used organic solvents include cyclic

esters (ethylene carbonate, propylene carbonate, y—butylrolactone), linear esters, amides

and sulfoxides.l A mixed solvent is sometimes preferred since the properties of the
electrolyte solution (conductance, viscosity, etc.) and its reactivity towards lithium can
often be “tailored” to give optimum performance. Choice of electrolyte salt is often
limited mainly by solubility which must be high so that the resulting solution will have
sufficient conductivity and hence the cell will have a relatively low internal resistance. In
non-aqueous solutions of lithium salts, the lithium cation is solvated strongly by the -
solvent, while the anion is less subject to this influence. In choosing an organic solvent, it
is important to consider the stability of the electrolyte in contact with lithium metal. It is
generally accepted that no solvent is thermodynamically stable towards Li. The success of

commercial Li battery systems is dependent on the kinetic stability of a passivating

protective film,2 consisting of insoluble reaction products of Li with the electrolyte. Its
nature depends largely on the composition and purity of the electrolyte. The film acts as a
solid phase between the metal and the solution. It has the property of a solid Li ion

conductor with high electrical resistance and is referred as the SE1 (solid electrolyte

interphase).3 The reaction of Li with propylene carbonate serves as a good example of

anode-electrolyte reactions and the formation of SEI phase (Li2C03 in this case, equation

7):

0

Lo + u ——> CszCH—CH3 + Ii2CO3 (7)

Film formation ensures long shelf-life for Li cells, but causes severe problems for
good electrode cyclability. This, along with relative poor conductivity, led to research on
cells with molten salts which offer true thermodynamic stability towards Li meta] due to
the absence of solvents. They also enable high power capacity to be achieved because of
their excellent ionic conductivity. Unfortunately, there are many problems concerning
reliability, safety, cycle life, corrosion and others connected with these advanced

technology systems.

b. Solid Polymer Electrolytes (SPEs)

An ideal battery cell should feature many characteristics, including the absence of
the problems associated with liquid electrolytes. An ideal battery would have a long life
time, a broad range of operating temperatures, no separator, and efficient construction and
packaging. An all solid-state battery system would possess a number of these desirable
characteristics. Unfortunately, high ionic conductivity in the solid state is still difficult to
achieve. Moreover, good contact between solid state electrolytes and electrodes is difficult

to maintain during successive discharge/charge cycles.

The discovery of a new class of solid state electrolytes, i.e., the solid polymer

electrolytes (SPEs)4 allowed many of the problems discussed above to be solved. Armand
reported that certain heteroatom-containing high molecular weight polymers,
poly(ethylene oxides) in particular, could conduct alkali metal ions to levels > 10“ S/cm.
The physical properties of polymer electrolytes are such that they can easily be produced
in thin high area sheets using standard thick film processing technologies. Although
physically solid, polymer electrolytes behave as viscous liquids at the microscopic level
and hence form an excellent interface with solid electrodes. Moreover, they can easily
accommodate electrode volume changes during cycling by microscopic polymer flow. In
other words, SPE makes it possible to manufacture all-solid-state cells and batteries
without the difficulties generally associated with the use of rigid or liquid/molten

electrolytes.

There has been much effort in developing devices using polymer electrolytes in the

past twenty years.5'7 Although the potential benefits of SPE are well recognized, few
such systems are used in applications. This is because the successful use of polymer
electrolytes requires them to be highly conductive, mechanically strong, yet processable.
NMR studies 8 have shown conclusively that the amorphous phase is responsible for the
ionic motion in polymer electrolytes. High molecular weight PEO is highly crystalline and
its mixtures with salt tend to contain either crystalline PEO or high melting polymer-salt

‘ complexes. At ambient temperature, PEG-based electrolytes are generally poor

conductors (o < 10‘8 S/cm) due to the high degree of crystallinity. Reasonable

conductivity can be achieved (0 > 10'5 S/cm) at temperatures of the order of 100 °C, but

the mechanical properties of the polymer are significantly poorer due to the loss of
crystallinity in the polymer hosts. Consequently, these electrolytes are themselves of

limited applicability in the development of practical devices.

A number of approaches have been used to prepare amorphous forms of PEOs.
This includes c0polymerization,9 introducing defects into the polymer chain that disrupt
the crystalline structure, 10 attaching PEO oligomers to polymers having low glass
transition temperatures,1 1‘15 and the synthesis of branched and hyperbranched
polymers. 16’19 A general review has appeared recently.20 Typical conductivities at 30 °C
for polymer-salt complexes derived from amorphous PEOs reached 5x 10'5 S/cm. Often,

however, the mechanical properties of these materials were disappointing, those with the

best conductivity often creeped and lacked mechanical stability.

Blends of a third component with polymer electrolytes can be used to increase the
mechanical property, and ionic conductivity as well in some cases. One type of blend is a
“gel electrolyte” which incorporates a plasticizer in addition to polymer and lithium salt. A

number of polar polymers and low molecular weight substances have been used to make
gel electrolytes and work in this area has been recently reviewed21 and explored further

by Abraham22 and Wieczorek23 and others.24 To be brief, polymers used include
poly(vinylidene fluoride) (PVdF), poly(ethylene oxide) (PEO), poly(vinyl chloride) (PVC),

poly(vinyl alcohol) (PVA), poly(acrylonitrile) (PAN), and polystyrene (PS). In addition,

electrolytes made of low molecular weight polymers supported by a interpenetrating
polymer networks (IPNs) have also been reported.25t26 Plasticizers can be classified into
two types: (1) low molecular weight organic solvents such as propylene carbonate (PC),
ethylene carbonate (EC), THF, DMF, and y-butyrolactone (yBL); (2) low molecular
weight polyethers including poly(ethylene glycol) (PEG) and derivatives and ethylene
oxide (E0) and propylene oxide (PO) copolymers (EO-PO). Organic solvents offer higher
conductivity whereas oligomers provide more reliability. PEGs were found superior to the
methylated counterparts in terms of solvation of salt and conductivity. Unfortunately, OH
groups on these substances are not stable towards lithium metal. These added plasticizers
solvate the salt easily, reduces the viscosity, and increase the conductivity of the system
due to their low molecular weight and high dielectric constant. The conductivity follows
an Arrhenius-type temperature dependence and the activation energy is independent of
polymer concentration, suggesting that ions migrate primarily through the solvent domain
which surrounds the polymer matrix. Thus the role of the polymer in a gel electrolyte is no
longer that of an ionic conductor, but rather as a binder or sponge that contains solvent

and provides favorable mechanical properties. Gel electrolytes are attractive since they

offer liquid-like conductivity with similar processability as solid state electrolytes.27»28
Electrochemically, however, gel electrolytes contain organic solvents and retain many of
the limitations of the corresponding liquid electrolytes such as diffusion and decomposition
Of the electrolyte at the electrode surface. Replacing organic solvents with low molecular
weight polymers as plasticizers seems promising since these oligomers are much less

volatile and more stable towards active metals.

Application of a polymer to PEO-salt complexes has been reported to improve

mechanical stability. Tsuchida et a129 studied systems of PEO-LiClOi (molecular weight =

400 - 20,000) supported by poly(methacrylic acid). Although the mechanical properties

were increased, the conductivity was poor. Gray et al30 reported a polystyrene supported
PEO-LiCF3803 electrolyte. Above the percolation limit, mechanical stability was
significantly improved at high temperatures while conductivity was decreased only slightly.
Similar studies were carried out with PEO-MX (MX = NaI, LiI, LiClOs, and LiBFs)
blended with poly(methyl methacrylate).31t32 A room temperature conductivity as high as
2.4x 10“ S/cm was realized for the polymer blend incorporating LiClOs. A copolymer of

styrene and methoxypolyethylene glycol methacrylate was also used to improve
mechanical stability and reduce crystallinity.16 A new blend system was reported
recently23 involving free radical polymerization of methyl methacrylate in a solution of
LiCF3SO3 and ethylene oxide-propylene oxide copolymer (molecular weight = 2600). An
ambient temperature conductivity of 3x 10" was obtained and the electrolyte-electrode

interface was improved.

Blends of polymer electrolytes and fillers have been studied with the goal of
improving the mechanical stability and conductivity of simple high molecular weight
polymer electrolytes. This is significant since such polymers are often susceptible to creep,

particularly at elevated temperatures. The earliest research in this area was done by

Weston and Steele33 who incorporated various proportions of an inert filler (or-alumina)

into PEO-LiClOt. Above 20 % by volume, the filler aggregated to form large insulating

regions. At 10 % filler, the conductivity was not significantly impaired but there was no

evidence of creep up to 120 °C. A number of studies of PEO mixed with ceramic powders

have been reported since then mainly by the Wieczorek34i35 and Scrosati15a36t37
research groups. One of the interesting findings was that hydrophilic fillers showed a
stronger interaction with the polymer than the hydrophobic material, and composites
formed with the hydrophilic material showed a reduction in polymer crystallinity and a
higher conductivity. More recently, Scrosati and coworkers38 reported a complex system
involving filler material, high and low molecular weight PEO, lithium imide, and either
propylene (PC) or ethylene carbonate (EC) in some cases. High conductivity was achieved
and attributed to the plasticizing effect from both low molecular weight substances and the
lithium salt, whereas good mechanical properties were thought to be due to the
contribution from the filler material. It is now generally accepted that introducing an inert
filler material is an effective means of increasing mechanical properties and stability of the

lithium/electrode interface. Further more, if the particle size of the filler was small, (i.e.,

lower than 5 um,36) the electrochemical properties of the electrolyte could be significantly
enhanced. In most cases, the highest increase was found at the concentration range of filler
up to 10-20 % by volume. NMR and X-ray diffraction studies35 confirmed that the
enhancement of conductivity was due to stabilized and increased PEO amorphous phase
content in the composite polymer electrolyte. The explanation for this behavior was that
the presence of highly dispersed particles in the polymer matrix afl‘ects the crystallization
rate by preventing the agglomeration of polymer chains. A small particle size is important

for assuring a highly dispersed distribution of the filler material throughout the polymer

10

matrix. An optimum concentration of the filler is necessary to suppress the crystallization
rate of the polymer host and to prevent phase discontinuity and filler-catalyzed

agglomeration of polymer chains.

There are problem associated with these composite polymer electrolyte system.
The room temperature conductivity is still low (< 10" S/cm) since high molecular weight
PEOs were used in most cases, and the long term stability of these composites are

questionable. Finally, the interpretation of conductivity data is rather complex.

' 4. New Polymer Electrolyte Systems

In 1992, Fritz and Stein reported a new system incorporating fumed silica as one

of the components of a novel solid electrolyte.39 This new type of electrolyte was
obtained by immobilizing solutions of lithium salts in aprotic solvents in hydrophilic fumed
silica. A room temperature conductivity as high as 10'2 S/cm was realized which is

comparable or even higher than the pure liquid electrolytes. Three years later, another

publication appeared40 on the rheology and conductivity of composite polymer
electrolytes that incorporate hydrophobic fumed silica and low molecular weight PEO. In
addition to a good conductivity, the composites were also mechanically stable and
processable - characteristics attributed to the branched, three-dimensional firmed silica '

network. Research with an emphasis on the electrochemistry of the same system was also

carried out.41 High room temperature conductivity, 3 large potential window, and a
relatively stable interface with lithium electrodes were demonstrated by these authors. To

better understand these new system, it is essential to know some of the firndamental

11

properties of fumed silica particles and their interactions among themelves and with

organic molecules.

(1. Basic Characteristics of F umed Silica

Representatives of firmed silica are those with the trade mark AEROSIL produced
by Degussa These fumed silicas are made through continuous flame hydrolysis of silicon
tetrachloride. During this process, SiCh is vaporized and then reacted with water vapor
formed from an oxyhydrogen flame to produce the desired silicon dioxide. The only by-
product is gaseous HCl which is separated from the solid product. Fumed silicas thus

produced are amorphous, nonporous, and thermally stable. Their primary particles are

almost spherical and extreme smalL ranging from 7 to 40 nm 42.

In general, the most important characteristic of silica is its surface chemistry. Since
the surface area is large in relation to the mass, the surface chemistry plays a significant
role. There are two types of functional groups on the surface of silica: hydrophilic silanol
groups and hydrophobic siloxane groups. The surface chemistry of pristine silicas is

dominated by silanol groups which are necessary for certain applications. However, for
other applications such as stationary phases for metal ion chromatography,43 catalysts,44

as well as immobilized enzymes,45, a hydrophobic surface is desirable. In this case, a
chemical after treatment process can be used to modify the silica surface. Typically, an
alkoxysilane or chlorosilane can be used to react with silica to replace silanol groups with
organic groups - a process often termed silylation. It is sometimes important to know the

density of silanol groups on the pristine and modified silicas. The most reliable method for

12

quantitative analysis so far is a lithium aluminum hydride method in which the amount of
hydrogen gas released from reaction between silanol groups and hydrides is measured and

the silanol group density is determined. IR spectroscopy, on the other hand, is a

convenient method for qualitative analysis.46 Table 1 lists some important IR absorption

bands of firmed silicas. There have been numerous reports on silica surface modification

and characterization.46'52

Table 1 Useful IR absorptions for characterizing silica surfaces.

 

 

Functional Groups IR Absorption (cm'l)
SiOH isolated 3747
SiOH bridged 3200-3300
SiOH bridged, proton acceptor 3720
SiOH bridged, proton donor 3520
SiOSi combination 1860
HOH 1 620

 

Two research groups, Blitz and co-workers and Tripp and co-workers, contributed
most to the understanding of interactions between Silanes and surface silanol groups. Blitz
and co-workers53 used diffuse reflectance FTIR spectroscopy to study the silylation of
Cab-O-Sil, a commercial fumed silica, with alkoxysilanes by and concluded that Silanes did
not bond to the silica surface after high-temperature curing in the absence of water. When

surface-adsorbed water was present, post-reaction curing increased hydrolysis and

13

condensation to form siloxane bonds. The same authors54 also conducted silylation
reactions in the presence of amines. They found that amines with exchangeable protons
catalyzed the reaction greatly and proposed two possible mechanism. Scheme 1 shows a
step-wise reaction mechanism The first step is adsorption of amine to the silanol site.
Hydrogen bonding between amine nitrogen and silanol hydrogen makes the silanol oxygen
more nucleophilic. The next step is nucleophilic attack of this oxygen to silane molecules
and the subsequent formation of a penta-coordinated intermediate. The leaving group
(OMe) abstracts a proton from the amine to form MeOI-I, and the amine takes away the
silanol proton and starts another catalytic cycle. The second silylation mechanism involves
concerted bond breaking and bond fonnation. This mechanism is expected to be

energetically favorable due to a 6-membered ring transition state shown as 1.

Scheme 1. Step-wise silylation mechanism

IOMe 1? / OM 1% /
\ - e
Sr N 5+ ,N 5+ _$_
/ {/5— H’ \ 5\- i\-" If \ (g1 + MeOH
—__. ’ ——->
0 / 0 8+ _1 HN/
_ _ an; \

 

 

 

14

 

More recently, Tripp and Hair55'57 developed a method by which chlorosilanes
can be effectively attached to the surface of AEROSIL silica through chemical bonding.
This process also uses an amine adsorbed onto the silica surface to promote the reaction.
Their proposed mechanism is similar to Blitz’s step-wise mechanism in that a penta-
coordinated intermediate was considered to be involved. The difference between when
chlorosilanes are used instead of alkoxysilanes is that the by-product is a strong acid and
needs to be neutralized by stoichiometric amounts of base. In other words, the base is no

longer a catalyst.

Scheme 2. Chlorination/reduction of silica surface

ilsr-on + socb —. §|Si—Cl + so2 + HCl

first.

§|Si—H + LiCl +AlH3

15

Using a different silylation reaction to modify silica surface, Sandoval et a158'6O
reported a novel synthetic approach to produce chemically bonded silica-based
chromatography stationary phases. The procedure involves the preparation of a silica
intermediate containing stable silicon hydride surface species, followed by the catalytic

addition of this hydride modified silica to organic compounds bearing a terminal vinyl

group - a reaction developed by Speier61 thirty years ago and often referred to as
hydrosilation. The preparation of the hydride intermediate was first carried out via

chlorination of silica followed by reduction with lithium tetrahydridoaluminate (Scheme

2).58t59 Since this chlorination / reduction sequence had limitations such as extreme
sensitivity to moisture during both steps, they later developed a new approach, referred by
these authors as “silanization” (Scheme 3), which involves the controlled deposition of the

hydrolysis product of triethoxysilane or related Silanes to preferentially form an Si-H

rnonolayer on silica.60 This new approach of making surface bonded materials was
believed to have advantages over the traditional method since it results in stronger surface
attachment. It has long been known that in the case of trifunctional Silanes reacting with
silica surface, the number of linkages range from 1 to 3. Such linkages result from
covalent attachment of the silane to the silica surface and from cross-linking between
adjacent Silanes. Using Silanes that contain a hydride instead of a bulky alkyl leads to
greater condensation with silanols because of the minimal steric hindrance of the hydride.
Products made from this method show greater hydrolytic stability, suggesting a larger.

number of siloxane linkages.

16

Scheme 3. Silanization/hydrosilation of silica surface

ilSi—OH + RO—ii—H EEO—HE i'Si-O—di—H + ROH
loxane

Pt Cat, CH2=CI'IR'

ilSi—O—ii—CH2CH2R

b. Thickening and T hixotropic Effects of Famed Silica

The interactions between silica particles themelves and with the dispersion phase
play a decisive role. Two important interactions are: (1) van der Waals attractive forces,
and (2) hydrogen bond interactions. In comparison with precipitated silicas, fumed silicas
have a lower silanol group density (2.5/nm2) and therefore little possibility exists for the
fomration of intra-particular hydrogen bridge linkages and only inter-particular hydrogen
bonding needs to be considered. Figure 2 shows hydrogen bonding linkages between silica
primary particles and with water. Hydrogen bonding becomes less important when silanol
groups are replaced chemically by organic groups, and the extent of this effect depends on
the degree of replacement and the length of organic groups. Long alkyl groups tend to
shield the functional groups on the silica surface. The lower the silanol group density (or

the higher the organic group coverage), or the longer the alkyl groups, the less hydrophilic

17

the silica surface. Silicas with significant amounts of alkyl groups show hydrophobic

characteristics and van der Waals attractive forces may become dominant.

 

Figure 2. Hydrogen bonding between AEROSIL particles and with water molecules.

If AEROSIL is dispersed in a liquid, the surface groups interact with each other
either directly or indirectly via the molecules in the liquid. The affinity can be attributed to
hydrogen bonding in the case of hydrophilic silicas, and to both van der Waals attractions
and hydrogen bonding between surface organic groups and residual silanol groups in the
case of hydrophobic silicas. Such interactions result in a temporary, three-dimensional
network structure and a thickening effect. Under a rrrechanical stress (intensive stirring or
shaking), the structure is broken down and the viscosity drops. In the static state, the
structure reform. This process is termed thixotropy and is represented schematically in a
viscosity-time plot shown in Figure 3. A more vivid picture showing the aggregation

patterns of silica particles during shearing and standing is given in Figure 4 where only the

18

 

 

 

Viscosity ——>

Static
period

Shearing Static period

 

 

 

 

Time ——>

Figure 3. Thixotropy: decrease in the viscosity during shearing.

case of hydrophilic silica is shown. The thickening and thixotropic effect of AEROSIL

silicas is the basis for many of their applications such as anti-setting agents and coatings.62
Such effects depend on the relative polarity of silica surface and the liquid. In the case of ‘
hydrophilic silicas in polar solvents, the silica particles are solvated/wetted by the solvents
and screened against external effects by hydrophobic alkyl groups fi'om liquid molecules.
Consequently, a destabilization of the thixotropic network occurs. For this reason, a gel is
only possible with large amounts of hydrophilic silica. In contrast, gelation is possible in
non-polar liquids by using small amounts of silica. To summarize, hydrogen bonding

between AEROSIL particles is maximized in non-polar fluids.

19

 

51s.: 1:;

Agglomerated aggregates Three-dimensional network

Figure 4. Schematic representation of the interaction

between AEROSIL particles in liquids.

c. Current Work
With the fundamental knowledge of firmed silicas, let us re-examine the new solid
electrolyte system mentioned earlier. In Fritz and Stein’s system,39 polar organic solvents

and hydrophilic firmed silica were employed. It is not surprising that they needed to use up

to 20 wt% of fumed silica to obtain a “paste-like” electrolyte. In the second system“),41
10 % by weight of hydrophobic fumed silica was enough to bring about a fi'ee-standing
gel. As discussed above for polar solvents, which have to be used in order to dissolve
reasonable amounts of lithium salt, a hydrophobic instead of a hydrophilic silica is a better '
choice. Another advantage of the second system was the use of low molecular weight
poly(ethylene oxide), which has a much lower volatility. Finally, PEOS contain only the C-
0 functional group - one of the most stable groups in organic chemistry, and therefore
should have favor electrochemical stability towards electrodes. We have known fi'om

previous discussions that complexes of high molecular weight PEO with salt are crystalline

20

at room temperature and the conductivity is too low for practical usage unless a high
temperature (> Tm = 60 °C) is reached. Poly(ethylene glycol)s and derivatives with
molecular weight lower than 500 are amorphous liquid at room temperature and therefore
would ofl‘er high ionic conductivity once mixed with lithium salt. Although these
oligomers are of little value by themselves as solid polymer electrolytes, they can be
incorporated into a fumed silica network. In doing so, both high ionic conductivity and

good mechanical properties will be ensured.

The major advantage of using filmed silica as a filler material is that it can be
modified in many different ways. In addition to tuning the polarity and hydrophobicity, one
can also attach a polymerizable group on firmed silica surface. This involves the use of a
class of compounds called silane coupling agents (SCA) with the following general

formula (equation 8):

R(4-n)San (3)

where X is a hydrolyzable group such as a chloro or alkoxy group; R is a non-
hydrolyzable functional group such as ethylene oxide, (meth)acrylate, hydroxyl or thiol. A
schematic representation of a silane coupling agent and its functions is shown in Scheme 4.
These compounds are effective linkers between two inorganic substrates or between an

inorganic substrate and an organic moiety. They play an important role in many areas in

industry.63

21

Scheme 4. Basic functions of SCAs.

 

 

For condensation with
OH terminated surfaces

 

 

W

_" X‘Ei’x

X = Cl, OR

 

 

For cross-linking with
polymerizable substrates

Loy

 

 

 

polyme riza ble
substrate

 

 

 

22

Silicas containing pendant polymerizable groups are useful in making composite
polymeric materials. They serve as a “monomer” and will co-polymerize with another
monomer. In a recent example, Ford and coworkers“,65 published papers on a
composite of colloidal crystals of silica in poly(methyl methacrylate) (PMMA).
Monodisperse silica particles were prepared using a sol-gel process by hydrolyzing
tetraethoxysilane in the presence of water and ammonia. 3-(Trimethoxysilyl)-propyl
methacrylate (TPM) was introduced on silica surface by basic hydrolysis and condensation
to form TPM oligomers which were then adsorbed and condensed on silica surface.
(Scheme 5). These TPM modified particles form colloidal crystals spontaneously once
dispersed in a proper solvent. Such dispersions are useful materials as optical rejection
filters and optical limiters. Addition of methyl methacrylate and copolymerization with
polymerizable silica (Scheme 6) formed robust films while the crystal structure was
retained. These novel materials are more stable and may find wider applications in

selective filters and other optical devices.

Another recent example involving SCA modified silica was reported by Guyot and

coworkers.66'69 They carried out emulsion polymerization of poly(ethyl acrylate) in the
presence of functionalized and non-functionalized fumed silica particles. SCA modification
of the silica surface was done in two steps using two different kinds of coupling agents, 3-
(trimethoxysilyl)-propyl methacrylate (TPM) and 3-(trimethoxysilyl)-propane thiol
(TESPT). The SCA was first hydrolyzed in water, the partly condensed silane was added

to a silica suspension in toluene (Scheme 7). The pre-hydrolysis step helps increase the

23

Scheme 5. TPM modified of silica.

hydrolysis condensation 1;
RSi(OEt)3 —> RSi(0H)3 —.-’ — l-
IhO/NH3 (5H
TPM oIgomer

EtOH
Si(OEt)4 —. ® OH
NH3 (25%) R

_§,_

.—C-—%i—R COM“ @ 1:115

hydrogen bonding

II
R = CH2CH2CH20C$=CH2
CH3

Scheme 6. Composite of colloidal crystals of silica in PMMA

CH3

0
.—O—dl—CH2CH2CH20EC=CH2 + HzC-T-é
l cl b02013

® 1 EVIL/3:: O OCH
0: CH3 DMPA: “-iCH
4:?an Eli-CH2)—

H3 02CH3

24

Scheme 7. Functionalization of fumed silica

(CH30)3—Si—(CH2)3—R
1 1. I120

O
l
R = odC=CH2 TPM-grafted silica
H3

R = SH TESPT-grafted silica

degree of attachment of the coupling agent. The TPM modified silica serves as a co-
monomer whereas TESPT modified silica is not polymerizable and acts only as a chain
transfer agent. In the first case, large amounts of polymer was grafted onto silica surface
through small sized “loops”. In the second case, although chemical bonding between
surface and polymer is also present, no such loops formed. Their studies showed that,
while both firnctionalized and non-functionalized silicas can result in significant
reinforcement effects as evidenced by an increase in Young’s modulus, the mechanical
properties at large strains are greatly improved only when silicas with polymerizable

groups were used.

25

An alternative way of making covalently surface-bound polymer is to carry out
polymerization using photo-initiators chemically attached to the surface - the so-called
core-active initiators. Koehler and Ohngemach70 were the first to introduce a photo-
initiator group to a surface. For example, triethoxylsilyl substituted (2-hydroxy-2-propyl)
phenone was prepared by hydrosilation followed by silylation with silica gel (Scheme 8).
Such core-active photo initiators were used under unstirred conditions to polymerize
dimethylarninoethylmethacrylate (DMAEMA) and n-vinyl-pyrrolidone (NVP). By
changing the monomer concentration and the initiator content on the silica surface, they
were able to regulate the amount of polymer bonded to surface. The intermolecular
distance of monomers was believed to determine the radical transfer which in turn affected

the photo-polymerization process.

We are interested in making composite polymer electrolytes with desirable

conductivities, permanent mechanical properties and dimensional stability, and

processability. This can be done by extending the work done by Khan 40 and Fan 41 by
using SCA or core-active initiator modified silicas. It is not difficult to. imagine that if a
combination of an appropriate polymerizable silica, PEO oligomer, lithium salt, free
radical initiator, and small amount of co-monomer (e. g., alkyl methacrylate) would result
in a gel-like material due to the thickening effect discussed above. Upon irradiation with
UV light, co-polymerization between this silica and the added monomer would take place,
yielding a cross-linked composite polymer electrolyte. In the case of core-active initiator

modified silica, a free radical initiator will no longer be necessary yet a similar material

26

Scheme 8. Preparation of coreactive initiators

CH3
(CH3CH20)38IH + CH2=CHCH20CH2CH20‘O'—E—é—OH
PtCat.
CH3

ll
(CH3CH20)3Si—CH2CH2CH20CH2CH20—O~—C—(J2—OH
A l
(I) CH3
.a—li—CH2CH2CH20CH2CHZO—.—d—d—OH
l l in,

hv
0 CH3

."O—%l—CH2CH3CH30CH2CH20 . ° i—OH

H3
1 monomer I

Surface-Bound Polymer

27

would be expected to be realized. The significance of cross-linking process lies in the fact
that, instead of a temporary three-dimensional network resulted from thixotropic effect, a

permanent robust network structure would be created which features long-term

mechanical and dimensional stability.

There are multiple issues involved in the successful preparation of such a cross-
linkable composite. (1) surface properties Coolarity/lozdrophobiciw) of polymerizable
silica. We have pointed out that hydrophobic silica would be the material of choice since
low molecular weight PEOs are polar solvents. On the other hand, most polymerizable
firnctional groups, e.g., methacrylate, are polar and methodology has to be developed to
compensate accordingly. (2) selection of co-monomer. This is crucial since the property of
formed polymer will directly influence the conductivity of electrolyte. (3) conditions
necessary to bring about a photo-polymerization. The next section will discuss certain

aspects regarding photo-initiation of polymerization.

28

II. Photoinitiation of Polymerization

Many chemical reactions are induced by external energy such as heat and light.
Light-initiated reactions, or photochemical reactions are attractive since light intensity can
be controlled conveniently and the light sources can be turned on or off very easily. Other
benefits of photochemical reactions include rapid reaction, low energy requirements, low
temperature operation, and non-polluting in cases of solvent-free systems. Since light can
only penetrate a thin layer, photo-polymerization is mainly used for thin-film applications.
In this section, we will first give a brief summary regarding interactions of light with
absorbing species, including photo-induced electronic transitions, and then move on to

photoinitiation of polymerization.

1. Theory

The relationship between absorbance, the sample thickness, and the concentration

of the absorbing species can be expressed by Lambert-Beer law:
A=L0g10(Io/I)=80b (9)

A: absorbance,

Io: incident light intensity,

I: Intensity of light exiting the sample,

c: concentration of solute in moles/liter,

29

b: optical path length in cm,

a: molar absorptivity in mol"cm".

If we define Lb. as the fraction of light absorbed by a solution, then

1....= 10( 1-10“") (10)

A UV spectrum is usually a plot of wavelength vs. molar absorptivity. The former
represents the energy necessary to bring about an electronic transition between ground
state and excited state; and the latter is an indication of probability of such transitions.
Figure 7 gives electronic transitions and their characteristics in the case of carbonyl
compounds (only one nonbonding orbital is shown). The n —-) 1" transitions of single
chromophoric groups are symmetry forbidden (non bonding orbital on oxygen is
orthogonal to it anti-bonding orbital), and as a result, the molar absorptivity is low (a <
100). T! —> rt* transitions, on the other hand, are allowed and therefore often show strong
absorptions (e > 10,000). In Figure 7, So, S1, and S; are ground, first and second excited
singlet states, respectively; T1 and T2 are first and second triplet excited states. Internal
conversion (IC) is a term often used to describe transitions between S; and S1, and

intersystem crossing (ISC) represents transitions between singlet and triplet states.

30

4:c=o.:

...... + +
Blair + il—
nit +r +

=+r++
:l:

.1.

T1 T2
yound n——.n “an. n__.n n__.n*
state forbidden allowed forbidden allowed
small a large a small a large a

Figure 5. Characteristics of photo-induced electronic transitions in carbonyl compounds.

31

2. Polymerization Initiators

As in other chain reactions, polymerization involves three basic steps: initiation,
propagation, and termination. In principle, any monomer will undergo photolysis upon
absorption of light with enough energy. But only conjugated monomers absorb above the
vacuum UV region (>200 nm). Often times, even these monomers themselves do not form
reactive enough radical species upon light absorption to initiate a chain reaction. In other
words, the initiation efficiency is too low. For this reason, other molecules are often used
that combine sensitivity to UV radiation and ability to provide a reasonable initiating
efficiency. In choosing an initiation system, one has to consider the availability of light
sources and reaction vessels. Quartz transmits light above 200 nm whereas glass does so
Only above 325 nm. As far as the light source is concerned, low pressure mercury lamps
only generate significant numbers of photons at 254 nm and therefore medium (or high)

pressure mercury lamps are often used that give light with a broader energy distribution
(254 , 280, 297, 303, 313, 334, 366 nm).71 Light filters can be used ifa light with a
certain wavelength is desirable. Since Pyrex reaction vessels are always more handy than
quartz ones, people tend to choose an initiator system which absorbs above 325 nm and

matches well one of the lines in the spectrum of medium (or high) pressure mercury lamps.

Scheme 9 show three pathways to generate reactive species. The simplest case is
when only one kind of molecule, the initiator (I), is involved. This initiator absorbs light
(hv), is excited (1*), and finally yields radicals (R-) to initiate polymerization. When a

hydrogen donor (HD) (often termed a co-initiator) is present, initiators in excited states,

32

Scheme 9. Pathways to generate initiating radicals

I\hV‘I*/R°
h. ‘/ \“Ds

Sen ——-> Sen“ HI‘ + 13'

typically those built on benzoyl chromophore, abstract a hydrogen from the co-initiator to
generate a pair of radicals (HI. and De). These radicals initiate polymerization directly or
do so after a few fragmentations. A photo-sensitizer (Sen), defined as a molecule capable
of absorbing light and transferring excitation to the initiator, is sometimes used. The idea
behind a photo-sensitizer is that when I “ cannot be produced directly by irradiation of I
with a certain wavelength, the excitation is accomplished by first photo-exciting the photo-

sensitizer, and then the excited sensitizer Sen“ transfers the energy to 1.

Many compounds are useful as photo-initiators for polymerization, including
peroxides, azo compounds, and especially carbonyl compounds. These substances
undergo homolysis by either or both of the two processes: fi'agmentation and hydrogen

abstraction. Benzoyl peroxide (BPO) is representative of peroxide initiators. This

molecule fragments into two benzol radicals (11) upon absorption of UV light:72

33

Hg-o—o—iz') ——h-"—» 2 .—i:')—o- (11)

The 0-0 bond energy is only 30 kcal/mol and radicals form easily. Disadvantages include

poor thermal stability, absorption below 300 nm, and low reactivity towards monomers.

Azo compounds are another family of useful initiators. An important member of
this family is 2,2’-azobisiso(butylronitrile) (AIBN) which generates radicals HI upon UV

absorption (equation 12):

H H
i 3 i 3 hv CH3
NC —-N=N —CN —» 2NC I. + N; (12)
H3 $113 CH3
III

The generated radicals are very effective in initiating polymerization since one molecule of
nitrogen gas is released from each AIBN molecule, leaving a small distance between the

two just-formed radicals and reducing the chance for them to recombine.

Carbonyl compounds, especially aromatic ketones, have found wide spread

application in industry. The main mode of fragmentation is through or-cleavage (Norrish I

photo-scission). 2,2’-Dimethoxy-2-phenyl-acetophenone (DMPA),73 a commercial

34

product (Irqacure 651), has a very high initiation efliciency due to the reactive methyl

radical (IV) produced through a two step process:
0 OCH3 hv O OCH3
g—C —> . + . (E (13)
OCH3 OCH3
o /
CH3- + CH30—g—© (14)

IV

A similar compound, 2,2’-diethoxyacetophenone (DEAP) gives a reactive ethyl radical

(V) in the same way:74

0 OCH2CH3 by O OCHZCH3
©t-d_s ... @Jl ..I-H (15)
OCH2CH3 0CH2CH3

/ (16)

O
CH3CH2 ° + CH3CH20—g—H

l-Phenyl-1,2-propanedione-2-O-benzoyloxime (PPO), a ketoxime ester of benzoin,

generates a very aggressive initiating species, phenyl radical V1.75 Small molecules, C02

35

Scheme 10. F ragrnentation of ketoxime

MEN-“j

36

and CH3CN, are released during fragmentation (Scheme 10). As in the case of AIBN,

recombination of formed radicals is suppressed.

Irradiation of ketones in the presence of a hydrogen donor solvent results in co-

initiation processes. Amines, alcohols, thiols, and ethers are common co-initiators. Below

is an example of benzophenone as initiator and isopropanol as co-initiator:76

O l
Ph—g-Ph —>

I0 It! 3 (I) It:
(Eh—d—Ph ) ———> (Ph—d-Ph > (17)

3 '0 4! CH3 OH CH3
(Ph—d—Ph > + H-d—OH ——> Ph—d—Ph + d—OH (18)
H3 H3

An initiation system containing anthraquinone and THF was reported by Taylor et
a1. (equation 19)77 Here, TI-IF provides hydrogen and also acts as the principal initiating
species (VII). Other initiation mechanisms include [B-cleavage,78 Norrish II

photoscission,79 and charge transfer.80

37

3. UV Induced Polymerization

Figure 6 is a simple view of UV initiated radical polymerization. UV light
absorption by the initiator causes electronic transition from the ground state to excited
singlet state. After intersystem crossing, a triplet excited state is realized which then
fragments (or abstract a hydrogen if present) to generate initiating radicals. These radicals
add to monomer to form the first monomeric radicals which propagate to result in polymer
chains. Propagating radicals terminate by coupling or by disproportionation. Chain transfer

processes are involved in many cases.

As mentioned earlier, most polymerization reactions are induced by initiators in
triplet states. The quantum yield for initiation, 4);, is defined as the number of polymer

chains initiated per absorbed photon. The rate of initiation R; can be described as:

R; = 2 d); 1.1,. (20)

where 1.... is the fraction of light absorbed by initiator. For any chain polymerization

reaction, the rate of polymerization R9 is written as:

Rc=kp[M](Rt/2kt)”2 (21)

Substitution of equations 10 and 20 into 21, we end up with

38

Rs= klel I¢i Io(1-10“b)/kt]m (22)

Using steady-state treatment, the kinetic chain length is obtained as:

V=Rp/Rt=Rc/Ra=[kp/2ktml[M1[¢ilo(l-10‘°b)/ktl"’2 (23)

the ratio of k9/ 2 k."2 is termed characteristic ratio of the monomer system which

measures the ability of monomer to support a radical chain reaction.

 

 

 

 

 

 

52
ll i mummgmneas
31 T—
A2

 

 

 

 

k __T2/ 1
E l
4'— first monomeric radical
2'1 T1 1
s0 IPOLvhumg

 

 

 

 

Figure 6. Schematic of UV induced photo-polymerization.

39

4. Inhibition and Retardation

Apart from reacting with monomer molecules to initiate and propagate the chain

polymerization, radicals also react with other substances present in reaction system. As a

result, polymerization is suppressed.81 Depending on the degree of suppression, these
substances are classified as inhibitors which stop the reaction until consumed, or retarders
which only slow down the reaction to some extent. Inhibitors or retarders are added
intentionally to stabilize commercial monomer products, for example, a tiny amount (<
100 ppm) of hydroquinone monomethyl ether is often added to methacrylate monomers,
and 4-t-butyl-catechol is often added to styrene. In such cases, the inhibitor is often
removed before use by distillation, chromatography, or extraction. If the monomer is
synthesized in the lab, impurities due to incomplete purification may act as an inhibitor or
retarder. Many kinds of compounds can act as inhibitor or retarder,. quinones are most
common inhibitors. As shown in Scheme 1 1, propagating radicals can attack either at
oxygen or at the ring to generate two different radicals: VIII and IX. VIII can terminate or

add monomers; IX may couple with another radical to form X or rearrange to XI.

Aromatic nitro compounds are another family of substances that are used as
inhibitors or retarders. The more nitro groups on the ring, the more polymerization is
suppressed. Radicals can attack either the aromatic ring or the nitro group on these

compounds.

40

Scheme 11. Benzoquinone as an inhibitor.

Cowling
disproportionation
add monomer

Scheme 12. Nitrobenzene as an inhibitor: attack on ring.

Mn‘NOZ + NIH‘

In the first case (Scheme 12), radical XII is formed which either couples with

M . Mn’
Mn’ 11
——~
H — M
XII

another radical or re-aromatizes by transferring a hydrogen radical to a monomer
molecule. In the second case (Scheme 13), radical XIII is generated which couples directly

or indirectly with another radical, or fragments through a-cleavage.

41

Scheme 13. Nitrobenzene as an inhibitor: attack on nitro group.

©N=O + MnOMn
Mn.
Mn° OM“ M .
0.
x111

The oxygen molecule reacts with most radicals and is a powerful inhibitor:
02 + radicals ———> nonreactive peroxy radicals (24)

This is one reason why an induction period is often observed in polymerization processes.
From the industrial applications standpoint, a long induction period is not tolerable and
both physical and chemical means have been reported to tackle this problem. Physical

means include inert gas protection and vacuum application. Chemical means involve
adding a Chemical substance to trap oxygen through a chain process 82 or transform

triplet oxygen into singlet oxygen.83 In systems where tertiary amines are added as

hydrogen donors, oxygen can be removed according to the following cycle (Scheme 14):

42

Scheme 14. Tertiary amines as anti-oxidants

-R + H3CN(CH3)2

w)

- CH2N(CH3)2 HOOCH2N(CH3)2
02 ( YmeflCHflz
. OOCH2N(CH3)2

This cycle is able to consume up to 12 molecules of oxygen by same number of amine
molecules with the loss of only one radical. This explains why tertiary amines are effective

antioxidants.

43

RESULTS

We were interested in making composite polymer electrolytes incorporating low
molecular weight PEO (PEGDME-SOO), lithium salt, and functionalized fumed silica.
Although a few firmed silicas are commercially available, they are limited in terms of the
organic chain length and kind, and organic group coverage. We know that both pristine
hydrophilic and hydrophobic silicas can be efl‘ective thickening agents, and while the
thickening mechanism seems quite clear for the former, it is still ambiguous for the latter.
In order to optimize mechanical properties of the composite polymer electrolyte system
and investigate the interaction between filler materials and conducting polymers in general,
we prepared a series of functionalized firmed silicas, including hydrophobic silicas with
different alkyl groups, PEO oligomer modified silicas, and most importantly, hydrophobic
and cross-linkable silicas. As we will see later, firmed silicas carrying both hydrophobic
and cross-linkable groups, combined with PEGDME-Li salts, monomer and initiator, are
extremely useful materials for realizing a permanent three-dimensional networks. In
particular, good mechanical properties are achieved without impairing the conductivity of
the composite electrolyte. All of these silica materials have been or are being used by our
collaborative partners at North Carolina State University to investigate rheology and
electrochemistry of fumed silica composites. Information from these studies will help us to
optimize the physical and chemical properties of composite polymer electrolyte systems

that incorporate fumed silicas as filler materials.

This section describes experimental conditions leading to good silylation products,

the preparation of various functionalized fumed silicas, and finally the formation and basic

characteristics of composite polymer electrolytes.

The terminology used to describe the modified silica samples in this report is as
follows: For all the samples, FS is used to represent Fumed Silica. For most samples, the
abbreviation for the silane used to modified the silica follows F S. For example, FS-TMCS-
1 is a trimethylchlorosilane modified fumed silica The number following the abbreviation
is simply the run number. Exceptions are: 2H and 13H which represent the time of
reaction in hours; C2, C8, and C18 denote the number of carbons in the hydrocarbon
chain. TOM denotes a mixture of Trimethoxysilypropyl and Octyltrimethoxy
Methacrylate. DM stands for DiMethyl, H for Hydride, E0. for number of Ethylene
Oxide units, and M for Methacrylate. A few other samples not covered above can be

found in the related reaction schemes.

45

l. Silylation Conditions

1. General

In order to obtain the desired silica products, a series of control experiments were
carried out to optimize the reaction conditions for silylation. We were primarily concerned

with the amounts of the amine catalyst, the reaction time, and the relative amount of the

silane compound. Based on Blitz’s finding54 that amines with exchangeable hydrogen(s)
show catalytic activity when alkoxysilanes are used in silylation reactions, diethyl amine
(DEA), a liquid at room temperature, was chosen for all experiments in this report. Ethyl
amine and ammonia have more exchangeable hydrogen atoms and should catalyze
silylation more efficiently, however, both are gases and are less convenient for lab use. For
silylation reactions involving chlorosilanes, the type of amine is less important since it

serves as a base instead of a catalyst.

FTIR spectroscopy was used to monitor changes during silylation reactions
between fumed silicas and silane compounds. Usually, the attenuation or disappearance of
the isolated silanol peak (3 744 cm") and the appearance of new C—H absorption bands
were observed. Since silica is not transparent to infrared below 1300 cm", only the 4000 -
1300 cm'1 portion of the spectra will be shown. Thermogravimetric analysis (TGA) and
lithium aluminum hydride titration techniques were used to supplement FTIR in some

C8388.

2. Amounts of Amine Catalyst

Infrared spectra of Aerosil modified with excess methyl triethoxysilane (relative to
silanol groups on silica surface) in the presence and absence of DEA are shown in Figure
7. When the reaction is carried out in the absence of DEA, the silane C-H stretch modes
are barely seen and the peak characteristic of isolated silanol groups remains strong. This
spectrum (curve a) is similar to that of the starting AEROSIL silica where the band at
3744 cm'1 is assigned to isolated silanol groups, the broad peak ranging from 3000 to
3800 cm'1 is attributed to the hydrogen bonded OH groups, the Si-O-Si combination band
appears at 1869 cm", and the peak at 1628 cm'1 arises fiom the deformation mode of the
adsorbed water. Addition of 1 mol % of DEA results in significant attachment of the silane

(curve b), suggesting that the DEA does catalyze the silylation reaction, as previously

reported.54 Bands in the 2800-3000 cm’1 region come from the C-H stretch modes of the
tethered silane. Qualitatively, addition of more amine does not seem to change the amount
silane attached to the silica, but condensation between the adjacent silane molecules might
be enhanced as implied by the relative increase of C-H asymmetric stretch of Si-CH; at

2980 cm".

3. Reaction Time

Reaction time also plays an important role in silylation reactions, especially when
using multifirnctional Silanes, since hydrolysis of the second and third functional groups

and the cross-linking reaction need longer time. In a study of silylation of Cab-O—Sil with

trimethoxysilane, Blit254 reported that a reaction time of two hours was enough for

complete reaction. Using triethoxysilane and firmed silica, we carried out silylation for 2 h

47

and longer times (Figure 8). We found that the silane loading became constant after 13 h -
a condition employed throughout our experiments. Although we found that silane
attachment could be increased by increasing the silylation temperature from ambient
temperature to 70 °C, we never applied this condition to our actual preparations since

satisfactory results could often be achieved at room temperature.

48

 

 

 

 

 

 

 

to
U
a
a
a
i-
e
'3 c
<
o
>
'5
a
3‘2“
b
a
3800 3300 2800 2300 1800 1300

-1
Wavenumber, cm

Figure 7. IR spectra of amine-catalyzed silylation of A200 with excess methyl
triethoxysilane: (a) FS-MTES-O, without amine; (b) F S-MTES-l, 1 mol % of amine based

on surface silanol content; (c) FS-MTES-Z, 25 mol %; (d) FS-MTES-3, 100 mol %.

49

 

 

Relative Absorbance

 

 

 

 

 

r

3800 3300 2800

L
I

2300

-r
Wavenumber, cm

Figure 8. IR spectra showing the effects of reaction time on silylation

1800

1300

of A200 with excess MeSi(OEt)3: (a) F S-MTES-ZH, 2h; (b) F S-MTES-IBH, 13h.

50

4. Amount and Type of Silane

Aerosil surfaces can be modified by either multifunctional or monofunctional
silanes. Multifirnctional Silanes are more effective in generating high density organic
surfaces due to reactions of their functional groups with multiple surface silanol groups
and condensation reactions between silanes. Often however, the modified surface is not
well—defined since there always exist some unreacted functional groups which makes the
characterization and interpretation of data more complex. An end-capping procedure is
sometimes used to solve this problem. In contrast, monofunctional Silanes yield better
defined surfaces which can be characterized more easily. A disadvantage is that a stable

surface with high organic group coverage is often difficult to achieve.

Silylation reactions between monofunctional Silanes (XSiMe3) and firmed silica
were carried out to prepare well-defined surfaces with known silanol coverage. The TGA
and titration results are shown in Table 2. When X = Cl, replacement of silanol groups by
TMS groups is quantitative until a maximum coverage of 73 % is reached. Further
increases in the silane/Si-OH ratio do not enhance the attachment as evidenced by titration
and also IR spectra (Figure 9). When X = OEt and one equivalent of the silane is used,

only 35% of the surface silanol groups reacted.

51

Table 2. Results of silylation of fumed silica with XSIMC3.

 

 

 

 

 

 

Entry Sample Name Silane/Si-OH TGA loss, Coverage of
Number Ratio' wt % TMS, %"
1 A200 - 0.59 0
2 FS-TMCS 1c 0.25 1.8 24
3 FS-TMCS 2 0.70 2.3 73
4 FS-TMCS 3 1.0 2.4 73
5 FS-TMES 1" 1.0 0.74 35.0

 

 

 

 

 

 

 

a) Molar ratio of silane to AZOOSilanol groups.
b) Calculated from titration experiments.
c) TMCS = trimethylchlorosilane.

d) TMES = trimethylethoxysilane.

52

Relative Absorbance

 

 

 

 

 

 

 

J
W I

3800 3300 2800 2300 1800 1300

 

-1
Wavenumber, cm

Figure 9. IR spectra of the silylation products of A200 and Me3SiCl
at three silane/Si-OH ratios: (a) F S-TMCS-l, 0.25;

(b) FS-TMCS-Z, 0.70; (c) FS-TMCS-3, 1.0.

II. Hydrophobic Silica

Based on the above silylation conditions, we first prepared hydrophobic silicas
with different alkyl groups (Scheme 14). Figure 10 shows IR spectra of these silica
products. The corresponding TGA curves are given in Figure 11. We did not try to
quantify the organic group coverage through titration, since it is not possible to
differentiate between residual surface silanols and those from the hydrolysis of silane
. molecules. IR spectra demonstrate a steady increase in the intensity of C-H absorption
bands in going fi'om short (C2) to long (Cts) chains, and TGA curves show a constant
increase in weight losses. The increase, however, is not linear - a consequence of the steric

efl‘ects of long alkyl groups.

Scheme 15. Silylations of A200 using modifiers of different lengths.

I-INEtz/‘tohene A
.—0H + RSiCl3 —» .—|—R

FS-CZ, R= C2H5; FS-C8, R = Can; FS-CIS, R = C13H37

54

Relative Absorbance

 

 

 

 

 

'=_—__—=:

 

 

 

 

 

 

; J

 

T I

3800 3300 2800 2300 1800 1300

Wavenumber, cm'1

Figure 10. IR spectra of silylation products of A200 with RSiCl3:

’ (a) FS-C2, R = Csz; (b) rscs, R = Can; (c) FS-C18, R = C13H37.

55

Weight, %

 

100 7

 

98‘

96*

94~

92~

90~

 

88 ‘-

 

 

 

86 i i f i i i
100 200 300 400 500 600 700 800

Temperature, °C

Figure 11. TGA curves of silylation products of A200 with RSiC13:

(a) A200; (b) FS-CZ, R = C2115; (c) FS-C8, R = CiHn; (d) FS-C18, R = C18H37.

56

A series of A200 derivatives featuring different alkyl coverages were prepared
using octyldimethylchlorosilane (ODMCS), Can(Me)2SiCl, a monofirnctional silane (see
Scheme 15). IR spectra and TGA data for the samples are given in Figures 12 and 13,
respectively, and the results of silanol titrations and TGA weight losses are listed in Table
3. Increasing the silane/SiOH ratio (i.e., increasing the silane amount) leads to an increase
in silanol group replacement as high as 42%, attenuation of the isolated silanol peak, and

the increase in C-H relative absorptiOn on IR spectra and in sample weight losses on TGA

CUTVCS.

Scheme 16. Silylations of A200 with varying amounts of ODMCS.

 

HNEt2/tohene Ire
.414 + C3H17Si(Me)2Cl » .—a—Ii—can
Me

FS-DM-C8-n (n = 1, 2, 3)

Table 3. Results of silylations of A200 with octyldimethylchlorosilane (ODMCS).

 

 

 

 

Sample Silane/OH TGA loss OH Replaced
Name Molar Ratio wt % mol %
F S-DM-C8-1 0.034 1.6 1
F S-DM-C8-2 0.24 2.4 31
FS-DM-C8-3 1.0 3.9 42

 

 

 

 

 

 

57

Relative Absorbance

 

 

 

 

 

 

J .L l I l
r r r T

3800 3300 2800 2300 1800 1300

-1
Wavenumber, cm

Figure 12. IR spectra of silylation products of A200 and Cal-117(Me)28iCl
obtained at difi‘erent silane/Si-OH molar ratios: (a) FS-DM—C8-l, 0.03;

(b) FS-DM-C8-2, 0.3; (c) FS-DM-C8-3, 1.0.

58

Weight, ”/6

 

 

 

 

101
100 J_
99 .-
a
98 s, b
97 u_
c
96 ~-
95 4 i : .L a : .
100 200 ' 300 400 500 600 700 800

Temperature, °C

Figure 13. TGA curves for silylation products of A200 and C3H17(Me)2SiCl
obtained at different silane/OH molar ratios: (a) FS-DM-CS-l, 0.03;

(b) FS-DM-C8-2, 0.3; (c) FS-DM-C8-3, 1.0.

59

Scheme 17. Silylations of A200 with dimethylchlorosilane.

 

HNEtz/toluene 1;“
.—OH + HSi(Me)2Cl » .—e——|i—H
Me

FS-DM-H

Hydride modified silicas, as envisioned by Sandoval et al.,58t59 are usefirl
intermediates in the synthesis of many surface bonded materials. Hydride modified silica
surfaces are hydrophobic in nature due to siloxane and Si-H linkages, and for some cases,
alkyl groups. There are several ways of introducing the Si-H firnctionality onto silica
surface. Both mono- and multi-filnctional Silanes can be used, and in most cases, a
catalyst/promoter (either an acid or a base) is added. We first tried silanization using
dimethylchlorosilane (Scheme 16) using a base promoted process. The product could be
filtered easily and by-products and excess silane could be removed by a combination of
solvent rinsing and vacuum drying at elevated temperatures. The yields were usually
quantitative. An IR spectrum of sample F S-DM-H is shown in Figure 14, curve d, where

the band at 2250 cm’1 is due to the Si-H absorption.-

Scheme 18. Silylations of A200 with triethoxysilane.

_l,_
b

.—eH + rrsi(OEt)3 __i_. .—e—ii-H
_§,_
1 I

—;i-— —bi— —-bi-—
d
ii-H -—> mad—(B —biH-H—->® .—O—£i-H

burs

PS-OEt-H FS-OH-H ‘ FS-TMS-H

(i) HCl/H20/dioxane, (ii) HNEtz/toluene, (iii) HCl/HzO/dioxane,

(iv) HNEt2/toluene/TMSC1.

61

We also used trialkoxysilane as a modifier. In this case, we tried two different
approaches. The first was an acid—catalyzed process (Scheme 17) in which ethoxy silane
molecules hydrolyze and condense with surface silanol groups. The advantage of this
method is that it gives a more stable surface due to cross-linking of adjacent Silanes and
the absence of non-hydrolyzed alkoxy groups. Two conclusions could be made based on
the IR spectrum of the product (Figure 14, curve e). Acid hydrolysis of alkoxy groups was
complete since the C-H absorption due to ethoxy groups nearly disappeared, but the
condensation between silanol groups was incomplete as evidenced by a strong absorption

at 3744 cm'1 due to isolated silanol groups.

The second method we employed was a base-catalyzed silanization process. In
contrast to the first method, these products show incomplete hydrolysis of ethoxy groups,
a lower Si-H density, and a lower Si-OH density (Figure 14, curve a). The residual ethoxy
groups were hydrolyzed in an acidic medium, and except for a lower density of hydrides,
the product generated resembles the one made using an acid catalyst (curves b and e). The
silanol groups generated from the hydrolysis reaction were end capped using a small silane
compound, TMSCI, in the presence of a base catalyst. As can be seen from curve c, the
silanol groups are replaced by TMS groups to a significant extent and the product is
similar to that obtained by direct silylation using dimethychlorosilane (curve (1). In both
cases, the C-H asymmetric and symmetric stretch modes of the methyl groups are clearly

seen (curves c and d).

62

Due to the oxidation of Si-H linkages with oxygen, hydride modified silica
products show thermal transitions at elevated temperatures which which can be used to
verify the presence of hydrides. The TGA curves for samples F S-DM-H and FS-H are
given in Figure 15. Three distinct stages are seen in TGA curves that can be rationalized in
the following way: (1) an initial weight loss due to the condensation of residual silanol
groups; (2) a weight gain due to the oxidative transformation of Si-H into Si-OH moieties;
(3) filrther weight loss from firrther condensation of new silanol groups. The presence of
hydrides on modified silica surfaces is also evident in their DSC traces, Figure 16. We
were unable to quantify the hydride coverage from DSC results due to the lack of
standards, but qualitatively, sample F S-DM-H shows a lower hydride coverage and lower
thermal stability compared to F S-H. The endothermic transition is weaker in DSC traces
and less weight was gained in TGA experiments at lower temperatures during the
oxidation period. The same conclusion can be drawn fiom IR spectra of the two samples

where F S-H shows a stronger Si-H absorption at ca. 2260 cm".

63

 

 

 

Relative Absorbance

 

 

 

 

3800 3300 28m 2300 1800 1300

- Wavenumber, cm'1

Figure 14. IR spectra of hydride modified A200 as defined by Schemes 16 and 17:

(a) FS-OEt-H; (b) FS-OH-H; (c) FS-TMS-H; (d) FS-DM-H; (e) FS-H.

Weight, %

100.5

100

99.5

99

 

 

F S-DM-H

F S-H

 

 

1
I

100 200

1
j

300 400

l
j

500

1
T

600

1
fi

700

Temperature, 0C

1
I

800

Figure 15. TGA curves of samples FS-H and FS-DM—H.

65

900

Heat Flow,—> endo

 

 

 

 

 

200

A; 4
I

r 1 L
I I I

250 300 350 400 450 500 550

Temperature, 0C

Figure 16. DSC traces of samples FS-H and FS-DM-H.

600

III. Hydrophilic Silica

In the Introduction, we discussed electrolyte systems that incorporate silica as filler
materials. We mentioned that hydrophobic silica is superior to hydrophilic silica in terms
of its thickening effects in polar liquids. However, the silanol groups on pristine fumed
silica are acidic and electrochemically unstable, and thus are not suitable for making
electrolytes for rechargeable lithium batteries. If however, a fraction of the silanol groups
are replaced with stable organic groups, the remaining silanols would be shielded and this
problem could be suppressed. We first decided to make silica fillers compatible with PEO
derivatives by tethering PEO oligomers onto the silica surface. By doing so, the
electrochemical stability of silica toward lithium metal would be enhanced, and the

compatibility of the PEO oligomer and the filler would be improved.

Scheme 18 shows reactions leading to chlorosilanes containing ethylene oxide
units and their attachment to fumed silica surfaces. The properties PEO modified silica
samples are listed in Table 4. As seen in the Table, 40 mol % of the silanol groups could
be replaced by short PEG chains. As the chain length increases, replacement of silanol
groups becomes more and more difficult and only 20% replacement was possible for the
chains with 17 ethylene oxide repeat units. This is mainly due to the steric effects. TGA
curves for these samples are given in Figure 17 which show that the percentage weight
losses increase with the chain length as expected. The same conclusion was made from
their IR spectra (Figure 18) which show a progressive increase in the intensity of the C-H

absorption of ethylene oxide units. The peak at 1752 cm'1 was not expected and has not

67

Scheme 19. Preparation of PEO tethered filmed silicas: First approach.

1 (n=1)
2 (n=2)
CH OH 3 (n=3)
3IO/\’l’n 4 (n=8)
_ “ 5 (n=12)
7 (n=l) 1. ll 6 (n=17)
8 (n=2) '
9 (n=3) CH O
10 (n=3) 31‘0N’fl \/\
ll (n=12)
12 (n=l7)
I“ 13 (n=1)
l4 (n=2)
0 S. l 15 (n=3)
n 17 (n=12)
rs (n=17)

CH3lO/\’l/:W?ie—O

FS-DM-EOn ( n = 1, 2, 3, s, 12, 17 )

(i) NaH/THF; (ii) CH2=CH-CH2Br; (iii) HSi(Me)2C1; Pt cat.,

toluene; (iv) A200, HNEtz/toluene.

68

Table 4. Modified silica fillers and their properties.

 

 

 

 

 

 

 

Sample TGA loss, Coverage of Si-OH,

Name wt % mol %

A200 0.6 100
FS-DM-EO. 2. 1 62
FS-DM-EO; 4.9 60
FS-DM-EO3 6.4 65
FS-DM-EOg' 8.8 76
F S-DM-EOI-r‘I 16.4 80

 

 

 

 

 

' Average number of ethylene oxide units.

yet been assigned. Although it is now difficult to explain why the weight loss occurs at a
lower temperature for sample F S-DM-EOn, our TGA experiments have shown that,

compared with pure PEO, the physical blend of PEG and pristine firmed silica starts to
lose weight at much lower temperature (ca. 100 °C lower), an indication of involvement

of silanol groups in decomposing the PEO chains, probably as a result of acid catalysis.

We also tried another approach to making PEO modified silicas (Scheme 19).
Poly(ethylene glycol) allyl methyl ethers were prepared as before and were allowed to
react with hydride modified silica intermediates FS-DM-H and FS-H. IR spectra of the

products are shown in Figure 19. When FS-H was used, the residual Si-H peak is still

69

Weight, %

 

102

1

100 i

98'“

96 ~-

94 ~-

92 7*

88 ‘-

86 ‘r

84 .-

=1

n=2

n:

 

n=17

 

 

82
100

.L
1

is

1
1

200 300 400 500 600

Temperature, 0C

Figure 17. TGA curves of samples FS-DM-EOn.

70

 

700 800

Relative Absorbance

 

{f

 

 

 

 

 

 

n=17

n:

 

=2

 

n=l

 

 

 

 

3800 3300 2800 2300 1800 1300

.1
Wavenumber, cm

Figure 18. IR spectra of samples FS-DM-EOn.

discernible after the hydrosilation reaction, not surprising since increased amounts of
attached ethylene oxide chains lead to increased steric shielding of the Si-H bond.
However, when F S-DM-H was used, the Si-H bond is no longer seen after the reaction. A
reasonable explanation for this result is that a portion of the silyl groups bearing Si-H
linkages left the surface during hydrosilation. This did not happen in the case of F S-H due

to its higher stability compared to F S-DM-H.

Scheme 20. Preparation of PEO tethered fumed silicas: Second approach.

CH3 OW 9 £11113
0 n=
T N; 10 (n=8)
FS-H or 11 (n=12)
FS-DM-H
Pt cat

CH3TO/\l/n0\/\_%i—O

FS'EOn or PS-DM-S'EOn
(n= 1,3, 8, 12)

72

Relative Absorbance

 

 

 

 

 

L I I

4000 3500 3000 - 2500 2000

Wavenumber, cm'1

Figure 19. IR spectra of (a) FS-DM-S-EOt; (b) FS-DM-S-EOs;

(c) FS-EOI; (d) FS-E03; (e) FS-EOs; (f) FS-EOIZ.

IV. Polymerizable Silica

1. Silylation Using Single Silanes

As mentioned earlier, silica particles can form three-dimensional networks, but this
structure is only temporary and is easily broken down on shearing. For long term
reliability, permanent dimensional stabilityis desirable. In composite polymer electrolytes
using filmed silicas as fillers, dimensional stability can be introduced as a cross-linking step
that binds adjacent filmed silica particles into a stronger network structure. Early work on
the preparation of polymerizable fumed silicas focused on using the coupling agent 3-
(trimethoxysilyl)propyl methacrylate (TPM), an economical commercial product with a
short hydrocarbon spacer (C3). As seen later, silylation of A200 with TPM can generate
products with very high coverage of functional groups. Unfortunately, this product gave
us a liquid-like gel upon mixing with oligomeric PEO due to hydrophilic character of the
ester functional group on the surface modified silica product. In order to impart both
polymerizability and hydrophobicity to firmed silica, we prepared two coupling agents with
longer hydrocarbon spacers (Cs and Cs), hoping that longer spacers would confer

hydrophobic characteristics to the surface.

Polymerizable silicas were initially prepared through several steps according to
Scheme 20, method 1. The needed 4-penten-1-ol (20) is commercially available while 7-
octen-l-ol (21) was prepared via its TI-IP ether (25) as outlined in Scheme 21. The THP

ethers of or,c0-alkenols (22 and 23) were allowed to undergo hydrosilation reaction with

74

Scheme 21. Preparation of cross-linkable firmed silicas: Method 1.

19 i l 20 (n=3)
21 (n=6)
0 CW
22 (n=3)
ii 23 (n=6)

0 Owl... @

m l FS-CS-P (n=3)
FS-C8-P (n=6)

n

,v FS-CS-OH (n=3)
FS-CS-OH (n=6)

FS-CS-M-l (n=3)
FS-C8-M-l (n=6)

(i) TsOH, dioxane; (ii) F S-DM-H or F S-I-I, Pt cat., toluene;

HCl/H20/ EtOH; (iv) methacryloyl chloride, pyridine, toluene.

75

Scheme 22. Synthesis of 7-octen-1-ol.

i
Home ——. (1
6 0 CMC]

24 25

21 23

(i) TsOH, DHP, dioxane; (ii) KO t-Bu, DMSO; (iii) HCl/HzO/ EtOH.

either FS-DM-H or FS-H. Deprotection of these intermediates afforded silica surfaces
containing alkyl chains temrinated with hydroxyl groups. Polymerizable silicas, FS-CS-M-
1 and F S-C8-M-1, were obtained through the base promoted acylation of surface hydroxyl
groups with methacryloyl chloride. Two representative IR spectra of the products thus
generated are shown in Figure 20. As seen, high surface attachment was not achieved,

probably due to both incomplete hydrosilation and acylation.

In Method 2, shown in Scheme 22, hydrosilation would take place in a
homogeneous phase and a higher yield would be expected. Initially, we chose acryloyl
chloride (26a, R= H) over methacryloyl chloride (26b, R = Me) for the acylation (reaction
i) because the former has higher reactivity towards radical polymerization. Much work

was devoted to optimizing the hydrosilation reaction, but satisfactory results were not

obtained. According to the literature,84 hydrosilation of terminal alkenes takes place

76

Relative Absorbance

 

 

F S-CS—M-l

FS-C8-M—1

1 AL I

  
   
 

 

   

 

 

 

I
T I I I I

3800 3300 2800 2300 1800

-1
Wavenumber, cm

Figure 20. IR spectra of samples F S-C5-M-1 and F S-C8-M—1

as defined in Scheme 20.

77

1300

Scheme 23. Preparation of cross-linkable silicas: Method 2.

O
R .
26a (R=H) ll 20 (n=3)
26b (R=Me) o 21 (n=6)

ii 11

R 27a (n=3,R=H)
27b (n=3, R=Me)
28a (n=6, R=H)

28b (n=6, R=Me)

Yioi’iy/Nix, .1. RH/OW%.O_.

29 (n=3, X=OEt, R=Me) _
30 (n=6, x=OEt, R=Me) FS'CS'M" (" ‘ 3)
31 (n=6, X=C1,R=Me) FS'CS‘M'Z 0‘ = 5)

(i)CC14, 3A mol. sieves, 80 °C; (ii) HSiX3, Pt cat., toluene, 60 °C overnight:

9

(iii) FS-H or FS-DM-H, Pt cat., toluene; (iv) A200, HNEtz, toluene.

78

at temperatures just above room temperature, but the conjugated double bond of the
acrylate group is less reactive and hydrosilation occurs in only moderate yields at
temperatures of 100 °C or higher. Based on this precedent, hydrosilation of our alkenyl
acrylate (27a and 28a, reaction ii) should occur selectively at the alkenyl group.
Unfortunately, hydrosilation at the acrylate double bond occurred under a range of
reaction conditions. The same results were obtained for hydrosilation of the same alkenyl
acrylate using hydride modified silica in heterogeneous phase (reaction iii). When
methacrylol chloride (26b) was used for the acylation reaction, hydrosilation reactions
occurred specifically at the alkenyl double bond. Clearly, the steric bulkof the methyl

group on the alkenyl methacrylate plays an important role in inhibiting hydrosilation.

Separation problems were common when triethoxysilane was used. Crystallization
was not successful, and chromatography and vacuum distillation gave either little or none
of the desired monomer product. Removal of the excess of triethoxysilane under vacuum
was not successfirl due to its high boiling point (bp 134 °C). By switching to the lower

boiling trichlorosilane (bp 32 °C), excess silane could be removed by vacuum.

The IR spectrum of a polymerizable silica made from 8-(trichlorosilyl)octyl
methacrylate (31), sample FS-C8-M-2, is shown in Figure 21 along with FS-C3-M, a
sample made using commercial TPM. Their schematic structures are shown in Figure 22.

The absorption bands characteristic of carbonyl and C=C double bonds are comparable in

79

Relative Absorbance

 

 

 

F S-C3-M

FS-C8-M-2

 

 

 

 

 

 

3800 3300 2800 2300 1800

-1
Wavenumber, cm

Figure 21. IR spectra of polymerizable silicas: FS-C3-M and F S-C8-M-2.

1300

0
HO OH to HO

HO 0 OH
OH
o

H o o H H ' OJY

OH
HO o OH .
HO H ) 4 HO
o

:V

o ‘3.
Yo
Figure 22. Schematic representations of A200 (top left),

FS-C3-M (top right), and F S-C8-M—2 (bottom).

81

intensity for both samples, and are stronger than for the silicas shown in Figure 20,
suggesting a higher coverage of methacrylates. The small peak at ca. 2260 cm'1 in the

spectrum of sample F S-C8-M-2 is due to the small amount of Si-H impurity.

The significance of the above experiments is that the thickening effect of the silicas
was enhanced as the hydrocarbon spacer was increased. The best results were obtained
when a Ca spacer was present between silica surface and methacryloyl group. We were
targeting a silica filler which, upon mixing with PEG oligomer and lithium salt, would
generate a physically stable gel thin enough to allow trapped air bubbles to be pumped out

by applying a moderate vacuum. FS-C8-M-2 meets the above requirements very nicely.

2. Silylation Using Mixed Silanes

The major difference between samples F S-C3-M and FS-C8-M-2 is the length of
carbon spacer that tethers the methacrylate group to the silica surface. The methacrylate
group by itself is polar in nature, whereas the hydrocarbon spacers are non-polar. The
direct result of increasing the the spacer length is to increase the hydrophobicity of the
silica, and in turn its thickening effect in polar solvents. Based on this rationale, we
developed a second route to polymerizable hydrophobic flamed silicas - silylation using
mixtures of silanes. One silane would cany a hydrolyzable group for attachment to silica
and a polymerizable group; while the other silane would only need to have a hydrolyzable
group. The advantage of mixed silylations is that a variety of commercial silane products
become usefial to us. Although only a few TPM-like coupling agents are available

commercially, there are many silanes such as octyltn'methoxysilane (OTMS) that bear

82

hydrolyzable groups. Ranging from short chains (methyl group) to longer chains (decyl
and octadecyl groups), each of these silanes can be either monofunctional or
multifunctional. Many combinations of silanes are possible, and the relative amounts of the
two silanes within each combination are also tunable. In summary, by using the mixed

silylation scheme, the polymerizable silica surface can be “tailored” in a delicate and

sophisticated way.

We selected TPM and OTMS for our mixed silylation reactions. The former offers
polymerizability through its methacryloyl group and the latter provides hydrophobicity due
to its long octyl chain. Table 5 lists the experimental conditions and results of such studies
(TOMn will be used in the following discussion to represent such products). Known
amounts of the two silanes were weighed to give the desired TPM mole fraction. The
relative and total degree of attachment of the silanes were determined from the mass and
1H NMR spectrum of the reaction residue collected through filtration and solvent rinsing
of the silica. A portion of the 'H NMR spectrum for the residue of sample TOM4 is shown
in Figure 22. As can be seen, the peak for the methyl group adjacent to the double bond
on TPM (1.95 ppm) and the terminal methyl group on OTMS (0.89 ppm) are well-
separated from other peaks and were used for integration. The calculated weight losses for
TGA experiments were derived using the following assumptions:

0 Each silane molecule reacts with one silanol group.

0 All of the methoxy groups convert to siloxane during silylation.

0 Upon heating, the silicon-carbon bonds first break to form silanol groups, and

two such silanol groups then condense to give siloxane linkage.

83

Table 5. Mixed silylation of A200 with OTMS and TPM.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample TPM/(TPM+OTMS), Total Silane Weight Loss Weight Loss,
Name TGA (expt.) TGA (calc.)
mol % mmol/g SiOz
% %
Feed Residue Surface
FS-OTMS 0.0 0.0 0.0 400 4.0 4.6
FS-TOM-l 9.0 0.0 7.2 330 3.9 4.0
FS-TOM-Z 20 1.3 31 510 5.3 5.8
FS-TOM-3 24 5.6 35 520 4.8 6.3
FS-TOM-4 31 13 37 490 6.0 6.2
F S-TOM-S 50 21 82 380 5.0 4.9
FS-C3-M 100 100 100 550 6.5 6.7
CH3 in OTMS

l

k JUAL

l I l l l l I 3 l : l E l 3 l l I

2.0 1.5 1.0 0.5

CH3 in TOM

 

 

 

 

Figure 23. A portion of the 1H NMR spectrum of the residue from sample TOM4.

84

 

Although the actual situation deviates somewhat fiom this model (some unreacted

methoxy groups present after the silylation), the calculated weight losses are very close to

those measured by TGA analyses.

When the two pure silanes were used separately for silylation, the TPM modified
product showed a slightly higher coverage than OTMS-modified silica (Table 5). The
situation changed dramatically when the silylation reaction was carried out using a mixed
silane solution. Compared to OTMS, TPM attaches to the silica surface preferentially.
Results on mass balance studies listed in Table 5 were plotted in Figure 23. Although the
TPM fractions on silica surface and in residue are all approximately proportional to the
TPM fraction in feed, the fraction on surface is always larger than that in the residue. In
addition, the rate of increase of TPM fraction on the surface is always higher (larger slope)
than that in the residue. Considering the structure difl‘erences in the two silane molecules,
the above observation can best be explained in terms of selective adsorption of TPM over
OTMS, since the former has ester fimctional groups. The schematic representations of
sample FS-TOM4 as well as a commercial hydrophobic silica R805 are given in Figure 24.
Compared with R805, the modified silica FS-TOM4 has not only hydrophobic groups
(octyl), but also cross-linkable groups. Thus, the TPM fiaction on silica surface can be
controlled by adjusting the feed composition. IR spectra of the selected samples from
Table 5 are shown in Figure 25 which clearly show the expected increases in the relative

absorbances of the carbonyl and double bond as the proportion of TPM is increased.

85

HO

OH
H
OH
HO
HO
OH
H OW/K
O
O OH

=24) 110

Figure 24. Schematic representations of R805 (top) and FS-TOM-4 (bottom).

TPM, mol %

 

 

 

 

 

 

 

 

90
I
80 -~
0 In Reisdue
70 " I On Surface
60 ~-
50 _,
40 4
30 «-
20 ~~
10 ~~
o . 0 °
‘10 I TL 41 l 1r
0 10 20 30 40 50
TPM in Feed, mol %

Figure 25. TPM fractions on silica surface and in residue.

87

 

Relative Absorbance

 

 

 

 

 

 

 

 

 

 

-1
Wavenumber, cm

Figure 26. IR spectra of representative samples from Table 5.

88

N ..
N TOM2
ML TOMS
OTMS

38:00 33:10 zsfoo 2500 11:00

1300

The chemical attachment of silane coupling agents to silica surfaces was also
verified by Cross-Polarization Magic Angle Spinning (CP/MAS) NMR spectroscopy. The
29Si CP/MAS spectra of samples TOM4 and FS-C8-M are given in Figure 25. The broad
peak from -80 to -110 ppm contains more than one component and is characteristic of
finned silicas without Si-C linkages. The resonances can be assigned to silicon atoms
surrounded by four oxygen atoms. The other broad peak, ranging from -30 to -60 ppm, is
also multi- component and is related to silicon atoms that have three oxygen atoms and
one carbon in their nearest neighborhood. The 13C CP/MAS NMR spectra of the same
samples are shown in Figure 26 where the carbonyl carbon is seen at 170 ppm. The two
peaks corresponding to the methacrylate double bond overlap due to the poor spectral
resolution and the peak at 50 ppm can be assigned to unreacted methoxy groups which are
not present in the spectrum of FS-C8-M due to the complete hydrolysis of chlorosilyl

groups.

89

 

 

 

 

 

 

IUUIIIIIIIUIIIIIUIIIIIIIli'lilllill'l'lllllIIIUIIIIIllllllIIIIIUIIII'IIIIIIIIIIITIUUIIIII[IIIIIIIIIIIFTIIIIII[IIII'IIIUIIll'l

40 20 O -20 -4O -60 -80 -120 -160 pm

 

 

A
'lIIIIUIIIlll'l'l'll'IIIIIll'l'[I'llVli'Ull'l'IIIIUIIIIIIIIIIIIIIIIIIIIIIIIIUIIIFUIIIIIIIIIUIUIIIIIIIIIIIIIIIIUIIIIIIIIUIIITT

40 20 0 -20 -40 -60 -80 -120 -160 pm

 

 

 

Figure 27. 29Si CP/MAS spectra of samples TOM4 (top spectrum) and
FS-C8-M (bottom spectrum). For each sample from top to bottom:

actual spectrum, full fit and individual component plots.

 

 

 

 

 

 

—I I I I I H l I I I I I I I I I I I I I I l I I I I I I I I I I I I I I I

250 200 150 " 100 so 0 -50 ppm

 

 

 

 

IIlIIIIIIIrITHTII'IIIIlIIIIlIIIIlfiTjIl

250 200 150 100 so 0 -50 ppm

Figure 28. '3 C CP/MAS spectra of samples TOM4 (top) and FS-C8-M (bottom).

91

V. Composite Polymer Electrolytes

We prepared composite polymer electrolytes using all the silica products prepared.
While both oligomeric PEG-bearing silicas and hydrophobic and cross-linkable silicas were
used, emphasis was given to the latter since the mechanical properties of the composite

were enhanced with the addition of a monomer and subsequent cross-linking.

1. Compatible Composite Polymer Electrolytes

Most composite polymer electrolyte samples were made using 15 wt % of silica,
but others required more filler to prevent the electrolyte from flowing. When the filler is a
pure silica or one modified with long PEO chains, more than 25 wt % of silica is needed

for gelation (samples A200 and F S-DM-EOn, Table 6). Conductivities for all electrolytes

 

Figure 29. Schematic of PEO modified silicas.

92

Table 6. Conductivity data of electrolytes with compatible fillers.

 

 

 

 

 

 

 

Sample Name Silica, wt % 0* 10‘, S/cm
A200 25 1 .01
FS-DM—E01 15 1.18
FS-DM-Eoz l 5 l .89
FS-DM-EO3 1 5 1 .49
FS-DM-Eos' l 5 l .66
F S--DM--E017‘I 28 1 .05

 

 

 

 

 

‘ Average number of ethylene oxide units.

fall in a narrow range. Electrolytes containing pure silica and FS-DM-EOu show similar
but lower conductivities than the others, probably due to a “diluting effect” from the filler
materials (25 and 28 wt %, respectively). All of these electrolytes are processable gels
with good mechanical stability. For example, no sign of phase separation or of filler
settling out has been seen over a period two years. It is interesting to note that the filler
with the longest PEG chain requires less PEGDME to form a gel. Therefore, it might be ‘
possible to prepare a two-component composite polymer electrolyte incorporating only
lithium salt and a “core-shell” material, featuring both good mechanical properties due to
the silica “core” and high ionic conductivity because of a “shell” of PEG chains. There are
many ways to realize such materials. For example, copolymerization of a polymerizable
silica with a monomer such as poly(ethylene glycol) ethyl methacrylate, or similar

monomers, would generate a core-shell structure.

93

2. Cross-linked Composite Polymer Electrolytes

The cross-linking experiments were carried out by adding a small portion of
organic monomer (10 wt %) initiator to the recipe used for making composite electrolytes
from filmed silica and polyethers. Cross-linking was achieved by exposure to UV light.
Scheme 23 shows how a cross-linked fumed silica composite was made starting from
commercial firmed silica. The four steps involved were:

0 Hydration of Aerosil 200 to increase the silane attachment;

0 Silylation at room temperature in an organic solvent;

0 blending of the electrolyte components;

0 UV light initiated cross-linking. .

The models of the target composite before and after cross-linking are shown in Figure 29.
In short, our goal was to prepared first a temporary three-dimensional network, and then
lock this network in place through a cross-linking step. In order to achieve this goal, the
following issues were crucial to us. (1) The modified fumed silica would have to bear both
hydrophobic and cross-linkable groups. The former would provide a physical gel through
van der Waals attractions, and the latter would wpolymerize with the added monomer to
turn the gel structure into a dimensionally stable material. (2) The monomer would have to
be able to access to the cross-linkable groups on the silica surface in order for the cross-
linking to occur. The most extreme and favorable case is where the monomer phase
separates fi'om the PEGDME-500 and forms a thin layer around the silica network. Upon

cross-linking, the monomer copolymerizes with silica functional groups

94

Scheme 24. Preparation of cross-linked composite polymer electrolytes.

Fumed silica

fl Half-saturated NH4NO3

 

 

Hydrated Fumed Silica
_ Et
EtO—§i(CH2)7CH3
OEt
OEt
v EtO—Si—CHZCHZCH20>_zH3
OEt

0 CH2

H36 O-Cflzcflzcflbs/l\o

/ on
"20 o :21; I
' .-:-',c-.’---.-‘ H
0 CH2
0
\ /
/8|

 

o OCH20I‘I20H2 \
CllzcflzCHzO 0":

I I 8:1(6 H2)1CI'I3

Fumed silica - TOM

LiClO4
butyl methacrylate
PEG-DM-500

AIBN, UV

Cross-linked Fumed Silica Composite

95

Temporary

 

 

Silica Network

=

.2

E

a: g

E

2'

c

a.

8

V 7
Permanent

3-D Network

 

 

 

 

Figure 30. Designed models for cross-linkable composite polymer electrolyte:

before (top) and after (bottom) cross-linking.

and the formed polymer grafis on silica surface to reinforce the strength of the network.

The conductivity of the electrolyte would remain the same because the polyether-Li salt

matrix would be unchanged.

Since PEGDME-500 is a polar substrate and the network is hydrophobic, a non-
polar monomer compatible with surface cross-linkable groups would meet the above
criteria. The criteria for the selection of polymerization initiator would be similar to that
for the monomer where an oil-soluble initiator would be desirable in order for the cross-
linking to take place selectively near the network surface. A few monomer were tested and
butyl methacrylate (BMA) was chosen as the monomer, and 2,2’-azobisisobutyronitrile
(AIBN) was selected as initiator. BMA was distilled under vacuum before use. PEGDME-
500 was treated with calcium hydride for three days at room temperature and the liquid
was then distilled under vacuum. A series of studies were carried out, including the
physical state of the electrolyte in relation to its composition, monitoring the cross-linking
process using IR spectroscopy, morphology studies of composites using Freeze-Fracture
TEM, thermal characterization of composites, and the temperature and salt-dependent

conductivities.

The physical state of the composite depends on the type of fumed silica used , as
well as the amount of fumed silica and LiClO. in the composite. In general, the viscosity
increases with the increase of .silica and lithium salt contents. This trend follows the well
known increase in T8 seen for PEG electrolytes with increasing salt concentrations. The ‘

silica content in the composite was 10 wt% in all preparations, since lesser amounts tend

97

not to provide an adequate thickening effect, and larger amounts usually results in a
dilution effect that leads to lower conductivities. The air bubbles trapped in the composites
prepared from methacrylates other than MMA during blending were removed by applying
a vacuum. The time required depends on the viscosity of the composite. MMA-containing
composites were not treated with vacuum. Instead, an aging process was used and the
composites were air-free spontaneously after a few days. At such a silica content, at least a
day is necessary to remove the air bubbles fully fi'om R805 composites due to its high
viscosity. Composites containing F S-C3-M show the lowest viscosity and FS-C8-M—2 is
an ideal choice for composite polymer electrolytes since it ofl‘ers a moderate viscosity

which allows for the bubbles to be removed readily and for the gel to be physically stable.

Figure 30 shows IR spectra of a composite polymer electrolyte before and afier
cross-linking. The absorption at 1640 cm'1 due to the methacrylate double bond (which
overlaps with water deformation mode) disappeared, and the carbonyl peak at 1705 cm'1
shified to 1724 cm'1 as a result of loss of the double bond conjugation. The UV-initiated
polymerization kinetics were monitored using m spectroscopy. The spectrum at time zero
was taken as the reference so that all of the spectra shown in Figure .31 are difference
spectra. As can be seen, the double bond and carbonyl peaks are consumed with
irradiation time. The relative intensity (peak height) of these peaks at different times
represents the conversion of monomer to polymer and were recorded in Table 8 and used

to make Figure 32. It is easy to see that polymerization was complete in less than 20 min.

98

 

 

 

 

 

Relative Absorbance

.l.
M

r
I

 

 

 

L
1

l

3800 3300 2800 2300 1800 1300

-1
Wavenumber, cm

Figure 31. IR spectra of composites before (top) and after (bottom) cross-linking.

99

 

Table 7 IR data for UV initiated polymerization.

 

 

 

 

 

 

 

 

 

 

 

 

 

Intensity Intensity Time Conversion, % Conversion, %
C=O C=C minutes C=O C=C
8.0 2.5 1.0 12.3 21.2
12.5 3.5 2.0 19.2 29.7
15.8 4.0 3.0 24.3 33.9
20.0 4.6 4.0 30.8 39.0
29.3 6.0 6.0 45.1 50.8
42.5 7.5 8.0 65.4 63.6
55.0 9.0 10.0 84.6 76.3
60.0 9.7 12.0 92.3 82.2
63.0 10.2 14.0 96.9 86.4
65.0 11.2 18.0 100.0 94.9
63.0 11.8 26.0 96.9 100.0

 

 

 

 

 

100

 

[0'11

loss of C=O at 1705 cm:1

.\
lzl
.....l;

4

 

26
18

14

12

10

 

 

 

 

. “l

{k 1 new C=O at 1724 cm'1

K

 

 

 

{ll

 

loss of C=C at 1640 cm'1

Figure 32. IR-generated cross-linking kinetics.Spectra are a series, and correspond

to the original spectrum-the spectrum taken at time t.

101

 

110

 

. _ A
100 A C=C A

 

 

I C=C I

 

Conversion, "/0

20""

10 r"

DI

 

 

 

0 I 1 P 1 1
I I I I T

0 S 10 15 20 25

Irradiation Time, minute

Figure 33. Conversion of polymerization vs irradiation time.

102

30

The conversion curve obtained from the carbonyl peak deviates slightly from that achieved
from the double bond peak, probably due to the overlapped absorption of carbonyl group

before and after cross-linking.

In order to study the effects of silica and subsequent cross-linking on the glass
transition temperature of polyether matrix, a series of DSC experiments were carried out.
Our DSC measurements of PEGDME/LiClO, mixtures without added silica are shown in
Figure 33 and thermal transition data are listed in Table 9. These data are second scans,

taken after flash quenching samples from 100 °C. At high O/Li ratios (low LiClO,

Table 8. Thermal properties of PEGDME-LiClO, system

 

O/Li 8 12 15 18 22 26 29 35 39 60 139

 

Tg (‘C) -50 -58 -67 -69 -73 -74 -55 -58 -59 -61 -66

 

TJ‘C‘) - -14 -4o -53 —61 -64 - '- - - -

 

 

 

 

 

 

 

 

 

 

 

 

 

T”, (°C) - 6 6 9 9 ll l3 14 13 13 14

 

 

103

content), DSC traces show a single broad melting peak that corresponds to the melting of
a pure PEGDME phase. Addition of LiCLO, causes both the melting point and the
enthalpy for the transition to decrease, and at an O/Li ratio of 8, the transition is no longer
detectable. A low temperature exothermic peak visible at intermediate salt concentrations
corresponds to crystallization of a portion of the sample, presumably a pure PEGDME
phase. A glass transition is also seen for these samples that, as expected, shifts to higher
temperatures with increasing salt content. No crystallization was observed by DSC for
samples with low salt contents, since for these materials, the PEGDME relaxation times

are short and the PEGDME chains crystallize while being quenched from 100 °C.

DSC scans of composites with silica (Figure 34) are remarkable in that the thermal
transitions (see also Table 9 and 10) for composites with and without silica appear at the
same temperatures. These results show that the mobility of the PEGDME, as measured by
Tg, is unaffected by the added silica. This is strong experimental verification of the notion
that the properties of the PEGDME phase are strongly decoupled fi'om the added silica,
and that the improved mechanical properties provided by the firmed silica matrix are

gained without significant losses in conductivity.
DSC traces for cross-linked composites containing LiClO, are shown in Figure 35.

Qualitatively, they resemble those taken for composites before cross-linking. Some slight

shifts in position and intensity are seen for the melting and crystallization transitions, but

104

Heat Flow -> endo

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

O/Li

139 M 4_
60 w?

_35 4/\ H
29
26 _ A
V'—
22
18
12 _ A
8 I
-100 -80 -60 -40 -20 0 20 40
Temperature, °C

Figure 34. DSC second heating scans for PEGDM-SOO/LiClO4 electrolytes.

Conditions: run under He at a heating rate of 10 °C/min.

105

60

 

O/Li

39
29 /\_ J

18

 

 

Heat Flow —> endo

 

 

 

 

-100 -50 0 50

Temperature, °C

Figure 35. DSC second heating scans for PEGDM-SOO/fumed silica/LiClO,

electrolytes. Conditions: run under He at a heating rate of 10 °C/min.

106

 

 

 

 

 

 

 

 

 

 

 

 

 

0/ Li
60 /\

1":
l 29
B
3
in
5
I 18

8

-100 ~50 0 50
Temperatu re, °C

Figure 36. DSC second heating scans for cross-linked PEGDM-SOO/fiimed silica/LiClO,

electrolytes. Conditions: run under He at a heating rate of 10 °C/min.

107

Table 9. Thermal transition data of composites before and after cross-linking.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

O/Li 8 18 29 39 60
Before T g (‘C) ~51 ~75 ~61 ~65 ~66
T c (‘C) ~ ~60 - - -
T ... (‘1’?) ~ 6.0 9.0 1 1 11
After T g (“(3) ~48 ~69 ~75 ~62 ~66
T. (‘C) - -46 ~64 - -
T ,,, (CC) ~ 5.6 5.4 11 1 1

 

the values of the glass transitions are unchanged. Some hints of the chemistry that takes
place during the cross-linking reaction can be inferred by monitoring changes in the IR
spectrum of composites, and the TGA data for cross-linked electrolytes. An electrolyte
was prepared from PEGDM-SOO, cross-linkable firmed silica LiClO, and butyl
methacrylate. As shown in Figure 36, the C=O and C=C vibrational bands characteristic of
methacrylates are found in the spectrum of this uncross-linked silica (middle scan), and
after irradiation, these bands shift to positions expected for a polymerized methacrylate.
The bottom scan in Figure 36 corresponds to the residue of a cross-linked sample that was
extracted exhaustively with acetone, a good solvent for poly(butyl methacrylate), and then
dried (Scheme 24). Prominent C=O and C-H stretching bands of the polymethacrylate
dominate the spectrum TGA scans of the same silica/polymer residue (Figure 37, bottom
scan) show a 52% weight loss. The predicted weight loss, assuming all of the butyl

methacrylate co-polymerized with the methacrylates bound to the fumed silica, was 56%.

108

Scheme 25. Solvent extraction experiments to analyzed surface-grafted polymer.

 

 

 

 

 

 

Cross-Linked A200
Composites
' solvent
E extraction
1"" \‘x
Polymer-Grafted PEGDME-MC 104
Silica Non-grafted Polymer

 

 

 

 

 

 

Thus, we conclude that nearly all of the added monomer was chemically bound to the

functionalized silica during cross-linking.

TEM imaging techniques were used to investigate changes in the structure and
morphology of the silicas and their composites. Figures 39, 40, and 41 are TEM images of
A200, cross-linkable A200, and the extraction residue of cross-linked composite. For
comparison purposes, images taken at two different magnifications are shown in each
case. As can be seen, the network structure of these three samples become more even and

denser in going fi'om most polar (A200) to least polar (residue). The same phenomenon

was also observed by others.‘42

109

 

4k A200 ”J
cross-linkable
A200

X-linked A200
composite
(after extraction) LJ

Relative Intensity

   
  

 

 

 

 

 

 

 

 

3800 3300 2800 2300 1800 1300

-r
Wavenumber, cm

Figure 37. IR spectra of surface modified silicas. Top scan: Aerosil A-200 fumed silica:

middle scan: A-200 modified with surface-bound methacrylates; bottom scan:

cross-linked composite electrolyte after exhaustive extraction with acetone.

110

 

 

 

 

 

 

 

 

1 10
A200
100 -
90 _ cross-linkable silica
.\e 80 ‘
I?
.9.”
0
a 70 .
6O .
cross-linked composite (after
extraction)
50 -~
40 1 . 4
100 300 500 700 900

Temperature, 0C

Figure 38. TGA scans of surface-modified fumed silicas. Top scan: Aerosil A-200 filmed
silica; middle scan: A-ZOO modified with surface-bound methacrylates; bottom scan:

cross-linked composite electrolyte afier exhaustive extraction with acetone.

Conditions: run in air at a heating rate of 10 °C/min.

lll

 

hs of A200

nucrograp

TEM

Figure 39

112

 

Figure 40. TEM micrographs of cross-linkable silica.

113

 

Figure 41.TEM micrograph of extraction residue of cross-linked composite.

114

 

Figure 42. F reeze-Fracture TEM images of composites before cross-linking.

115

.100 mu 7'

 

Figure 43. Freeze-Fracture TEM images of composites after cross-linking.

116

In order to study the morphology of composites in situ, replicas of electrolytes
were made using Freeze-Fracture techniques. Typically, The sample was placed on a
copper cap and fi'acturedto generate a fresh surface which was replicated. The silica

network structure is clearly seen before cross-linking (Figure 42), but more featureless

after cross-linking (Figure 43). Two possible explanations might be possible. First, before
cross-linking, adhesion between the silica and PEGDME-500 is poor, and fracture
preferentially occurs at the silica-PEGDME-SOO interface. Polymerizing the added
methacrylate binds the silica particles together, strengthens the silica-PEGDM-SOO
interface, and leads to fracture within the PEGDME-500 phase. Second, before cross-
linking, silica network is only temporary and is easily broken down. Upon cross-linking, all
the particles are bound together and the network resists to the fi'acture. Thus, uncross-
linked composites looks particulate, while the cross-linked samples appear smooth. The

featureless morphology of polymer latex films has recently been correlated with the

physical properties of the two polymer phases.85 Additional experiments are needed to

corroborate this model.

Conductivity measurements show that the cross-linking reaction has little effect on
the ionic conductivity of the composites. As shown in Table 10 and Figure 44, the room
temperature data for a sample before and after cross-linking are nearly identical, 2-4 x10"
S/cm. These results are consistent with the thermal analysis data that show little change in
the T3 of composites after cross-linking. The conductivity depends on the lithium salt

content and reaches a maximum at GM ratios of 20-40. The conductivity data were also

117

analyzed using an Arhennius plot (Figure 45) and fiom these plots, an interesting trend
emerges. At high O/Li ratios, the slopes of the plots (activation energies) are similar, but
as the O/Li ratio further decreases, the apparent activation energies steadily increase. In
addition, the data taken at a given salt content are not linear, and are suggestive of a phase
transition near 55 °C. The nonlinearity is most pronounced for materials with high salt
contents, which would be consistent with a crystalline-amorphous phase transition.

However, to date we have been unable to verify such a transition by DSC or optical

microscopy.

Table 10. Conductivity data (x10‘) for Figure 43.

 

 

 

 

 

 

 

O/Li 368 K 348 K 328 K 318 K 298 K 298 K
8 12.6 6.62 3.62 2.16 0.62 1.14
18 13.2 11.4 7.29 5.08 2.25 2.93
29 13.5 11.9 8.13 6.32 4.15 2.63
39 8.13 6.94 5.08 4.15 2.13 2.17
60 5.04 4.12 3.16 2.53 1.44 0.9

 

 

 

 

 

 

 

118

 

 

 

 

 

 

 

 

 

16
14 g _ 368 K after
. 348 K after
12 w . A328 Kafter
’ 0318 K after
10 J _ D 298 K after
E I 298 K before
3 cross-linkinL
m. s -_ 4
V
S a
a: . *
b 6 .
0 A
4 -_ D O a
A
- A
l o
2 4_ O D I
D
g u
0 r .
0 20 40 60 80
O/Li Ratio

Figure 44. Conductivity of composite electrolytes at selected

temperatures before and after cross-linking.

119

 

 

 

 

3 n
I
0 I
. A I
A I
’8 ° . -
00 a . A I
I
'3 a n A I
D O I
'35 ’ Cl A
. O/Li=8 f z
. a
o 18 D
O
I 29
A 39
O
’ n 60
~45 1* i 1’ i
2.6 2.8 3 3.2 3.4
rooo/r,l(‘

Figure 45. Arrhenius plots of the conductivity of cross-linked composite electrolytes.

120

DISCUSSION

1. Current System

1. Composite Reinforcing mechanism: phase separation

We have shown that highly processable fumed silica-based electrolytes, consisting
of PEGDME, LiC104, butyl methacrylate, and firmed silica with hydrophobic and cross-
linkable groups, can be converted via photopolymerization into mechanically robust
electrolytes with no apparent loss in conductivity. These results are in stark contrast to the
data typically obtained with cross-linked polymer electrolytes. Usually cross-linking or the

formation of interpenetrating network stmctures are accompanied by a large drop in

conductivity, often by an order of magnitude.26’86’87 The principal difi‘erence between
the firmed silica system and traditional polymer electrolytes is that the mechanical
properties of the composite and the ion conducting characteristics are decoupled, and thus

these two important characteristics can be optimized independently.

A simple model to understand these results can be based on phase separation
processes. For example, the addition of firmed silica to polyethers-LiClO, mixtures yields
a two-phase system, the silica network, structure that provides mechanical reinforcement
of the composite and the polyether—LiClO, phase that results in high conductivity. Single
phase (or at least highly dispersed) polyether/silica mixtures can be prepared by
appropriate surface modifications. For example, as discussed in the Results section, using

standard chemical techniques, we attached short polyether chains to A200. As long as its

121

content is low enough (<10% by weight), the modified silica completely dispersed in
PEGDME-500 to give a clear viscous liquid. Similariy, there are many reports of the
preparation of silicas by hydrolysis of tetraethoxysilane followed by surface modification

to match the chemical characteristics of a particular matrix. These also form homogeneous

solutions“,65 Mechanical properties appreciably better than the matrix itself requires a
silica network structure, which in turn requires a fumed silica that does not disperse, but

instead aggregates. This structure must then be preserved in the cross-linked electrolyte.

Cross-linked composites based on polymerizable fumed silica alone were
mechanically weak and fractured easily, a sign of minimal chemical connectivity between
adjacent silica particles. For strong composites, added monomer is needed to knit the silica
particles into a mechanically stable framework. Given thatthe starting point for our model
is a two-phase system of silica dispersed in a polyether, the most important question is the
fate of the monomer added to the composite before photopolymerization. We knOw from
test polymerizations of butyl methacrylates in PEGDM-SOO that monomer/polyether
solutions are initially homogeneous, but phase separate when lithium salt is present. When
fumed silica is added to the above mixture, phase separation is not visible on a
macroscopic level. Upon cross-linking, the slurry becomes cloudy, an obvious indication
of phase separation, suggesting that phase separation might have occurred even before
cross-linking on a microscopic scale. Moreover, polymerizations using octyl methacrylate
yield the most turbid products while the more polar methyl methacrylate and poly(ethylene

glycol) ethyl methacrylate appears to be the most clear. These results are in line with the

122

relative differences in solubility parameters for the methacrylate polymers relative to
PEGDME-500. Since none of these polymers are completely soluble in PEGDME-500,
these results are not surprising, but they do point to a phase separation model. At least
three limiting cases can be considered. (1) The monomer phase separates from the
PEGDME-500 phase to form monomer droplets which develop into distinct polymer
phases upon cross-linking (Figure 46, A). if this model were true, the glass transition
temperature of polyethers would remain the same, but the guest polymer phase would be
extractable. (2) The monomer mixes with polyethers and develops into a homogeneous
phase on macroscopic level after polymerization (Figure 46, B). In this case, not only
would the guest polymer be extractable, but the glass transition of polyethers would be
affected. (3) The monomer separates from polyethers to form a thin layer surrounding the
silica network where copolymerization between monomer and surface functional groups
takes place (Figure 30). The first two scenarios are ruled out by extraction and DSC
experiments carried out on cross-linked composites. Fully cross-linked samples were
extracted to remove the free polymer and the PEO-LiClO4 and the insoluble residue was
analyzed. prure homopolymer formed a distinct polymer phase, then it should be easily
removed, and the weight loss in the TGA experiment should be comparable to that seen
for the surface-modified silica alone (i.e., 6%). The results, however, point to nearly all (>
90 %) of the monomer being chemically connected to the silica. The scenario (2) is firrther
ruled out by the DSC data which show that the glass transition temperature for polyethers
in the composite was unchanged before and after cross-linking. Thus, some version of

scenario (3) must be at play. This model (Figure 30) can be rationalized by the

123

 

Figure 46. Scenario 1 (top) and 2 (bottom) as defined in the text.

characteristics of the components of the composite and supported by additional evidence

obtained using FT IR, impedance spectroscopy, and TEM.

In order to test the model, further experiments were carried out using three
different silicas FS-C3-M, F S-C8-M, and R805. The first two are polymerizable silicas
made in this laboratory and R805 was a hydrophobic conunercial product with about 50 %
of silanol groups replaced by octyl groups. Four different monomers were employed:
methyl methacrylate (MMA), butyl methacrylate (BMA), octyl methacrylate (OMA), and

poly(ethylene glycol) ethyl methacrylate (PEGEMA).

During UV irradiation, the characteristics of the composites evolved fi'om being
gel-like to those more typical of a cured silicon rubber. More detailed mechanical and
rheological measurements are in progress and will be reported elsewhere. The cured
composites were translucent and scattered light, with those cross-linked using MMA and
PEGEMA the most clear and those prepared using octyl methacrylate the most cloudy.
These results are in accord with increasing degrees of phase separation during crosslinking

as the monomer used becomes less polar.

Conductivities of this series of composite polymer electrolytes were measured and
the results are listed in Table 12. Solubility parameters for each polymethacrylate was

calculated based on its stnrcture and group molar attraction constants according to

equation 25:88

125

6=pZGlM (25)

where p represents the density and M is the mer molecular weight of the polymer. Group
molar attraction constants have been calculated and are readily available in many
textbooks. The solubility parameter for PEO was calculated to be 9.9. and the solubility
parameter difference between PEO and the guest polymer should determine the miscibility
of the two polymers. A large difference would mean a large driving force for phase
separation and since only the PEO phase is responsible for ionic conductivity, a strong

phase segregation would correspond to a high conductivity.

The conductivity data listed in Table 11 do not differ appreciably, but Figure 47
shows that the the conductivity increases with the difl‘erence in solubility parameters of the
two polymers. In addition to using solubility parameters, we also tried to correlate the

conductivity with the carbon to oxygen ratio in an individual polymer since it is a rough

Table 11. Conductivity data for Figures 47 and 48.

 

 

 

 

 

PMA C/O 8,...A 59,508.... 8*10‘S/cm
Type Ratio Data Data FS-C3-M F S-C8-M R805
PEGEMA 2.4 9.8 0.1 1.94 2.00 2.15
PMMA 2.5 9.1 0.8 2.15 2.07 2.13
PBMA 4 8.8 1.1 2.32 2.23 2.28
POMA 6 8.4 1.5 2.9 2.76 2.64

 

 

 

 

 

 

 

 

 

126

 

 

 

 

 

 

 

 

2.9 -- 9
o F S~C3~M
I FS-C8-M-2 I
2.7 AP A R805
A
2.5 _.
V
O
T
b 2.3 1r :
I
‘ 2
2.1 --
I
I
o
1.9 ~~
1.7 i i f
0 0.5 1 1 5
8I’IIO"81’NIA

Figure 47 . Conductivity vs. solubility parameters difference between PEO and PMA.

127

but independent estimate for the polarity of that polymer. The plot of C/O ratio vs
conductivity is shown in Figure 48. Again, composites prepared using polymers that have
a high C/O ratio (less polarity) show high conductivity, in good agreement with solubility

parameter results.

As before, a combination of solvent extraction and TGA analysis was used to
determine the amount of grafted polymer on the silica surface. Attempts to extract R805-
containing composites afier cross-linking were not successful since the fine silica particles
dispersed in the liquid phase and did not settle out ~ a good indication of no chemical
bonding between silica surfaces and the polymer. However, when cross-linkable silicas
(F S-C3-M and FS-C8-M-2) were used. large solid particles settled out as soon as the
stirring was terminated. TGA measurements were used as before to analyze grafting
efficiency. Two extremes could be considered. If none of the monomer copolymerized
with the silica surface, the predicted TGA weight loss after extraction would be the same
as the starting silica, i.e., ca. 6 - 9% (Figure 49); if, however, all of the monomer
molecules cross-linked with silica, the predicted weight loss would be 56 - 59 %. In the
case of composites made with F S-C8-M and when BMA is used, the actual weight loss is
ca. 50% ~ nearly 90% of the predicted value. When a polar monomer, poly(ethylene glycol
ethyl methacrylate (PEGEMA), was used, the corresponding weight loss dropped to 40%
- 70% of the predicted value. Thus, ca. 30% of the PEGEMA monomer formed
homopolymer in the PEGDME phase which could account for the drop in conductivity.
Interestingly, the composite made from F S-C3-M showed the same characteristics as

composites fi'om FS-C8—M. This is not surprising considering the fact that they both share

128

(mo4

 

 

 

 

 

 

 

2.9 a o
o FS-C3-M
I FS-C8-M—2
A R805 '
2.7 ~-
A
2.5 r
o
2.3 - ‘
I
A.
A
2.1 r
I
I
O
1.9 r~
1.7 a 1 i i i
l 2 3 4 5 6
C/O Ratio in PMA

Figure 48. Conductivity vs. carbon to oxygen ratio in PMA.

129

 

Weight, %

110

100

90*

80-

70

60‘

50

40

 

R805

ll

 

 

\ FS'C3'M \ FS-C8_M_2

' ~ FS-C8-PEGEMA

 

/ FS-C8-BMA

 

 

-_ /

FS~C3~BMA

 

L 1 ' 1 1 4
7

 

 

1
I T Y I l

100 200 300 400 500 600 700 800

Temperature, 0C

Figure 49. TGA weight losses for samples as labeled.

130

900

the same polymerizable group as the guest monomer, and although the C3 chain is more
hydrophobic than the C3 chain, as discussed in the previous section, the difference was
overcome quickly upon copolymerization with their nearest layer of monomer molecules.

When the cross-linking is complete, the two composites are all rubbery materials.

131

2. Cross-Linking and Emulsion Polymerization: A Comparison

Our cross-linking system is related to an emulsion polymerization regime. To
understand this point better, a little background material regarding emulsion
polymerization formulations will be helpfiil. A typical formulation for a traditional
emulsion polymerization contains monomer, a dispersing medium - usually water, an
emulsifier or surfactant, and a water-soluble initiator. Emulsifier are typically amphiphilic
molecules and their physical state depends on concentration. The critical micelle
concentration (CMC) is the concentration beyond which emulsifier molecules aggregate to
form micelles. Monomer is dispersed in the water phase to form monomer droplets whose
sizes are much larger and whose total surface areas are much smaller than the micelles.
Since the initiator is water-soluble, the initiating radicals are generated in water phase and
diffuse into micelles where the monomer and the initiating radicals meet and

polymerization takes place.

In recent years, novel composite polymeric materials have been prepared using

emulsion polymerization techniques in the presence of various inorganic particles such as
calcium carbonate, barium sulfate,89'93 silica, titanium dioxide, quartz,94:95 limestone,96

calcium sulfite,97 silica gel, activated alumina, white carbon, zeolite, et al.98 In systems
where the filler material does not carry polymerizable firnctional groups, the filler particles
are encapsulated in situ by the growing polymer chains during emulsion polymerization.
The amount of encapsulated polymer and the extractable portion of polymer depend on

the surface property and the composition of the filler used, the kind and amount of the

132

surfactant, and the monomer type. A series of experiment589‘97,99 showed that surface
modification of inorganic particles with anionic surfactants or titanates led to
polymerization localized at the surface of the particle. Presumably, polymerization takes
place in the hydrophobic chemical environment defined by the adsorbed surfactant. Some
homopolymer also forms in the aqueous phase (the expected emulsion polymerization
product), but by keeping the surfactant level below the critical micelle concentration, the
amount of monomer dispersed in the aqueous phase and the amount of free polymer was
minimized. In most cases, however, the adhesion between filler and polymer is poor due to
the lack of chemical bonding. After exhaustive extraction with solvent, only a small

portion the encapsulated polymer can be retained on the particle surface.

Attempts have been made to increase the interaction between polymer matrices

particle surfaces to improve the physical properties of structural composite. For instance,

Hasegawa and coworkers 96 reported a soapless emulsion polymerization system
incorporating newly ground limestone particles. The strength of adhesion between the
encapsulated polymer and solid surface was improved - a consequence attributed to the
existence of active centers such as defects and disturbances in the crystal lattice. In
addition, a report closely related to the chemistry of our work describes the emulsion

polymerization of ethyl acrylate in the presence of fumed silica firnctionalized with acrylate
groups.67 The results reported are in accord with the added monomer separating from the

continuous (aqueous) phase, and copolymerizing with surface-bound acrylates.100

Composites prepared from these latexes show improved mechanical properties.

133

Their method of making composite polymeric material and properties of the
resulted products are closely related to ours. The major difi‘erence is that they were aiming
at grafting polymer on separate silica particle, while we wanted to bind silica particles
together to create a permanent three-dimensional network. Table 13 lists the two
formulations. The polymerizable silica they used was made by modification with either
TPM or hydroxyethyl methacrylate, in other words, the hydrocarbon spacer between silica
surface and methacryloyl group is either a C2 or C3. As we discussed in our previous
section, a surface like this is characteristic of ester firnctional group which lacks the
hydrophobicity needed to attract the monomer molecules. A surfactant had to be used to
increase the driving force of the monomer moving from water phase to silica surface. In
our system, low molecular weight PEO containing dissolved lithium salt serves as
dispersant. In addition, a surfactant was not necessary since the monomer we chose,
BMA, is much more hydrophobic than the one they used (hydroxyethyl methacrylate) and
will be driven to the silica surface even more as the polymerization proceeds. As far as the
polymerization site is concerned, we do not expect it to occur in the PEO phase due to the
phase separation between BMA and PEO, and also due to the poor solubility of our

initiator (AIBN) in PEO.

Additional evidence for strong segregation of the polymer and PEGDM-500 1
phases in our system comes from comparisons of the mechanical and electrical properties

of cross-linked electrolytes with those prepared without added monomer. As noted in the

134

Table 12. Formulations of our and Bourgeat-Lami’s systems.

 

 

 

 

 

 

 

 

 

Bourgeat-Lami’s ours
silica cross-linkable‘ cross-linkable"
dispersant water LiClO, in PEO
emulsifier SDSc none
monomer ethyl acrylate butyl methacrylate
initiator potassium peroxodisulfate AIBN

 

 

" Modified with TPM or ethyl methacrylate.

" Modified with methacrylate having Cs hydrocarbon spacer or by mixed

silylation.

° Sodium dodecyl sulfate.

Results section, data sensitive to the segmental mobility of the PEGDM-SOO phase such as

conductivity and T, values, show little or no change with cross-linking.

Bulk mechanical measurements coupled with impedance spectrocopy best illustrate
the independence of mechanical properties and conductivity. As shown in Figure 50, the
composite is a physical gel prior to cross-linking as evidenced by the elastic modulus (G’)
being larger than the viscous modulus (G”). Following cross-linking, the elastic modulus
increases by almost two orders of magnitude to a value comparable to that of nibbery

solids. At the same time, the conductivity showed only slight declines. Note that the

135

conductivity for this sample (Figure 51) is high because of switching from LiClO4 to

lithiul imide as the salt.

To summarize, we have achieved our goal to synthesize composite polymer
electrolytes with good mechanical stability and high conductivity and the ability to tune

these two factors independently.

 

 

 

A : I I U I U T l ' I I O Q ......
w h ' ,,,,,,,,,,,, g ............
9: “ .§_.. ...... g ...... . ...... Q Q
‘3 b after crosslinking ' 9
_ ,,,,,,,,,,,,,,, 9 ......

E 9 ..... § ...... 9 ..... .9. Q .9-
? 105 E"
S9. :
‘3 3 G' ..
8 . ...... g ...... . ...... g ...... . ...... g ...... g ...... . ...... . ...... ...n
.9 _
:- before crosslinking
i9, 10‘ :—
.9 ...... 9 ...... 9 ...... ¢ ...... 0 ...... Q ...... ® ......
‘65
g I-
w .-

L l 4 L 1 1 1 1 l J

0.1 1

 

Frequency, 0) (rad/s)

Figure 50. Dynamic moduli (G’ and G”) before and after cross-linking for
composites containing 10% cross-linkable silica and 10% butyl methacrylate in a

PEGDME-LiClO, matrix.

136

T/ 00
1 1 1 84 60 39 21

 

1 0'2

  

 

10% TOM + 10% BMA + 80% Sol.
(before crosslinking)

10% TOM +10% BMA + 80% Sol.
(after cross-linking)

20% TOM + 10% BMA + 70% Sol.
(before cross-linking)

20% TOM + 10% BMA + 70% Sol.
(after cross-linking)

10'3
2.4 2.6 2.8 3.0 3.2 3.4 3.6

1000rr (K4)

1 o/(Scm’l).

I

 

 

 

 

 

 

Figure 51. Conductivity of cross-linkable composite polymer electrolytes before

and after cross-linking. Solution (Sol) in legend is Li imide + PEGDME. Proportions are

given in weight percentage.

137

11. Future Work and Suggestions

1. Optimization of Current System

We have successfully achieved our goal, i.e., a composite polymer electrolyte
system with high conductivity and good mechanical stability and processability. Further
efforts could be devoted to optimizing the system. For example, the mechanical stability
could be enhanced by increasing the amount of cross-linkable fumed silica up to 20% by
weight. Further increases would result in decrease in conductivity due to the formation of
an insulating silica region. As far as the monomer content is concerned, it is likely that

smaller amounts (5-10% by weight) would be enough to bring the same mechanical

stability to the system.

2. Core-active Initiators

In the Introduction, we introduced the use of coreactive initiators to make

covalently bound polymer (Scheme 7).70 This method is well-suited for our system and
provides us with an alternative way of making our target products. Since the initiating
radicals are generated in situ, it is possible to limit the polymerization to silica surface. In
order for the silica particle to form a network structure and for the non-polar monomer to
localize around this network, it is necessary to attach organic groups carrying both

hydrophobic and light- or heat-sensitive groups to the silica surface. This can be done by

modifying Koehler and Ohngemach’s initiator,7O i.e., by replacing their spacer with a

suitable hydrocarbon spacer. The composite polymer electrolytes thus formed would

138

expected to show similar behavior to our reported products in terms of conductivity and

mechanical stability, yet provide us with an alternative to achieve our goals.

3. Two-Component Core-Shell Composites

We discussed in the Results section that when fumed silica is modified with PEO
oligomers, it becomes more compatible with the PEGDME matrix. The extreme case
would be when the surface coverage of PEO chain on silica surface is high enough, the
PEG-grafted fumed silica itself will be a good ionic conductor upon adding lithium salts.
By controlling the molecular weight and chain length, the PEO could be kept amorphous
and therefore the conductivity of the composite should be high. The main advantage of
this composite would be that the core-shell structure provides both good conductivity and
mechanical stability. The preparation of this composite should be fairly easy. For example,
Figure 52 shows a schematic route to such a product. Cross-linkable firmed silica can be
made by silylation reactions using either a single silane incorporating both a cross-linkable
group and a hydrophilic spacer, or a mixture of a coupling agent (e. g., TPM) and a second
silane containing hydrophilic group. By changing the length of the spacer or the ratio of
the two silane molecules, the desired silica surface can be realized. By dispersing this silica
in a mixture containing lithium salt, a polar initiator, and a monomer carrying ethylene
oxide units (e. g., PEGEMA), followed by polymerization, a composite polymer electrolyte

will be generated. This system would resemble a hyperbranched PEO system but without

problems such as poor conductivity and creep. .

139

 

fumed
silica

 
   
 

Li+

PEGEEM = poly(ethlene glycol) ethyl ether methacrylate
CLG = cross-linkable group
TG = tunable group, used to adjust polarity

Figure 52.A proposed route to a core-shell composite polymer electrolyte.

140

4. Highly Dispersible & Cross-Linkable Silicas for Emulsion Polymerization

We showed above that emulsion polymerization has been used to prepare polymer
encapsulated inorganic particles. The ideal case is that, instead of a network structure, one

would like to disperse the inorganic particles completely in the dispersant, usually water.
This has proved to be difficult to achieve so far, since particles such as silicas“:68 and

quartz powders95 show a strong tendency to form large aggregates in aqueous media.
Through our research, we have gained knowledge about interactions between particles
and particle-solvent molecules. For instance, if a polar solvent is chosen, a network will
form with the addition of a hydrophobic filler material, while a good dispersion can be
achieved when a hydrophilic filler, e. g., 8 PEG or polyoxymethylene modified silica, is
used. Two examples of hydrophilic and cross-linakble firmed silicas are shown in
Figure53. We believe that by using such filler materials, an improved dispersion of
particles in aqueous media could be achieved, the therefore “ideal” silica-polymer
encapsulated composites could be realized. This methodology applies not only to silica
system, but also to any other inorganic system where MOH ( M = Al, Ti, B, etc.) groups

are present on its surface.

141

“work

acme

Figure 53.Schematic of hydrophilic fumed silicas: (a) silica modified with a single silane;

(b) silica modified with a mixture of two silanes

142

111. Summary

We have prepared various surface-modified fumed silicas and explored their
behavior in polyether solutions. Results show that PEO modified hydrophilic silicas tend
to disperse well in PEO derivatives due to the compatibility between surface groups and
the dispersant. When a three-dimensional network is desired, the use of such fillers is not
favorable since a large amount of filler material is needed. However, firmed silicas bearing
both hydrophilic and cross-linkable groups would likely be suitable as seeds for emulsion
polymerizations to make polymer encapsulated inorganic materials owing to their good
dispersibility. Hydrophobic silicas, on the other hand, tend to form aggregates and
network structures at a small loadings and are usefirl materials for controlling the

mechanical properties of the inorganic-organic composite materials.

Modified firmed silicas featuring both hydrophobic and cross-linkable
characteristics are the center of this research. These materials were prepared using either a
single silylation or a mixed silylation process. When a polyether-Li salt solution was gelled
using the above silica, the hydrophobic portion of the silica, through van der Waals
attractions, formed a temporary open structured three-dimensional network, giving the
composite good mechanical stability and processability. The polyether-Li salt phase was
responsible for the good ionic conductivity (10" S/cm) via the large open spaces provided
by the network. Therefore, the mechanical properties and conductivity of the composite
came from two different phases, the silica and polyether-Li salt phases, respectively. These

two properties were decoupled and could be tuned independently.

143

The possibility of tuning the mechanical properties was provided by the cross-
linkable groups on the silica surface. We were able to lock the temporary three-
dirnensional network in place by adding a non—polar monomer, BMA, and an oil-soluble
initiator, AIBN, to silica/polyether-Li salt composites. The monomer and the initiator, due
to their non-polar characteristics, phase separated from the polyether-Li salt phase and
formed a thin monomer layer near the hydrophobic silica surface. Upon photo-initiation,
the added BMA monomer and the cross-linkable group on the silica surface
copolymerized to lock the network structure in place, converting a temporary network
into a permanent network, and a physical gel into a solid rubbery material. Our extraction
experiments combined with TGA analyses show that nearly all of the polymer had been
grafted onto firmed silica surface. The conductivity was not afl‘ected by the above cross-
linking process, suggesting that the electrochemical and mechanical properties had been
decoupled. In addition to the mechanical properties, we were also able to tune the
conductivity independently. By replacing lithium perchlorate with lithium imide, we were
able to realize a conductivity as high as 10'3 S/cm, a value which is competitive with gel

electrolytes.

144

EXPERIMENTAL

I. Materials

Unless otherwise specified, ACS reagent grade starting materials and solvents
were used as received fi'om commercial suppliers without firrther purification. Aerosil 200
(A200) was a gift fiom Degussa A.G., Frankfurt, Germany. This silica has a surface silanol
content of 1 mmng, and was pre-treated in two difi‘erent ways. The term “evacuated
A200” is used for samples dried under vacuum at room temperature for 5 min; and the
term “hydrated A200” is used for samples stored over a half-saturated NH4NO3 solution in

a desiccator for at least one week.

145

11. Analytical Methods

1H and '3 C nuclear resonance analyses were carried out at room temperature in
deuterated chloroform (CDCl3) on a Varian Gemini-300 spectrometer with the solvent

proton and carbon signals being used as chemical shift standards.

13C and 29Si solid state NMR spectra were obtained at 79.5 and 100.6 MHz,
respectively, on a Varian 400 spectrometer under the conditions of 1H and 29Si cross-
polarization, magic angle spinning. Samples (ca 100 mg) were packed in 7.0 mm zirconia
rotors. Chemical shifts are reported in ppm and are referenced to talcum for 29Si and

adamantane for 13C. A spinning frequency of 4 kHz was used.

A Nicolet IR/42 spectrometer purged with dry nitrogen was used to obtain
infrared spectra. Samples used were 1 cm2 pressed pellets prepared fi'om ca. 10 mg of the
various pure silicas. All spectra reported were acquired by signal averaging 32 scans at a

resolution of 4 cm".

Differential scanning calorimetry (DSC) were performed under a helium
atmosphere at a heating rate of 10 °C/rnin on a Perkin Elmer DSC 7. Thermogravimetric
analyses (TGA) were performed in air atmosphere at a heating rate of 10 °C/min on a

Perkin Elmer TGA 7 instrument. For DSC measurements, the temperature was calibrated

146

with an indium standard and samples were initially heated to 100 °C to erase the sample’s
thermal history, and then were quenched to ~100 °C prior to starting the run. Samples for
TGA measurements were first dried in vacuum at 120 °C overnight. Dried samples (~5-10
mg) were held at 115 °C in the TGA until a stable weight reading was obtained, and then

the run was started.

Images of pristine and modified silicas were obtained as follows. One drop of a
silica dispersion in acetone was placed on a carbon-coated grid. The solvent was allowed
to evaporate for a few minutes and the grid was observed in the microscope. Images of
composite polymer electrolytes were based on replicas prepared using freeze-fiacture
techniquess. The sample was placed on a copper cap and fractured to generate a fresh
surface which was then replicated The sample was fractured at ~90 °C at 10'6 torr and
replicated using unidirectional shadowing with platinum and carbon at an angle of 45°.
The replicas and the grids were observed in a JEOL 100 CX H transmission electron

microscope.

Silanol group analyses were performed by LiAlI-h titration using literature

procedures as guides“,101 All samples were dried at 150 °C under vacuum to 10'3 mm
Hg before being tested. About 1 g of sample and an excess amount (20 mL, 0.4 wt%) of
lithium aluminum hydride in diglyme were used for each run. The volume of hydrogen gas
released from the reaction was measured and the silanol groups were calculated according

to the ideal gas equation.

147

The electrical properties of the samples were measured by ac impedance
spectroscopy over the fiequency range 6 Hz - 11 MHz using an HP 4192A LF Impedance
Analyzer. An applied voltage of 0.1 V was used to collect the data. Cross-linked samples
were prepared by sandwiching uncross-linked electrolyte between a glass slide and a
stainless steel disk, with the desired thickness maintained using a Teflon spacer. The
composites were polymerized as described previously, and after polymerization, the glass
slide was peeled off gently and replaced with another stainless steel disc. The electrolyte
thickness was obtained by subtracting the thickness of the steel discs from the total
thickness measured using a micrometer. Conductivity measurements were taken fiom
high to low temperatures, with the samples held at each temperature for at least 15

minutes under a nitrogen gas purge before being measured.

148

III. Hydrophobic silicas

FS-TEMS-l. The experimental procedure employed in this example is typical of
those used to make FS-TEMS-l, FS—TEMS-Z, FS-TEMS-3, FS-TEMS-O, FS-TEMS-
2H, and FS-TEMS-13H. To 2 g of evacuated A200 were added 40 mL of toluene
containing 2 11L (0.02 mmol) of diethylarnine (DEA) and a mixture of 10 mL of toluene
and 1.4 g (8.0 mmol) of methytriethoxysilane. The slurry was shaken until homogeneous
after each addition. The reaction was allowed to proceed in a stoppered Schlenk flask for
2 h. The slurry was filtered, washed with 100 mL of dry toluene and evacuated at room

temperature and then at 150 °C until a 10‘3 mm Hg vacuum was reached.

Table 13. Reaction conditions for FS-TEMS series.

 

 

 

 

 

 

 

Sample Name DEA (mmol) Time (h)
FS-TEMS-l 0.02 2
FS-TEMS-z 0.50 2
FS—TEMS-3 2.0 2
FS-TEMS-O 0 2
FS-TEMS-ZH 1.0 2
FS-TEMS-13H 1.0 13

 

 

 

 

 

149

FS-TMCS- 1. The experimental procedure employed in this example is typical of
those used to make FS-TMCS-Z, TMCE-3, and TMES-l.To 2 g of evacuated A200
were added 40 mL of toluene containing 0.2 mL (2.0 mmol) of diethylamine and a mixture
of 10 mL of toluene and 54 mg (0.50 mmol) of trimethychlorosilane (TMCS). The slurry
was shaken to homogeneous each time of addition. The reaction was allowed to proceed
in a stoppered Schlenk flask for 2 h. The slurry was filtered, washed with 100 mL of dry
toluene and evacuated at room temperature and then at 150 °C until 10'3 mm Hg vacuum

was reached.

Table 14. TMS modified silicas.

 

Sample Name Silane (mmol)

 

FS-TMCS-l TMCS (0.5)

 

FS—TMCS-Z TMCS (1.4)

 

FS-TMCS-3 TMCS (2.0)

 

FS-TMES-l TMES (2.0)

 

 

 

 

FS-C2. The experimental procedure employed in this example is typical of those
used to make FS-C8 and FS-C 18. Using a transfer needle, 300 mL of toluene containing
1.7 mL (16 rnrnol) of diethylamine was added to 16 g of evacuated A200 followed by a
mixture of 2.1 mL (16 mmol) of ethyltrichlorosilane (ETCS) in 10 mL of toluene. The
slurry was shaken until homogeneous after each addition. The reaction was allowed to

proceed in a stoppered Schlenk flask for 2 h. The slurry was separated by centrifugation

150

and washed sequentially with two portions of toluene, ethanol/water (50:50), ethanol, and

diethyl ether. The diethyl ether was evaporated and the solid was transferred into a

Schlenk flask and dried under vacuum at 150 °C for at least 2 h.

Table 15. Reaction conditions for F S-Cn series.

 

Sample Name A200 (g) DEA (mmol) Silane (mmol)

 

 

 

FS—C2 12 16 ETCS (16)
FS—C8 12 13 orcs (13)
FS-C18 16 15 ODTCS (15)

 

 

 

 

 

 

‘ OTC S = octyltrichlorosilane.

ODTCS = octadecyltrichlorosilane.

FS-DM-C8-1. The experimental procedure employed in this example is typical of
those used to make FS-DM-C8-2 and FS-DM-C8-3. Toluene (300 mL) containing 25 111.
(0.25 mmol) of diethylamine followed by a mixture of 0. 10 mL (0.41 mmol) of
chlorodimethyloctylsilane in 10 mL of toluene was added through a transfer needle to 12 g
of evacuated A200. The slurry was shaken to homogeneous each time of addition. The
reaction was allowed to proceed in a stoppered Schlenk flask for 2 h. The slurry was
separated by centrifugation and washed sequentially with two portions of toluene,
ethanol/water (50:50), ethanol, and diethyl ether. The diethyl ether was evaporated and

the solid was transferred into a Schlenk flask and vacuum dried at 150 °C for at least 2 h.

151

Table 16. Reaction conditions for F S-DM-C8-n series.

 

 

 

 

Sample Name A200 (g) DEA (mmol) CDMOS (mmol)
'FS-DM-Cs-l 12 0.25 0.41
FS-DM-C8-2 14 2.9 3.3
FS-DM-C8-3 12 12 12

 

 

 

 

 

 

FS-DM-H. To 60 g of the evacuated A200 were added 600 mL of toluene
containing 7.5 mL (71 mmol) of diethylamine. After the silica was well dispersed through
shaking, a mixture of 10 mL of toluene and 8.2 mL (72 mmol) of chlorodimethylsilane
was added. The slurry was shaken again to assure a good dispersion of added silane. The
reaction was allowed to proceed overnight in a stoppered Schlenk flask. The slurry was
separated by filtration and washed sequentially with two portions of toluene, ethanol/water
(50:50), ethanol, and diethyl ether. The diethyl ether was evaporated and the solid was

transferred into a Schlenk flask and vacuum dried at 150 °C for at least 2 h.

FS-H. Evacuated A200 (25g) was replaced in a 1 L RBF equipped with a
condenser, an addition fiJnnel with equalizing tube, an oil bath, and a magnetic stirrer.
Dioxane (600 mL) containing 35 mL of 2.3 M HCl was then added. The mixture was.
heated to 75 °C, and asolution of 5.8 mL triethoxysilane in 100 mL dioxane was added
dropwise over a period of 1 h. After gently refluxing for two h, the product was

centrifirged and washed consecutively with 20:80 water/THF, THF, and diethyl ether

152

(twice with each solvent). The final product was dried at 60 °C and then at 150 °C under

vacuum.

FS-OEt-H. To 2 g of the evacuated A200 were added 40 mL of toluene
containing 0.10 mL (1.0 mmol) of diethylamine and a mixture of 10 mL of toluene and 0.8
mL (4.1 mmol) of triethoxysilane. The slurry was shaken until homogeneous after each
addition. The reaction was allowed to proceed for 4 h in a stoppered Schlenk flask. The
slurry was separated by filtration and washed 3 times with toluene and diethyl ether. The

diethyl ether was evaporated and the solid was transferred into a Schlenk flask and

vacuum dried at 150 °C for at least 2 h.

FS-OH-H. FS-OEt-H (0.50 g) was added to a mixture containing 20 mL of
dioxane, 1.0 mL of 2 N HCl, and 1.0 mL distilled water. The slurry was stirred and
brought to reflux for 1 h. The solid product was collected through centrifirgation and
washed consecutively with water/THE (20:80), THF, and diethyl ether (twice with each
solvent). The diethyl ether was evaporated and the solid was transferred into a Schlenk

flask and vacuum dried at 150 0C for at least 2 h.

FS-TMS-H. FS-OH-H (0.20 g) was dispersed in 20 mL of toluene containing
0.05 mL (0.50 mmol) diethylamine. Chlorotrimethylsilane (0.10 mL, 0.78 mmol) was
added, and the reaction proceeded for 2 h. The solid product was collected through

centrifugation and washed consecutively with water/THF (20:80), THF, and diethyl ether

153

(twice with each solvent). The diethyl ether was evaporated and the solid was transferred

into a Schlenk flask and vacuum dried at 150 °C for at least 2 h.

154

IV. PEO-Modified Silicas

Ethylene Glycol Allyl Methyl Ether (EGAME) (7). The experimental procedure
described in this example was used to make diethylene glycol allyl methyl ether (8),
triethylene glycol allyl methyl ether (9), and poly(ethylene glycol) allyl methyl ether (10,
ll, 12). In a 500 mL round bottom flask was loaded 25 g NaH (1.1 mol, 60% dispersion
in mineral oil), 50 g (0.42 mol) ethylene glycol monomethyl ether (1) diluted with 250 mL
dry THF was added to the flask through a dropping funnel under nitrogen. After refluxing
for 1 h, 56 g allyl bromide was added dropwise and refluxing continued for 3 hours. After
cooling in ice water for 0.5 h, the mixture was quenched with ethanol and then water and
was refluxed for 0.5 h. The liquid portion was separated through decantation and
concentrated. The residue was extracted with methylene chloride and ice water 3 times.
The organic layer was dried over anhydrous magnesium sulfate. Afier removing the
solvent, the mixture was distilled. 41 g (69%) product was collected; bp 61 °C/70 mm Hg
(lit102 56 °C/60 mm Hg). 1H NMR: 5 3.37 (3H, s), 3.50-3.60 (4H, m), 4.05 (2H, d),

5.14-5.28 (2H, m), 5.85-5.98 (1H, m).

Di(ethylene Glycol) Allyl Methyl Ether (DEGAME) (8). DEGAME was
prepared as above using 50 g (0.42 mol) of diethylene glycol monomethyl ether (2), 25 g

(1.1 mol) of NaH (60% dispersion in mineral oil), and 40 mL (0.46 mol) of allyl bromide.

Yield: 85 %; bp 111 °C/60 mm Hg (lit102 86 °C/ 15 mm Hg). 1H NMR: 6 3.37 (3H, s),

3.5-3.7 (12H, m), 4.0.4.4(211, dt), 5.15-5.30 (2H, m), 5.84-5.98 (1H, :1).

155

Tri(ethylene Glycol) Allyl Methyl Ether (TEGAME) (9). DEGAME was
prepared using 52 g (0.30 mol) of triethylene glycol monomethyl ether (3), 16 g (0.63
mol) of NaH (95%), and 30 mL (0.34 mol) allyl bromide. Yield: 83 %; bp 119 °C/ 19 mm
Hg. lH NMR: 6 3.37 (3H, s), 3.5-3.7 (12H, m), 4.0-4.4 (211, (it), 5.15-5.30 (2H, m), 5.84-

5.98 (1H, 11).

Poly(ethylene glycol-350) Allyl Methyl Ether (PEGAME—350) (10). PEGAME-
350 was prepared as above using 274 g (0.78 mol) of poly(ethylene glycol-350)
monomethyl ether (4), 47 g (2.0 mol) of NaH (95%), and 88 mL (1.0 mol) of allyl
bromide, except that, instead of distillation, the product were extracted with methylene
chloride, washed with NaHCO3 solution and water, and dried over anhydrous MgSO4.

After removing the solvent, the products were evacuated at 120 °C for 1 h.

Poly(ethylene glycol-550) Allyl Methyl Ether (PEGAME-SSO) (l 1).
PEGAME-SSO was prepared as for 10 using 274 g (0.50 mol) of poly(ethylene glycol-
550) monomethyl ether (5), 50 g (1.3 mol) of NaH (60% dispersion in mineral oil), and 61

mL (0.65 mol) of allyl bromide.

Poly(ethylene glycol-750) Allyl Methyl Ether (PEGAME-750) (12). PEGAME-
750 was prepared as for 10 using 200 g (0.27 mol) of poly(ethylene glycol-750)

monomethyl ether (6), 17 g (0.67 mol) of NaH (95%), and 35 mL (0.40 mol) of allyl

bromide.

156

Chlorodimethyl(4,7-dioxaoctyl)silane (13). The experimental procedure
employed in this example was also used for preparing Chlorodimethyl(4,7,10-
trioxaundecyl)silane (l4) and chlorodimethyl(4,7,10, 13-tetraoxatetradecyl)silane (15). A
l3-mm Pyrex tube was charged with 5.0 g (43 mmol) of EGAME (7), 6.1 g (65 mmol) of
chlorodimethylsilane, and 5 1.1.1.. of Speier’s catalyst (0.12 M HthCls 6H20 in isopropyl
alcohol). The tube was sealed and heated for 0.5 hour at 100 0C. The product was
collected by vacuum distillation. Yield: 6.9 g (76%); bp 92 °C/ 10 mm Hg (lit103 110
°C/ 15 mm Hg). 1H NMR: 6 0.4 (6H, s), 0.79-0.84 (2H, t), 1.64-1.74 (2H, m), 3.47 (3H,

s), 3.44 (2H, t), 3.50-3.58 (4H, m).

Chlorodimethyl(4,7,10-trioxaundecyl)silane (l4). Silane 14 was prepared as
above using 10 g (62 mmol) of diethylene glycol ally methyl ether (8) and 10 mL (92
mmol) of chlorodimethylsilane. Yield: 9.0 g (56%); bp 135 °C/40 mm Hg. 'H NMR:
(CDCl3). 6 0.4 (61-1, 5), 6 079-084 (2H, t), 6 1.64-l.74 (2H, m), 6 3.47 (3H, s), 6 3.44

(2H, t), 6 3.50-3.68 (8H, m).

Chlorodimethyl(4,7,10,13-tetraoxatetradecyl)silane (15). Silane 15 was
prepared as above using 5.8 g (28 mmol) of triethylene glycol ally methyl ether (9) and 5.0
mL (43 mmol) of chlorodimethylsilane. Yield: 5.5 g (66%); bp 77 °C/0.4 mm Hg. 1H
NMR: 6 0.4 (6H, s), 079-084 (2H, t), 1.64-1.74 (2H, m), 3.47 (3H, s), 3.44 (2H, t),

3.50-3.68 (12H, m).

157

Poly(ethylene glycol-350) methyl 3-(chlorodimethylsily)propyl ether
(PEGMPE-350) (l6). Silane 16 was prepared as above using 7.7 g (16 mmol) of
PEGAME-3 50 (10), 5 1.1L of Speier’s catalyst, and 3.0 mL (26 mmol) of
chlorodimethylsilane. The product was difi'rcult to distill; vacuum and heat were therefore

applied to remove volatile impurities.

Poly(ethylene glycol-550) methyl 3-(chlorodimethylsily)propyl ethers (l7).
Silane 17 was prepared as above using 9.2 g(13 mmol) of PEGAME-SSO (ll), 5 11L of
Speier’s catalyst, and 3.0 mL (27 mmol) of chlorodimethylsilane. The workup procedure

was the same used for (16).

Poly(ethylene glycol-750) methyl 3-(chlorodimethylsily)propyl ethers (l8).
Silane 18 was prepared as above using 11 g (12 mmol) of PEGAME-750 (12), 5 11L of
Speier’s catalyst, and 3.0 mL (26 mmol) of chlorodimethylsilane. The workup procedure

was the same used for (16).

FS-DM-EOl. The experimental procedure employed in this example was used to
make FS-DM-EO; through FS-DM-EOS. To 15 g of evacuated A200 in a 500 ml.
Schlenk flask was added through a transfer needle 300 mL of toluene containing 1.7 mL
(16 mmol) of diethylamine followed by a mixture of 3.5 g (17 mmol) of
Chlorodimethyl(4,7-dioxaoctyl)silane (13) in 50 mL of toluene. The slurry was shaken
until homogeneous after each addition. The reaction was allowed to proceed in a

stoppered Schlenk flask overnight. The slurry was separated by filtration and washed

158

sequentially with two portions of toluene, ethanol/water (50:50), ethanol, and diethyl
ether. The diethyl ether was evaporated and the solid was dried at 150 °C in a vacuum

oven for at least 2 hours.

Table 17. Preparation of PEO attached silicas: Method 1.

 

 

 

 

 

 

 

 

Sample Name A200 Silane (mmol) DEA (mmol)
FS-DM-EO, 15 13 (17) 16
FS-DM-EO; 15 14 (12) 16
FS-DM-EO; 15 15 (17) 16
FS-DM-EOs 12 16 (20) 12
FS-DM-EOu 14 17 (14) 14
FS-DM-EOn 12 18 (14) 12

 

 

 

 

 

FS-Eol. The experimental procedure employed in this example was used to make
FS-EO3, FS-EOs, FS-EOu, FS-DM-S-EOl, and FS-DM-S-EOs. To a solution
containing 2.3 g (20 mmol) of EGAME (7), 0.10 mL Speier’s catalyst, and 50 mL dry
toluene was added 4.5 g of F S-H. The mixture was stirred and refluxed for 24 hours. The
solid sample was collected through centrifugation and washed with toluene and methylene

chloride three times each and dried at 120 °C under vacuum overnight.

159

Table 18. Preparation of PEO attached silicas: Method 2.

 

 

 

 

 

 

 

 

Sample Name Silica (g) Organic Substrate (g)
FS-EO, FS-H (4.5) 7 (2.3)
FS-EO; F S-H (3.0) 8 (2.5)
FS-EO; _ FS-H (4.5) 10 (7.0)
FS-EOu FS-H (3.0) 11 (7.0)

FS-DM-S-EO, F S-DM-H (5.0) 7 (2.2)
FS-DM-S-an F S-DM-H (5.0) 10 (7 .0)

 

 

 

160

 

V. Polymerizable Silicas

5-(Tetrahydropyran-2-yloxy)pent-l-ene (20). To a magnetically stirred solution
of 2.7 mL (0.025 mol) of 4-penten-1-ol and 0.35 g of p-toluenesulfonic acid monohydrate
(18 mmol) in 35 mL anhydrous dioxane at room temperature was added 7.3 mL (0.08
mol) of dihydro-4H-pyran (19) dropwise over a period of 3 min. After an additional 5 min,
half saturated methonolic ammonia was added until the products were slightly basic.
Removal of the solvent gave an oil which was dissolved in 60 mL chloroform and the
solution was washed with 20 mL of 15% NaHC03 The chloroform layer was dried over

anhydrous MgSO, and then concentrated. Distillation gave 2.2 g (51%) of 20; bp 93

°C/33 mm Hg (111104 98-100 °C/10 mm Hg). 1H NMR: 6 0.65 (2H, t), 1.23 (9H, t), 1.35-

1.85 (8H, m), 3.34—3.55 (2H, m), 3.72 ( 2H, t), 3.78-3.86 (6H, q), 4.80 (1H, t).

2-(8-Chlorooctyloxy)tetrahydropyran (25).105 To a magnetically stirred
solution of 8-chlorooctanol (25 g, 0.15 mol) and reagent grade p-toluenesulfonic acid
monohydrate (1.9 g, 9.8 mmol) in anhydrous 1,4-dioxane (280 mL) at room temperature
was added drop-wise 3,4-dihydro-2H-pyran (45 mL, 0.52 mol) over a period of 15 min.
After stirring an additional 10 min, half-saturated methonolic ammonia was added until the
mixture was slightly basic. Removal of solvent gave a viscous solution which was washed
with 15% aqueous sodium bicarbonate extracted and extracted into chloroform. The
organic layerwas dried over MgSO, for 2 h, filtered and the solvent removed to yield 2-

(8-chlorooctyloxy)tetrahydropyran. This material was used without further purification.

161

2-(7-Octenyloxy)tetrahydropyran (23)106 To a 1 L round bottomed flask were
added 600 mL DMSO and 34 g (0.29 mol) KOt-Bu. The mixture was stirred until
homogeneous. The crude 2-(8-chlorooctyloxy)tetrahydropyran from the previous
preparation was added through a dropping firnnel over a period of 30 min. After stirring
for 2 h, the mixture was extracted with pentane and half-saturated NaHC03. The organic
layer was washed with distilled water and dried over MgSO... Concentration and

distillation gave the desired product as a clear oil; yield: 27.4 g (88%); hp: 94-96 °C/0.4

mm (lit. 107 130 °C/12 mm). ‘H NMR: 6 5.87-5.74 (1H, ddt), 5024.90 (211, m), 4.57

(1H, t), 3.90-3.34 (2H, m), 2.07-2.01 (2H, m), 1.87-1.33 (16H, m).

FS-C8-P. To a solution containing 4.0 mL (19 mmol) of 2-(7-
octenyloxy)tetrahydropyran (23), 0.23 mL Speier’s catalyst, and 50 mL dry toluene was
added 10 g of FS-DM-H. The mixture was stirred and refluxed for 24 h. The solid sample
was collected through centrifirgation and washed with three portions of toluene and

methylene chloride, and dried at 120 °C under vacuum overnight. FS-CS-P was prepared

in the same way.

FS-C8-OH. To 120 mL of acidified ethanol [prepared from 500 mL ethanol and 5
mL 0.10 M HCl] was added 7.0 g of FS-C8-P. The slurry was heated on an oil bath under
reflux for 4 hours. The product was centrifuged, washed 3 times with ethanol and diethyl
ether, and dried at 120 °C under vacuum for at least 2 h. The same procedure was used to

prepare FS-CS-OH.

162

FS-C8-M-l. FS-C8-OH (1.0 g) was dispersed in 14 mL of toluene containing 2
mL of pyridine. Methacryloyl chloride (1.0 mL, 9.2 mmol) in 4 mL of toluene was added
dropwise during a period of 15 min while the slurry was cooled in an ice water bath. The
mixture was then heated to and kept at 60 °C for 2 h. The solid portion was separated
through centrifugation and washed with water/ethanol (50:50), ethanol, and finally diethyl
ether. The sample was dried at room temperature and at 120 °C in vacuum for at least 2 h.

The same procedure was employed to prepare FS-CS-M-l.

7-Octen-l-ol (21).Using a literature procedure as a guide, 105 7.2 g (34 mmol) of
8-tetrahydropyranyl-l-octene was added to 200 mL of acidified ethanol, and the mixture
was heated with an oil bath to the reflux temperature for 30 min. The ethanol was
removed and the product was diluted with CH2C12 and dried over MgSO4. After

concentration and distillation, 4.0 g (90% yield) of the deprotected product was obtained;
bp 40 °C/0.2 mm (111103 64-66 °C/7 mm). 1H NMR: 6 5.88-5.74 (ddt,lH), 5.03-4.91 (m,
2H), 3.62 (t, 2H), 2.08-2.01 (dt, 2H), 1.61-1.52 (m, 2H), 1.44-1.28 (m, 6H). 13(3 NMR: 6

139.1, 114.2, 63.0, 33.7, 32.7, 28.9, 28.8, 25.6.

Acrylic acid pent-4-enyl ester (27a). To a slurry of 7.7 g (0.088 mol) of 4-
penten-l-ol and 24 g of dry 3A molecular sieves in 70 mL CC14 was added dropwise 9.7 g
(0.11 mol) of acryloyl chloride in 10 mL CCL. The mixture was heated at the reflux

temperature for 4 h. The mixture was filtered and product was distilled to afford 7.4 g

163

(60%) of 27a; bp 71 °C/20 mm Hg. ‘H NMR: 6 1.72-1.84 (2H, m), 2.11-2.20 (2H, m),

4.15-4.21 (2H, m), 4.98-5.08 (2H, m), 5.76-5.89 (3H, t), 6.38-6.44 (1H, dd).

Methacrylic acid pent-4-enyl ester (27b). Ester 27b was prepared from the same
procedure as in (27a) using 8.5 g (0.073 mol) of 4-penten-l-ol and 12 g (0.10 mol) of
methacryloyl chloride. The product was purified using a column chromatography
(SiOzlhexane). Yield: 10.4 g (92%). 'H NMR: 6 6.08 (1H, s), 5.86-5.73 (1H, ddt), 5.53
(1H, s), 5.06-4.95 (2H, m), 4.14 (2H, t), 2.17-2.10 (2H, dt), 1.92 (3H, s), 1.80-1.71 (2H,

m). IR (6111"): 3079, 1642 [v(-CH=CH2), v(-CH3C=CH2)]. 1713 [v(OCO)].

7-Octenyl methacrylate (28). To a 100 mL round bottomed flask were added 30

mL CC14, 4.0 g (31 mmol) of 7-octen-l-ol, and 8 g of crushed 3A molecular sieves. The
mixture was heated to reflux and 4.4 g (47 mmol) methacryloyl chloride in 10 mL CCl.
was added over a period of 15 min under dry N2. Heating was continued overnight. The
molecular sieves were removed by filtration, and after removing the solvent, the product
was purified using column chromatography (SiOzlhexane); yield: 5.8 g (95%). 1H NMR: 6
6.09 (1H, s), 5.88-5.74 (1H, ddt), 5.54 (1H, s), 5.03-4.91 (2H, m), 4.13 (2H, t), 208-201
(2H, dt), 1.94 (3H, s), 1.71-1.60 (2H, m), 1.44-1.30 (6H, m). 13C NMR: 6 167.5, 138.9,
136.5, 125.1, 114.3, 64.7, 33.6, 28.7, 28.7, 28.5, 25.8, 18.3. IR (cm") 3079, 1642 [v(-

CH=CH2). v(-CH3C=CH2)], 1713 [v(ocon. MS m/z 1W 196.

164

8-(Trichlorosilyl)octyl methacrylate (31). In a dry box, a 13-mm Pyrex tube was
charged, with 5.8 g (30 mmol) of 7-octenyl methacrylate, 0.1 mL Speier’s catalyst61 (0.12
M HthCl6 in isopropanol), and 4.8 g (35 mmol) HSIC13. The tube was sealed and heated
at 60°C overnight. The product was evacuated under vacuum at room temperature and
stored under vacuum until use. ‘H NMR: 6 6.09 (1H, s), 5.54 (1H, s), 4.13 (2H, t), 1.94
(3H, s), 1.69-1.53 (4H, m), 1.44-1.30 (10H, m). 13C NMR: 6 167.5, 136.5, 125.2, 64.7,
31.7, 290,289, 28.5, 25.9, 24.3, 22.2, 18.3. IR (6111"): 1640 [v(-CH3C=CH2)]. 1712
[v(OCO)], 590, 567 [v(Si-Cl)]. MS m/z 1W 331. 5-(Triethoxysilyl)pentyl methacrylate
(29) and 8-(triethoxysilyl)octyl methacrylate (30) were prepared similarly using

triethoxysilane.

FS-C8-M-2. To 30 g of hydrated A200 in a 1L round bottomed flask were added
600 mL of toluene containing 3.0 mL (1.5 mmol) of diethylamine. The flask was attached
to a mechanical shaker, and a mixture of 9.8 g (29 mmol) 8-(trichlorosilyl)octyl
methacrylate (31) in 30 mL of toluene were added. The reaction was allowed to proceed
at room temperature overnight. The product was separated by filtration, and washed with
three portions of both toluene and diethyl ether. The residual diethyl ether was evaporated
and the solid was transferred into a Schlenk flask and dried under vacuum at 120 °C

overnight.

FS-C3-M. FS-C3-M was prepared by the same procedure as FS-C8-M-2 except
15 g of A200, 1.5 mL (15 mmol) of diethylamine, and 3.1 g (12 mmol) of TPM were

used.

165

FS-OTMS. FS-OTMS was prepared by the same procedure as FS-C3-M except

2.9 g (12 mmol) of OTMS was used.

FS-TOM-l. To 15 g of hydrated A200 in a 500 mL round bottomed flask was
added 450 mL of toluene containing 1.5 mL (15 mmol) of diethylamine. The flask was
attached to a mechanical shaker, and a mixture of 2.7 g (11 mmol) octyltrimethoxysilane
(OTMS) and 0.28 g (1.1 mmol) trimethoxysilylpropyl methacrylate (TPM) in 30 mL of
' toluene were added. The reaction was allowed to proceed at room temperature overnight.
The product was separated by filtration, and washed with three portions of both toluene
and diethyl ether. The residual diethyl ether was evaporated and the solid was transferred

into a Schlenk flask and dried under vacuum at 120 °C overnight.

166

Table 19. Products of mixed silylation reactions.

 

 

 

 

 

 

 

Sample Name A200 (g) TPM (g) OTMS (g)
FS-TOM-l 15 0.28 2.7
FS-TOM-ZV . 15 0.61 2.3
FS-TOM-3 30 1.5 4.4
FS—TOM-4 9.0 0.46 i 0.98
FS-TOM-S 1 5 l .5 1.5

 

 

 

 

167

 

IV. Preparation of composites

Compatible composites. All compatible electrolytes were prepared using the
same procedure. A generic preparation is described below. To a blender cup were added
12 g of l M LiClO, in PEGDME-500, small portions of PEO modified fumed silica. The
mixture was blended in a Waring blender for 2-5 minutes with each addition. The addition
of silica was terminated when the gel stops flowing upon tilting. After evacuation to
remove air bubbles at 50 °C in vacuum, the sample was stored in a desicator until use. A
small portion of the electrolyte was sandwiched between two blocking stain-less steel

electrodes separated by a thin Teflon ring.

Cross-linkable composites. All cross-linkable electrolytes were prepared using
the same procedure. A generic preparation is described below. To a blender cup were
added 1.5 g of butyl methacrylate, 15 mg of AIBN initiator, 1.5 g of methacrylate-
modified fumed silica, and 12 g of a PEGDME-SOOILiClO4 solution with the desired O/Li
ratio. The mixture was mixed in a Waring blender for 2-5 minutes and then was
transferred to a vial. After evacuation to remove air bubbles, the sample was stored in a
desicator until use. Polymerization of the composite electrolyte was initiated with a 450W
medium pressure UV lamp. Samples were placed z 3 cm from the lamp, and were
exposed until rubbery. Cross-linking was performed under nitrogen and in air; curing

under nitrogen allowed shorter reaction times.

168

10.

11.

BIBLIOGRAPHY

Herr, R. Electrochimica Acta 1990, 35, 1257.
Dey, A. N. Thin Solid Films 1977, 43, 131.
Peled, E.; Yamin, H. IsraelJ. Chem. 1979, 18, 131.

Armand, M. B.; Chabagno, J. M.; Duclot, M. Second International Meeting on

Solid Electrolytes St. Andrews, Scotland, 1978.

MacCallum, J. R.; Vincent, C. A. Polymer Electrolyte Reviews-1; Elsevier Applied

Science: New York, 1987; Vol. 1.

MacCallum, J. R.; Vincent, C. A. Polymer Electrolyte Reviews-2; Elsevier Applied

Science: New York, 1989; Vol. 2.

Gray, F. M. Solid Polymer Electrolytes: Fundamentals and Applications; VCH

Publishers, Inc.: New York, 1991.

Berthier, C .; Gorecki, W.; Minier, M.; Armand, M. B. Solis State Ionics 1983, 11,

91.

Cameron, G. C.; Ingram, M. D. "Polymer Electrolyte Reviews-2"; MacCallum, J.
R. and Vincent, C. A. ed.; Elsevier Applied Science: New York, 1989; Vol. 2, pp

157.
Nicholas, C. V.; Wilson, D. 1.; Booth, C. PolymerJ. 1988, 20, 289.

Blonsky, P. M.; Shriver, D. F .; Austin, P.; Allcock, H. R. J. Am. Chem. Soc. 1984,

106,6854.

169

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

Fish, D.; Khan, 1. M.; Wu, E.; Smid, J. Brit. PolymerJ. 1988, 20, 281.

Allcock, H. R.; Kuharcik, S. E.; Reed, C. S.; Napierela, M. E. Macromolecules

1996, 29, 3384.

Allcock, H. R.; Napierala, M. E.; Cameron, C. G.; O’Connor, S. J. M.

Macromolecules 1996, 29, 1951.

Borghini, M. C.; Matragostino, M.; Passerini, S.; Scrosati, B. J. Electrochem. Soc.

1995, 142, 2118.
Xia, D. W.; Soltz, D. Solid State Ionics 1984, 14, 221.
Cowie, J. M. G.; Martin, A. C. S. Polymer 1987, 28, 627.

Yarnaguchi, Y.; Aoki, S.; Watanabe, M.; Sanui, K.; Ogata, N. Solid State Ionics

1990, 40/41, 628.

Hawker, C. J .; Chu, F.; Pomery, P. J .; Hill, D. J. T. Macromolecules 1996, 29,

3831.
Qiao, J. Ph.D. Dissertation; Michigan State University: E. Lansing, 1996.

Mixed polymer systems Gray, F. M. "Polymer Electrolyte Reviews-1"; MacCalum,
J. R. and Vincent, C. A. ed.; Elsevier Applied Science: New York, 1987; Vol. 1,

pp 139.

Choe, H. 8.; Carroll, B. G.; Pasquariello, D. M.; Abraham, K. M. Chem. Mater.

1997, 9, 369.

Wieczorek, W.; Stevens, J. R. J. Phys. Chem. B 1997, 101, 1529.

170

 

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

Wang, Z.; Huang, 8.; Wang, S.; Xue, R.; Huang, X.; Chen, L. J. Electrochem.

Soc. 1997, 144, 778.

Chiang, C. K.; Bauer, B. J .; Briber, R. M.; Davies, A. T. Polymer Commun. 1987,

28, 105.
Hudson, M. J .; Sequeira, C. A. C. J. Electrochem. Soc. 1995, 142, 4013.

LeNest, J. F .; Cheradame, H.; Ghandhini, A.; Cohen-Addad, J. P. Polym.

Commun. 1987, 28, 302.

Ichino, T.; Matsumato, M.; Rutt, J. S.; Nishi, S. J. Polym. Sci., PartA: Polym.

Chem. 1993, 31, 1701.

Tsuchida, E.; Ohno, H.; Tsunemi, K.; Kobiyashi, N. Solis State Ionics 1983, 11,

227.

Gray, F. M.; MacCallum, J. R.; Vincent, C. A. Solid State lorries 1986, 18/19,

252.

Such, K.; Florjanczyk, Z.; Wieczorek, W.; Przyluski, J. "Proceedings of the
Second International Symposium on Polymer Electrolytes"; Scrosati, B. ed;

Elsevier: London, 1990, pp 119.

Florjanezyk, Z .; Such, K.; Wieczorek,pW.; Przyluski, J. 7th International

Symposium on Solid State Ionics Hakone, 1989.
Weston, J. E.; Steele, B. C. H. Solid State Ionics 1982, 7, 75.

Plocharski, J .; Wieczorek, W. Solid Sate Ionics 1988, 28-30, 979.

171

 

 

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

Wieczorek, W.; Such, K.; Wycislik, H.; Plocharski, J. Solid State Ionics 1989, 36,

255.

Capuano, F.; Croce, F .; Scrosati, B. J. Electrochem. Soc. 1991, 138, 1918.

Gang, W.; Ross, J .; Brinkmann, D.; Capuano, F.; Croce, F .; Scrosati, B. Solid

State Ionics 1992, 53-56, 1102.

Appetecchi, G. B.; Dautzenberg, G.; Scrosati, B. J. Electrochem. Soc. 1996, 143,

6.

Fritz, H. P.; Stein, K.; Herr, R. .1. Power Sources 1992, 37, 315.

Khan, S. A.; Baker, G. L.; Colson, S. Chem. Mater. 1994, 6, 2359.

Fan, 1.; Fedkiw, S. J. Electrochem. Soc. 1997, 144, 399.

Michael, G.; F erch, H. "Technical Bulletin Pigments," Degussa AG, 1993.
Jezorek, J. R.; Freiser, H. Anal. Chem. 1979, 51, 366-373.

Tundo, P.J. Chem. Soc., Chem. Commun. 1977, 641-642.

Weetall, H. H. Immobilized Enzymes, Antigens, and Antibodies, and Peptides

New York, 1975.

Mathias, J .; Wannemacher, G. J. Colloid Interface Sci. 1988, 125, 61.
Morrow, B. A.; McFarlan, A. J. Langmuir 1991, 7, 1695.

Tripp, C. P.; Hair, M. L. Langmuir 1992, 8, 1961.

Tripp, C. P.; Hair, M. L. Langmuir 1992, 8, 1120.

172

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

Liu, C. C.; Maciel, G. E. J. Am. Chem. Soc. 1996, 118, 5103.
Odlyha, M.; Scott, R. P. W.; Simpson, C. F. J. Thermal Anal. 1993, 40, 1197.
Blitz, 1. P.; Murthy, R. S. S.; Leyden, D. E. J. Am. Chem. Soc. 1987, 109, 7141.

Blitz, J. P.; Murthy, R. S. S.; Leyden, D. E. J. Colloid Interface Sci. 1988, 121,

63.

Blitz, J. P.; Murthy, R. S. S.; Leyden, D. E. J. Colloid Interface Sci. 1988, 126,

387.

Tripp, C. P.; Hair, M. L. Langmuir 1991, 7, 923.

Tripp, C. P.; Hair, M. L. J. Phys. Chem. B 1993,97, 5693.
Tripp, C. P.; Hair, M. L. Langmuir 1995, 11, 149.
Sandoval, J. E.; Pesek, J. J. Anal. Chem. 1989, 61, 2067.
Sandoval, J. E.; Pesek, J. J. Anal. Chem. 1991, 63, 2634.

Chu, C.-H.; Jonsson, E.; Auvinen, M.; Pesek, J. J .; Sandoval, J. E. Anal. Chem.

1993, 65, 808.

Speier, J. L.; Webster, J. M.; Barnes, G. H. .1. Am. Chem. Soc. 1957, 79, 974.
Ferch, H. "Technical Bulletin Pigments," Degussa AG, 1993.

Plueddemann, E. P. Silane Coupling Agents; Plenum Press: New York, 1982.
Sunkara, H. B.; Jethmalani, J. M.; Ford, W. T. Chem. Mater. 1994, 6, 362.

Joseph, R.; Zhang, 8.; Ford, W. T. Macromolecules 1996, 29, 1305.

173

 

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

Bourgeat-Lami, E.; Espiard, P.; Guyot, A. Polymer 1995, 36, 4385.

Bourgeat-Lami, E.; Espiard, P.; Guyot, A.; Gauthier, C.; David, L.; Vigier, G. Die

Angewandte Malaomolekulare Chemie 1996, 242, 105.
Espiard, P.; Guyot, A. Polymer 1995, 36, 4391.
Espiard, P.; Guyot, A.; Perez, J .; Vigier, G.; David, L. Polymer 1995, 36, 4397.

C oreactive Photoinitiators for Surface Polymerization Koehler, M.; Ohngemach,
J. "Radiation Curing of Polymeric Materials"; Hoyle, C. E. and Kinstle, J. F. ed.;

Americam Chemical Society: Washington, DC,1989; Vol. 417, pp 106.

Photocurable Coatings Hoyle, C. E. "Core-active Photoinitiators for Surface
Polymerization"; Hoyle, C. E. and Kinstle, J. F. ed.; American Chemical Society:

Washington, DC, 1989; Vol. 417, pp 1.

Sen Gupta, P. K.; Bevington, J. C. Polymer 1973, 14, 527.

Borer, A.; Kirchmayer, K.; Rist, G. Helv. Chim. Acta 1978, 61, 305.
Sandner, M. R.; Osborn, C. L. Tetrahedron Lett. 1974, 415. .
Delzenne, G. A.; Laridon, U.; Peeters, H. Eur. Polym. J. 1970, 6, 933.
Suppan, P.; Porter, G. Trans. Faraday Soc. 1965, 61, 1664.

Ledwith, A.; Ndaalio, G.; Taylor, A. R. Macromolecules 1975, 8, 1.
Bemer, G.; Kirchmayer, R.; Rist, R. .1. Oil Chem. Assoc. 1978, 61, 105.

Harnity, M.; Scaiano, J. C. J. Photochem. 1975, 4, 229.

174

 

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

Sadhir, R. K.; Smith, J. D. 3.; Castle, P. M. J. Polymer. Sci. A: Polym. Chem.

1983, 21,1315.

Odian, G. Priciples of Polymerization; 3rd ed.; John Wiley and Sons, Inc.: New

York, 1991.

Bartholomew, R. F.; Dividson, R. S. J. Chem. Soc., Chem. Commun. 1971, 2347.

Decker, C. Makromol. Chem. 1979, 180, 2027. . '

Organosilanes Armitage, D. A. "Comprehensive OrganometallicChemistry: The
Synthesis, Reaction and Structures of Organometallic Compounds"; Wilkinson, G.,

Stone, G. A. and Abel, E. W. ed.; Pergarnon Press: New York, 1982; Vol. 2, pp 1.

Permeation and Morpholog» in Polymer Latex Films Containing Leachable
Additives Steward, P. A.; Hearn, J .; Wilkinson, M. C .; Wilson, A. J .; Roulstone, B.
J. "Film Formation in Waterbome Coatings"; Provder, T., Winnik, M. A. and
Urban, M. W. ed; American Chemical Society: Washington, DC, 1996; Vol. 648,

pp 359.
Nazri, G. A.; Meibuhr, S. G. J. Electrochem. Soc. 1989, 136, 2450.

Wieczorek, W.; Such, K.; Chung, S. H.; Stevens, J. R. J. Phys. Chem. Chem.

1994, 98, 9047-55.

Sperling, L. H. Introduction to Physical Polymer Science; 2nd ed.; John Wiley &

Sons, Inc.: New York, 1992.

Hasegawa, M.; Arai, K.; Saito, S. J. Polym. Sci., Polym. Chem. Edn. 1982, 20,

1021.

175

 

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

Hasegawa, M.; Arai, K.; Saito, S. J. Polym. Sci., Polym. Chem. Ed. 1987, 25,

3117.
Hasegawa, M.; Arai, K.; Saito, S. J. Chem. Eng. Japan 1988, 21, 30.

Janssen, A. W. G.; Van Herk, A. M.; German, A. L. ACS Polym. Prepr 1993, 34,

532.
Konno, M.; Arai, K.; Saito, S. J. Polym. Sci., Polym. Chem. Ed 1987, 25, 223.
Hergeth, W.; Starre, P.; Schmutzler, K.; Wartewig, S. Polymer 1988, 29, 1323.

Hergeth, W.; Steinau, U.; Bittrich, H.; Simon, G.; Schmutzler, K. Polymer 1989,

30, 254.

Hasegawa, M.; Arai, K.; Saito, S. J. Appl. Polym. Sci. 1.987, 33, 411.

Arai, M.; Arai, K.; Saito, S. J. Polymer Sci., Polymer Chem. Ed. 1982, 1021.
Fukano, K.; Kageyama, E. J. Polym. Sci., Polym. Chem. Ed 1975, 13, 1309.

Hasegawa, M.; Arai, K.; Saito, S. J. Polym. Sci., Polym. Chem. Ed. 1987, 25,

3231.

Wu, P.-W.; Holm, S. R.; Duong, A. T.; Dunn, B.; Kaner, R. B. Chem. Mater.

1997, 9, 1004.

Kramer, G. W.; Levy, A. 8.; Midland, M. M. "Organic Synthesis Via Borane";

Brown, H. C. ed.; John Wiley & Sons: New York, 1975, pp 243.

Mitchell, T. N.; Kerstin, H. W. J. Organomet. Chem. 1992, 436, 43.

176

 

103.

104.

105.

106.

107.

108.

Steward, O. W.; Lutlcus, A. G.; Greenshields, J. B. J Organomet. Chem. 1978,

144, 146.

Bestmann, H. J .; Koschotzky, K. H.; Schaetzke, W.; Sness, J.; Vostrosky, O. L.

Liebigs Ann. Chem. 1981, 9, 1705.
Van, J. H.; Hershied, J. D. M. Synthesis 1973, 169.
Wood, N. F. ; Chang, F. C. J. Org. Chem. 1965, 30, 2054.

Matsuda, H.; Yamamoto, A.; Iwamoto, N.; Matsuda, S. .1. Org. Chem. 1978, 43,

4567.

Tolstikov, G. A.; Odinokov, V. N.; Galeeva, R. 1.; Bakeeva, R. S.; Akhunova, V.

R. Chem. Nat. Compd (Engl.Transl.) 1982, 18, 219.

177

 

illiiiiliiliir