ENORGANIC CONDENSATEGN REACHONS
I. STANNOSHLOXANE FORMATION

Thesis for We Dsq'rcu of DH. D.
MECHEGAN STATE UNWERSITY
Curt Thies

1962

THESIS

L

.-

~-— --

z===w _ i
Michigan State -'
University

 

a ' P". r... -~--- cwm‘

 

 

FMPV_:H'\~§| C‘T'\TC H5*\‘{‘.;r:r‘j71
’ ' ‘QAHH. bunnuoil

' n . _.. .
6|.vlanua=.1

EhS L:’-‘~.IVSMG,£\1;Ci*-aléil‘xi‘4

 

ABSTRACT

INORGANIC CONDENSATION REACTIONS
I. STANNOSILOXANE FORMATION

by Curt Thies

' .- A series of stannosiloxane compounds has been prepared by the
condensation of sodium triphenylsilanolate with various organotin
chlorides in benzene. Infrared and elemental analyses were employed
to characterize the reaction products.

The stoichiometry of the reaction of sodium triphenylsilanolate
with several organotin chlorides in benzene has been established.
It has been found that such reactions go rapidly to completion and are
free of significant side reactions. This implies the reaction of pure
(NaOhSiCbz with organotin dichlorides in an aprotic solvent will yield
high molecular weight stannosiloxane polymers having a regular
alternating (-Si-O-Sn-) structure.

Several stannosiloxane polymers have been prepared by the

i_r_i situ condensation of (NaO)zSi¢z with dibutyltin dichloride. The

 

polymers were not fully characterized, but it has been established that
the polymerization process is rapid. Highly purified (NaO)zSi¢z, which
is soluble in absolute ethanol, has been isolated and will be used in

future stannosiloxane polymerizations .

INORGANIC CONDENSATION REACTIONS

I. STANNOSILOXANE FORMATION

BY

Curt Thies

A THESIS

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

DOCT OR OF PHILOSOPHY

Department of Chemistry

1962.

ACKNOWLEDGMENTS

The author wishes to express his sincere appreciation
to Dr. J. B. Kinsinger for his guidance and encouragement
during the course of this investigation.

Grateful acknowledgment is also extended to Dr. R. N.
Hammer and Dr. M. E. Russell for many stimulating dis-
cussions and to the United States Air Force whose research

grant provided personal financial assistance from December
1959 to December 1961.

***>l<****>l<>§<*>l<*

ii

TA BLE OF CONT ENTS

I. INTRODUCTION .......... . ..........

A.

C.
D.

II. EXPERIMENTAL .............

A.

Heewwvoe

III. RESULTS AND DISCUSSION

A.
B.

C.

IV. SUMMARY ...........................

BIBLIOGRAPHY ............

APPENDICES

General . . . ...................
B. Principles of Polycondensation ...........
Organosiloxane Formation . . ..........
Stannosiloxane Formation . . ..........

Preparation of Reactants for Stannosiloxane

Formation ......... . ...........
l. Hexaphenyldisiloxane ...........
2. SOdium Triphenylsilanolate ........
3. Sodium Diphenylsilanolate. . . . . . . . .
4. Organotin Chlorides . . . . . .......
Preparation of Stannosiloxane Compounds . . . .
Solvent Purification ................
. Infrared Spectra ..................
. Analytical. . . . .................
Stoichiometry . . ................
Polymer Formation ................
Polymer Fractionation ..............
. Solubility Studies .................

Stannosiloxane Synthesis and Characterization . . .

Stoichiometry and Rate of Stannosiloxane

Formation .....................
Polymer Formation ................

iii

11
14

20

20
20
20
21
22
22
22
24
24
27
36
42
43

45
45

49
64

e
68

72

LIST OF TA BLES

TABLE Page
I. Postulated Structures of Reported Organometallosiloxane
Polymers .......................... 2
11. Rate Constants for Esterification of Homologous Series of

Monobasic and Dibasic Acids

III. Tabulation of Known Stannosiloxane Compounds ......
IV. Organotin Chloride Analytical Data .......... . . .
V. Measured and Literature Refractive Indices for Solvents

VI.

VII.

VIII.

XI.

XII.

XIII.

XIV.

XV.

XVI.

. Material Balance Closure Data ....... .

Used in the Study of Stannosiloxane Formation . . .
Stannosiloxane Combined Oxides ..............

Tin Analytical Data . .

Standard Chloride Analys es

.......

. Scatter of Data Used to Plot Beer's Law Calibration

Curves ............................
Duplicate (_r_l-C4H9)zSn(OSi¢3)z Yield Determinations . . . .
Polymer Material Balance ................

Material Balance Data for Stannosiloxane Polymer
Fractionations ........................

Stannosiloxane Compounds Prepared by Silanolate Conden-
sation Reaction In Benzene .......... .

. Infrared Absorption Peaks of Stannosiloxanes in Carbon

Disulfide at Room Temperature . . ...........

Stannosiloxane Stability In Benzene at Room Temperature

iv

19

23

25

26

28

31

37

38

39

4O

43

46

48

50

LIST OF TABLES - Continued

TABLE

X VII. T riphenyl (triphenylsiloxy) stannane Stoichiometric
Results ...........................

XVIII. Bis (triphenyl siloxy)di- n.- butylstannane Stoichiometric
Results 0 O 0 O O O O O O O O O O O Q C O O C .....

XIX . Bis (triphenylsiloxy) dim ethylstannane Stoichiometric
Results ........................

XX. Stoichiometric Results for (2-C4H9)zSn(OSi¢3)z Formed
Using a Known Imbalance of Reactants . . . .......

XXI. Stoichiometric Results for (3-C4H9)ZSn(OSi¢3)z Formed
inReagentGrade Benzene. . . . . . . . . . . . . ..

XXII. Comparison of Interfacial and Melt Polycondensation

XXIII. Analytical Data for Methanol-Insoluble Stannosiloxane
Polymer Fractions ..................

Page

51

52

53

56

59

63

65

LIST OF FIGURES

FIGURE Page
1. Plot of Xn vs r calculated from Equation 11 ....... 12
2. Flow .diagram of stoichiometry procedure . ....... 29

3. (ID3SiOSnO3 Beer's Law calibration plot, C6le = solvent .. 33

4. (CH3)ZSn(OSi¢3)2 Beer's Law calibration plot, CS; =
solvent .............. . . ........ ‘ . . . 34

5. (2-C4H9)zSn(OSid)3)z Beer's Law calibration plot,

C6le = solvent. , ..................... 35
6. Histogram showing frequency of experimentally

determined stn(05i<1>3)Z yields . ............. 55
7. Theoretical stannosiloxane yields, ‘70, vs. NaOSiR3

deficiency, %, for R3SiOSnR3' and Rz'Sn(OSiR3)z ..... 57
8. Sodium Chloride Recovery vs Reaction Time . . . . . . . 60

vi

LIST OF APPENDICES

APPENDIX Page
A. Reaction Conditions and Stoichiometric Data for
¢35iosn¢,, (CH,)zsn(05icf>3)z, and (2-C4H9)ZSn(OSi¢>3)z

Formation......... ....... . ....... 73

B. Spectrophotometric Data for ¢3SiOSn¢>3, (CH3)zSn

(05169,, and (2-C4H9)zSn(OSi(l)3)z Formation ...... 78
C. Solubility Characteristics of Stannosiloxane Polymers. 82
D. Infrared Spectra of Stannosiloxane Compounds. . . . . 84
E. Sodium Chloride Rate Data . ............ . . 96

vii

I. INTRODUCTION

A. General

In recent years a need has arisen for materials having both stability
and flexibility at high temperatures. This has stimulated work on polymers
containing inorganic elements in the backbone structure. The organo-
siloxane polymers have served as a starting point for much of the work
since they combine good thermal stability with mechanical versatility (1).
The use of these polymers at extremely high temperatures is limited
principally because of decomposition and a tendency to form cyclic struc-
tures under such conditions (2). It has been proposed this tendency would
be reduced by replacing some of the Si-O units in the polymer chain with
Ti-O, Al-O, or generally, metal-oxygen units (3). Several authors have
postulated the more ionic bond character caused by the introduction of
metal-oxygen units would itself increase the thermal stability of the polymer
structures (3, 4). t For these reasons many attempts have been made to
prepare organometallosiloxane polymers containing silicon and various
elements from the second, third, fourth, and fifth groups of the periodic
table (5, 6). All of these have been prepared by simple condensation
processes as few monomers are known which polymerize by addition
reactions (7). Proposed structures and nomenclature for several polymers
which have been synthesized are listed in Table I.

Although high molcular weight organosiloxane polymers (the silicones)
have been prepared, most organometallos iloxane polymers reported to
date have had number average molecular weights (Mn) ranging from 1000
to 5000. The low Mn's are believed due to various side reactions which
in part lead to formation of cyclic structures. As no quantitative studies

have been carried out to establish the course and kinetic mechanisms of

Table I. Postulated Structures of Reported Organometallosiloxane

 

 

 

Polymers
Polymer Structure Reference
‘n‘ I.“
Organoarsonosiloxane flSi-O As -O 8
I
R n_>_l R x
R
Organoantimonosiloxane S'l-O Sb - O 9
R n>1 OR' x
1?
Organoborosiloxane S|i-O B ~04} 10
. R n>l R' x

l iR'
Si-O Ti-O 12

R n21 OR' x

(Triorganosiloxy)alumino-. %§i-— O A1-03}3
siloxane

Or ganotitano siloxane

1: I."
Organostannosiloxane Si-O Sn-O 4, 11, 12, 13,
. | I 14
R n>1 R. x

 

the condensation reactions, the purpose of this study was to investigate
a prototype inorganic condensation reaction. The stannosiloxane system
was selected for the following reasons:

1. Organotin monomers are readily available.

2. The Sn-C bond is stable (15).

3. Organotin compounds can be easily purified and handled without
elaborate techniques.

4. It has been established Si-O-Sn bonds can be formed (16, 17).

5. Several qualitative studies of low Mn stannosiloxane polymers
have been made (4,11,12,13,14).

6. Organotin additives are known to increase thermal stability
of some organosiloxane polymers (4).

. Since no quantitative studies of stannosiloxane reaction systems
have been made, the present investigation was undertaken to obtain
stoichiometric and, if possible, rate data for the formation of well-
defined compounds. The results will later be applied to polycondensation

reactions .

B. Principles of Polycondensation

Polycondensation occurs when substances with two or more reactive
groups undergo a simple chemical reaction with each other. The first
reaction step produces a dimer which by subsequent reactions with
additional monomers gives a trimer, tetramer, etc. However, this
isn't the only possibility. Two dimers may react to form a tetramer,
two trimers may unite to form a hexamer, etc. This process eventually
leads to a high polymer as shown in the following simplified scheme:

X-R-Y -F X - R - Y —>X-—R—R—-—-Y +XY

X—R—R—Y + X--R—Y —-—> X—R—R—R——Y‘ + XY

‘X-—'-}(R)rfi-—-—Y + X——-(-R-)fi—Y —-> X—+R)mY + XY

The following discussion of the above process is based on Flory's
classical treatment of polycondensation reactions (18). I

A fundamental assumption of polycondensation theory is that the

reactivity of a functional group is independent of chain length and hence,

all functional groups in the system react at random. It has long been
known reaction rate constantsi for a homologous series of monomeric organic
compounds are independent of chain length except for very short chains.
This is illustrated by the results obtained-for the acid catalyzed conden-
sation of mono- and dibasic acids with ethanol shown in Table II (18, 19).
Except for the first members of both series, there is no significant
variation of the rate constant with chain length. Likewise, differences
in the rate constants for esterification of mono- and dibasic acids
disappear as chainlength separating the carboxyl groups increases.

Perhaps more important is the observation that the same rate
constant can be applied to each of the consecutive steps in the conversion

of the dibasic acid to a diester:

O 0
II I k H g
HO'C ‘R‘C 'OH + CzHSOH ‘—> HO'C‘R" 'O'Csz + H20
0 O

O O

HO-g-R-g-O-CZHS + CZHSOH -£9 CZHSO-H -R-|C|-O-CZH5 + HZO
No variation in k was observed throughout the course of the reaction.
Similar results were obtained from hydrolytic studies of ethylene glycol
di and triacetate as well as diethyl succinate. This demonstrates that
ordinarily the reactivity of a functional group in a polyfunctional molecule
can be assigned a definite value which does not change during the course
of a reaction. Neighboring substituents may markedly affect reactivity,
but the reaction of one functional group will not noticeably affect its
influence on the other. Any effect of this nature decreases rapidly with
the distance of separation in the molecule. These generalizations are

for reactions which do not involve interaction between a pair of charged

Table II. Rate Constants for Esterification of Homologous Series of
Monobasic and Dibasic Acids (18, 1‘9)

 

kAxlo‘ at 25°C ka104 at 25°C
Chain (Esterification of (Esterification of
Length monobasic acids)* dibasic acids)*
1 22.1 -__
2 15. 3 6. 0
3 7. 5 8. 7
4 7 . 4 8 . 4
5 7.4 7. 8
6 --- 7. 3
8 7. 5 ---
9 7. 4 ---
Higher 7. 6 ---

 

l

>:< -l ..
All rate constants express in (gram equivalents/liter) sec. 1.

reactants. When charged Species are involved, functional group reactivity
may not be entirely independent of the status of other groups in the
molecule.

Although the reaction rate of a functional group does not depend on
the size of the molecule to which it is attached, this does not necessarily
mean the :rate, is independent of size in polymers of high molecular
weight. Several arguments for decreased reactivity with increased chain
length have been made based entirely on the mechanics of interaction of
two functional groups attached to large molecules. It was postulated the
collision rate for large molecules must be small because of their low
kinetic velocity and the high viscosity of the liquid medium would further
reduce reactivity. Shielding of the functional group within the coiled chain
was offered as another reason for lowered reactivity.

However, Flory has shown from a theoretical viewpoint that none
of these arguments is valid. Large molecules are known to diffuse
slowly, but this must not be confused with the collision rates of the
functional group. A Mobility of this group is much greater than the macro— A
scopic viscosity would indicate. Even though, its range is limited by
attachment to the polymer, the reactive group may diffuse readily over
a considerable region through rearrangements in the conformations of
nearby segments of the chain. Its actual oscillations against immediate
neighbors may occur at a frequency comparable to, or not much less than,
that existing in simple liquids. Consequently, the actual collision
frequency will be essentially independent of the mobility of the molecule
as a whole and the macroscopic viscosity.

As Rabinowitch and Wood have shown (20), a pair of functional
groups which diffuse together in the liquid state may undergo repeated
collisions before diffusing apart. A decrease in the diffusion rate will
prolong this series while proportionately increasing the time between

successive encounters. In this way decreased mobility will change the

time distribution of collisions experienced by a functional group, but it
shouldn't significantly affect the collision rate averaged over a period
that is long compared to the time for diffusion from one partner to the
next. Even if it does, the duration of the collided state will be
proportionately increased with the net effect that concentration of pairs
of functional groups close enough for reaction (assuming the necessary
activation energy is available) is independent of mobility.

The independence of reaction rate and molecular size may be
qualitatively shown from transition state theory where the rate constant,

K is given by

k = K* (LT/h) (1)

and 16: Boltzmann's constant, T = absolute temperature, h = Planck's
constant, and K* = equilibrium constant for the activated complex.
For a bimolecular reaction

I
>,<

K = Fab*/Fan exp (-Eo/.kT) (2)

' >:<
where Fab , F and Fb are the partition functions of the activated com-

a!
plex and the two reactants respectively; E0 is the energy of the activated
complex at absolute zero. If the complexity of one or both of the

>3:
reactants is increased, Fab and the product Fan change by nearly
identical factors for the added degrees of freedom unless this increase

alters the molecule in the immediate vicinity of the functional group.

Thus, mobility within the liquid will not affect the equilibrium

reactants 2:? activated complex.

The rate of reaction is proportional to the activated complex concentration
so it will not be influenced by mobility of the molecules, diffusion rate,

or viscosity.

Exceptions to these conclusions result when molecular mobility is
exceedingly low or when the reaction rate constant is exceptionally large
and the mobility is low. In such cases equilibrium concentration of
reactant pairs may not be maintained and the reaction would become
diffusion controlled. As condensation polymerizations generally involve
reactions of moderate rate, the concentration of reactant pairs should
be readily maintained even at highest molecular weights and viscosities.

The only remaining argument for low polymer reactivity is the
belief that the reactive group of a large polymer will be shielded by the
coiling of the chain. This may be a real effect in very dilute solutions
where the randomly coiled polymer molecules are more or less independent
of each other. However, in more concentrated solutions the polymer
chains are randomly intertwined. A reactive group has no preference
for units in its own chain and will generally be encircled by units of other
molecules. These units act as diluents and are taken into account by
writing the rate expression in terms of concentrations of functional groups
participating in the reaction per unit volume.

Flory verified his theoretical arguments for the independence of
the reaction rate and molecular size by comparing the kinetics of homo-
geneous monofunctional and bifunctional esterification reactions (21).
Assuming the velocity constant, k, is the same for all functional groups,

the rate expression for uncatalyzed polyesterification is

-d [COOH]/dt = k [coonflon] (3)

where concentrations are expressed as equivalents of the functional
groups. Assuming equal carboxyl and hydroxyl concentrations, this

may be integrated to give:

2kt = T - Const. (4)

where c = reactant concentration at time t.

By defining p as the fraction of functional groups initially present that

have undergone reaction at time t, equation 4 can be replaced by

l
2kt = W - Const. (5)

where c = (l - p) co.

Flory found that data for both monofunctional and polyfunctional
esterifications fit equation 5 except during the initial stages of the
reaction. The exact cause for this behavior during the earlier stages of
the reaction is not important since it is characteristic of simple esterifi-
cations as well as those which lead to polymer formation. The similarity
in mono- and polyesterifications is direct evidence for nondependence of
reactivity on molecular size.

The principle of equal reactivity of all functional groups is
extremely important because it allows the kinetics of a polycondensation
process to be treated as a simple chemical reaction and makes possible
the analysis of polymer constitution using simple statistical methods (18).
It also permits calculation of the number average degree of polymeri-
zation, Xn, from the extent of reaction p.

If a polycondensation reaction is free of side reactions and involves
pure bifunctional monomers of the type X-Y, or equivalent proportions
of X-X and Y-Y, the number of unreacted groups is equal to the number
of molecules present in the system. The number of molecules per unit
volume is given by co(l-p). Define a structural unit as each residue
from an X-T Y, or X-X and Y-Y unit so that the total number of structural
units is equal to the total number of bifunctional monomers initially used.

. Xn will then be given by

_ Number of units _ Co = 1 (6)
n Number of molecules (l-p) Co (1-P)

 

 

10

The number average molecular weight is

fin: Mo/(l-p) (7)

where M0 is the mean molecular weight of a structural unit.

Equations 6 and 7 assume reactive groups X and Y are present in
equivalent amounts and that no side reactions occur. , Mn values calculated
for cases where these stringent conditions are not met will be erroneous.
Flory (18) has developed a relation between theoretical Mn values and
nonequivalence of reactants, monofunctional reactants, or imbalance in
the stoichiometric proportions. Assume a small amount of a reactant
Y+Y is added to either pure monomer X—Y or stoichiometric proportions
of X—X and Y—Y. Y—l—Y may or may not be identical with Y—Y.

Define Nx as the total number of X groups initially present, NY as the
total number of initial Y groups, and r as the ratio Nx/Ny so that the total

number of units is
1
(NX+Ny)/2=Nx(l+-r-)/2 (8)
while the total number of ends of chains is given by
2Nx(1 -p)+ (NY-Nx)=NX[2(l-p)+ (1-r)/r] (9)

The total number of molecules is half the total number of chain ends
so that Xn is expressed by

Number of units _ l + r

32 : —.
n Number of molecules 2r(l-p) + l - r

 

 

(10)

This relation is applicable to polymers containing small amounts of a
monofunctional reactant Y+-if r is defined as Nx/'(Nx+ 2NY+).

Assuming a complete reaction (p = 1), theoretical maximum values
of Xn can be calculated for the case where small amounts of monofunctional

reactants are present or where stoichiometric imbalance of the bifunctional

ll

monomers exists. For p = 1, equation 10 becomes

1+r
l-r

 

Xn = (1 1)

A plot of 33,, for various r values in Figure 1 shows Yn is markedly
affected by small deviations from r = 1. This graphically shows why high
molecular weight linear condensation polymers are formed only from
reactions involving extremely BEE. bifunctional monomers present in
stoichiometric proportions. If these stringent conditions are not met and
if side reactions occur which create monofunctional units or destroy the
chain growth process, it will not be possible to synthesize high polymers

by a simple condensation process.

C . Organosiloxane Formation

Since organosiloxane polymers are prototypes of organometallo-
siloxane polymers, the mechanism of siloxane bond formation is of interest
in the present study. These linkages can’be formed by the condensation

of organosilanols
+ -
- l l
H °r OH: -Si-O-Si- + H20
A I I

This reaction has been studied by Grubb (22) who examined the acid and

 

4
‘

l l
-S'i-OH + HO-S|i-

base catalyzed condensation of trimethylsilanol in methanol at 25°C.

The proposed mechanism for acid catalyzed condensation is -.

H (QHals H+ +
(CH3),si-o: + Si—O——CH3——> (CH3)3SiOSi(CH3)3+ CH3OHZ

where all protonated species were assumed to be 'n non- rate determining
equilibrium. Since the concentration of (CH3)3SiO CH3 is proportional to
acid catalyst concentration, the rate is first-order with respect to the

catalyst. A similar mechanism was proposed for base catalyzed

12

 

 

 

400 -
300 -
200 *—
in
100 l—
p = 1
0 ' l ' l l l l J J J L l
1.00 0.98 0.96 0.94 ‘ 0.92 0.90

1‘

Figure l. Alplot of X}, vs. r calculated from equation 10.

l3
condens ation:

(CH3)3Sio‘: + (CH3)3Si-OCH3 ,———> (CH3)3SiOSi(CH3)3 + CH3o'

The Silanolate ion concentration is proportional to the base catalyst
concentration and was regarded as the active species.

It is presumed the above reactions would follow similar routes
in aprotic solvents except that two molecules of silanol would now be
involved in the condensation process (23);

.H+ +

(CH3)3SiOH + (CH3)3SiOH (_—_-_-_-> (CH3)3SiOSi(CH3)3 + H3O

(CH3)3Sio' + (CH3)3SiOH ——> (CH3)3SiOSi(CH3)3 + OH-

The mechanistic discussion up to this point has involved the con-
densation of monofunctional silanols. Although siloxane polymers can
be formed in principle by the condensation of bifunctional silanols, such
a reaction actually yields a mixture of low molecular weight polymers
and cyclics: -

R
, 1
(n + x) stuomz C—a-t—> H(-O-Si-) OH + (RzSiO)
n X

X: 3-8

High molecular weight siloxane polymers are generally prepared by
catalytic opening and polymerization of cyclic siloxanes. Grubb and
Osthoff have investigated this ring-opening process using octamethyl-
cyclotetrasiloxane as the monomer and potassium hydroxide as the
catalyst (24). Their results indicate a basic catalyst like potassium
hydroxide re’acts quantitatively with siloxane bonds to form potassium

Silanolate

l4

(EH3 ('3H3 (fHa (EH3
KoH+-5i-o-5i- —>- -Si-OK+HO-Si-
I I e— l

CH3 CH3 CH3 ' CH3

The Silanolate partially ionizes to give Silanolate anions
C|3H3 ‘ CH3
-Si- OK ——a- -Si—o' + K+
<——
CH3 ' CH3

which act as the active Species in the polymerization reaction. The

propagation step of the reaction may be represented as

CH3 CH3

C \S/ 1 CH CH CH
. CH‘ CH H i
. I .3 3 .4 \3 /O/ \O\ 3 l 3 l 3
KO -( 81- O )- 1 - O + i Si -—? K04 i—O Si—-—O
n ' \0 O/ n+4|
CH3 CH3 CH3 \SK/ CH3 3 CH3
/
CH, CH,

It is possible a similar catalytic polymerization of cyclic compounds could

be employed to prepare high molecular: weight metallosiloxane polymers (7).

D . Stannosiloxane Formation

Syntheses of several stannosiloxane polymers have been reported
(4, 11, 12, 13, 14). Andrianov first prepared such polymers by cohydrolysis
of diorganodichlorosilanes and diorganotin dichlorides in the presence of

excess base (12):

(mx)R S'Cl + st Cl st——-—"'——>O s o + x +1)MC1
.2 1 z z n z———> MOH 1"- (m

The initial hydrolysis products were transparent liquids soluble in organic
solvents. They were transformed upon heating into insoluble, infusible

solids. The polymers were not fully characterized. Koenig and co-workers

15

repeated the original cohydrolysis process using dimethyltin dichloride
and dimethyldichlorosilane as reactants-((4.13). A rubbery plastic was
isolated but not characterized except for silicon-tin analysis.

Koenig has also synthesized stannosiloxane polymers by the
reaction of silanols with organotin oxides (l3).

. R' R
. A .
(RzSnO)x+(mx)Rv281(OH)z —9 %1:O}m>—SH—_O ‘ + (mx)HZO

1. R x
The products were for the most part brittle resins of low softening pOint
(SO-70°C). They were appreciably soluble in organic solvents and rapidly
hydrolyzed by dilute mineral acids. Mn ranged from 1000 to 5000.

The cohydrolysis and silanol-organotin oxide reactions described
above are not suited for the preparation of well-defined-linear conden—
sation polymers. They bring about a random integration of tin and silicon
into a polymer chain and a (Si—O-Sn-O)x repeating chain is not obtained.
For example, the cohydrolysis of trimethylchlorosilane with a dialkyltin
dichloride yielded tetraalkyl-l, 3-bis-(trimethylsiloxy)-distannoxanes

rather than bis(trimethylsiloxy)dialkylstannanes (25) .
_ .xs NH . .
CH381C1 + RZSnC1z ——r-——-L> (CH3)351O(SnRzO)251(CH3)3

Because of this, the cohydrolysis and silanol-organotin oxide reactions
were not considered suitable for the present study.

Two reactions which in principle can form stannosiloxane polymers
with chains having an alternating Si-O-Sn-O structure are transesterifi-

cation

.x RzSn(OR')z + 'x RZSi(00CCH;,)z\A3 R R
—ESi-o-Sn-o+,-c + ZLXRTOOCCH3

x RzSn(OCCH3)z + x RZSi(OR')2 A/' R R

and the condensation of alkali metal silanolates with diorganotin dihalides.

16

A

M = alkali metal.

O'Brien has prepared stannosiloxane polymers by .the.tr.ans-
esterification method (11). Waxy solids were obtained at reaction
temperatures in excess of 200°C while viscous fluids were formed at
lower temperatures. The polymers were not characterized beyond noting
the physical appearance so it is not known whether or not copolymers
with a 1:1 silicon-tin ratio were formed.

Henglein, Long, and Schmack obtained polymers by prolonged
heating of dimethyldiethoxysilane with diisobutyldiacetoxytin at 120°C (26).
From the low molecular weight products they isolated l ethoxy-3 acetoxy-
tetrai s obutyldi stannoxane .

(icqu) (ic.H9)

CZHSO— n—O— n—OCOCH3

(iC4H9) (iC4H9)

This clearly indicates side reactions do occur in the transesterification
reaction under these conditions.

It is interesting to note that Okawara, gt a_._l. failed to obtain
(CH3)3SiOSn(CH3)3 and compounds of the type (CH3)3SiO-{Sn(CH3)zO};Si(CH3)3
by transesterification of trimethylethoxysilane with trimethyltin formate
or dimethyltin diformate (25).

Koenig and Hutchinson reported the synthesis of stannosiloxane
polymers by the condensation of (NaO)ZSi¢z with ¢zSnClz in dioxane (14).
The (NaO)zSi4)z was prepared by reacting sodium metal with (HO)zSi(’)z
and used i_1_1 §i_tu_._ They refluxed the reaction mixture one hour and then
let it stand several days at room temperature before isolating by fil-
tration a precipitate which had formed. This precipitate was not pure
sodium chloride, but otherwise was unidentified. A low molecular weight

oily material was isolated from the dioxane filtrate. This contained

17

20. 88% Sn and 7.16% Si. These values are not in good agreement with
the theoretical values expected for a 1:1 silicon—tin copolymer (24. 3%
Sn and 5. 76% Si). However, nothing was known about the (NaO)zSi4)z
purity, so it is not surprising a 1:1 Si-Sn copolymer was not obtained.

Although both the'transesterification and silanolate condensation
reactions have been used to prepare low molecular weight stannosiloxane
polymers, the nature of these reactions and the products obtained have
not been thoroughly examined. A kinetic study of such reactions would
be particularly difficult for the case of bifunctional systems since
nothing is known about stannosiloxane chemistry. Only a few stanno-
siloxane compounds have been reported while no studies of the reactions
involved in their formation have been made. Thus, it was concluded
that the present study would deal principally with low molecular weight
stannosiloxane compounds. The main purpose is to establish the
stoichiometric and kinetic nature of reactions which can eventually be
used in the formation of stannosiloxane polymers. The silanolate
condensation reaction was selected since it has been established stanno-
siloxane compounds can be formed via this technique (l6, l7).

Sommer, et 11. appear to have been the first to use silanolates in
organic syntheses. They prepared several organosiloxanes by the
condensation of silanolates with organochlorosilanes (27). Gilman,
Benedict, and Hartzfeld found this a generally useful method for prepar-
ing disiloxanes in high yields (28). Summers showed this reaction
could be used to form a Si-O-metal linkage by condensing NaOSi¢3 with
~¢3PbC1 to give ¢3SiOPb¢3 (29). The silanolate condensation reaction

has since been used to prepare
{(CZH5)3SiO} 3,13(30), {(CH,),Sio} 4Ti(31), {(CH3)3SiO} 3Sb(35).

Tatlock and Rochow (46) reported the synthesis of {(CH3)3SiO}ZSn,
(CH3)3SiO 48:1, and {(CH,),Sio:)zsn(CH3)z, but did not characterize the

18

product outside of recording the infrared spectra. Since this original
synthesis, several investigators have prepared various other stannosiloxane
compounds using the silanolate condensation reaction. Table III is a
tabulation of known stannosiloxane compounds. Few analytical data have
been reported for these and nothing is known about the reactions involved.

in their preparation.

19

Table HI. Tabulation of Known Stannosiloxane Compounds.

 

 

 

 

Compound m. p. b. p. Reference
1°C. ) (°c .1
(CH3)3SiOSn(CH3)3 -59 141 32
[(CI-L3)SSiQ]z$n(QH3)z 48 160(decomp.) 16, 32
[(CH3)§SiOLSn(QH3) 34 155(decomp.) 32
[(csz),SiO],sn --- 200-2/4 mm 33
¢,Siosn¢, 138- 139 - - 17
(¢3SiO)ZSn(CH3)Z 166- 167 -- 13
154-155 —- 34

(¢)3Si0)4Sn 322(decomp.) -- 33

 

II. EXPERIMENTAL

A. Preparation of Reactants for Stannosiloxane Formation

1. Hexaphenyldisiloxane: ¢3SiOSi¢3 was prepared by the base

 

catalyzed condensation of HOSi¢3 in absolute ethanol. For example, a
30. 0 g (0. 109 mole) sample of HOSi¢3 was refluxed with 0. 28 g sodium
hydroxide in 150 ml of absolute ethanol for 12 hours. The ¢3SiQSi¢5
precipitated during the reaction and after standing overnight at room
temperature. The yield of crude product (m.p. 217-2200C) was 23.0 g
(79. 1%). After recrystallization from cyclohexane, the melting point
increased to 224-2250C. All reported melting points are uncorrected

and determined by the capillary method.

2. Sodium Triphenylsilanolate: NaOSi¢3 was initially prepared

 

by the reaction of HOSi¢3 (Dow-Corning Purified Grade) with excess
sodium metal in anhydrous ether (27). However, soluble products with
neutralization equivalents corresponding to a purity greater than 95%

1 could not be consistently obtained. A much more satisfactory method
for NaOSi¢3 preparation is due to Hyde, e_t ’11. , which involves the
cleavage of ¢3SiOSi¢3 with stoichiometric amounts of sodium hydroxide

(36. 37).

(1135105413 + ZNaOH __. 2Na051¢3 + H20

The equilibrium is forced to the right by removing the water formed from
the reaction mixture.

In a typical preparation, 1. 94 g. of sodium hydroxide (0. 0485 mole)
was pulverized in a dry box (38) and dissolved in 20 ml. of methanol.
A stoichiometric amount of ¢3SiOSi¢3 (12. 93 g. , 0.0243 mole) was added
to this solution together with 10 ml. of isopropanol. The slurry was

heated to reflux temperature and 15 ml. of toluene added. Approximately

20

21

15 ml of solvent were slowly removed from the refluxing mixture via

a Dean-Stark trap before 10 more ml of isopropanol were added to the
reaction mixture. This concentration and addition procedure was
repeated, usually two or three times, until a clear or slightly hazy solu-
tion was obtained. At this point, 5 m1 isopropanol and 5 ml of toluene
were added and approximately 15 ml of solvent once again removed.
Finally 5 ml toluene were added and the now viscous solution concen-
trated another 5 or 7 ml. The reaction mixture was then cooled to room
temperature. After overnight crystallization, the NaOSi¢3 crystals were
isolated by filtration and dried approximately 12 hours 111 mat 50-550C.
The yield was 7. 63 g (56. 3%). The product was readily soluble in organic
solvents. Its neutralization equivalent was determined by titration in
absolute ethanol to a bromthymol blue end point. Calculated neutrali-
zation equivalent: 298. 7. Found neutralization equivalent: 307. 9, 308. 0.
-A11 silanolates used in this study were 96 to 98% pure as determined by
neutralization equivalents .

3. Sodium Diphenylsilanediolate. (NaO)zSi¢z was prepared by a
slight modification of Hyde's technique for synthesizing NaOSi¢3 (36, 37).
A 1.900 g (0.0475 mole) sample of powdered sodium hydroxide was
dissolved in 20 ml of methanol. A stoichiometric amount of (HO)zSi¢z
(5.154 g, 0. 0238 mole) and 10 ml of isopropanol were added to this
solution. . The slurry was raised to reflux temperature and 15 ml of
toluene added. Thirty-three ml of solvent were then removed via a
-Dean-Stark trap over a three to four hour period before the clear reaction
mixture was cooled to room temperature. After six hours, the (NaO)zSi¢z
crystals were removed by filtration and dried approximately 12 hours

_i_n vacuo at 65°C. The yield was 1. 126 g (18. 2%). .Neutralization equi-

 

valents were determined in absolute ethanol as for NaOSi§D3. Calculated

neutralization equivalent: 260. . Found neutralization equivalent: 265. 8,

266.6.

22

Gibbs and co-workers (39), synthesized (NaO)zSi¢z by slowly
adding an ether solution of (HO)zSi¢z to a sodium dispersion, and removing
the insoluble product by filtration. No solvent could be found for the
product. In contrast, the (NaO)zSi¢)z prepared in this study was soluble
in ethanol, but insoluble in benzene and toluene. It also dissolved in
unpurified dimethylformamide, but was insoluble in dimethylformamide

purified by the method of Thomas and Rochow (40).

4. Organotin Chlorides: All organotin chlorides were obtained from

 

Metal and Thermit Corporation. Triphenyltin chloride [4)3SnC1], diphenyltin
dichloride [¢zSnClz],and dimethyltin dichloride [(CH3)ZSnC12] were
recrystallized from toluene at -14OC. Dibenzyltin dichloride [(¢CH3)zSnClz]
was recrystallized from toluene at -40C while dibutyltin dichloride
[(n-C4H9)zSnC12] was recrystallized from 30-600 petroleum ether at -4OC.
Analytical data for all organotin chlorides used in this study are shown

in Table IV.

B. Preparation of Stannosiloxane Compounds

A series of stannosiloxane compounds was prepared by dissolving
NaOSi¢3 (0. 006-0. 010 mole) in 50 m1 of benzene and adding to this solution
a stoichiometric amount of the appropriate organotin chloride dissolved
in 10-15 ml benzene. A slight increase in temperature was always observed
indicating the condensation reaction is slightly exothermic. .After mixing,
the reaction mixture was allowed to stand five to fifteen minutes at room
temperature before the sodium chloride precipitate was removed by
filtration. The clear filtrate was taken to dryness in wand the product

recrystallized.

C . Solvent Purification

All solvents were distilled over calcium hydride through a 70 cm

column packed with 5 mm glass helices. Thiophene-free benzene, the

Table IV. Organotin Chloride Analytical Data

23

 

 

 

 

Analytical
r51. p. 70 Calculated % Found
Compound ( C.) C H Cl C H C1
¢,snc1 104.5-105.5 56.07 3.89 9.22 56.30 4.00 9.35
(CH3)zSnClZ 106-108 10.92 2.73 32.32 10.87 2.71 32.57
(i_i-c,,H.,)_,_snc1z 41-42 31.61 5.93 23.38 31.79 6.12 23.13
(fizsnc12 41.5—43.5 41.90 2.91 20.66 41.71 3.09 20.65
(dongzsnmz 44.96 4.28 19.00 45.06 4.30 18.75

 

24

solvent for stannosiloxane formation, was fractionally crystallized
before drying and distillation. The purified benzene, methanol, and
isopropanol were stored in a desiccator when not in use. The measured
refractive indices are compared with the literature values and shown

in Table V.

D. Infrared Spectra

A Perkin-Elmer Model 21 spectrophotometer equipped with
Connecticut Instrument cavity cells was used to obtain all infrared
spectra recorded during this study. The spectra were recorded at
concentrations below 2. 0% (by weight). Such low concentrations were
employed because of the intensity of the Si-O-Sn peak. Wright and
Hunter have attributed this in the case of siloxanes to the ionic nature

of the Si-O-Si bond (42).

E . Analytical

Carbon, hydrogen, and chlorine analyses were carried out by
Spang Microanalytical Laboratory, Ann Arbor, Michigan, and Micro-
tech Laboratories, Skokie, Illinois.

Combined oxides were determined by the method of Gilman and
King (43). The stannosiloxane compound (0. 3 to 0.4 g) was first treated
with a four percent solution of bromine incarbon tetrachloride. The
sample was then decomposed with a mixture of nitric, fuming nitric,
and sulfuric acids before being ignited to give a white SiOz and SnOz
residue. Silicon carbide formation was averted by igniting the decom-
posed sample slowly over a Bunsen burner and finally at 900°C in a
muffle furnace. The average results shown in Table VI agree within

:1; 1. 0% of the theoretical value.

25

Table V. Measured and Literature Refractive Indices for Solvents
'Used in the Study of Stannosiloxane Formation

 

 

Solvent

Measured n-’-

0

Literature n20 (41)

 

D D
Benzene 1.5008 1.5011
Toluene 1.4959 1. 4969
Cyclohexane 1. 4260 1. 4262
Methanol 1. 3289 1. 3288
Isopropanol 1.3768 1.3776

 

26

 

 

 

Table VI. Stannosiloxane Combined Oxides
Sample Sample Theo. oxide Found oxide Recovery
Weight (g) Weight (g) Weight (g) (96)
((1,5i05n4), 0.4084 0.1390 0.1377 101.1
0.4020 0.1357 0.1357 100.0
0.3705 0.1250 0.1247 99.7
0.3652 0.1232 0.1241 100.8
Av.= 100.4
(CH3)zSn(OSi¢3)Z 0. 3957 0.1531 0.1532 100. 0
0.4156 0.1610 0.1589 98.7
Av = 99.4
(n-c,H,)zsn(OSi¢,)z 0. 3571 0.1235 0.1222 99.0
0.3509 0.1214 0.1201 98.9
Av = 99.0
()>,,sn(05i<)>,)2 0-3591 0.1180 0.1175 99. 6
0.3776 0.1243 0.1233 99.2
Av = 99.4
(dongzsnmsubg, 0. 3475 0.1103 0.1102 100.0
0.3541 0.1124 0.1127 -100.1
Av.= 100.1

 

27

An attempt was made to carry out an independent tin analysis so
that specific silicon and tin values could be reported- Samples were
decomposed either by potassium iodate in sulfuric acid (12) or a mixture
of fuming nitric and sulfuric acid (44). - In both cases, the decomposed
sample was reduced over lead (45) for 3 to 4 hours, cooled in an ice
bath to approximately 40C, and then titrated with standard iodine in a
carbon dioxide atmosphere. - Standard precautions were taken to exclude
air from the reduced samples. Theoretical tin values were obtained
for several standard organotin samples (Table VII). However, the
stannosiloxanes gave erratic results which were consistently low as
shown in Table VII. This pronounced variation is believed due to occlusion

of small amounts of tin by the silica.

F. Stoichiometry

Figure 2 is a flow diagram of the procedure used to obtain stoichio-
metric and qualitative rate data for the formation of various stannosiloxane
compounds in benzene. With this technique a closed material balance
was established about a system containing semi-micro quantities of
reactants.

In order to obtain sufficient material for analysis, 5 ml. aliquots
(10 ml. for solutions below 0. 02 1:1.) of the reactant solutions were pipetted
into the reactor, a 50 m1. Erlenmeyer flask. The solutions ranged in
concentration from 0. 007 to 0. 05 11. After mixing, the reaction mixture
was kept in the reactor for a known period at room temperature before
being filtered. into a tared 50 or 125 m1. filter flask. The clear filtrate
was temporarily set aside.

All traces of benzene in the reactor and filter were removed by
drying at 900C. . Upon cooling, the reactor was extracted with 10 ml. of
distilled water. This solution was transferred to the filter and drawn

into a clean 125 m1. filter flask. A second extraction with 5 m1. of water

28

Table VII. Tin Analytical Data

W

 

Decomposition Theo. tin, Found tin, Recovery
Sample Method (milliequi- (milliequi- (%)
valents) valents)
(CH3)ZSDCIZ HNO3-stO4 1. 351 1. 374 100. 2
()5,an12 HNO3-HZSO4 0.999 1.020 100. 9
HNO3-HZSO4 1.063 1.074 101.0
4),,Siosncb, HNO3-HZSO4 1. 280 1. 171 91. 5
HNO3-HZSO4 1. 360 1. 324 97. 4
HNo,-H,_so, 0. 956 0.933 97.6
K103 1.042 1.027 98. 5
K103 1.220 1.183 97.0
K103 1.289 1.213 94.1
(2-C4H9)zSn(OSi¢3)z HNO3-HZSO4 0.891 0.868 97.4
Kio3 0. 928 0. 864 93. 0

(¢CH,)zsn(08i¢,),_ mo, 0. 840 0. 835 99.4

 

29

 

Naosiité3

 

R35I1C1 or stnCIZ Reactor

 

 

 

 

 

 

Dry, extract Eggndg Filtration of

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

with water reaction mixture
filtrate
Filtration of ‘——I- Drycombined
extraction ' filtrate in vacuo,
mixture '
filtrate
. Spectrophotometric
$01163 Chloride analysis of
analysis solid residue

 

 

 

 

 

 

 

A

Dry, extract
with benzene, filtrate
and filter ’ A

 

 

 

 

 

 

Figure 2. Flow diagram of stoichiometry procedure.

30

was used to insure complete removal of sodium chloride from the

reactor and filter. Chloride present in the combined water extract (15 ml.)
was determined by titration with 0. 0200 1_\I_ standard silver nitrate using
dichlorofluorescein as the indicator.

Several standard chloride analyses were carried out to establish
the effectiveness of the above extraction and titration method. . Known
amounts of reagent grade sodium chloride were pulverized, dried at
105°C, and suspended in the reactor with 10 ml. aliquots of benzene solu-
tions containing ¢3SiOSn¢3 or various organotin chlorides. The results
in Table VIII indicate that in the presence of ¢3SiOSn¢3 the average
recovery is 99. 0% with a standard deviation ( G’) of_+_ 0. 2%. The presence
of organotin chlorides did not appear to significantly affect chloride
recovery except in the case of (CH3)ZSnClz where the effect is small.

After the water extraction, the reactor and filter were dried at
1050C. Upon cooling, the reactor was extracted with 10 ml. benzene.
The extract was transferred to the filter and drawn into the tared 125
m1 filter flask containing the original benzene filtrate. A second 10 m1.
benzene extraction was used to assure complete removal of solids from
the walls of the reactor and filter. The combined benzene filtrate was
taken to constant weight i_n wand the weight of the solid residue
determined by difference. The amount of stannosiloxane compound
present in this residue was measured spectrophotometrically using the
characteristic Si-O-Sn peak for identification. : Reaction conditions and
material balance data obtained in this manner are tabulated in Appendix A.

Beer's Law is expressed as

X i i). .1. (12)

:P
u

x absorbancy at wavelength X; b = cell path length, cm;

concentration of i_th component in solution,, .M;

0
ll

molar absorptivity of _i_th component at )t, (l/M cm).

at

p

7"
ll

31

Table VIII. Standard Chloride Analyses

 

_ 4”

 

Theo. Sodium Recovered
Suspension chloride Sodium chloride Recovery
Medium“ (milliequi- ~ (milliequi- (%)
valents) valents)
20.0169 1_4 (5381054), 0.2649 0.2628 ‘ 99.2
0. 2495 0. 2474 99.1
0.1880 0.1864 99. 2
0.2513 0.2464 98.1
' 0.2444 0.2438 99.8
0.2821 0.2802 99.3
0.2325 0.2272 97.7
0.2154 0.2134 99.1
0. 2444 0. 2422 99.1
Average: 99. 0,
o’= i 0. 2%
0. 0127 M (2—C4H9)ZSnClz 0. 2239 0. 2230 99.
0. 2701 0. 2666 98.
0.2735 0.2786 101.
0. 0254 124. f¢,snc1 0. 2034 0.2030 99. 8

 

*
Ten ml. of solution used for each run.

32

Examination of infrared spectra of. the reactants and products established
that only the intense Si-O-Sn peak in the various stannosiloxane com-
pounds absorbed significantly in the 10. 3 to 10. 6 11 region. NaOSi433 has
a strong narrow peak at 10. 25 11, but this falls off sharply so that by

10. 40 (.1 it contributes negligible absorption over the concentration range
employed in this study. Hence, all 6.1 is in this region except that for

the Si-O-Sn peak are essentially zer; and equation 12 may be rewritten

for a given stannosiloxane compound as:

A = €~bc (13)
A = absorbancy of Si-O-Sn peak; b = cell path length, cm;
c = (concentration “of stannosiloxane compound, M; ‘5 =

molar absorptivity of Si-O-Sn peak, 1/1\_/I cm.

In order to carry out spectrophotometric determinations of various
stannosiloxanes, it was first necessary to prepare calibration plots for
each compound studied. Solutions of known concentrations ranging from
0. 0055 to 0.0091 M were prepared using purified (1)3SiOSn4D3, (2-C4H9)ZSn
(OSi433)z, and (CH3)ZSn(OSi4>3)z samples. Cyclohexane was used as the
solvent for ¢3SiOSn¢3 and (2-C4H9)ZSn(OSi¢3)2, while carbon disulfide
was employed for (CH3)zSn(OSi¢3)Z. The infrared spectra for each
solution were recorded on 10 x 10 cm chart paper. The operating pro-
cedure followed the Perkin-Elmer Instruction Manual (46). Scatter of
the data due to solvent evaporation from the cells was minimized by
sicanning only the region from 8. 5 to 11. 0 11. . Eight and one-half p. was
selected as the 100% transmission reference point because the the
stannosiloxanes have relatively little absorption at this point over the
concentration range used.

The calibration plots for ¢3SiOSn¢3, (_r_1_-C4H9)ZSn(OSi¢3)Z, and ’7 .
(CH3)ZSn(OSi¢3)2 are shown in Figures 3, 4, and 5. In all cases, Beer's
Law was obeyed over the concentration range investigated. The method
of least squares was used to draw the best straight line through the

experimental points .

Absorbancy

0.08

0.04

Figure 3. ¢3SiOSn¢3 Beer's Law Calibration Plot, C6le = Solvent.

33

 

 

I l

 

2.

0

420 6.0
Concn. x 103 (M)

8.

0

10.

0

 

Abs ovrbancy

0.32

0.28

0.. 24

0.20

0,16

0.12

0.08

0.04<

Figure .4. '(CH3‘)zSn(OS’i¢3)z“Beer's Law Calibration Plot,

34

 

 

J

 

 

2.0 4.0 ’ !
Concn. x 103 (119

C52. = Solvent.

'6‘. 0

8.0

110'.

0

Absorbancy

35

 

 

l l I l

 

 

2.0 ~4.0 6.0 8.0 10.
Concn. "x '103 (1\_/1)

Figure 5. (3C4H9)zSn(OSi¢3)z Beer's Law Calibration Plot,
C6H12 = Solvent.

36

The yield of a stannosiloxane compound in a given reaction
mixture was determined by preparing a solution (cyclohexane or carbon
disulfide) of the solid residue with a known concentration (co, M).
Absorbancy of the Si-O-Sn peak was then measured as outlined above
and the concentration (c,. M) of the stannosiloxane compound actually
present in the solution taken from the appropriate Beer's LaWtcalihjration

plot. The yield (%) was calculated from the relation
Y = (c/co) x10‘2 (14)

A, co, c, and Y values for each of the reaction mixtures investigated
are reported in Appendix B.

When equation 14 is used, it is assumed that complete material
balance closure had been achieved in the above stoichiometric procedure.
Representative data from Appendix A (Table IX) show the average closure
is 99.8%, G) = i 1.4%. and hence, the assumption is justified.

A measure of the error inherent in the spectrophotometric analysis
of stannosiloxane compounds is given by the scatter of the data about
the least squares Beer's Law line. Table X shows that a deviation from
this line of up to j; 3. 7% can occur for a given c/co value. However,
most errors fall within i 2. 0%. Much of this scatter is due to instrumental
errors as shown by the duplicate (n-C4H9)zSn(OSiO3)Z determinations

obtained for several reaction mixtures (Table XI).

G. Polymer Formation

Several stannosiloxane polymers were synthesized by the i_n 's_it_u
condensation of (NaO)zSi¢z with (3-C4H9)zSnClz. Specific reaction con-
ditions and material balances for each experiment are presented in
Table XII. The reaction procedure followed is outlined below.

In order to carry out a polymerization reaction, (NaO)zSi(t),_ was

prepared in a 100 m1 round bottom reactor by dissolving 0.04 to 0. 05

 

fia H "\b
:New .00 omnuo><

37

 

 

se.~oz momm.o zsdz.o NeNo.o smowso.o
se.zoz ~o-.o neaz.o osmo.o mawso.o
ms.aa ~o-.o smdz.o zsmo.o mamso.o
ov.ss eezz.o czon.o sszo.o cmo.o
sm.¢o osza.o azoz.o sszo.o s~o.o 1.
mm .2: 2.: .0 mm: .c 3.8 .o c8 .o aaosmsxameoév
me.ss msmm.o ossa~.o Hemo.o mswdo.o
sc.sa msmm.o HasN.o ~s~o.o mando.o
za.wa smez.o Hemz.o cszo.o smao.o
os.oo2 cmsa.o momz.o wszo.o_ smzo.o .
sz.sa cmsz.o memz.o mszo.o emzo.o Hosmmv
m~.ss, sasz.o stz.o Hemo.o msmso.o
om.oos asaz.o «msa.o ms~o.o msmso.o
em .8 :62 .o :2 .o 38 .0 names .o aaosmaxamov
33 Amy mpflom Amy @0995qu Amy pouo>ooom 315 .9500 opfiuofinu
canoe/000m adofiohooAH. vapwmom pflom opfiuoHAU 5560M “denuded quocdmno

 

mqu 0053mm 3:332 o>3mpaomoumom .5 3nt

38

Table X. Scatter of Data Used to Plot Beer's Law Calibration Curves

 

w

 

Run ‘ 60(M) C(M) c/co x 102
(n- C4H9)ZSn(OSi¢3)Z 0. 00756 0. 00758 100. 3
0.00506 0.00491 97.0
0.00556 0.00554 99.6
0.00587 0.00598 101.9
0.00600 0.00608 101.3
0.00631 0.00635 100.6
0.00646 0.00651 100.8
0.00690 0.00717 103.9
0.00705 0.00716 101.6
0.00756 0.00755 99.9
0.00790 0.00763 96.6
0.00805 0.00782 97.1
0.00671 0.00669 99.7
0.00738 0.00738 100.0
0.00802 0.00802 100.0
range 103.9 to 96.6%
4D3SiOSn¢3 ‘ 0.01012 0.01022 101.0
0.00784 0.00767 97.8
0.00893 0.00887 99.3
0.00719 0.00720 100.1
0.00836 0.00849 101.6
range 101.6 to 97.8%
(CH3),zSn(OSi¢3)z 0. 00790 0. 00797 100. 9
0.00687 0.00687 100.0
0.00667 0.00670 100.4
0.00598 0.00690 100.3
0.00501 0 97.7

. 00548

 

range 100.9 to 97.7%

 

39

Table XI. Duplicate (£1-C4H9)zSn(OSid>3)z Determinations

 

Run c.0021) . A c(L_/I) c/co x 102
21 0.00662 0.339 0.00641 96.8
0.00662 0.326 0.00617 93.2
22 0.00811 0.404 0.00762 94.0
0.00811 0.409 0.00771 95.1
24 0.00664 0.312 0.00590 88.9
0.00664 0.311 0.00589 88.7
26 0.00636 0.305 0.00578 90.9
0.00636 0.309 0.00584 91.8
29 0.00659 0.352 0.00666 101.1
0.00659 0.355 0.00670 101.7
30 0.00613 0.321 0.00608 99.2
0.00613 0.325 0.00615 100.3
38 0.00770 0.319 0.00610 79.2
0.00770 0.317 0.00590 76.6
39 0.00698 0.263 0.00497 71.2
0.00698 0.275 0.00519 74.4
40 0.00809 0.387 0.00730 90.2

(3

0.00809 0.390 0.00736 91.

 

40

 

 

v Kw H .vo . xsflou ow: come .0 >H 5m-
...... Rho 3.. Nwom ovmod HHH amt
N60 «1R0 daemon om oNNod E Sm:
N60 043 *1 mom ommod H 9m:
A§v pouch/ooh: gov >uo>ooou AOovouBMHomEou 7:85 08$ 73on popmlm magnum
amazon x15 H.308 opwnodfiu Swpom :oflomom doflomom ~HUnm~Aam¢Ow5
““3““ .Ilflll %

 

 

 

muogzom ocmxofiwmogfifim mo moflmfimmonm gem mm HOH damn oocgdm Eton—m2 . .HHK 3an

41

mole of powdered sodium hydroxide in 20 m1 of methanol. A stoichio-
metric amount of (HO)zSi¢z and 10 ml of isopropanol were added and
the slurry raised to reflux temperature. At this point 15 ml of toluene
were added. Twenty ml of solvent were then removed from the reflux-
ing reaction mixture via a Dean-Stark trap over a two hour period.
The clear silanolate solution was finally brought to the desired polymer-
ization temperature and a stoichiometric amount of (2-C4H9)zSnClz,
dissolved in 18 ml of toluene and 8 m1 of isopropanol, added. . It was
assumed that a 100% (NaO)zSi§Dz yield had been obtained. After the reaction,
the products were filtered into a tared filter flask and the clear filtrate
was temporarily set aside.

The reactor and filter were dried at 1100C before the former was
extracted with 25 ml of water. This solution was transferred to the
filter and drawn into a clean 125 ml filter flask. A second 25 ml water
extraction was made before the reactor and filter were dried at 1100 C
and then extracted twice with 25 ml aliquots of hot toluene.

After drying at 110°C, a third extraction with 25 ml of water was
made. . The combined water extracts were transferred quantitatively to
a 100 (or 250) m1 volumetric flask and diluted to proper volume. Three
five-ml aliquots were pipetted from the volumetric and each titrated
with 0. 1 11 standard silver nitrate using dichlorofluorescein as the indicator.
Total chloride recovery was calculated from the average of these runs.
The 3 runs deviated from this mean by (:0. 5%. . The clear filtrates were
combined and taken to constant weight in min order to determine the
yield of bulk polymer. A small amount of solid material (~1-2% of total
solids) insoluble in hot toluene was discarded.

Polymers (Z—Bu‘II and O-Bu III were prepared exactly by the
technique outlined above. . A slight modification was made for polymer
¢-Bu IV. After the original reaction mixture was filtered, the clear

filtrate was concentrated to half its original volume and allowed to stand

42

overnight. The precipitate which formed was removed by filtration and
the c1ear filtrate taken to dryness i_n 3323. Thus, the 4).-Bu 1v bulk
polymer was divided into two fractions. In the case of polymer 4D-Bu I
the (NaO)zSi¢z preparation was modified by adding 10 m1 of benzene

along with the 15 m1, of toluene. Approximately 30 m1. of solvent were
slowly removed from the refluxing reaction mixture via the Dean-Stark
trap before 10 ml. of isopropanol were added. Another 15 m1. of solvent
were removed and 5 ml. of toluene added. The (_r_1-C4H9)zSnClz was then

added and the previously outlined procedure was followed.

H. Polymer Fractionation

Polymers prepared by the i_n s_i_t_u_ condensation of (NaO)zSi¢z with
(2-C4H9)ZSnClz were extracted with methanol and thereby separated into a
resilient solid and a viscous liquid fraction. The room temperature
fractionation was carried out by extracting 4. 1 to 6. 9 g of polymer with
25 ml of methanol in a 125 ml Erlenmeyer. Large clumps of the bulk
polymer were broken up and the extraction mixture agitated occasionally.
After 48 hours, a fine solid material was removed by filtration. The
clear filtrate and the solid fraction retained on the filter were dried to
constant weight i_r_1 3m Material balances for each of the fractionation

experiments are shown in Table XIII.

I. Solubility Studies

Solubility characteristics of the bulk and fractionated stannosiloxane
polymers were determined in a number of solvents at room temperature.
A 0.01 .1; 0.001 g. polymer sample was placed in a 10 x 75 mm. test
tube together with one ml. of solvent. After standing several days with
occasional agitation, the sample was examined and classified as soluble
(clear solution), partly soluble (> 50%. soluble), slightly soluble (< 50%
soluble), and insoluble (no visible dissolution). Results are shown in

Appendix C .

43

Table XIII. Material Balance Data for Stannosiloxane Polymer
Fractionations

w

 

Sample Bulk po1ymer Methanol-insoluble Methanol-soluble
sample (g) 3- fraction (%) fraction (%)
-BuI 4.10 78.5 21.5
-Bu 11 5.72 71.5 28.5

—Bu III 6.89 54.6 45.4

 

44

None of the bulk polymers or methanol-insoluble fractions were
soluble at room temperature in the solvents investigated. However, all
formed clear solutions in hot (110°C) _r_1-amyl acetate and no precipitation
occurred upon cooling to room temperature. The methanol-soluble
fractions readily dissolved in every solvent investigated except isopropanol
and acetonitrile.

The bulk stannosiloxane polymers have peculiar solubility character-
istics. The polymers were soluble in the original reaction mixture, but
after being dried, they were not soluble in a wide variety of solvents at
room temperature. The resilient solids isolated by methanol extraction
of the bulk polymers were insoluble at room temperature in all solvents

examined. This insolubility may be due to crystallization.

III. RESULTS AND DISCUSSION

A. Stannosiloxane Synthesis and Characterization

A series of stannosiloxane compounds has been synthesized by the
procedure outlined in the experimental section of this dissertation.
Pertinent data are listed in Table XIV. Only carbon-hydrogen analyses
are reported here; combined Si-Sn oxides have been reported in Table VI.
All of the compounds are white, crystalline solids with low to moderate
melting points. Their solubility characteristics vary considerably.
(_r_1-C4H9)zSn(OSi(b3)2 is very soluble in most organic solvents at room
temperature. ¢3SiOSn¢3, (CH3)ZSn(OSi¢3)z, and (OCH3)ZSn(OSiq)3)2 are
only moderately soluble at room temperature, but dissolve readily when
heated. ¢zSn(OSi¢3)z is relatively insoluble in most solvents even when
heated, but does dissolve in hot g-xylene. All of these compounds are
sparsely soluble in alcohols. Because of their solubility characteristics,
the stannosiloxanes were relatively easy to purify. Diethyl ether (~4OC)
was used for recrystallization of ¢3SiOSn¢3 while g-xylene (-14OC) was
employed for ¢ZSn(OSi¢3)2. (_r_1_-C4H9)ZSn(OSi¢3)Z,. (CH3)zSn(OSi¢3)Z and
(OCH3)zSn(OSi4)3)Z were recrystallized from n-heptane (-4OC). No attempt
was made to increase the reported yields by successive concentrations
and crystallizations of the mother liquors.

The compounds, (B-C4H9)zSn(OSi¢3)z and (OCH3)zSn(OSi4>3)z, are
new and have not been reported previously. The ¢3SiOSh¢3 preparedtin
this investigation appears to be the same material reported by Post and
Papetti (17). Likewise, the (CH3)ZSn(OSid)3)2 reported in the present study
appears to be similar to the material obtained by Chamberland from the
same reaction (34). Chamberland gave no analytical data, but reported

only the melting point. Koenig (4) claimed to have prepared

45

 

46

38 3.8. $8 3.2 one. 37%; $256 Nxaesmovsmeqmoev

 

«on 3.3 $6 8.2. New 903-83; mmsood 248954
$8 8.3 see 3.3 miss Eds Esood exaehmovsmaxamdhv
Nan $58 88 3.3 sec 37.6.2 wosood Namehmoesmeamov
$4. 21$ 84. miss 5% 03-82 2256 aesmoamas
8an save seam $30 3236; Goings sons usaoaasoo
osaoh - . sousgoano €302
3033.924 mo ago:

 

E

ocousom :H £03000“ GowummcmVpGoU oudaocmfim >3 poummohnm mUCSOQEoU oamxodmoccmuw (SN 0305

47

(CH3)ZSn(OSi¢3)z (m.p. 165-166°C) by cohydrolysis of H0510, with
dimethyltin oxide, but no carbon-hydrogen analyses were reported and
his Si-Sn data differ from the theoretical value by 34%. The Si-Sn
analytical procedure was not described. Koenig and Crain (13)
attempted to prepare 4);Sn(OSi4)3)z by cohydrolysis of HOSi¢3 with
(¢zSnO)x, but a product correSponding to Sn(OSi¢3)4 was isolated. Thus,
it appears the cohydrolysis procedure is not a satisfactory method for
preparing well-defined stannosiloxane compounds. In contrast, such
compounds have been obtained in good yield by the condensation of
NaOSi¢3 with various organotin chlorides. This synthesis is rapid,
easy to carry out, and employs mild reaction conditions so that side
reactions which might become significant at elevated temperatures are
minimized. These factors made the silanolate condensation reaction
ideally suited to the synthesis of well-defined stannosiloxane compounds.

Infrared spectra for the stannosiloxane compounds prepared. in this
study are shown in Appendix D while characteristic peaks are listed in
Table XV. As in the case of organosiloxanes (42, 47,’ 48), the most
interesting feature of these spectra is the great intensity of the bands
lying in the 8 to 14 (.1 region. The intensity may be due at least in part
to the partial ionic character of the Si-O-Sn, Si-R, and Sn—R bonds.
This explanation is analogous to that offered by Wright and Hunter for
the organosiloxanes (42).

Stannosiloxane compounds are readily identified by a broad and
intense absorption peak in the 10. 3 to 10. 7 p. region which is attributed
to the Si-O-Sn bond. This peak lies at 10. 3-10. 4 u in ¢3SiOSn¢3, but
is shifted to approximately 10.7 (.1 in (CH3)zSn(OSi¢3)z, (szn(OSi(()3)z, and
(¢CH3)zSn(OSi¢3)z. It lies at 10.6 n in (n-c,H,)ZSn(OSi¢,)2. Other
identifying peaks used in this study are those lying at 9.0, 9.7, 13. 5-13.6,
and 14. 2 11. These are bands characteristic of phenyl on silicon (47).
The peak lying at 13. 8 u in ¢3SiOSn¢3 and OZSn(OSi(I)3)2 is probably due
to the ¢-Sn linkage.

48

.9830 u Amv “Smooch n A65 3003

"momfimqgcm x00nm
ow

 

34.3 .3~.3 .33.: .0392 .3e.2 .354. .3046 enema
34.3 .30.: .33.? .353 .354. .305 .336

34.24.: 3...: .392 .3... .53.. .3... .3...

axaeamovsmaxamoov
2.000540

enshmovsmaxameo -5

 

34.3 603.2 .0312 .352 .3io .3m.m axaeamovsmaxamov

34.3 .3... .3 .
30.2 .2392 .345792 .3392 .35“. .398 .304. .....3m..m sosmoamno
1 5030004 x00nm 0559800

 

0p§520 c0900 E 0000033000005 mo 0.000% ”539800344 00H0HHGH .>X 030B

49

Relatively little is known about the general stability of stannosiloxane
compounds. Schmidtbauer and Schmidt have reported (CH3)3SiOSn(CH3)3
was hydrolytically unstable (32) while Koenig (4) found stannosiloxanes
were readily cleaved by dilute mineral acids. Although no Specific
attempt was made to examine stability in the present study, it was possible
to show ¢3Siosn¢3, (3-C4H9)zSn(OSi(P3)z, and (CH3)ZSn(OSi¢3)2 are not
highly labile. Several stoichiometric runs involving these compounds
were allowed to stand for extended periods at room temperature before the
reaction products were isolated and characterized. The yields from these
runs (Table XVI) are essentially quantitative. There is'no evidence the
stannosiloxane compounds are unstable under the reaction conditions
employed in this investigation. The high yields of purified products
obtained by recrystallizations from hot n-heptane and g-xylene are further

evidence the stannosiloxanes are stable compounds under mild conditions.

B. Stoichiometry and Rate of Stannosiloxane Formation

Stoichiometric results for the reaction of NaOSi¢3 with ¢3SnC1 in
benzene are listed in Table XVII. Similar data for the reaction of NaOSi¢3
with (2-C4H9)ZSnC12 and (CH3)ZSnClz in benzene are shown in Tables XVIII
and XIX. Two to five runs were usually made at each initial reactant
concentration. Equivalent proportions of reactants were always used.

The, reactions were carried out at room temperature.
The ¢3SiOSn¢3, (2-C4H9)zSn(OSi¢3)z, and (CH3)ZSn(OSi¢3)2 yields

are consistent with the over-all equations:

NaOSi¢3 +¢3sn01 ———> ¢3Siosn¢3 + NaCl
2Na051¢3 + (3-04H9)25nc12——> (3-C4H9)28n(05i¢3)z +ZNaC1

ZNaOSi¢3 + (CH3)zSnC12—‘—> (CH3)zsn(OSi¢3)z+ ZNaCI

50

Table XVI. Stannosiloxane Stability in Benzene at Room Temperature

 

 

Initial Na0$i¢3 Reaction Stannosiloxane
Run . Concentration, time yield
Number (_I\_I) (min. ) (%)
¢35105n¢3
7‘ 0.0127 2796 103.2

10 0.0134 1447 97.2

18 0. 0345 2994 98. 5

20 0. 0490 2225 96. 3
(2-0.119)an(051493)Z

23 0.0127 2978 96.6

26 0.0134 8878 91.4

30 0.0147 1207 99.8

33 0.0260 2153 93.7
(CH3)zSn(OSi¢3)Z

59 0.0127 2887 88.5

62 0.0134 8811 82.7

69 0. 0260 2795 96.4

78 0. 0490 2190 97.0

 

51

Table XVII, 413SiOSn413 Stoichiometric Results

 

 

Initial NaOSi¢3 Sodium
and 95351101 chloride ¢3SiOSn¢3
Run concns . yield yield
Number (31) (%) (%)

1 0.00783 95.3 97.5
2 0.00783 95.9 98.7
3 0.00783 95.1 96.3
4 0.00783 96.3 99.1
5 0.0127 96.2 96.7
6 0.0127 99.6 100.2
7 0.0127 98.0 103.2
8 0.0134 93.4 93.6
9 0.0134 94.3 99.3
10 0.0134 94.9 97.2
11 0.0299 96.3 97.9
12 0.0299 97.0 98.2
13 0.0299 96.3 98.1
14 0.0299 96.8 98.5
15 0.0299 97.2 101.2
16 0.0345 97.4 99.4
17 0.0345 97.6 98.5
18 0.0345 96.0 98.7
19 0.0490 95.2 96.9
20 0.0490 94.7 96.3

 

 

av. = 96.2, av. =98.3,
6:3; 1.4% Q"=_+_z.o%

 

52

Table XVIII. (2-C4H9)ZSn(OSi¢3)Z Stoichiometric Results

 

 

 

= 1: = =
Initial
. Naosub, and Sodium
Run (2-C4H9)ZSnC1Z chloride (2-C4H9)zSn
Number concns. yield, (08:10,)z yield,
(8. equiv. /1) (%) 1%)
21 0.0127 97.6 95.4
22 0.0127 --- 97.1
23 0.0127 98.0 96.6
24 0.0134 -—- 88.8
25 0.0134 94.6 90.2
26 0.0134 95.6 91.4
27 0.0147 95.2 96.9
28 0.0147 100.9 96.2
29 0.0147 99.8 101.4
30 0.0147 99.7 99.8
31 0.026 95.8 89.5
32 0.026 96.8 89.3
33 0.026 96.7 93.6
38 0.0490 94.6 76.1
39 0.0490 95.1 71.4
40 0.0490 94.9 90.6
41 0.0491 96.8 96.2
42 0.0492 97.3 97.1
43 0.0492 97.9 100.0
44 0.0492 98.3 97.5
av. = 97.0, av.= 92.8,
G’=.t 1.9% 6:: 7.6%

 

 

 

53

Table XIX. (CH3)zsn(08i<f>3)z Stoichiometric Results

 

 

Initial ‘ - Sodium

 

NaOSi¢3 and Chloride (CH3)zSn(OSi¢3)Z
Run (CH3)7_SnCIZ Concns. yield, yield.
Number (g. equiv./l) (%) ' (%)
57 0. 0127 ' -- 88. 5
60 0. 0134 -- 93. 2
61 0.0134 95.41 94.2
62 0.0134 ' -- 82.7
64 0.0136 99.01 97.8
65 0.0136 98.63 98.7
66 0.0136 90.91 87.1
68 0.0260 96. 92 96.4
69 0.0260 97.23 96.4
78 0.0490 94.67 97.0

av. = 93.2, G’=iS.3

 

54'

There is no evidence significant side reactions occurred. Although a
few relatively low (3-C4H9)zSn(OSi¢3)z and (CH3)zSn(OSi¢3)z yields were
observed, a frequency plot of data from 42 runs (Figure 6) shows the
results cluster about a central value of 96%. Since the NaOSi¢3 used
was only 96-98% pure, this is the expected range for total consumption
of reactants. A reasonably symmetrical spread exists around the
central value with small deviations occurring more frequently than large
ones. Most of the relatively small deviations can be attributed to errors
in the Spectrophotometric procedure. A few runs (~20%) had yields
of less than 90%. Thus, a systematic error appears to be responsible
for the larger variations in (_r_1—C4H9)zSn(OSi¢3)z yields. . In view of the
experimental procedure employed in this study, it is postulated that the
error results from stoichiometric imbalance of reactants.

The above hypothesis was tested by making several runs in
which 0. 859 milliequivalent of NaOSi¢3 was reacted with 0. 926 milli-
equivalent of (2-C4H9)ZSnC12. Results in Table XX show the average
sodium chloride recovery was 90. 5% while the average (2-C4H9)zSn(OSi¢3)Z
yield was 85. 3%.“ These yields are based on the (_r_1_-C4H9)ZSnC1z added
and are what would be expected if (2-C4H9)zSn(OSi¢3)z is formed by a two-

step reaction as

(_r_1_-C4H9)zSnClz + Na05i<f>3 ———9 (2-C4H9).zSn(OSi¢3)C1 + NaCl
(2-C4H9)25n(051¢3)C1 '1' NaOSi¢3 __’ (2' C4H9)zsn(05i¢3)z ‘1" NaCI

The relation between R'zSn(OSiR3)2 yield and NaOSiR3 deficiency for such
a process (Figure 7) shows that an 85. 7% (RESn(OSiR3)Z yield should be
obtained for a 7. 2% NaOSiR3 deficiency. The experimentally determined
85. 3% (2-C4H9)zSn(OSi4)3)z yield is in excellent agreement with the
theoretical value. . Likewise, the 90.5% sodium chloride recovery agrees

well with the expected value of 92. 8%.

55

 

 

 

 

 

 

 

 

 

28-

24,.

20-

.t.’ 161...
C1
:1
DD
.5
a.
:5
O
H

U 12—
H
O)
Q.
>~
U
G
3

3 8*-
H
Li.

4—

o J ' I I I l .1

 

106 98 90 82 74 66
Perc ent ‘RZSn(OSi(P3')z .Yield

Figure 6. Histogram Showing Frequency of Experimentally
Determined RZSn(OSi¢3)z Yields.

Table XX. Stoichiometric Results for (_r_1-C4H9)zSn(OSi4)3)z Formed Using
A Known Imbalance of Reactants

 

 

Initial NaOSi¢3 Initial (3—C4H9)ZSnC12 Sodium chloride (2-C4H9)zSn

 

(concn. concn. yield (OSi¢3)z yield,
(g. equiv. /1) (g. equiv. /1) (26) . (%)

0.0430 0.0463 89.4 84.0

0.0430 0.0463 90.3 87.1

0.0430 0.0463 90.6 83.5

0.0430 0.0463 91.6 86.6

 

av. =90.5 av. = 85.3

 

Yield of Stannosiloxane Compound, 70

57

100.0

80 . 0 —
R3SiOSnR3'

60.0—

R'zsn( OSiR3)z
40 . 0 —

20.0 -

 

0 I I l I J—
10.0 20.0 30.0 40.0 50.0

NaOSiR3 Deficiency (%)

Figure 7. Theoretical Stannosiloxane Yield, %, vs. NaOSiR3
Deficiency, %, for R3SiOSnR3'andRz'Sn(OSiR3)z.

58

, Included in Figure 7 is a plot of R3SiOSnR3' yield versus NaOSiR3
deficiency. This shows that Rz'Sn(OSiR3)z yields are twice as sensitive
to NaOSiR3 deficiencies as R3SiOSnR3' yields. . A five percent NaOSi433
deficiency will cause a ten percent decrease in RzSn(OSi¢3)z yield and
only a five percent drop in ¢3SiOSn4>3 yield. . From Figure 6, it can be
seen that errors of this magnitude could account for an overwhelming
majority of the reported low yields. Since such errors could easily
arise in the preparation and handling of the reactant solutions and by
variation in NaOSi¢3 purity (96-98%), it was concluded stoichiometric
imbalance of the reactants caused the major observed variations in
(_n-C4H9)zSn(OSi¢3)Z and (CH3)ZSn(OSi¢3)z yields. Particularly poor runs
38, 39, and 62 are attributed to gross procedural errors. — If these runs
are neglected, the standard deviation for (2-C4H9)zSn(OSi(I)3)Z and
(CH3)ZSn(OSi¢3)Z, :1; 4.6 and i 4. 1% respectively, are approximately'twice
as great as that for ¢3Si05n¢3 (: 2. 0%). This is further evidence that
variation of stannosiloxane yields is due to stoichiometric imbalance of
reactants.

The stoichiometric data discussed to this point were obtained for
NaOSi¢3-organotin chloride reactions carried out in reagent grade benzene
which had been fractionally crystallized and distilled over calcium hydride.
In order to determine if such highly purified benzene was actually
necessary, several runs were carried out in untreated reagent grade
benzene. Results in Table XXI show that quantitative yields were obtained
in all cases. This indicates that extremely pure benzene is not required
for quantitative stannosiloxane formation.

The data in Appendix E show that sodium chloride formation is rapid
when stoichiometric amounts of NaOSi¢>3 are mixed in benzene with
¢3SnCl, (CH3)ZSnClZ,. (B-C4H9)zSnC12, or ¢ZSnC12. . Representative data
have been plotted in Figure 8. Since the NaOSi¢3 used in this study was

96-98% pure, the observed sodium chloride recoveries correspond to

59

Table XXI. Stoichiometric Results for (2-C4H9)ZSn(OSi¢3)Z Formed in
Reagent Grade Benzene

 

 

 

Initial Sodium
reactant chloride (31: C4H9)zSn(OSi¢3)z
Run concn . yi eld, yield,
Number (g. equiv. /l) (%) (%)
45 0.0302 97.2 95.3
46 0.0302 98. 2 100.4
47 0.0302 98.2 97.5
48 0.0302 98.4 98.4
49 0.0359 92.6 94.9
50 0.0359 96.7 95.5
51 0.0359 96.6 96.7
52 0.0359 96.9 94.7

 

60

 

 

1. 00...
(D
0. 98— (D (D Q
'o 6
H O
x 8 $
28 Ce
3 o. 96-e
o
>
8
<0 6
as
'8
,1 0. 94-
.9.
,0
U
U) .0. 92
o, 901 I ! I I I
10 20 30 40 50
Reaction Time (Minutes)
Figure 8. Sodium Chloride Recovery vs. Reaction Time 9, 0.0299g.

equivalent/l NaOSi¢3 and ¢3SnC1; 0, 0.0492 g. equivalent/l
NaOSi¢3 and (n-0,H,)zanlz; 0, 0.0299 g. equivalent/1
Naosio3 and (CH3)zSnClz; 0, 0.0299 g. equivalent/1 Na08i¢,
and Q;SnC1z.

6.1;

total NaOSi¢3 consumption. Variation in the sodium chloride yields is
within experimental error.

The rapid rate of sodium chloride formation found in this study
indicates NaOSidh-organotin chloride condensation processes carried
out at room temperature in benzene are extremely fast . Such reactions
are complete in less than a minuteso quantitative rate data could not be
obtained. Although steric and thermal factors may slow the rate of
reaction, rapid reaction techniques originally developed by Hartridge and
Roughton (49) seem to offer the best approach to obtaining quantitative
rate data.

Since quantitative rate measurements were not made, elucidation
of a reaction mechanism for the Na-OSiC’h-organotin chloride condensation
process in benzene is not possible. However, certain inferences may be
drawn from the stoichiometric and qualitative rate data which were
obtained. One of these is that pure NaOSi¢3 will react in benzene with
a stoichiometric amount of R3SnCl or RzSnClz to give quantitative yields
of the corresponding stannosiloxane compound. Side reactions do not
occur under the mild reaction conditions employed in this study. This is
significant since conventional polycondensation processes, as discussed
in the introduction, must utilize reactions which are free of undesirable
side reactions. Extrapolation of the above results to bifunctional systems
which are capable of polymer formation might seem precarious, but
syntheses of organic condensation. polymers are based on simple
reactions of multifunctional groups. The fact that these reactions were
free of secondary processes was established from studies of the corres-
ponding monofunctional systems. The lack of such knowledge has been
one of the handicaps in syntheses of organometallosiloxane polymers.
Results of the present investigation indicate the reaction of pure (NaO)éSi(Pz
with RzSnClz in an aprotic solvent should yield stannosiloxane polymers

having a regular alternating (Si-O-Sn-O) structure. 7 If cyclic compounds

62

are formed, it should be possible to isolate and identify these. The
important point is Si-O-Si or Sn-O—Sn bonds should not be formed by
various side reactions under normal reaction conditions if pure reactants
are employed.

It is significant that the reaction of NaOSi4>3 with organotin chlorides
in benzene is extremely fast and irreversible. The sodium chloride by-
product precipitates immediately thereby forcing the reaction to com-
pletion. This distinguishes such reactions from those commonly employed
to synthesize organic condensation polymers. Organic polycondensation
processes, in general, have utilized slow, reversible organic reactions
which require elevated temperatures and reduced pressure for high
polymer formation. A notable exception is the interfacial polycondensation
technique which has recently been developed (50, 51). In this method a
fast, irreversible polymerization of two fast-reacting intermediates
occurs near the interface between phases of a heterogeneous liquid system
(50, 51). Characteristic features of ordinary melt and interfacial poly-
condensation processes are compared in Table XXII (50). The interfacial
method is obviously the simplest since impurities, stoichiometric
imbalance of reactants, and even side reactions do not necessarily limit
polymer formation. All of these factors drastically limit the molecular
weight of polymers formed by the melt condensation process.

Since reactions ordinarily used in interfacial polycondensations are
similar to the NaOSi¢3-organotin chloride reaction (fast and irreversible),
it appears that regular alternating stannosiloxane polymers could be
prepared by the interfacial method. , The only difficulty seems to be lack
of a suitable heterogeneous solvent system since relatively pure

(NaO)zSi4)z has already been isolated.

63

Table XXII. Comparison of Interfacial and Melt Polycondensation

 

Interfacial

Melt

 

Intermediates
Purity
Balanc e

Stability to heat

Polymerization conditions
Time
Temperature
Pressure

Equipment

Products
Yield
Structure

moderate to high (>90%)
unnecessary

unnecessary

s eve ral minutes
0-4o°c
atmospheric

simple, open

low to high

unlimited

high (>98%)
necessary

necessary

several hours
>zoo°c
high and low

special, sealed

high

limited by
stability to
heat and
fulsibility

 

64

C . Polymer Formation

Several stannosiloxane polymers were synthesized by the i_n s23,
condensation of (NaO)ZSi((>z with (2-C4H9)zSnClz. The polymers have not
been fully characterized as to their hydrolytic and thermal stability,
molecular weight, or various other physical properties. < Attention was
focused on determining if the silanolate-organotin chloride reaction
involving bifunctional reactants proceeded in the same manner as the
monofunctional systems. The polymerizations were carried out under a
variety of reaction conditions. Complete material balances were estab-
lished about each reaction system and the sodium chloride recovery data
taken as a measure of the extent of reaction.

The results (Table XII) show there is essentially no difference in
the extent of reaction of each polymerization process. Similar sodium
chloride recoveries were obtained for the polymerization carried out at
-l4oC for two hours and that carried out at reflux temperature for 24'
hours. These dataimply the i_n _s_i_tl_i_ reaction of (NaO)zSi¢z with
(2-C4H9)zSnClz is rapid .

Undesirable side reactions in the i_n §i_t1_i (NaO)ZSi4)z-(2-C4H9)ZSnClz
polycondensation process were detected by fractionating the bulk _
stannosiloxane polymers into a methanol-soluble and methanol-insoluble
fractions as outlined previously. Carbon, hydrogen, and chlorine.
analyses for each methanol-insoluble fraction are shown in Table XXIII.
Calculated carbon-hydrogen‘values are for the [O-Si¢z-O-Sn(_n_-C4H§)z']
unit. .

Although none of the fractions had carbon-hydrogen analyses which
agreed well with the calculated values, carbon-hydrogen analyses for the
¢-Bu I and ¢-Bu II fractions are extremely close. This indicates, but
does not prove, the ¢-Bu I and (IL-Bu II polymerizations were relatively
free of side reactions. In contrast, carbon-hydrogen analyses for the

¢-Bu-III and ¢-Bu .IV fractions (differ significantly from the calculated

65

Table XXIII. Analytical Data for Methanol-Insoluble Stannosiloxane
Polymer Fractions

 

 

 

 

 

 

== =: =
Analytical

Calculated . Found

Polymer C(%) H(%) C(%) H(%) c1(%)

—BuI 53.73 6.27 53.33 6.20 0.0
53.22 6.24

-Bu II 53.73 6.27 53.43 6.36 0.0
53.37 6.25

-Bu III 53.73 6.27 48.42 6.60 0.0
48.62 6.52

-Bu IV 53.73 6.27 44.64 6.86 0.0
44.72 6.92

 

66

values. In these cases the Si-Sn ratio is < 1 indicating significant side
reactions have occurred. It is interesting to note reaction times for
(P-Bu III and (ID-BuiIV are much greater than those for (b-Buxl and (b-Bull.

Infrared spectra for all methanol-soluble fractions were essentially
identical (a representative spectra is shown in Appendix D). Strong
absorption from 9. 0 to 10. 5 p. is due to several overlapping peaks. . The
peak at 9. O p is the Si-O stretch bond while that at 10. 4 p. is attributed to
the Si—O-Sn bohd. Absorption at 9. 3-9. 6 uis probably due‘to the Si-O-Si
bond.

The above results imply the i_n sit-3 (NaO)zSi4)z-(2-C4H9)zSnClz
polymerization process is rapid and yields polymers which contain Si-O-Sn
bonds. ,However, further. conclusions about these polymers can not be
made at present since they have not been properly characterized. . Such
characterization will be the subject of a future study. In addition it will
be necessary to perform the polymerization with highly purified monomers
under conditions more conducive to high polymer formation than those

attempted to date.

IV. SUMMARY

A series of stannosiloxane compounds has been prepared by the
condensation of sodium triphenylsilanolate with various organotin chlorides
in benzene. - Infrared and elemental analyses were employed to' characterize
the reaction products.

The stoichiometry of the reaction of sodium triphenylsilanolate with
triphenyltin chloride, dimethyltin dichloride, and dibutyltin dichloride in
benzene has been established. Sodium chloride by-product yields were
obtained by standard silver nitrate titration. . Stannosiloxane yields were
determined spectrophotometrically using the characteristic Si-O—Sn peak
for identification. Results obtained using these techniques show the
sodium chloride-organotin chloride condensation reaction in benzene goes
to completion and is free of significant side reactions.

Sodium chloride formation has been found to be rapid when sodium
triphenylsilanolate is caused to react at room temperature in benzene
with a stoichiometric amount of an organotin chloride. Such reactions
are complete in less than a minute, so quantitative rate data could not be
obtained. For this reason, a mechanism can not be postulated for the
sodium triphenylsilanolate-organotin chloride condensationyprocess in
benzene. However, the lack of side reactions in such processes is; 1
significant since this implies the reaction of pure (NaO)ZSi4)z with organotin
dichlorides in an aprotic solvent will yield high molecular weight stanno-
siloxane polymers having a regular alternating (-Si-O~Sn-) structure.

Several stannosiloxane polymers have been prepared under a wide
range of reaction conditions by the i_nflcondensation of (NaO)zSi¢z with
dibutyltin dichloride. The polymers were not fully characterized, but it
has been established that the polymerization process is rapid. . Highly
purified (NaO)zSi 2: which is soluble in absolute ethanol, has been isolated

and will be used in future stannosiloxane polymerizations.

67

1.

2.

4.

10.

11.

12.

13.

14.

BIBLIOGRAPHY

R. P. Anderson and M. M. Sprung, WADC TR 59-61 (1960).

E. G. Rochow, An Introduction to the Chemistry of the Silicones,
2nd ed. , Wiley, New York, 1951, p. 92‘. .

 

. J. B. Rust, and H. H. Takimoto, Research On High Temperature

Polymers, " Technical Report No. 1, Contract No. Nonr 2540(00),
Task No. NR 356-395, Hughes Aircraft Company, September 1958.

 

P. E. Koenig, In Conference on High, Temperature Polymer and Fluid
Synthesis, WADC TR 57-657 (1958).

 

. K. A. Andrianov and A. A. Zhdanov, J. Polymer Sci. 22, 513 (1958).
. K. A. Andrianov, Uspekhi Khim. _2_;7_, 1257-1303 (1958).

. K.IA.- Andrianov, Polymers With Inorganic Molecular Chains,

 

Russian translation from Wright Air Development Division, December

19, 1960.

. R. M. Kary and K. C. Krisch, J. Am. Chem. Soc. 19, 2140 (1957).

. F.-A..Henglein, R. Lang, and K. Scheinost, Makromol. Chem. 18-19,

102-118 (1956).

F. A. Henglein, R. Lang, and K. Scheinost, Makromol. Chem. ii,
177 (1955).

J. T. O'Brien, WADC TR 57-502, ASTIA No.-AD-142100, (1957).

K..A. Andrianov, T. N. Canina, and E. N. Khousteleva, Izvest. Akad.
Nauk S.S.S.R., Otdel. Khim. Nauk 1956, 798-804.

P. E. Koenig and J. H. Hutchinson, WADC TR 58-55, Part I, May
1958.

P. E. Koenig and R. D. Crain, WADC TR 58-44, Part II, June 1959.

68

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

69
R. K. Ingham, S. D. Rosenberg, and H. Gilman, Chem. Revs. _6_0_,
459-539 (1960).
W. S. Tatlock and E. G. Rochow, J. Org. Chem. _1_'_l_, 1555-63 (1952).
S. Papetti and H. W. Post, J. Org. Chem. _2_2, 526-528 (1957).

P. F. Flory, Principles of Polymer Chemistry, Cornell University
Press, Ithaca, New York, 1953, pp. 69-105.

 

B. V. Bhide and J. J. Sudborough, J. Indian Inst. Science SA, 89
(1925).

E. Rabinowitch and W. C. Wood, Trans. Faraday Soc. _3_2_, 1381 (1936);
E. Rabinowitch, Trans.. Faraday Soc. fl, 1225 (1937).

P. J. Flory, J. Am. Chem. Soc. _6_1_, 3334 (1939); 6_2_, 2261 (1940).
w. T. Grubb, J. Am. Chem. Soc. 7_6_, 3408-3414 (1958).

C. Eaborn, Organosilicon Compounds, Butterworths Scientific Publi-
cations, London, 1960, p. '250.

 

W. T. Grubb and R. C. Osthoff, J. Am. Chem. Soc. 21, 1405-1411
(1955).

R. Okawara, D. G. White, K. Fujitani, and H. Sato, J. Am. Chem.
Soc. 83, 1342-1344 (1961).

F. A. Henglein, R. Lang and L. Schmack, Makromol. Chem. _2_2_,
103-114 (1957).

L. H. Sommer, E. W. Pietrusza, and F. C. Whitmore, J. Am. Chem.
Soc. _6_8_, 2282 (1946).

H. Gilman, H. N. Benedict, and H. Hartzfeld, J. Org. Chem. _1_9_,
419 (1954).

L. Summers, Iowa State Coll. J. Sci. _2_6_, 292-294 (1952).
E. Wiberg and U. Kruerke, Z. Naturf. 8_b, 610 (1953).
K. A. Andrianov,.A. A. Zhadanov, N. A. Kurasheva, and V. C.

Dulova, Doklady Akad.. Nauk S.S.S.R. 112, 1050 (1957); C. A. _5__1_,
14544 (1957).

32.

33.

34.

35.

36.

37.

38.

39.

70
H. Schrnidbaur and M. Schmidt, J. Am. Chem. Soc. _8_3_, 2963-2964
(1961).

K.~A. Andrianov and A. A. Zhdanov, Izvest.. Akad. NaukS.S.S.R.,
Otdel. Khim Nauk 1958, 779-80; C. A. 22, 19916 (1958).

M. M. Chamberland, Inorganic Macromolecules Containing a Si-O-M

 

Linkage, Technical Report No. 1, Contract No. Nonr 1439 (07),

Project NR 052-419, Western Reserve University, September 23, 1960.

K. A. Andrianov, A..A. Zhdanov, and E. A. Kashutina, Doklady Akad.

.Nauk S.S.S.R., 126 (1959).

J. F. Hyde, and O. K. Johannson, U.S. 2,453,092 (August 5, 1947).

J. F. Hyde, O. K. Johannson, W. D. Dandt, R. F. Flemming, H. B.
Laudenslager, and M. P. Roche, J. Am. Chem..Soc. 7_5, 5615-5618

T. L- Brown, I. The Preparation and Physical Properties of Crystalline
Lithium Alkyls, II. The Integrated Intensity of the Infrared Absorption

 

(Band Due to O-H Stretchingin Aliphatic Alcohols, Ph. D. Thesis,

Michigan State University, 1956, 116 pp.

C. F. Gibbs, H. Tucker, G. Shkapenko, and J. C..Park, WADC
Technical Report 55-453, ASTIA Doc. No. 131036, September 1957.

40. A. B. Thomas and E. G. Rochow, J. Am. Chem. Soc. 12, 1843-1848

41.

42.

43.

44.

45.

(1957).

InternationalCritic'al Tables, Vol.1 VII, 'McGraw-Hilerook Cd. , .N.' Y. ,
1930., pp. 34-62.

 

N. Wright and M. J. Hunter, J. Am. Chem..Soc. 62, 803 (1947).
H. Gilman and W. B. King, J. Am. Chem..Soc. §_1_, 1213 (1929).
M. . Farnsworth and J. Pekola, In Treatise On Analytical Chemistry, ‘

Part II, 'Analytical Chemistry iof the Elements, " Vol. 3, Interscience
Publishers, New York, 1961, pp. 327-371.

 

 

W. F. Hillebrand, Applied Inorganic AnalysisJ ‘ 2nd ed. , J. Wiley
and Sons, Inc., New York, 1953. p. 293.

 

46.

47.

48.

49.

50.

51.

71
Instruction Manual for Perkin-Elmer‘lnfrared Equipment, Volume 3B,
Perkin-Elmer Corporation, Norwalk, Connecticut, 1952.
A. L. Smith, Spectrochim.-Acta1_6_, 87-105 (1960).

L. J. Bellamy, "The Infrared Spectra of Complex Molecules, 2nd
ed., J. Wiley and Sons, Inc., New York, 1958, pp. 337-338.

H. Hartridge and F. J. W. Roughton, Proc. Roy. Soc. London, A104,
376 (1923).

E..L. Wittbecker and P. W. Morgan, J. Polymer Sci. _4_0_, 289—297
(1959).

P. W..Morgan and S. L. Kwolek, J. Polymer Sci. :12, 299-327 (1959).

APPENDICES

72

APPENDIX A

Reaction Conditions and Stoichiometric Data
for @SiOanh, (CH3)zSn(OSi¢3)z, and
(2-C4H9)zSn(OSi¢3)z Formation.

73

poficflaou

 

 

so 66 ms: 6 38 6 $36 3.3 6 6 .38 S3 6:.emmovsmexemsomv mm
2 66 2.: 6 $666 $3 6 3.3 6 m .3 S36exeehmovsmexemdé mm
-- 2.: 6 366 6 $3 6 .: m .2 S36€eomovsmexemdév 3
2 .3 mom... 6 ,,_ 66$ 6 £3 6 I So 6 o 636 663 6 "6303.6 on
3.3 336 336 6.636 2.36 v.3. 6636 neemoomee . S
6666 $.26 3:6 336 3636 6.36... 336 ... 303.6 m:
3.3 3.36 62: 6 336 I 3636 66... $36 neemosmee S
S 66 82.6 3.36 3.36 336 66 336 nesmosmne f
-- -- -- n2.36 336 mg... 2.36 nesmoane S
- .. .. -- A336 236 m6 $36 @3036 S
-- -- -- .2.36 $36 66 6636 neemoemee 2
4 -- -- -- 336 E36 66 $36 neemoemee 2
7 -- -- .. 336 $36 66 2.36 nesmosmne :
$66 326 2.26 -336 2.36 6.5.3 336 . emosmee S
5.66 326 $36 $.36 3.36 4.... 336 e smoomne 6
$66 32 6 3.: 6 $36 $.36 36 336 6303.9 6
666 3: 6 3...; 6 $36 $36 6633 3.36 6303.6 e
s62 em: 6 .32 6 $36 $36 m .3 S36 nesmoemee 6
H66 on: 6 8.2 6 $36 2.36 2; S36 nesmommee m
.. -- -- 336 3.36 N23 3366 nesmohmee s
-- -- -- 836 E36 6.... 3366 eesmoomne m
-- .. -- $36 336 «6 $366 nesmosmse N
-- -- -- $36 $36 26 $366 sesmoomne. H
s: 3 as V E E Asses Ex is. .3 2838 .282
>so>ouon mUSOm posse/0.60.“ opwuofigu opwnoflmo .683 .0650 ~H0nm~m cam
mpSOm H.309 mpSOm oBSHOm Gunman Sacco :oflomwm no HOGmnm was
H.308 .ponO ocoucom .ponO posse/000m mfiflmOmZ 33%;.

k1

 

II“

6652232200

 

wv

 

3.3 6326 3226 336 2.2.36 6.3.22 336 263932.226m2
3662 6326 3226 336 336 3.22. 336 2. 30232.2.26dw 3
3.3 6326 3226 336 336 362 336 2. 28232.36fi2 2.
3.3 6326 3226 336 336 66 336 263932.263 2.
3.3 236 3326 336 336 2.36 336 2.630232.2.26-m2 2.2.
3.3 236 .326 336 336 22m 336 2.9393266é 2..
3.3 236 2326 336 336 6.. 336 26302322296 2.
3.3 236 326 336 336 2.2 336 263932.398 22.
3.32 336 2.326 336 22.36 6632 336 263032.226m2 62.
3.32 336 2326 .336 336 2.63 336 263932.396 3
3.3 336 2326 336 336 36 336 263932.395 3
I. -- -- 336 336 6.32 2.336 2630232....26m2 3
-- .. .. 336 336 2.3.- .336 22628232.m6-m2 3
n -- -- -- 336 6636 36 3.36 263932.396. 3
-- -- -- 336. 336 6.6 2.336 263932.295. 2.3
622.3 6.226 2626 336 2.2.3.6 6.32m $36 263932.226MV 3
3.3 6.226 3326.. 336 2.2.3.6 22.2... 336 263232.263 3
3662 2.226 3326 336 2.36 2.2 6636 263932.263 3
3.3 2326 A326 336 336 6.3.: 336 2.2630232.m6-m2 o...
3.3 232.6 :26 336 336 2.32 2.2.36 263932.226fi2. 3
3.3 2326.. 3326 .336 336 6... 2.2.36 263932.26..E 3
23.3 2326 3.26 .x 336 2.266 6.2. 2.2.36 263232.263 3
3.3 326 3626 336 6266 6.33 2.266 2639.32.22.96 3
3662 32226 3626 336 336 2.3 2.266 263932.26-E m...
3. 32.26 -- 336 I. .6 2.336 2.63023266222. 22...
3ch Amy Amv Amv E 72222»: A.H\ .3236 .3 6222209200 Honasz
ensue/006.2 wUSOm _ 66.29602: 63.2030 63.2030 053. .6280 NHOGmmm 2222M
mUSOm 2.33 mpSOm 3930a nausea EDHUOm :oflomom .8 20253.32 p23
Han—OH. .26on0 ocmncmm .ponU . pouo>oomm nfiflmOmZ H2332;

 

 

 

 

252232200 30.2%?vaan {mum m moved“...
-- -- -- 33 .6 $3 .6 m .3 336 2639322209 3
.. -- -- 336 336 m2 336 26393269 22.
-- .. -- 3266 336 36 336 263932.39 6.
3 .3 362 .6 336 . 336 336 6.33 6336 26393269 3
3 .3 362 .6 366 .6 33 .6 $3 .6 3 .2}. 633 .6 263932329 3
3.3 362 .6 .36 .6 326 .6 $3 .6 3 .6 6636 263932229 3
62 .3 322 .6 233 .6 326 .6 .33 .6 6 .33 33.6 263932229. 3
3662 3226 336 3.36 3266 3.3 3.36 263932229 3
Z 662 322 .6 236.6 3.36 326 .6 m 62 23266 26393269 3
3. 662 322 .6 6366 33 .6 3.36 223 3.36 263932229 3
. .. 362 .6 336 336 -- 6.223 336 263932£9 3
3.3 362 .6 33 .6 $3 .6 6.3 .6 6 .3. 326 .6 263932.229 3
6 -- 362 .6 336 33 .6 .. 3.2 336 263932.229 63
7 -- .32 .6 33 .6 33 .6 -- 6 .33. 326 .6 263932429 3
3.662 .32 .6 336 33 .6 32.3 .6 2.43 33 .6 263932239 3
-- .32 .6 33 .6 33 .6 -- m .2 33 .6 263932229 3
3.3 M32 .6 N32 6. 336 336 6.3 6366 26.393.222.93 3.
3.3 .32 .6 232 .6 336 336 3.3 636.6 263932.396 3
3.3 M32 .6 n32.2 .6 336 2.2.3.6 262 6366 263932.396 2}.
3.3 M32 .6 3.2 6 336 336 36 .3366 263932663 3
3.3 2.32 .6 362 .6 6236 336 6.3m 3.36 2639232.m6-m6 3
3.3 332 .6 2.32 .6 6236 336 2.2.3 336 2630.32.22.96 3
3.3 2.32 .6 2.32 .6 6236 336 2.22 3.36 26393266-...» 62..
3.3 2.326 326 6236 33.6 .6 3.36 263932663. 3
23 23 23 23 23 2.2.23 2.2\ $262.... .3 6.522860 3.28872
3826092 36203 vmuv>oomn 35.20230 03.2030 @2223 62200 NHUammm and
36:03 2.30» 33203 035203 83.003 E22503 Gowuommm .20 202.53% wad
22309 .26on0 mcmucmm .UUHMU Umum>oomm nfiflmOmZ 2.32:2:

I‘

 

 

77

 

 

osmucmm

mm .3 32 .o 2: .o 33 .o. :3 .o o .8: 83 .o wawmovnmfinmuv m:
cm .2: 32 .o $2 .0 £86 2.35 ~13. 2:56 N1 5932nt 2.
cm .2. 32 .o :2 .o 3.8 .0 SS .0 o; 83 .o «Aneaovnmxnmov 3
.,- $3 .o 32 .o 3.8 .o 1.. o 4:3 38 .o Aéfimovnmxnmov E.
S .3 SE .0 82 .o Nome .o oomo .o o .5 38 .c Aéflmovnmxnmov i
S .3 $3 .o 2.: .o 8.8 .0 mas .o o .o ,mvmo .o :éwmovamNAnmov 2
GAL Amy Amy Amy Amy TGMEV A.H\ Sing .3 wGSOQEOU umnEdZ
>um>oUon mUSOm mung/cum.“ oguogo mEnOEu was .350 fiucmnm and
mUEOm #30» mUSOm m330m ESfiUOm EdwUOm Gofiumom no HUGmmm find
305 .3me .316 v BESSm ”6502 SEE

 

 

APPENDIX B

Spectrophotometric Data for (#35105an3, (CH3)zSn(OSi¢3)Z ,
and (2-C4H9)28n(OSi¢3)z Formation.

78

79

 

 

 

Run co (M) A c. (M) c/co x 107‘
¢3Si03n¢32
1 0.00872 0.262 0.00850 97.5
2 0.00906 0.268 0.00869 95.9
0.00906 0.270 0.00875 96.6
3 0.00848 0.257 0.00837 98.7
4 0.00855 0.261 0.00847 99.1
5 0.00813 0.242 0.00786 96.7
6 0.00858 0.266 0.00860 100.2
7 0.00842 0.268 0.00869 103.2
8 0.00845 0.244 0.00791 93.6
9 0.00762 0.233 0.00757 99.3
10 0.00842 0.253 0.00818 97.2
11 0.00780 0.236 0.00765 98.1
12 0.00708 0.213 0.00693 97.9
13 0.00818 0.249 0.00806 98.5
14 0.00829 0.251 0.00814 98.2
15 0.00740 0.230 0.00749 101.2
16 0.00841 0.256 0.00830 98.7
17 0.00855 0.262 0.00850 99.4
18 0.00810 0.246 0.00798 98.5
19 0.00844 0.252 0.00818 96.9
20 0.00880 0.261 0.00847 96.3
(2-04149)zsn(051<b3)2:
21 0.00662 0.339 0.00641 96.8
0.00662 0.326 0.00617 93.2
0.00608 0.309 0.00582 95.7
22 0.00811 0.404 0.00762 94.0
0.00811 0.409 0.00771 95.1
0.00669 0.353 0.00667 99.7
23 0.00662 0.331 0.00626 94.6
0.00626 0.327 0.00618 98.7
24 0.00664 0.312 0.00590 88.9
0.00664 0.311 0.00589 88.7
25 0.00703 0.336 0.00634 90.2
26 0.00636 0.305 0.00578 90.9
0.00636 0.309 0.00584 8

91.

 

Continued

80

 

Run c0781) A c. (M) c/co x 107'
(_r_1-C4H9)2Sn(OSi(P3)Z cont'd:

27 0.00423 0.217 0.00410 96.9
28 0.00603 0.307 0.00580 96.2
29 0.00659 0.352 0.00666 101.1

0.00659 0.355 0.00670 101.7
30 0.00613 0.321 0.00608 99.2

0.00613 0.325 0.00615 100.3
31 0.00620 10.294 0.00555 89.5
32 0.00590 0.281 0.00530 89.3
33 0.00629 0.310 0.00588 93.5

0.00629 0.311 0.00590 '93.8
38‘ 0.00593 0.233 0.00440 74.2

0.00770 0.319 0.00610 79.2

0.00770 9.317 0.00590 76.6
39 0.00662 0.246 0.00463 69.9

0.00698 0.263 0.00497 71.2

0.00698 0.275 0.00519 74.4
40 0.00809 0.387 0.00730 90.2

0.00809 0.390 0.00736 91.0
41 0.00728 0.370 0.00700 96.2
42 0.00692 0.356 0.00672 97.1
43 0.00681 0.361 0.00681 100.0
44 0.00731 0.379 0.00713 97.5
45 0.00741 0.374 0.00706 95.3
46 0.00687 0.355 0.00670 97.5
47 0.00687 0.366 0.00690 100.4
48 0.00759 0.395 0.00747 98.4
49 0.00685 0.344 0.00650 94.9
50 0.00694 0.351 0.00663 95.5
51 0.00692 0.354 0.00669 96.7
52 0.00758‘ 0.380 0.00718 94.7
53 0.00774 0.345 0.00650 84.0
54 0.00689 0.317 0.00600 87.1
55 0.00729 '0.321 0.00609 83.5
56 0.00669 0.306 0.00579 86.6

 

Continued

81

 

:— L

 

Run (:0. (121.) A c. (M) c/co x10z

(CH3)zSn(OSi¢3)2:
57 0.00583 0.201 0.00516 88.5
60 0.00577 0.210 0.00538 93.2
61 0.00584 0.215 0.00526 94.2
62 0.00653 0.211 0.00540 82.7
64 0.00726 0.278 0.00710 97.8
65 0.00607 0.234 0.00579 98.7
66 0.00627 0.212 0.00546 87.1
68 0.00575 0.216 0.00554 96.4
69 0.00580 0.217 0.00559 96.4
78 0.00696 0.263 0.00675 97.0

 

APPENDIX C

Solubility Characte‘zristiés of
Stannosiloxane Polymers

82

83

 

 

 

 

 

 

 

 

Solvent Polymer Sample
(b- Burl
(bulk) (methanol- soluble) (methanol-insoluble)
Carbon disulfide ps* 6 ss
Isopropanol i- i i
N-heptane i s i
Nitrobenzene ss 5 ps
Benzene ps 3 ss
Ethylene dichloride i s i
Tetralin ps 3 ps
(15—Bun
(bulk) (methanol-soluble) ~ (methanol-insoluble)
Carbon disulfide ps 3 ss
Isopropanol i i i
~ N-heptane i s i
Nitrobenzene ss 8 ps
Methyl ethyl ketone ps 5 i
Anisole ps 3 ps
Acetonitrile i 33 i
(3)—Bum
(bulk) (methanol- soluble) (methanol-insoluble)
Methyl ethyl ketone ps 8 i
Anisole ps 3 ps
Ac etonitrile i s s i
N-amylacetate i s i
DMF 33 s i
Carbon tetrachloride ps 3 i
Ethyl ether ps 3 i
(b-BuIV
(precipitated) (non-precipitated)
N-amyl acetate -- ss
DMF 1 ps
Carbon tetrachloride i ps
Ethyl ether 1 ps
Benzene 38 ps
Ethylene dichloride i ps
Tetralin ss 88
*s = soluble; pa = partly soluble; as = slightly soluble; i = insoluble.

APPENDIX D

Infrared Spectra

84

85

«L

.3

63:5me confine GM aficmommnfl mo 8.30on vohmumcH

o

Amnouowfiv #333283

w

h

m

 

 

 

 

 

 

 

 

 

 

 

 

fi

 

 

 

A

 

—

4w

 

‘l

 

 

 

OH

on

on

0*

cm

co

or

ow

om

ooH

(%) aouen'suzsu'ex .I.

 

 

2 5::

86

 

 

.0633me conumno aw finnvflmovnmfiv mo muuoomm 60.8.35

Annoy owns 5323983

A: o m N. c m w m
_ q _ _

O

 

-1 om

om

 

ow

 

 

om-

 

 

 

 

 

on:

(%) aouentmsuex .L

87

a;

NH

52:5va 89:8 E Anewmovnmfinmo? mo 938% 33:3
Amnou UMEV Auwcgmkrmg

0H

o

w

b

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_

 

 

q

 

-

_

 

 

 

 

OH

on

hm

ow

om

co

ON.

cm

00

00H

(%) a onenymsuax 1,

88

MA

MA

62:86 3980 E «AnAYmovcmAnmuv mo 233m @223

Amcou 39$ fiwnoaosk?

HA

CH

o

m

 

 

 

 

 

q

—

 

 

 

d

o
1

 

L

 

 

OH

ON

on

ow

om

co

on.

om

oo

on:

(%) aauenguxsu'ezl

89

.mvwwflamflu Conrado cw NAnGmmOVGmNAaEquMV mo «300mm vmhmHmcH

Amcouuflhv aumaofimima
3 2 NH 2 S o m N. o m v m

D

_ _ _ _ _ _ _

 

J

 

 

 

 

 

 

 

 

 

 

.. ow

 

 

 

 

 

 

OOH

(%) aouentmsu'ex .L

9O

. .. .mwflgmfiv GOnTHmo a:
HHH 5m. fl mo coflumum QBEOmJOGmafimE mo wuuummm 60.333
Amcouowgv Quwcmfimkrm?

 

. N! A 1

 

 

 

 

 

 

 

 

_

 

 

 

s1

 

 

OH

ON

on

ow

om

oo

on

ow

o¢

OOH

(%) 9 DU‘BnEmsuex 1,

633.9va conumo GM HUGmnnv mo mun—ovum Uvnmumfi

Ammo.“ 35v Auwcofimkwm?

: 0H 0 m b o m. vm

 

 

 

91

 

 

O

4 fl . J _ _ .

 

 

 

 

 

 

 

 

OOH

92

.2

NH

633.9me conumo cw «Huanmnfl mo «.30QO 33‘9“de

Hg

OH

Amcou 3G5 gumcmfimkwm?

o

w

 

 

 

 

 

 

q

 

 

 

_

 

—

 

 

 

 

 

OH

ON

on

ow

om

co

2..

cm

oo

9.:

(0A) aouenguxsu'ex .I.

632236 cognac GM NHUGmNAanV mo «.30QO @933de

Annoy 35v sumcmfimkfim?

vd Md NH HA o“ .
I 4 _ q _ 4® . W m. N m V m

 

_
-
d

93
I

 

 

 

 

 

OH

om

om

ow

om

co

S.

cm

om

00H

(%) aéu'engtnsu'ex‘];

94

Ma

Ma

.wgfiamflv Gomumd Gfi -Ucm~A¢E«Uum.v we «.30QO vmumnfi;

:

Annoy away ”$983?me

a

m

N.

 

 

 

 

 

 

 

fi

 

 

 

 

_

 

 

 

 

 

 

 

—1

 

 

OH

on

om

ow

.0m

,3

on.

ow

.ow

 

 

OOH

(%) a auengmsuex .1.

.mgfismfiw Conrado GM nfifimOmZ mo «.30QO UmuduwaH

Amconuflhv “images?

“L ma NH 2 0H mu m N.

\0
Ln
VF
M

 

95

 

 

0

)d 4 A _ .3_ )— _ d 4 fi

 

 

 

 

 

 

 

 

 

 

 

 

moo

 

 

on:

(%) a ou’entwsu'ex .L

APPENDIX E

Sodium Chloride Rate Data

96

97

 

 

 

' Initial Reaction Sodium
Run NaOSi¢3 time Chloride
No . Compound conc. (121) (min.) yield (%)

1 063510511033 0. 00783 2.1 95. 3

2 ¢3s105n¢3 0. 00783 2.4 95.1

3 (035105110), 0.00783 5.6 95.9

4 (1135105114)3 0. 00783 24. 7 96. 3

5 41351os11¢3 0. 0127 1.1 96.2

6 (035105110), 0. 0127 51. 3 99. 6

7 (1135108110)3 0.0127 2796.0 98.0

8 ¢3SiOSn¢3 0. 0134 0. 9 93.4

9 ¢3SiOSn§b3 0. 0134 5. 4 94. 3
10 (035105114). 0. 0134 1447. 0 94. 9
11 (133510511013 0. 0299 0. 8 96. 3
12 (1)3510s11cb3 0. 0299 o. 8 96. 3
13 (b3510sncb3 0.0299 5.6 96.8
14 (113510511cb3 0. 0299 9 9. 3 97. 0
15 (11.510560)3 0. 0299 27. 5 97. 2
16 (113510511613 0. 0345 0. 9 96. 0
17 (03510511 3 0.0345 27.8 97.4
18 (03510511 3 0.0345 2994.0 97.6
1821 ¢3SiOSnd>3 0. 0490 0. 8 94. 2
19 <1>3s1os11<1>3 0.0490 54. 7 95.2
20 (b351osncb3 0.0490 2225. 0 94.7
22 (11- (3.11.,)zs11(o151<1>,,)z 0. 0127 54. 3 97. 6
23 (n-C4H9)zSn(OSi¢3)z 0. 0127 2978.0 98. 0
25 (11 <3,,11.,)2511(os1<t>3)2 0.0134 54.5 94.6
26 (n- C4H9)zSn(OSi¢3)2 0.0134 8878.0 95.6
27 (n-C4H9)zSn(OSi¢3)z 0. 0147 4. 0 95. 2
28 (n- C4H9)zSn(OSi¢3)z 0 . 0147 9. 0 100. 9
29 (n-04119)zsn(051¢3)2 0. 0147 148.1 99. 8
30 _ (11..c4111.,)25n(os1<b3)2 0.0147 1207.0 99.7
31 (n-C4H9)zSn(OSi¢3)z 0. 026 1. 1 95. 8
32 (n c4,11.,)zs11(os1<b3)z 0.026 44.8 96.8
33 (n- C4H9)zSn(OSi¢3)z 0 . 026 2153. 0 96. 7

 

Continued

98

 

 

 

Initial Reaction Sodium
Run NaOSi¢3 time . chloride
No. Compound conc. (N) (min.) yield (%)
34 (E- c,‘11.,)zs11(051<1>.,)Z 0. 0354 0. 6 97. 2
35 (E-C4H9)Zs11(051¢3)2 0.0354 0. 9 97. 5
36 (E c4119)._s11(os1d>_.,)z o. 0354 28. 1 98. 1
37 (E-c4119)zs11(051d>3)2 0. 0354 165.0 98.8
38 (11- c4111.,)zs11(os1<1>3)Z o. 0490 0. 8 94. 6
39 (E.c.1119)3814081413)z 0. 0490 50. 4 95. 1
40 (E.c,‘11.,)zs11(os1q>3)Z 0. 0490 1560. 0 94. 9
41 (E.c411,)zs11(051g1>3).z 0. 0492 1. 1 96. 8
42 (n- C4H9)zSn(OSi 3)z 0. 0492 7. 0 97. 3
43 (n- C4H9)zsn(051 3); O. 0492 21. 5 97 o 9
44 (E c4H9)zs11(051¢3),_ o. 0492;. 65. 4 98.3
45 (11.c4111.,)zs11(051<1>3)z o. 0302 0. 6 97. 2
46 (E.c4111.,)._s11(051d>3)Z 0. 0302 10. 3 98. 2
47 (E.c4,111.,)zs11(051(1):.)2 0. 0302 41 . 8 98. 2
48 (E- c4111.,)zs11(os1<1>3)z 0 . 0302 853 . 0 98. 4
49 (11..c411.,)zs11(os1<1>3)Z 0.0359 0. 7 92.6
50 (n- C4H9)25n(051¢3)2 0.0359 14.7 96.7
51 (11 c.,.11.,)zs11(051<1>3)Z 0. 0359 64. 4 96. 6
52 (E.c.,11.,)zs11(os1<1>_.,)z 0. 0359 352. 0 96. 9
53 (n-C4H9)zSn(OSi(b3)z 0. 0430* 0. 9 96. 4
54 (11 c4119)zs11(051<1>1)1 0. 0430 10. 1 97. 4
55 -(11..c4111.,)zs11(os1<b3)z 0 . 0430 43. 6 97. 6
56 (E.c4,14.,)._s11(os1<1>3)Z o. 0430 825. 0 98.8
58 (CH3)zSn(OSI¢3)Z 0.0127 8554 97.8
61 (0111312811(o.31<1>3)z 0 . 0134 58.18., 95.4
63 (c113),_s11(051¢3),_ 0.0136 3.8 98.4
64 (01113,)2811(os1cb3)z 0. 0136 12. 5 99. 0
65 (CH3)zsn(OSI¢3)Z O. 0136 61. 8 98. 6
66 (c113)zs11(os1¢>3)2 o. 0136 682. 0 90. 9
67 (c111;,)zs11(051d>3)z 0. 0260 0. 9 96. 4
68 (CH3),_s11(os1 3).2 0. 0260 54. 6 96. 9
69 (CH3)zSn(OSi 3)z 0. 0260 2795. 0 97. 2

 

*
In1t1al (2-C4H9)ZSHC12 COHC.

is 0.0463 g. equii'r./l.

Continued

99

 

 

 

= M

Initial Reaction Sodium '

Run NaOSiQS3 time chloride

No. Compound conc. (Ii) (min.) 7 yield(%)
70 (CH3)zSn(OSi(b3)z 0. 0299 0.9 98. 7
71 (c113,)._s11(os1<1>3)Z 0. 0299 1. 3 98. 4
72 (CH3)ZSn(OSi(b3)2 0. 0299 29.5 99. 2
73 (c3113)zs11(os1d>3)z 0. 0345 0. 9 97. 9
74 (c1113)._s11(051<1>3)Z 0. 0345 87. 9 98. 9
76 (CH3)zs11(os1d>3),_ 0. 0490 1. 0 95.0
77 (CH3)ZSn(OSi¢3)z 0. 0490 50. 2 95. 5
78 (c113)..,s11(081c1>3)Z 0. 0490 2190. 0 94. 7

 

CHEMISTRY LIBHA RY

JUN 20 ’62