Citation
Solution and gas phase rearrangements of chloromethylsilanes

Material Information

Title:
Solution and gas phase rearrangements of chloromethylsilanes
Creator:
Yost, Vernon Edward
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
103 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Chloromethyl group ( lcsh )
Chloromethyl group ( fast )
Genre:
theses ( marcgt )
non-fiction ( marcgt )

Notes

General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, Department of Chemistry
Statement of Responsibility:
by Vernon Edward Yost.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
20960197 ( OCLC )
ocm20960197
Classification:
LD1190.L46 1988m .Y67 ( lcc )

Downloads

This item has the following downloads:


Full Text
SOLUTION AND GAS PHASE REARRANGEMENTS OF
CHLOROMETHYLSILANES
by
Vernon Edward Yost
B.A., University of Colorado at Denver, 1981
A Thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Department of Chemistry
1988


This Thesis for the Master of Science Degree by
Vernon Edward Yost
has been approved for the
Department of
Chemistry
by
Date.


ABSTRACT
CHAPTER I: The gas phase reactions of R-substituted trimethylsilanes with
hydroxide and fluoride ion have been shown to produce the pentacoordinate anions
Me3RSiO" and Me3RSiF' respectively. Me3RSiO' was found to undergo immediate
migratory rearrangement of methyl or R groups to the hydroxyl proton to form
Me2RSiO" and methane or Me3SiO" and R-H. The ratios of R/Me migration were
shown to be strongly dependent on the gas phase acidities of the respective R-H
hydrocarbon. Me3RSiF* was found to be most stable when R was alkyl, but
decomposes to Me3SiF and R when R was benzyl, phenyl, alkenyl or alkynyl.
CHAPTER II: Solutions of R-substituted chloromethyldimethyl silanes in aromatic
solvents have been found to react with inorganic fluoride anion via phase transfer
catalysts such as 18-crown-6 to effect two primary migrations; that of an R group, to
produce RCH2Si(F)Me2, and that of either of the methyl groups, to produce
EtSi(F)(Me)R. The ratios of the R/Me migration have been correlated with the gas
phase acidities of the respective RH hydrocarbons and with various a type substituent
constants in the Hammett fashion. These have made it clear that the cleaved carbon
fragment undergoing migration has substantial carbanionic character.
CHAPTER III: Solutions of substituted arylchloromethyldimethyl silanes in
appropriate aromatic solvents have been found to react with fluoride anion via an


IV
18-crown-6 phase transfer catalyst to produce two interesting side products
(dimethyldifluorosilane and substituted toluenes) in addition to two sets of Ar/Me
migration products:
1 ArCH2Si(F)Me2 and EtSi(F)(Me)Ar
2 (ArCH2)2SiMe2 and ArCH2Si(Et)(Me)PhX
Hammett studies using ctx were found to correlate very well with a variety
of experiments and led to the conclusion of a pentacoordinate (at silicon) reaction
mechanism.
CHAPTER IV: The simple (drop by drop tube furnace) and vacuum pyrolysis of
substituted arylchloromethyldi(and mono)methyl silanes were found to produce
chlorosilanes which have undergone R/Me and H/Me migration respectively. The
relative migratory ratios obtained in the tube furnace studies were found to give crude
estimates of kinetic parameters used in the true gas phase, vacuum pyrolytic method (-
aka Stirred Row Pyrolysis). The rearrangement mechanisms were found to be
dependent on the technique of pyrolysis used.


V
ACKNOWLEDGEMENTS
Much credit for the research contained herein belongs to :
Dr. Robert Damrauer, Thesis Supervisor
Stephen E. Danahey
Bonnie K. O'Connell
Roger Simon


CONTENTS
INTRODUCTION............................................... 1
CHAPTER I GROUNDWORK FOR REARRANGEMENT STUDIES
1.1 Gas Phase Acidities of Hydrocarbons.......... 3
1.1.1 Review....................................... 3
1.1.2 Experimental Results........................... 4
1.2 Pentacoordinate Silicon Anions............... 6
CHAPTER H THE RELATIVE MIGRATORY ABILITIES OF ALKYL,
VINYL AND ARYL GROUPS IN SOLUTION PHASE
FLUORIDE INDUCED REARRANGEMENTS OF
CHLOROMETHYLSILANES
2.1 Introduction................................. 8
2.2 Summary of Optimization Experiments.......... 9
2.3 Results and Discussion.......................... 12
2.4 Alkoxide Rearrangements of Chloromethylsilanes. 17
2.5 Synthesis of Starting Materials................. 19
2.5.1 Preparation of EtSiMe2CH2Cl..................... 19
2.5.2 Preparation of n-BuSiMe2CH2Cl................... 20
2.5.3 Preparation of i-PrSiMe2CH2Cl................... 21
2.5.4 Preparation of (CH2)2SiMe2CH2Cl................. 24
2.6 Optimized RSiMe2CH2Cl + KF Reactions............ 26
- Detailed Experimental -
2.6.1 Reactionof EtSiMe2CH2Cl with KF................. 26
2.6.2 Reaction of n-BuSiMe2CH2Cl with KF.............. 26
2.6.3 Large Scale Preparation of n-CsHj]SiMe2F


vn
and Et(n-Bu)SiMeF............................ 27
2.6.4 Reaction of i-PrSiMe2CH2Cl with KF................. 28
2.6.5 Reaction of cy-PrSiMe2CH2Cl with KF................. 29
2.6.6 Reaction of H2C=CHSiMe2CH2Cl with KF................ 30
2.6.7 Reaction of PhSiMe2CH2Cl with KF.................. 31
2.6.8 Reaction of l-(chloromethyl)- 1-methyl-sila-
cyclopentane with KF................................ 33
2.6.9 Additional Reactions of Mechanistic Interest........ 33
2.6.9a Reaction of Me3Si(CH2)3Cl with KF................. 33
2.6.9b Reaction of (ClCH2)2SiMe2 with KF................... 33
CHAPTER m MECHANISTIC KINETIC STUDIES OF ARYL-
(CHLOROMETHYL) -DIMETHYLS ILANES
3.1 Introduction....................................... 35
3.2 Aquisition of Primary and Secondary Products from
the Reaction of substituted-aryl(chloromethyl)
dimethyl silanes with KF............................ 37
3.2.1 General Experimental................................ 37
3.2.2 Results and Discussion.............................. 37
r
3.2.3 Detailed Experimental............................ 44
3.2.3a Reaction of p-Me-PhSiMe2CH2Cl with KF............... 44
3.2.3b Reaction of PhSiMe2CH2Cl with KF.................... 45
3.2.3c Reaction of p-F-PhSiMe2CH2Cl with KF................ 46
3.2.3d Reaction of p-Cl-PhSiMe2CH2Cl with KF............... 47
3.2.3e Reaction of m-CF3-PhSiMe2CH2Cl with KF.............. 49
3.3 Competitive Relative Rates of Starting Material
Consumption......................................... 52
3.3.1 General Experimental................................ 52
3.3.2 Results and Discussion............................. 53
3.3.3 Detailed Experimental............................... 55


Vlll
3.3.3a PhSiMe2CH2aandp-Me-PhSiMe2CH2Cl................... 55
3.3.3b->d p-F-PhSiMe2CH2Cl, p-Cl-PhSiMe2CH2Cl, and
m-CF3-PhSiMe2CH2Cl vs. p-Me-PhSiMe2CH2Cl....... 57
3.4 Appearance Rates of X-sustituted toluenes
(ArCH3)........................................... 58
3.4.1 General Experimental............................ 58
3.4.2 Results and Discussion............................ 59
3.4.3 Detailed Experimental............................. 63
3.4.3a Appearance Rate of p-Xylene from the Reaction of
p-Me-PhSiMe2CH2Cl with KF......................... 63
3.4.3b->e Appearance Rates of p-F, p-Cl, m-CF3 and
Normal Toluenes from the Reactions of p-F, p-Cl,
m-CF3 and H-PhSiMe2CH2Cl with KF.................. 64
3.5 Synthesis of Starting Materials................... 64
3.5.1 Preparation of p-Me-PhSiMe2CH2Cl................. 64
3.5.2 Preparation of p-F-PhSiMe2CH2Cl................ 66
3.5.3 Preparation of p-Cl-PhSiMe2CH2Cl.................. 66
3.5.4 Preparation of m-CF3-PhSiMe2CH2Cl................. 67
CHAPTER IV PYROLYTIC REARRANGEMENTS OF SUBSTITUTED
ARYL- CHLORODI(AND MONO) METHYL SILANES
4.1 Introduction and Review........................... 69
4.2 Isolation of Sealed Tube Pyrolysis Products from
Phenylchloromethylmethylsilane................ 70
4.3 Argon Flow Pyrolysis of ArSiHMeCH2Cl and
ArSiMe2CH2Cl.................................... 72
4.3.1 General Experimental.............................. 72
4.3.2 Results of the Argon Flow Pyrolysis of Substituted
Arylchloro methylmono (and di-) methylsilanes..... 74
4.4 Stirred Flow Pyrolysis of ArSiHMeCH2Cl and
ArSiMe2CH2Cl...................................... 79
4.4.1 General Experimental.............................. 79
4.4.2 Results of Stirred Flow Pyrolysis of


IX
ArSiHMeCH2Cl................................... 79
4.4.3 Results of Stirred Flow Pyrolysis of
ArSiMe2CH2Cl................................... 80
4.5 Conclusions from Both Pyrolitic Methods........ 80
4.6 Synthesis of Aiylchloromethylmethylsilanes..... 87
4.6.1 Preparation of PhSiHMeCH2Cl.................... 87
4.6.2 Preparation of p-Me-PhSiHMeCH2Cl............... 90
4.6.3 Preparation of p-F-PhSiHMeCH2Cl................ 91
4.6.4 Preparation of p-Cl-PhSiHMeCH2Cl............... 92
4.6.5 Preparation of m-CF3-PhSiHMeCH2Cl.............. 93
REFERENCES.................................................. 94
APPENDIX
A Second Order Polynomial Least Squares
Obtained Fit for Reaction 3.4.3a->e Profiles.......- 96
B Argon Flow Pyrolysis Apparatus......................... 102
C Stirred Flow Pyrolysis Apparatus................. 103


TABLES
TABLE
I. Relationship of and Electron Affinities of R- to the Relative
Migratory Abilities of Alkyl, Alkenyl and Aryl Groups in the Gas
Phase Reaction of Me3SiR with OH".................... 4
n. Relationship of the Relative Migratory Abilities of Alkyl, Alkenyl and
Aryl Groups to Taft a* and Gj constants for the Solution Phase
Reactions of RSiMe2CH2Cl with KF.............. 12
HI. GC-FID Analysis of EtSiMe2CH2Cl + KF Reaction Mixture with
respect to time........................................ 26
IV. GC-FID Analysis of n-BuSiMe2CH2Cl + KF Reaction Mixture
with respect to time................................. 27
V. GC-FID Analysis of large scale n-BuSiMe2CH2Cl + CsF Reaction
Mixture with respect to time......................... 28
VI. GC-FID Analysis of i-PrSiMe2CH2Cl + KF Reaction
Mixture with respect to time......................... 28
VH. GC-FID Analysis of cy-PrSiMe2CH2Cl + KF Reaction
Mixture with respect to time......................... 29
VUL GC-FID Analysis of H2C=CHSiMe2CH2Cl + KF Reaction
Mixture with respect to time........................... 30
IX. GC-FID Analysis of PhSiMe2CH2Cl + KF Reaction
Mixture with respect to time........................... 31
X. Generalized Product Distribution from Reactions of
ArSiMe2CH2Cl with KF................................... 38
XI. Grouping and Ratios of Primary and Secondary Products from
Reactions of ArSiMe2CH2Cl with KF...................... 38
XII. Detailed Gravimetric Preparation of PhSiMe2CH2Cl vs.
p-Me-PhSiMe2CH2Cl + KF Competition Reaction Mixture.. 55
Xm. Detailed Gravimetric Preparation of Hexadecane/PhSiMe2CH2Cl,
p-Me-PhSiMe2CH2Cl Reference Solution................... 55
XTV. Example of GC-FID Response Factor Calculation from Reference


xi
Solution Injection...................................... 56
XV. Example of GC-FID Quantification of Competition Study 3.3.3a
Results.................................................. 56
XVI. Summary of Competition Study 3.3.3a->d Results........... 57
XVn. Summary of ArCH3 Appearance Rate Study (3.4.3a->e) Results.. 64
XVm. Results of the 560C Argon Row Pyrolysis of ArSiHMeCH^Cl.. 75
XEX. Results of the 700C Argon Row Pyrolysis of ArSiMe2CH2Cl... 76
XX. Results of the 560C Stirred Row Pyrolysis of ArSiHMeCH2Cl.. 79
XXI. Results of the 700C Stirred Row Pyrolysis of ArSiMe2CH2Cl..
80


FIGURES
FIGURE
l. Plot of log ([Me3SiO_] / [Me2RSiO]) (statistically corrected)
vs.AH^CofR-H)............................................. 5
n. Plot of log ([R migration] / [Me migration]) vs. AH^ (of R-H)
for Solution Phase Reactions (2.6.1-2.6.7) of RSiMeoCHoCl
withKF..................................................... 13
m. Plot of log ([R migration] / [Me migration]) vs. Taft a* for Solution
Phase Reactions (2.6.1-2.6.7) of RSiMe2CH2Cl with KF...... 13
IV. Plot of log ([R migration] / [Me migration]) vs. Hammett Oj
Constants for Solution Phase Reactions (2.6.1-2.6.7) of
RSiMe2CH2Cl with KF................................. 14
V. Simple Pentacoordinate Mechanism for Solution Phase
Rearrangements of RSiMe2CH2Cl Induced by Fluoride Ion
Attack at Silicon......................................... 15
VI. 13 -elimination Mechanism of the Reaction of (ClCH2)2SiMe2 with
KF to form Me2SiF2 and Ethylene........................... 16
VII. Illustration of the Proton NMR Splitting of Cyclopropanes. 25
Vm. Plot of log ([Me2SiF2]x / [Me2SiF2]n) vs. Hammett ox Constants
for Solution Phase Reactions (3.2.3a->e) of ArSiMeoCHoCl
with KF...............................................:.... 39
IX. Plot of log (Z2/Zl ) Products vs. Hammett Gx Constants for
Solution Phase Reactions (3.2.3a->e) of ArSiMe2CH2Cl with KF., 40
X. Plot of log ([X-Ph-Me] / [p-F-Ph-Me]) vs. oF Constants for
Solution Phase Reactions (3.2.3a->e) of ArSiMe2CH2Cl with KF. 41
XI. Detailed Mechanism of Primary and Secondary Product Formation
for Solution Phase Reactions (3.2.3a->e) of ArSiMe2CH2Cl with
KF........................................................ 43
XII. Plot of log Kx / Kh vs. Hammett ax Constants for the Relative Rates
of Starting Material Consumption Study (3.3).............. 53
XHL Reaction Profiles from the Rate of A1CH3 Appearance Study (3.4). 60


xin
XTV. Plot of log Kx / Kh (calculated at three hours reaction time) vs.
Hammett ax Constants for the Relative Rates of ArCH3 Appearance
Study (3.4)........................................... 62
XV. Two Step Radical Mechanism for the Pyrolytic Rearrangement of
PhSiMe2CH2Cl to ClSiMe2CH2Ph and CISiPhMeEt............ 82
XVI. Possible Carbene Elimination Mechanism of ArSiQ2Me formation
from the Sealed Tube of Argon Flow Pyrolysis of
ArSiMe2CH2Cl........................................... 83
XVH. Test for a Carbene Elimination Mechanism................. 84
XVm. Unimolecular Mechanism for the Gas Phase Rearrangement of
HMe2SiCH2Cl to Me3SiCl................................. 85


INTRODUCTION
In the field of organosilicon chemistry, the determination of a reaction
mechanism is a complex task because of silicon's ability to exist as a positive
"siliconium ion (A)1;" in neutral, tetracoordinate species (B); as well as in singly and
doubly charged penta and hexacoordinate complex anions (C and D).
v R4 R3 V/ Si R ^ NR2 nl 2 R V'- /Sl'-Rs r3 _ R, ri/S|1Vxr6 Ro
A B c D
In order to arrive at a valid mechanism, one must either perform a very large
number of experiments, or more preferably, study the recent literature concerning the
reaction of interest and develop a logical series of experiments which examine the
reaction under varying conditions. Doing so may elucidate different mechanisms
effecting different or identical reactions on the same molecule.
This was the case as seen by our research team working under Dr. Robert
Damrauer in studying the base-induced solution phase and gas phase rearrangements
RR'R"SiCH2Cl + B -----------> BSiR'R"CH2R + BSiRR"CH2R' +
BSiRR'SiCH2R" + Cl' (eq.l)
RR'R"SiCH2Cl
A
> ClSiR'R"CH2R + ClSiRR"CH2R'
+ ClSiRR'Cf^R" (eq.2)


2
of chloromethyl substituted silanes (eq. 1 and 2). We found that these compounds are
unique in that alterations of reaction conditions result in a shift of reactivity between
silicon and the chloromethyl carbon.
The solution phase rearrangements (eq. 1) induced by base or anionic attack at
silicon have been shown by us2 and others3 to occur only in aprotic media aided by
phase transfer catalysts.
The use of protic media was found4 to shift the reactive site to the
chloromethyl carbon and produce only the Sn1/Sn2 nucleophilic displacement product
RR,R"SiCH2B + Cl'.
This change of reactivity may or may not result in alterations of the product
formed (e.g., back in 1965, R.W.Bott, C. Eabom and B.M. Rushton found that
Me3SiCH2Cl reacts with AICI3 a Lewis acid to give ClSiMe2Et5, the same product
which would now be recognized to form from various pyrolytic reactions performed
on the same compound).
The scope of this thesis covers the mechanistic details of rearrangements (to
the chloromethyl substituted carbon) of various R groups attached to a chloromethyl
substituted silane. These rearrangements were brought about in two ways: (1) by
solution phase reaction with fluoride anion, and (2) by thermal or pyrolytic methods.


CHAPTER I
GROUNDWORK FOR REARRANGEMENT STUDIES
1.1 Gas Phase Acidities of Hydrocarbons
1.1.1 Introduction and Review
The work performed by C.H. DePuy, V.M. Bierbaum and R. Damrauer6 on
the gas phase reaction of substituted trimethylsilanes ((CH3)3SiR) with hydroxide ion
provided data which supported an anionic cleavage mechanism forming the products
listed in equations 3 and 4.
(CH3)3SiR + OH-------------> (CH3)3SiO- + R-H (eq. 3)
(CH3)3SiR + OH" ------------> (CH3)2RSiO- + CH4 (eq. 4)
The mechanism was found to involve the formation of a pentacoordinate
hydroxy-silane anion, (CH3)3SiOHR", which transfers one of the three methyl
groups* or the R group as a carbanion to the hydroxyl proton, forming CH4 or R'H in
addition to the siloxide anions which were observed and quantified. *The true R/CH3
migratory ratios were statistically corrected to be three times the observed siloxide
ratios. (= 3 [(CH3)3SiO-] / [(CH3)2SiRO-])
These R/CH3 migration ratios were used in this preliminary study and in our
subsequent studies to probe the mechanistic details of carbanionic rearrangements of
chloromethylsilanes.
Four R groups (phenyl (Ph), H, vinyl (H2C=CH) and methyl (Me)) with


4
known AH^ values for the gas phase reaction (eq. 5), were used to generate a linear
free
*%)---------------> R(g)' + H(g)+ (1- 5)
energy relationship between the log of the siloxide ratio (eq. 3 / eq. 4) measured
(theoretically equivalent, after statistical correction*, to the log of R/Me migration) and
AHacy. This line was used to predict approximate AHacid values for five other RH
hydrocarbons (R = ethyl (Et), isopropyl (i-Pr), t-butyl (t-Bu), cyclopropyl (cy-Pr) and
methoxymethyl (CH3OCH2)). These data are represented in Table I and Figure I.
Electron affinities (see eq. 6) of R" were calculated from AHacid and AHbond-dissoc.-
R-(g) + e" ------------> R" (g) (eq. 6) EA = AH for process
The correlation coefficient of the entire data set for log
[siloxide(3)/siloxide(4)] vs. AH^y was seen to be -0.998 with a slope of -0.0828.
Table I
Relationship of AHnd and Electron Affinity of R- to the Relative Migratory Ailities of
Alkyl, Alkenyl and Aryl Groups in the Gas Phase Reaction of Me3SiR with OH'
R t(CH3)3SiO-]* [(CH3)2RSiO'] AH^tRH) kcal/mole EA(R-) kcal/mole
Et 0.44 421 -9 weakest
i-Pr 0.61 419 -11 1
Me 1.0 416.6 1.8 1
t-Bu 1.7 414 -7 1
cyPr 2.4 412 8 1
CH3OCH2- 6.0 407 0 1
ch2ch 7.0 406 17 1
H 20 400.4 17.4 1
Ph 24 400.7 25 strongest
To investigate further electronic effects, the same log-ratio set was plotted


5
against Taft o* and Oj constants.2 Slopes (correlation coefficients) obtained
respectively were 1.55 (0.876) and 3.21 (0.601).
The correlation of log [siloxide(3)/siloxide(4)] with AHacid clearly indicated
the importance of acid strength in determining the relative cleavage aptitude of R
groups. The fair correlation with a* suggested that the amount of negative charge
localization on R", due to electronic dipole effects, is moderate. Inductive effects, as
shown by the rough correlation of log [siloxide(3)/siloxide(4)] with Hammett
derived Oj constants, seemed to outweigh polar effects (slope of 3.21 vs. 1.55) but
this was only due to the restricted range of Oj constants (0.1064) compared with that
of c* constants (0.6019). Nevertheless, the positive slope did indicate that electron
withdrawing groups lead to more ready cleavage.
Figure I
Plot of log ([Me3SiO*]/[Me2RSiO-]) (statistically corrected) vs. AH^ (of RH)


6
1.2 Pentacoordinate Silicon Anions
Four years prior to the work summarized above, the same research team aided
by five others7, was able to isolate gas phase pentacoordinate silicon anions by the
reactions of tetra-methyl, methoxy, vinyl and trimethyl-R-substituted silanes with
fluoride ion in a flowing afterglow apparatus (eq. 7).
R3siR;9>+
(eq. 7)
The purpose of this study was to effect the generation of stable gas phase
carbanions. This was possible only when R or R' was aryl, benzyl, alkenyl or
alkynyl. The reaction which transpired in these cases was the decomposition of the
pentacoordinate anion (eq. 8) into a free gas phase carbanion (R- or Ru) and a
fluorosilane (R3SiF or R2R'SiF).
R
"s'- ,
R -----------> - + R'" + R-iSiF + R.SiR'F (eq-8)
R (9) (9) 3 (9) 2 (9)
R
Mass spectral verification of gas phase pentacoordinate silicon anions led our
research team to speculate as to whether or not they were implicated as intermediates in
the solution phase reactions of chloromethylsilanes with fluoride and alkoxide anions.
This question was answered affirmatively through the work of S.E. Danahey8
with the reactions of various organofluorosilanes with stoichiometric amounts of
potassium fluoride and 18-Crown-6 (phase transfer catalyst) in aromatic solvent (eq.
9). The typical product isolated from these reactions was a crytalline 'caged'
potassium difluorosiliconate.


7
RR'R-SiF
toluene
lS-Crvm-4
0
0
K
+
F
R\|- ,
+ ^ Si R1 (eq.9)
|
R"
Eight of these compounds were characterized by multmuclear cryogenic FT
NMR (lH, and 29 si) as to their geometries and tendencies to undergo
psuedorotational processes similar to pentavalent phosphorous.


CHAPTER n
THE RELATIVE MIGRATORY ABILITIES OF ALKYL, VINYL AND ARYL
GROUPS IN SOLUTION PHASE FLUORIDE INDUCED REARRANGEMENTS
OF CHLOROMETHYLSILANES
2.1 Introduction
To find out if the migration of weakly acidic R groups to a chloromethyl
carbon in the solution phase reactions of chloromethylsilanes with bases (see eq. 2)
takes place through a pentacoordinate silicon anion similar to those discussed in the
gas phase groundwork studies, we chose to study the reactions of
dimethyl-R-substituted chloromethylsilanes (Me2RSiCH2Cl) with fluoride (eq. 10) so
that the ratios of various R to methyl migrations could be obtained and correlated with
gas phase (AH^a) and solution phase (a and Oj) parameters.
VV
Me2R$iCH2Cl - > Me2$iFCH2R + MeRSiF(Et) (eq.10)
18-Crown-6
R migration Me migration
Fluoride anion, a very hard base, was chosen because the strength of the Si-F
bond (154 kcal/mole)9 in the products enables their easy identification by GC/MS with
electron impact ionization. Phase transfer catalysts such as 18-crown-6 (I), silacrown
ethers (II) or quaternary ammonium salts (III) were found necessary to release fluoride
anion in aprotic solvents, the medium which gave the cleanest reaction mixture.


9
I
II
(n-Bu)4N+/F_
(octyl )3NMe/F
III
The source of the fluoride ion used in the optimized experiments with 18-Crown-6
was anhydrous potassium fluoride.
2.2 SUMMARY OF OPTIMIZATION EXPERIMENTS
Reactions of RSifCEtyoCFbCl in the condensed phase with
fluoride anion/18-crown-6 system in aromatic solvents
Introduction:
The optimization of the reaction conditions (temperature, fluoride source,
phase transfer catalyst, solvent and apparatus) was conducted for R = Me
(chloromethyltrimethylsilane) by S.E. Danahey10. This material has only one
possible migratory product, dimethylethylfluorosilane, resulting from the reaction in
equation 11.
Me3SiCH2Cl + F -----------> Me2SiF(Et) + Cl' (eq. 11)
An interesting alteration in the relative rate (and yield) of product formation
was noted for R = Me by S.E. Danahey11 and with R = Et as the alkali fluoride source
changed from potassium to cesium. In both cases the cesium reactions caused a 250
fold increase in the rate of starting material consumption and a 50 % increase in the
total product yield. These reactions were conducted in 20 ml serum cap vials at 60C


and 74C in 0.5 ml toluene and o-xylene respectively. The stirring and temperature
control of these and further reactions was controlled by a thermostated/magnetically
stirred oil bath manufactured by Sybron, Inc. The usual amount of reactants used in
these and further reactions of quantitative interest in chapter 2 was:
RSi(CH3)2CH2Cl = 5 mmol; KF (oven dried, Aldrich reagent grade) 0.55 g (10
mmol); 18-Crown-6 (Aldrich) = 0.13 g (0.5 mmol) in 0.5 to 2.0 ml of aromatic
solvent, the choice of which depended on the volatilities of starting materials and
products. The molar ratio of reactants was held constant throughout the study.
Additional Notes:
Cesium fluoride reactions were found to proceed almost to completion in the
absence of phase transfer catalyst and solvent12, but GC/MS analyses of reaction
mixtures were found to contain vast amounts of siloxanes [(RMe2Si)20]*, some of
which obscured the migration products of interest These are due to the hygroscopic
nature of even oven dried cesium fluoride. The release of its moisture during reactions
of Si-F bonds brings about the formation of Si-O-Si bonds. Its use apart from an
additional large scale preparation (7X the usual amounts of reactants) of the n-butyl
products (2.5.3) was discontinued. Sodium fluoride was found to give no reaction
with Me3SiCH2Cl13, whereas, potassium iodide gave the substitution product
Me3SiCH2I14.
Additional fluoride/phase transfer systems such as (n-Bu)4N+/F-/THF and
(octyl)3NMe+/Cl* + KF/toluene were tried but gave poor yields of migration products
and complex reaction mixtures.15
Pyridine and the non-aromatic solvents dimethylsulfoxide (DMSO),
acetonitrile (MeCN) and dimethylformamide (DMF) were also tried, but also resulted


in complex product mixtures16.
The chloromethyldimethyl-R silanes studied in section 2.6 were ethyl,
methyl, isopropyl, cyclopropyl, n-butyl, vinyl, phenyl and silacyclopentano. The
source of the R = methyl, vinyl and phenyl compounds was Petrarch Systems, Inc.;
l-(chloromethyl)-l-methyl-silacyclopentane was prepared by R. Damrauer via a
di-Grignard reaction:
Cl
Syntheses of the ethyl, n-butyl, isopropyl and cyclopropyl are contained herein.


2.3 Results and Discussion
It here again must be noted that the true (statistically corrected) R/CH3
migration ratio for the reaction of one mole of chloromethyl dimethvl-R-substituted
silane is two times the observed product ratio on the GC-FID. The results contained in
Table II were optimized for each reaction before the time (ORT) when secondary
product formation was first noted and the peak heights of the primary products were
maximized.
Table n
Relationship of the Relative Migratory Abilities of Alkyl, Alkenyl and Aryl Groups to
Taft a* and oj Constants for the Solution Phase Reactions of RSiMe2CH2Cl with KF
R ORTfhrs'i R/CH-2 AHVid a a:
Et 264 1.5 421^ -0.1019 -0.05518
Me 336 1 416.617 0.0019 -0.04618
Cypent 9.0 OO
i-Pr 168 0.67 41917 -0.1919 -0.06418
cyPr 264 0.57 41217 0.1118
n-Bu 168 0.87 42019 0.1319 -0.06018
Vy 48 20 40617 0.5620 0.0519
Ph 1.5 16 39917 0.6019 0.1019
The correlations of log(R/Me) migration with AH0^, (J* and constants are
given below in Figures II to IV. (It might be mentioned here that the AHacid values
contained in reference 6 included t-butyl instead of n-butyl. The synthesis of
chloromethyl- dimethyl(t-butyl)silane was attempted from t-butyl lithium and
chloromethyldimethylchlorosilane in pentane, but failed. We were fortunate to find the
AH0a^d value listed in reference 19 for n-butane).
The cr* constants utilized depict the relative abilities of R group anions to


[R migration] log [R migration!
[Me migration] o ^S^ation]
Figure II
t of log ([Me3SiO-] / [Me2RSiO']) vs. AH^ (of R-H) for Solution Phase
Reactions (2.6.1-2.6.7) of RSiMe2CH2Cl with KF
Figure m
Plot of log ([R migration] / [Me migration]) vs. Taft a* for Solution Phase Reactions
(2.6.1-2.6.7) of RSiMe2CH2Cl


1 4
localize a negative charge via dipole field effects (field effects); a negative a* indicates
a highly localized electron pair on a non-polar substituent, and vice versa for
Figure IV
Plot of log ([R migration] / [Me migration]) vs. Hammett Constants for Solution
Phase Reactions (2.6.1-2.6.7) of RSiMe2CH2Cl with KF
a positive o*. The ability of an R group to stabilize a negative charge through
inductive processes are represented by the Oj constants; a negative ctj indicates
electron flow away from the R group (relative anionic instability) and vice versa for a
positive Oj.
The slopes (correlation coefficients) listed in this thesis for loglO(R/Me)
migration vs. the a* and AHacid were correctly reported, unlike the 1.81 (0.943) and
-0.0674 (0.904) reported in a paper submitted to JACS in 198421. This was a result
of an erroneous report of the ratio of cyclopropyl/methyl migration as 7.0 instead of
0.57. The error was reported in a subsequent paper submitted to Organometallics in


198522, but as 0.66.
The difference (1.82 vs 9.06) in the slopes of the a*, CTj correlations is only
due to the restricted range of Oj constants (0.1064) compared with that of a* constants
(0.619). Such a comparison serves no purpose because a constants are methyl
based whereas aj constants are proton based. The small slope obtained with the
AH0.^ correlation in both the solution and gas phase experiments is only indicative of
the 5.5 % variance between the AHacjd values of the R groups studied. The
0.82-0.95 correlations of all three parameters with the R/Me migration ratio combined
with the similar results from the gas phase study mentioned in section 1.1 led us to
postulate the mechanistic model below:
Me2RSiCH2Cl ---->
F
Si CHoCl
/\ T
Me Me
Me2SiFCH2R
> MeEtSiFR
Figure V
Simple Pentacoordinate Mechanism for Solution Phase Rearrangements of
RSiMe2CH2Cl Induced by Fluoride Ion Attack at Silicon
This involves the formation of a pentacoordinate anionic intermediate where
Me and R migrate competitively to displace chloride ion. The behavior of this solution
phase intermediate is analogous to that of real gas phase fluorosilicate anions (section
1.2) in that aryl and vinyl groups are more likely to separate as freer carbanions than
are alkyl groups, thus effecting a higher R/Me migration ratio. The driving force for
pentacoordination is the thermodynamic stability of the Si-F bond (20-40 kcal/mol)23,
and the migratory efficiency is controlled largely by anion stability. Pentacoordination
at silicon is easily accomplished through bonding with fluorine. The thermodynamic


stability of the Si-F bond apparently lowers the energy of the unoccupied 3d orbitals
sufficiently to cause formation of a resultant trigonal bipyramidal dsp^ hybrid orbital.
This is not suprising in view of the fact that the hexacoordinate SiFg^- anion has been
known for years, uses d^sp^ hybrid orbitals and is very stable.
The GC-MS identification of Me2Si(CH2R)2 in the R = vinyl and phenyl
cases and Me2SiF2 in the phenyl case was the basis for the further mechanistic studies
of substituted arylchloromethyldimethyl silanes in chapter 3 of this thesis.
A mechanism for the H2C=CH2 and Me2SiF2 forming reaction (2.5.10) of
bis(chloromethyl)dimethylsilane and KF is postulated below:
Me2Si(CH2Cl)2
&
Cl _ci
- Me2$i.

Me?Si --
V-*
Cl
ch2
Me2SiF2CH2CH2Cl
I-
--- Me2(ClCH2CH2)SiF
d elimi nation
Me2$iF2 + h2C = CH2
Figure VI
B-elimination Mechanism of the Reaction of (ClCH2)2SiMe2 with KF to form
Me2SiF2
This involves a total of 2 moles of fluoride per mole of starting material, each reacting
to form a different pentacoordinate adduct.The first adduct undergoes a chloromethyl
migration with loss of chloride and the second undergoes a B elimination of chloride
from the chloromethyl migratory product to form ethylene and dimethyldifluorosilane.



1 7
2.4 Alkoxide Rearrangements of Chloromethvlsilanes
An interesting parallel to the fluoride reactions contained herein is found in the
work of R.L. Kreeger, P.R. Menard, E.A. Sons and H. Shechter with methoxide
anion.24
Part of their study involved the reaction of methoxide anion with
phenylchloromethyldimethylsilane in either methanol (eq. 12) or dioxane (eq.13), with
and without the addition of 18-Crown-6. Analogous phenyl (a) and methyl (b)
migration products were formed in somewhat similar (at least in magnitude) ratios
when NaOMe/dioxane was used at 30C and 60C. The use of methanol (a polar
protic solvent) with sodium, potassium and cesium methoxide resulted in products
formed via: chloride displacement by MeO' (c) (in decreasing quantities respectively);
chloromethyl displacement by MeO" attack at Si (d) and MeO' attack at Si, displacing
benzyl anion from (a) found in the form of toluene (e) (both in increasing quantities
respectively). No migratory products were found in the MeOH system whereas no
chloride displacement was found in the aprotic dioxane system.
PhSiMe2CH2Cl + MeO'/dioxane-----> PhCH2SiOMeMe2 (a) + EtSiOMePhMe (b)
+ small quantities of (d) and (e) (eq. 12)
PhSiMe2CH2Cl + MeO'/MeOH--------> PhSiMe2CH2OMe (c) + PhSiMe2OMe (d)
+ 'CH2CI (observed as CH3CI) + Me2Si(OMe)2 (e)
+ "CH2Ph (observed as MePh) (eq. 13)
The formation of Me2Si(OMe)2 and toluene is analogous to that of Me2Sil?2
and the substituted toluenes contained in the next chapter. The effect of the crown


ether used in both solvents was to increase and decrease the yields of Me2Si(OMe)2
and PhCH2SiOMeMe2 (in the dioxane system) respectively.
The yield of Me2Si(OMe)2 with respect to time (in the methanol system) was
found to increase drastically when CsOMe was used. This is entirely analogous to our
findings with CsF reactions and suggests a higher extent of ionic dissociation with
cesium salts in both protic and aprotic media. That Me2Si(OMe)2 was found in the
methanol system, is suggestive of the instability of benzylic silanes in protic ionic
media and perhaps that the pentacoordinate intermediate [PhSiOMeMe2CH2Cl]" is
rather short lived in such media. This is not unlikely in view of the large proportion of
the chloride displacement product (c) formed in the methanol system; hydrogen
bonding decreases the "hardness" of the methoxide base.
Although no RSiMe2F or RSiMe2CH2F analogues were detected by S.E.
Danahey16 in polar solvents, it is clear that H > F bonding from hygroscopic CsF or
polar solvents creates messier reaction mixtures by providing alternate reactions (esp.
siloxane formation from trace moisture).


2.5.
Synthesis of Starting Materials
2.5.1 Preparation of CEfoCF^SiCCHg'bCF^Cl
I. C2H5Br + Mg -----------> C2H5MgBr
Into a dry, 3 necked 500 ml flask equipped with an addition funnel, a helium
supplied reflux condensor and magnetic stirrer, was placed magnesium turnings (3.06
g, 0.1259 mol, Fischer) and one crystal of iodine. 100 ml anhydrous ether was
added, and the system was purged with helium. Into the addition funnel was placed a
solution of ethyl bromide (15.6 ml, 0.21 mol, Fischer) in 30 ml ether. Excess
ethylbromide was used to compensate for its volatility. The halide solution was added
slowly at a 0.5 drop/sec rate. Grignard formation began in fifteen minutes and was
complete after fifty minutes. The mixture was refluxed gently for an additional hour.
fl. C2H5MgBr + ClSi(CH3)2CH2Cl----> CH3CH2Si(CH3)2CH2Cl +
MgBrCl
Using an identical set-up, 50 ml of Grignard supernatant was syringe
transferred into a helium purged (septum sealed) addition funnel. The 500 ml flask
contained a solution of ClSi(CH3)2CH2Cl (13.6 g, 0.095 mol, Petrarch) in 40 ml
ether. The Grignard was added dropwise over a fifteen minute period, during which
time a gentle reflux of ether and MgBrCl precipitation was observed. 70 extra ml ether
was added and the mixture was refluxed under gentle heat overnight.
The unreacted ClSi(CH3)2CH2Cl and products were extracted and
hydrolyzed with 200 ml saturated, aqueous NH4CI solution. The salts dissolved as
the double ammonium salt (NH4Cl.MgBrCl) in the aqueous layer. Unreacted
chlorosilane starting material hydrolyzed to the siloxane((Si(CH3)2CH2Cl)20) and
dissolved in the ether layer along with the main product. The flask contents were
transferred into a separatory funnel; the lower, aqueous layer was discarded and the


20
upper ether layer was dried over anhydrous Na2S04. The drying agent was filtered
through a Buchner funnel and most of the ether was removed under gentle vacuum
through a rotary evaporator. Nine grams was collected (15 ml) of a light yellow fluid,
shown by GC-FID to be a 50:50 mix of product and siloxane.
Distillation of this fluid at atmospheric pressure through a 20 cm Vigreux
column afforded 2.83 g, (21.8 % yield) of 99.5 % pure CH3CH2Si(CH3)2CH2Cl:
MW: 136.7; bp: 90C/630 mmHg; NMR (CDCI3/CHCI3): 5 = 0.12 (s, 6H,
Si(CH3)2), 0.6-0.76 (t, 3H, CH3CH2Si), 0.9-1.0 (q, 2H, CH3CH2Si), 2.8 (s,
2H, SiCH2Cl)*.
2.5.2 Preparation of n-C^pSifCHg'bCHoCl
I. n-C4H9Cl + Mg -------------> n-C4H9MgCl
In a dry 500 ml 3-necked flask equipped with an addition funnel, a nitrogen
supplied reflux condensor and magnetic stirrer, was placed magnesium turnings (10 g,
0.41 mol, Fischer), a couple of iodine crystals and 50 ml anhydrous diethyl ether.
The system was purged with nitrogen. Grignard formation was initiated by adding 5
ml pure n-butyl chloride through the funnel. This was followed by addition of a
solution of 40 ml n-butyl chloride in 60 ml diethyl ether over a 90 minute period. The
total amount of n-butyl chloride used was 45 ml (39.88 g, 0.43 mol, Fischer). The
reaction was complete within two hours of initial Grignard formation.
H. n-C4H9MgCl + ClSi(CH3)2CH2Cl----> n-C4H9Si(CH3)2CH2Cl +
MgCl2
Using an almost identical set-up, 55 ml (0.15 mol; 40 % excess) of
n-butylmagnesium chloride solution was syringe transferred to a septum sealed
addition funnel atop a nitrogen purged 3 necked flask which contained
ClSi(CH3)2CH2Cl (15 g, 0.105 mol, Petrarch) dissolved in 50 ml diethyl ether. The


21
Grignard was added over a period of one hour. No immediate refluxing without
external heat was observed, so gentle heating was used to reflux this reaction over a 48
hr period, during which time most of the ether evaporated.
At this time a resultant huge cake of MgCl2 was broken up with 800 ml ether
(in small portions at a time). The ether/MgCl2 sludge was removed and hydrolyzed
with 200 ml saturated aqueous NH4CI solution to decompose unreacted Grignard and
dissolve magnesium salts. The aqueous layer was separated from the organic layer
with a separatory funnel. The organic layer was dried over anhydrous sodium sulfate.
After drying, the sulfate was filtered, and ether evaporated under gentle vacuum with a
rotary evaporator. Crude n-C4H9Si(CH3)2CH2Cl (21 ml) was obtained. The crude
product was distilled under vacuum through a 20 cm Yigreux column.
n-C4H9Si(CH3)2CH2Cl: 11.76 g (68.1 % yield; 97 % pure by GC-FID); MW:
164.75; bp: 67C/30 mm Hg; NMR (CDCI3/CHCI3): 8 0.11 (s, 6H,
Si(CH3)2), 0.65 (t, 3H, CH3CH2CH2CH2Si), 0.90 (m, 4H, CH3CH2CH2CH2Si
), 1.32 (t, 2H, CH3CH2CH2CH2Si), 2.8 (s, 2H, SiCH2Cl).
2.5.3 Preparation of (CH^loCHSifCHgloCFbCl
I. (CH3)2CHC1 + 2 Li --------------> (CH3)2CHLi + LiCl
Lithium dispersion (19.85 g, 30 wt % in mineral oil, 0.858 mol, Aldrich) was
washed with 800 ml dry pentane in a three neck, one liter flask equipped with an
addition funnel, overhead stirrer, Friedrichs condensor with argon gas inlet and a
filter/drain system designed by myself. This consisted of a simple teflon stopcock
drain fused to the bottom center of the flask via 1/4 inch glass tubing. Directly above
the stopcock was placed 2 inches of fine glass wool filter. Above this was placed a
layer of sand on top of which was a layer of coarse, crushed borosilicate glass. This,
in conjunction with the overhead stirrer, was used to scrape the surface of lithium


22
particles to increase the rate of formation and yield of active lithium reagent. Gentle
argon pressure was used for draining the mineral oil, pentane wash.
Through the funnel, 200 ml pentane was added to the clean lithium along with
10 ml pure (CH3)2CHC1 to start reaction. The balance of the chloride (30 ml), in 200
ml pentane under argon, was added over a period of two hours with overhead stirring.
Total chloride used was 40 ml (33.7 g, 0.429 mol, Kodak). Heat from reaction
refluxed the pentane for two hours after addition.
All but 4 ml of the Li reagent solution (300 ml) was drained under gentle
argon pressure through a septum atop into a 500 ml addition funnel attached to an
argon purged one liter, three necked flask equipped with a Friedrichs condensor and
magnetic stirrer. The remaining 4 ml was titrated under argon with 1.35 ml sec-butyl
chloride in a 100 ml three necked, flask containing the lithium reagent, 20 ml toluene,
0.5 ml of a solution of 100 mg 2,2-biquinoline indicator in 50 ml toluene.This
indicator changes the solution color to yellow when the lithium reagent is completely
reacted. The concentration of the lithium reagent was found to be 0.34 M.
D. ClSi(CH3)2CH2Cl + 0.33 SbF3 -----> FSi(CH3)2CH2Cl + 0.33
SbCl3
A 100 ml single neck flask with a magnetic stirring bar was charged with
SbF3 (11.45 g, 0.0641 mol, Aldrich). This was cooled to -78C with a dry
ice/acetone bath. A reflux condensor with nitrogen inlet was attached and the flask
was pinged. Under slow agitation of the fluoride, ClSi(CH3)2CH2Cl (25.0 g, 0.175
mol, Petrarch), via syringe, was added. The solution fumed at first, and then
solidified within five minutes. After removal of the cooling bath, the reaction mixture
was then allowed to warm up to ambient temperature over a period of 4 hours.
During this period, a gentle reaction occurred with a color change of light brown to


23
purple and the precipitation of antimony trichloride.
A GC of the supernatant fluid showed an almost quantitative conversion to
FSi(GH3)2CH2Cl. The crude product was distilled through a 20 cm Vigreux column
to yield FSi(CH3)2CH2Cl: 19.11 g (19 ml, 86.4 % yield, 98.2 % pure); MW:
126.6; bp: 76C/630 mmHg.
m. (CH3)2CHLi + FSi(CH3)2CH2Cl-----> (CH3)2CHSi(CH3)2CH2Cl
+ LiF
To the purged flask in I., containing FSi(CH3)2CH2Cl (12.7 g, 0.10 mol) in
20 ml pentane, was added the lithium reagent solution (0.34 M, 300 ml, 0.102 mol Li)
with magnetic stirring. Gentle reflux began and white clouds of LiF were observed
indicating reaction. The reaction mix was stirred for 96 hours, at which time there was
no further increase in product or decrease in starting material as observed by GC-FID.
The reaction mixture was hydrolyzed with 40 ml cone. HC1 in 200 cc crushed
ice. When all clouds of LiF had dissolved, the hydrolysis mixture was transferred to a
one liter separatory funnel. The layers were separated, and the aqueous layer was then
washed with two 100 ml portions of pentane. The organic phase was dried over
anhydrous sodium sulfate. After filtration of the drying agent, pentane was removed
from the crude product via distillation through a 40 cm Vigreux column. 20 ml of a
fluid were obtained. Two major products (close in volatility) in addition to 24.6 % of
the starting material were obtained. Distillation of this fluid at atmospheric pressure
through a 45 cm vacuum jacketed teflon annular spinning band column (mfd by
Nester/Faust Corporation) afforded fair purification of (CH3)2CHSi(CH3)2CH2Cl;
the best fraction collected was shown to be 81.2 % pure by GC.
(CH3)2CHSi(CH3)2CH2Cl: 1.28 g (6.49 % yield); bp: 129C/630 mmHg;
NMR (CDCI3/CHCI3): 8 = 0.10 (s, 6H, Si(CH3)2), 1.0 (s, 7H, (CH3)2CHSi),


2.5.4 Preparation of Dimethvlchloromethvlcvclopropvlsilane
I. H2C=CHSi(Me)2CH2Cl + CH2I2 -> (CH2)2CHSi(Me)2CH2Cl + Znl2 +
Cul
Charged into a dry 500 ml three necked flask were zinc dust (18.8 g, 0.288
mol, Mallinkrodt), copper (I) chloride (2.85 g, 0.029, Mallinkrodt) and 50 ml
anhydrous diethyl ether. An overhead stirrer, Friedrichs condensor with nitrogen inlet
and an addition funnel were then attached to the three necked flask. The system was
purged with nitrogen, and H2C=CHSi(CH3)2CH2Cl (15.0 g, 0.111 mol, Petrarch)
was added through the funnel, with stirring. This addition was exothermic. Over a
fifteen minute period, diiodomethane (38.8 g, 0.145 mol, Baker) was added to the
reaction mixture. Ether reflux, without external heating, was then observed for 2
hours. More ether (30 ml) was added, and the solution was maintained at reflux for
24 hours.
The reaction mixture was then vacuum filtered through a Celite bedded
Buchner funnel into a 500 ml flask. The precipitate was washed with 200 ml ether.
This ether mixture was then hydrolyzed with 200 ml saturated, aqueous NH4CI
solution. The mixture was transferred to a separatory funnel, and the aqueous layer
discarded. The organic layer was dried over sodium sulfate. After filtration from the
drying agent, the ether was removed by rotary evaporation. Fifteen ml of a dark fluid
containing four major products was obtained.
Vacuum distillation of this fluid through a 20 cm Vigreux column afforded
(CH2)2CHSi(CH3)2CH2Cl: 2.73 g (16.6 % yield; 96.3% pure); MW: 148.5; bp:
40C/30 mmHg; !h NMR (CDCI3/CHCI3): 8 = -0.35 (s, 1H, (CH2)2CHSi), 0.07
(s, 6H, Si(CH3)2), 0.25 (sxt, 2H, (CE2)2CHSi ), 0.60 (m, 2H, (CH2)2CHSi),


25
2.80 (s, 2H, SiCH^Cl). Note: Cyclopropanes give complex spectra since Ha and Hb
are not equivalent and not only have slighdy different chemical shifts, but also split
each other (fig.7):
Si(CH2Cl)Me2
Figure 7
Illustration of the Proton NMR Splitting of Cyclopropanes


26
2.6 Optimized RSifCFh'bCHoCl/KF Reactions -Detailed Experimental
General Experimental, see section 2.2.
2.6.1 EtSi(CH3)2CH2Cl + KF------> PrSi(CH3)2F + Et(Et)Si(CH3)F + KC1
Time of first appearance of migratory products: 48 hrs.
Initial Ratio Et/Me migratory products: 0.75 at 168 hrs.
Subsequent GC-FID analysis (additive peak height based) performed
on a Perkin Elmer 3020B with a 6', 1/8 inch OV-l/chromosorb
column.
Table in GC-FID analysis
Reaction Time (hrs) % completion wrtS.M. Et/Me product ratios
168 <5 0.75
264 7.4 * 0.75
456 10.0* 0.60
600 10.0* 0.60
768 48* 0.60
*
appearance of secondary products
GC/MS analysis: The ethyl migration product was found to elute first, from
an identical column in a Finnegan 3200 with a Teknivent data system. This was
followed by the adjacent methyl migration product, starting material, solvent, three
side products and 18-Crown-6.
Relevant ions and (% abundances) for the main products:
I. PrSi(CH3)2F M+ 120 (5); M+- Me 105 (12.5); M+- Pr 77 (100)
H. Et(Et)Si(CH3)F M+ 120 (8); M+- Me 105(4); M+-Et 91(100)
2.6.2 n-BuSi(Me)2CH2Cl + KF > PeSi(Me)2F + Et(n-Bu)Si(Me)F + KC1
n-Bu migration Me migration
Time of first appearance of migratory products: 72 hrs.


27
Initial Ratio n-Bu/Me migratory products: 0.541 at 72 hrs.
Table IV GC-FID analysis
Reaction Time fhrsl % completion wrt S.M. n-Bu/Me product ratios
168 10* 0.429
336 ~10 0.456
912 31** 0.438
* formation of two major side products
** formation of four major side products
GC/MS analysis: The n-butyl (or R') migration product was again found to
elute first and in smaller quantity than the adjacent methyl migration product. These
were followed by mesitylene solvent, starting material, four major side products and
18-Crown-6.
Relevant ions and (% abundances) found for the main products were:
I. PeSi(CH3)2F M+ 148 (1.67); M+- Me 133 (13.3); M+- Pr 105
(22); M+- Bu 91 (7.5); M+- Pe 77 (100);
n. Et(n-Bu)Si(CH3)F M+ 148 (5.0); M+- Me 133 (5.0); M+- Et
119(100); M+-Pr 105(3.33); M+-Bu 91(100);
2.6.3 Large Scale Preparation of n-CgHnSiCCTh'bF and Et(n-Bu)Si(CH3)F from
n-BuSiCCHg^CHoCl + CsF/18-Crown-6 in mesitvlene solvent at 80C
A 25 ml round bottom flask was charged with n-BuSi(CH3)2CH2Cl (5.8 g,
0.0352 mol), mesitylene (4.23 g, 0.0352 mol, Fischer), cesium fluoride (10.7 g,
0.0704 mol, Aldrich) and 18-crown-6 (0.93 g, 0.00353 mol, Aldrich). A magnetic stir
bar was added along with a nitrogen supplied reflux condensor. This system was
placed in an 80C oil bath.


28
Observations:
Time of first appearance of migratory products: 3.5 hrs.
Initial Ratio n-Bu/Me migratory products: 0.463 at 3.5 hrs.
Table V GC-FID analysis
Reaction Time fhrsl % completion wrt S.M. n-Bu/Me product ratios
24 42* 0.463
48 <55.4** 0.453
72 90.6*** 0.468
* initial production of three main low volatile side products
** % of total products being transmuted to side products = 39.5
*** stopped reaction by freezing to -20C at this point
The cold reaction mixture was filtered under gentle vacuum with a few extra
ml of mesitylene. The filtrate (20 ml) was distilled through a 45 cm annular teflon
spinning band column (vacuum jacketed). This afforded mesitylene free fractions of
PentylSi(CH3)2F and Et(n-Bu)Si(CH3)F (1.6 g mixture; 10.8 mmol; 30.7 % yield;
bp 113C / 630 mm Hg). Structure verifications were performed, again, by GC/MS.
2.6.4 i-PrSi(Me)2CH2Cl + KF > i-BuSi(Me)2F + Et(i-Pr)Si(Me)F + KC1
(81.2 % purity) i-Pr migration Me migration
Time of first appearance of migratory products: 24 hrs.
Initial Ratio i-Pr/Me migratory products: 0.35 at 24 hrs.
Table VI GC-FID analysis
Reaction Time (hrsl % completion wrt S.M. i-Pr/Me product ratios
72 10 0.288
96 10.5 0.30
168 11.1 0.30
240 76* 0.36
* rapid appearance of three major side products in about 5X the
quantity of the main products. Reaction mixture was frozen and
stopped at this point.


29
GC/MS analysis: As expected, the isopropyl or R migration product was
found to elute before the larger adjacent methyl migration product. These were
followed by starting material (+ impurity), mesitylene solvent, three major side
products and 18-Crown-6.
Relevant ions and (% abundances) found for the main products were:
I. i-BuSi(CH3)2F M+134 (2.00); M+- Me 119(15.0); M+-i-Pr 91
(24); M+- i-Bu 77 (100)
H. Et(i-Pr)Si(CH3)F M+ 134 (5.00); M+- Me 119 (7.50); M+- Et
105 (100); M+- i-Pr 91 (100)
2.6.5 cy-PrSi(CH3)2CH2Cl + KF------> cy-PrCH2Si(CH3)2F + KC1
+ Et(cy-Pr)Si(CH3)F
cy-Pr migration Me migration
Time of first appearance of migratory products: 96hrs.
Initial Ratio cy-Pr/Me migratory products: 0.230 at 168 hrs.
Table VII GC-FTD analysis
Reaction Time cy-Pr/Me
fhrsl % comnletion wrt S.M. nroduct ratios
192 5 0.250
264 17,3 0.286
456 40* 0.270
600 66.8* 0.286
768 -100** 0.270
* appearance of two main side products in 2/3 to 3/4 the relative
^ quantity of the major products.
* Reaction mixture was frozen and stopped at this point.
GC/MS analysis: The cyclopropyl (R) migration product once again preceded
the methyl migration product in elution. These were followed by starting material,


30
mesitylene solvent, two major side products and 18-Crown-6.
Relevant ions and (% abundances) found for the main products were:
L cy-PrCH2Si(CH3)2F M+ 132 (8.33); M+- Me 117(6.50); M+-
cy-PrCH2 77 (100)
H. Et(cy-Pr)Si(CH3)F M+ 132 (2.60); M+- Me 117 (10.0); M+- Et
103 (100); M+- cy-Pr 91 (100); M+- (cy-Pr + Me; w/proton
rearrangement?) 77 (72)
2.6.6 C2H3Si(CH3)2CH2Cl + KF---------> C2H3CH2Si(CH3)2F +
Et(C2H3)Si(CH3)F + KC1
C2H3 migration Me migration
Reaction was carried out by S.E. Danahey25 on a scale 20X the usual amount
of starting materials
Time of first appearance of migratory products: 24 hrs.
Initial Ratio C^oJMe, migratory products: 10.0 at 24 hrs.
Table VIII GC-FID analysis
Reaction Time (Tits') % completion wrt S.M. C2H3/Me product ratios
24 66.7 10.0
48 85.7 10.0
72 100* 10.0
* one major side products reported at this time; Reaction mixture
was frozen and stopped at this point.
GC/MS analysis: The large vinyl (R) migration product again preceded the
methyl migration product in elution. These were followed by a small starting material
peak, o-xylene solvent, the lone side product (ID.) and 18-Crown-6.
Relevant ions and (% abundances) found for the main products were:


31
L C2H3CH2Si(CH3)2F M+118 (100); M+- Me 103(92); M+-C2H3
91 (0.5); M+- C2H3CH2 77 (100)
E. Et(C2H3)Si(CH3)F M+ 118 (48); M+- Me 103 (100); M+- C2H3
91 (100); M+- Et 103 (100);
m. (CH3) 3Si(C2H3CH2)2 M+ 140 (2.5); M+- Me 125 (100); M+-
C2H3 113(0.5); M+- C2H3CH2 99(100);
Isolation of products:
Distillation of the filtered reaction mixture through a Vigreux column gave 0.5
g of a liquid whose GC indicated that it was a mixture of both C2H3CH2Si(CH3)2F
and Et(C2H3)Si(CH3)F in a ratio of 5:1. The peaks were very close in retention time
and were not separated.
2.6.7 PhSi(CH3)2CH2Cl + KF------ PhCH2Si(CH3)2F + Et(Ph)Si(CH3)F + KC1
Ph migration Me migration
Time of first appearance of migratory products: 1.5 hrs.
Initial Ratio Ph/Me migratory products: 8.0 at 24 hrs.
Table IX GC-FID Analysis
Reaction Time ('hrs') % comoletion wrt S.M. Ph/Me product ratios
1.5 18.12 8.0
23 41.4* 6.37
48 58.3** 3.70
* rapid appearance of 2-3 side products
* reaction mixture was frozen and stopped at this point.
GC/MS analysis: The phenyl (R) migration product was the fourth substance


32
to elute, followed by the adjacent methyl migration product. These were preceded by
two highly volatile side products found to be (CH3>2SiF2 and (ClCH2)2Si(CH3)2,
and toluene solvent in order, respectively. The main, primary (1) products were
followed by starting material, the secondary (2) products (PhCH2)2Si(CH3)2,
Ph(CH3)2SiOSi(PhCH2)(CH3)2, PhCH2(CH3)2SiOSi(Ph)(Et)(CH3) and
18-Crown-6.
Relevant ions and (% abundances) found for the main products were:
l. (CH3)2SiF2 M+96 (10.1); M+- Me 81(100)
H. PhCH2Si(CH3)2F M+168 (32.5); M+- Me 153(11.5); M+-Ph
and/or CH2Ph+ 91 (23); M+- CH2Ph and/or Ph+ 77 (100)
m. Et(Ph)Si(CH3)F M+ 168 (26.5); M+- Me 153 (9.5); M+- Et
139 (100); M+- Ph 91 (28.5); Ph+ 77 (100)
IV. (PhCH2)2Si(CH3)2 M+240 (7.0); M+- Me 225(1.5); M+-CH2Ph
91 (23)
V. Ph(CH3)2SiOSi(PhCH2)(CH3)2 M+ 300 (0.5); M+- Me 285 (1.0);
M+- CH2Ph 209 (100); M+- Me2SiPh 165 (7.5); M+-
OSi(PhCH2)(CH3)2 135(20); Ph(CH3)2SiOSi+(CH3)2 193
(37.0) ;
VI. PhCH2(CH3)2SiOSi(Ph)(Et)(CH3) M+314 (0.5); M+- Me 299
(1.0) ; M+- Et 285 (4.2); M+- Ph 237 (trace); M+- CH2Ph 223
(100); M+- MeEtSiPh 165 (6.0); M+- OSi(PhCH2)(CH3)2 149
(2.0) ; CH2Ph+ 91 (21.5); (CH3)2Si+SiMeEtPh 207 (28.5)


33
2.6.8 1 -chloromethyl-1 -methyl-silacyclopentane + KF-------->
1-fluOTO-l-methylsilacyclohexane + 1-fluoro-l-ethylsilacyclopentane + KC1
Ring opening/ R migration - Me migration
Ring closed product
Time of first appearance of migratory products: 9 hrs.
Initial Ratio R/Me migration:
Short path distillation gave an initial fraction rich in R migration product with
mesitylene as the only other detectable component (~41 %). This solution was suitable
for proton and carbon-13 NMR analyses, which revealed that only
1-fluoro-l-methyl-silacyclohexane was formed.
Additional Reactions of Mechanistic Interest:
2.6.9a Cl(CH2)3Si(CH3)3 + KF ----------> BuSi(CH3)2F + KC1
Rearrangement within 336 hrs (2 weeks) at this temperature; the only reaction
with this system was a possible dimerization of the starting material with loss of
chloride.
2.6.9b (ClCH2)2Si(CH3)2 + KF------------> (CH3)2SiF2 + C2H4 + KC1
This reaction, conducted for three days in a septum-sealed serum cap vial
resulted in so much internal pressure that the septum-cap assembly was found to have
blown off the vial (along with the reactants).
It was conducted again at room temperature for four days after which both gas
and solution phase samples of the reaction mixture were analyzed by GC-FID and
GC/MS.


34
The solution phase GC-FID revealed two highly volatile products in addition
to a very small amount starting material peak indicative of an almost complete
reaction. Subsequent GC/MS revealed the least volatile of the first products to be
(CH3)2SiF2: ions and (% abundances): M+96(81); M+-Me 81(100); M+-F
77 (52).
Retention time comparative GC-FID analysis of the gas phase revealed the
most volatile of the two products to be ethylene.


35
CHAPTER IE
MECHANISTIC KINETIC STUDIES OF
(ARYLCHLOROMETHYL)DIMETHYLSILANES
3.1 Introduction:
The anionic migration mechanism postulated in chapter 2 is further probed in
this chapter for its validity. Chloromethyldimethyl-R-silanes with readily migrating R
groups were chosen for further study because they (esp. aryl) are used for classic
probes of electronic effects. Anionic stability was found to be the predominant
parameter responsible for the relative migratory abilities of these and of the alkyl
anions in the previous chapter. The presence of secondary products of the
Me2Si(CH2R)2 type in the R = Phenyl and Vinyl cases and of Me2SiF2 (eventually
found to be present in all of the solution phase fluoride reactions via GC/MS) led us
to speculate about what happens mechanistically when a second mole of fluoride
reacts. [*The fluoride/silane stoichiometry in all reactions contained in study 1 and 2
was kept at 2 : 1, but perhaps only < 0.05 % of the total fluoride was freed by the
18-crown-6 at any given time26 for reaction in the aromatic medium.]
The series of reactions chosen for chapter 3 utilized five
aryl(chloromethyl)dimethyl silanes rather than vinyl because a great number of
Hammett o constants are known and were derived for substitutions on the benzene
ring. These were X = p-Me, H, p-F, p-Cl, and m-CF3. All except the phenyl (X =


36
H)** were synthesized from a substituted arylmagnesium halide Grignard reaction
with chloromethyldimethylchlorosilane. [section 3.5] [** Obtained from Petrarch
Systems, Inc.]
The first of these reactions (3.2.3) concentrated on the acquisition of the total
primary (1) and secondary (2) product distributions (eq. 14). Next, a competition for
fluoride between the p-Me-Ph-silane and all the others yielded relative ratios of
reaction rates based on starting material disappearance.
The last of these solution studies concerned itself with the rate of appearance
of X-substituted toluenes the major secondary product observed in the first series and
suspected of forming secondary products of the Me2Si(CH2R)2 type.


37
3.2
Acquisition of primary Cl0') and secondary (2) products for reactions of the type:
ArSiMe2CH2Cl + KF -----------> ArCH2SiMe2F + EtSi(Ar)Me2F (1)
+ Me2SiF2 + ArCH3 + Me2Si(CH2AR)2
+ EtMeSi Ar(CH2Ar) (2) (eq. 14)
3.2.1 General Experimental:
These reactions were conducted in 20 ml serum cap vials with miniature
magnetic stir bars at 55C maintained by a thermostatted oil bath manufactured by
Sybron Inc. Toluene (1.5 ml) was the solvent chosen. The scale of the reactions was
cut in half from those in Study 1: 5 mmol silane, 10 mmol KF and 0.5 mmol
18-Crown-6; the 10:20:1 reactant molar ratio was unchanged.
The product and migration rados were determined by GC-FID peak heights
on a Perkin Elmer 3020B chromatograph with a 6' x 1/8" OV-l/chromosorb stainless
steel column. Fairly reproducible* ( 5 %) 0.3 pi injections were used to sample the
reaction mixtures which were prepared gravimetrically equivalent on a Cenco balance
with a sensitivity of 0.1 mg. Qualitative peak identification was performed on a
Finnigan 3200 GC-MS with a Teknivent data system. (* As determined by the
18-crown-6 GC-FID peak heights).
3.2.2 Results and Discussion:
The generalized product distribution from reactions of KF and
ArSiMe2CH2Cl (based on peak heights relative to the Me migration product (D)
EtSiArMeF found in a reactions) after 23 hours reaction at 55C in toluene is shown
below in table X. [does not include siloxanes resulting from hydrolysis of Si-F


38
bonds.]
Table X
Generalized Product Distribution from Reactions of ArSiM^CH^Cl with KF
compound Ar
phenyl
p-tolyl
p-fluorophenyl
p-chlorophenyl
m-(CF3)phenyl
A B c D E F
1.13 * 6.8 1.0 0.4 U.D.
0.24 1.7 3.9 1.0 0.06 U.D.
0.16 0.95 U.D. 1.0 0.15 0.016
0.45 1.4 U.D. 1.0 0.58 0.065
1.43 2.5 U.D. 1.0 1.8 0.59
legend: A = Me2SiF2; B = ArCH3; C = AiCH2SiMe2F; D = EtSiArMeF;
E = (AiCH2)2SiMe2; F = EtMeSiAr(CH2Ar)
* one secondary product, toluene, and solvent were identical in this case so it was not
included in calculations, performed below in Table XI.
peak ratios of A/D were not reported in reference 22 but are included in the
calculation of the relative ratio of £2/Xl products contained below in table XI.
U.D. = undetermined presence even by capillary GC-FID
Table XI
Grouping and Ratios of 1 and 2 Products
comnound Ar Y2 = A + B + E + F y.i = c+D Y2/Y.r
p-tolyl 2.0 5.6 0.357
p-fluorophenyl 1.276 1.0 1.276
p-chlorophenyl 2.495 1.0 2.495
m-(CF3)phenyl 6.32 1.0 6.32
The solvent phase concentrations of Me2SiF2 (bp 2-3C/760 mm Hg) relative
to the X = H reaction (3.2.3b), calculated from GC-FID peak hieghts using 0.3 |il 5
% volumetric injections of reaction 3.2.3 supernatants after 23 hours reaction at 55C
were: X = p-CH3 (0.62); X = p-H(1.0); X = p-F(1.77); X = p-Cl (2.54); X =
m-CF3 (7.69).
The concentrations* of substituted toluenes (A1CH3 produced relative to the
X = p-F reaction (3.2.3c), calculated from such volumetric injections after the same
reaction time were: X = P-CH3 (0.71); X = p-F (1.0); X = p-Cl (1.29); X = mCF3


39
(2.22).
* Assuming equal GC-FID repsonse factors.
** The X = p-F reaction was used in place of the X = H reaction as a
reference because the toluene produced by reaction 3.2.3b was identical to the solvent.
Considering the error inherent in the quantitative method used for reactions
3.2.3 a -> e, Figure Vm shows a remarkable correlation between log([Me2SiF2]X/
[Me2SiF2]H) and the Hammett ox constant for the respective substituents. (an
defined as 0.0 of course). The ax values used were28: X = p-Me (-0.15); X = H
(0.0); X = p-F (0.06); X = p-Cl (0.23) and X = m-CFg (0.49). These represent a
sum of both resonance Or and inductive Ofc effects inthe classic Hammett expression
log (k/ko) = pox.
Plot of log ([Me2SiF2]x / [Me2SiF2]n) vs. Hammett ox Constants for Solution Phase
Reactions 3.2.3a->e of ArSiMe2CH2Cl
Similar plots were made for log (22/Xl) products vs. ox (figure IX), and


40
for log ([ArCH3]X/[ArCH3]F) vs. aF after shifting the proton based cx values
downward by -0.06 to obtain cF values (figure X)
The only R/Me migration ratios obtained to correlate with ax belong to X =
p-Me and X = H. Statistically corrected, they were 16.76:1 and 16.0:1 repsectively.
Plot of log (Z2/£l) Products vs. Hammett ctx Constants for Solution Phase
Reactions 3.2.3a->e of ArSiMe2CH2Cl
Data obtained for X = p-F and m-CF3 indicate only one primary product, even on a
capillary GC-FID.27 GC-MS data however, reveal this one peak to contain both
migratory products (C and D, Table X). The shockingly low ratio for the p-Cl silane
(0.046) may indicate a steric problem for p-CIPh migration in the previously
postulated pentacoordinate adduct (Figure V).


41
Figure X
Plot of log ([X-Ph-Me] / [p-F-Ph-Me]) vs. Constants for Solution Phase Reactions
(3.2.3a->e) of ArSiMe2CH2Cl with KF
The excellent correlations and especially the high slopes (rho values) shown
in figures VIII and IX suggest that a large degree of negative charge stabilization in
some secondary (peihaps another pentacoordinate) transition state is necessary for the
formation of the secondary products Me2SiF2, ArCH3, Me2Si(CH2Ar)2, and
EtSiMeAr(CH2Ar). The lower slope (0.791), shown in figure X for the approximate
rate of ArCH3 formation with respect to p-fluorotoluene vs aF could be due to the
assumption made that all substituted toluenes possess equal FID response factors. The
0.995 correlation coefficient would then be due to chance. To eliminate this type of
uncertainty, all subsequent Hammett studies were performed with an actual unreactive
internal standard hexadecane.
Nevertheless, a mechanism for 1 and 2 product formation consistent with the
observations in section 3.2.3 is presented below in figure XI (see next page).
This mechanism explains the lowering of the aryl/methyl migration ratio as the


42
X-substituent on the aryl group becomes increasingly electron-withdrawing because
the attack of a second mole of fluoride on the aryl migration 1 product should form
from Me2SiF2 and resonance stabilized benzyl anions which would then react with
any available acidic protons to form ArCH3. The secondary migration products
Me2Si(CH2Ar)2 and EtSiMeAr(CH2Ar) would then arise from the same
rearrangement effected by P reacting with the starting material, except that the reacting
anion is now A1CH2", competing with any P remaining. As mentioned in figure XI,
the alternate route to the secondary migration products involving benzyl anion attack
and fluoride displacement is thermodynamically unfavorable in view of the strength of
tetracoordinate Si-F bonds (154 kcal/mole) compared with that of Si-C bonds (88
kcal/mole). See Figure XI.
The 'semi'-quantitative information obtained by the 0.3 pi 5 % injections of
reaction 3.2.3 supernatants provided the insight necessary to give the go-ahead
X
2
Cl
Cl
Cl
aryl migration methyl migration
products


43
- KCI
Me
1
2nd mole F'
Me-" |'}KD-X
Kjc"2
{
Et -si
Me
Me2SiF2 + X-^^- CH2
+ H1
- H
/x
'* /--X
Mesi
X--
CH-
/ MD-x m/ MD"X
methyl migration aryl migration
secondary products
toluene products
relative amount depends
on stability of benzyl anion
Possible but not probable route to secondary migratory products:
Me $iC5^"X ^ /-(o}-X
I Me + X-^0)-CH2 Me-Si + F

MeSi^ +
I Et
F
x --_ch2
Me
/-X
Et-Si + F
Me/
Figure XI
Detailed Mechanism of Primary and Secondary Product Formation for Solution Phase
Reactions (3.2.3a->e) of ArSiMe2CH2Cl
decision for performing two actual (quantitatively controlled by a GC-FID internal
standard) Hammett studies on these solution phase reactions: 3.3. The relative overall
rates of reaction as determined by a competitive starting material consumption between
the X = p-Me silane and the other four (two at a time of course); 3.4 A kinetic study
of the rate of appearance of the X-substituted toluenes chosen over the rate of


44
Me2SiF2 formation becuase of the volatility of the latter.
3.2.3 Detailed Experimental:
3.2.3a
p-CH3-PhSiMe2CH2Cl + KF-----> p-CH3-PhCH2SiMe2F + EtSi(p-CH3-Ph)MeF
p-MePh migration Me migration
GC-FID Observations with Respect to Reaction Time of 0.3|il Supernatant Injections
R/Me primary % completion
Rxn Time ('1') product ratio wrt to S.M.
1.5 hrs 8.38 11.5
3.5 hrs 10.0 33.0
23 hrs 3.88 57.0
Remarks
Both p-xylene and Me2SiF2
present
Relative amounts of Me2SiF2 and
p-xylene had doubled and
tripled respectively. Slight
amounts of secondary products
and one other side product had
formed.
Relative amounts of Me2SiF2 and
p-xylene had tripled and
octupled respectively from
the time of the last observation.
Substantial amounts of
secondary products had
formed. The reaction was
halted by freezing.
GS-MS Data: Relevent ions and (% abundances) for the main products found were as
follows (listed in elution order):
l. Me2SiF2: M+ 96 (10.1); M+ Me 81 (100)
H. p-xylene: M+ 106 (33.5); M+ H 105 (18.5); M+ Me 91 (100)
m. pMePhCH2SiMe2F: M+ 182 (18.0); M+ Me 167 (5.0); M+ -
CH2Ph(p-Me) 77 (100); p-MePhCH2+ 105 (23.5)


45
IV. EtSi(p-MePh)MeF: M+ 182 (25.0); M+ Me 167 (8.5); M+ Et 153
(100); M+ p-MePh and /or p-MePh+ 91(44.0)
V. (pMePhCH2)2SiMe2: M+ 268 (3.5); M+ Me 253 (8.0); M+ -
CH2Ph(p-Me) 163 (100); p-MePhCH2+ 105 (23.0)
VI. Me3SiOSiMe2(p-MePh): M+ 238 (22.0); M+ H 237 (100); M+ H
and .0. (Me3SiOSiMe2(Ph-p-CH2+)) 221(27.5); M+-p-MePh 147
(4.0); p-MePh+ 91 (5.2); M+ SiMe3 165 (2.0); Me3Si+ 73 (26.0)
3.2.3b
PhSiMe2CH2Cl + KF ----------> PhCH2SiMe2F + EtSi(Ph)MeF
Ph migration Me migration
GC-HD Observations with Respect to Reaction Time of 0.3|il Supernatant Injections
R/Me primary % completion
RxnTime (Tj product ratio wrt to S.M. Remarks
1.5 hrs 8.0 18.12 Trace of Me2SiF2; (any toluene
formed was blanketed by the
solvent)
23 hrs 6.37 41.4 Slight amounts of 2 product
forming. About 6 fold the
amount of Me2SiF2 present at
initial analysis.
48 hrs 3.70 58.3 Substantial quantities of
secondary and one other side
product had formed. Twice the
amount of Me2SiF2 present as in
the previous analysis. The
reaction was halted by freezing.
[GS-MS Data: Relevant ions and (% abundances) for the main products found were
as follows (listed in elution order):
I. (CH3)2SiF2 M+96 (10.1); M+- Me 81(100)
H. (ClCH2)2Si(CH3)2 35C12 M+- H 155(100); 35C1/ 37ci M+ H 157


46
(30.0); 37Cl2 M+ H 159 (9.0); 35Cl2 M+ Me 141 (5.0); 35C1/
37C1M+ Me 143(1.5); 37Cl2M+-Me 145(0.45)
m. PhCH2Si(CH3)2F M+168 (32.5); M+- Me 153(11.5); M+-Ph and/or
CH2Ph+ 91 (23); M+- CH2Ph and/or Ph+ 77 (100)
IV. Et(Ph)Si(CH3)F M+ 168 (26.5); M+- Me 153 (9.5); M+- Et 139
(100); M+- Ph 91 (28.5); Ph+ 77 (100)
V. (PhCH2)2Si(CH3)2 M+ 240 (7.0); M+- Me 225 (1.5); M+- CH2Ph 91
(23)
VI. Ph(CH3)2SiOSi(PhCH2)(CH3)2 M+ 300 (0.5); M+- Me 285 (1.0); M+-
CH2Ph 209 (100); M+- Me2SiPh 165 (7.5); M+- OSi(PhCH2)(CH3)2
135(20); Ph(CH3)2SiOSi+(CH3)2 193(37.0);
VH. PhCH2(CH3)2SiOSi(Ph)(Et)(CH3) M+ 314 (0.5); M+- Me 299 (1.0);
M+-Et 285(4.2); M+-Ph 237 (trace); M+-CH2Ph 223(100); M+-
MeEtSiPh 165 (6.0); M+- OSi(PhCH2)(CH3)2 149 (2.0); CH2Ph+ 91
(21.5); (CH3)2Si+SiMeEtPh 207 (28.5)]
3.2.3c
p-FPhSiMe2CH2Cl + KF -----------> p-FPhCH2SiMe2F + EtSi(p-FPh)MeF
p-FPh migration Me migration
GC-FID Observations with Respect to Reaction Time of 0.3p.l Supernatant Injections
R/Me primary % completion
Rxn Time (11 product ratio wrt to S.M. Remarks
1.5 hrs undetermined 32 Slight amount of Me2SiF2
(see GC-MS data) present. (It was very hard to
determine the presence of
p-fluorotoluene at this point
because of near identical
retention times between it and


47
the solvent)
3.5 hrs undetermined 41.3
Twice the amount of Me2SiF2
present as at the initial
analysis.
23 hrs undetermined 51.0
Enough p-F toluene present to
elute just past the solvent peak.
Amount of Me2SiF2 had 10 fold
increased since the 3 1/2 hour
analysis. Large amounts of
secondary products had also
formed.
49 hrs undertermined 55.7
Secondary products present in
almost equal quantities as the
primary products. The reaction
was halted by freezing.
GS-MS Data: Relevant ions and {% abundances) for the main products found were as
follows (listed in elution order):
l. (CH3)2SiF2 M+ 96 (10.1); M+- Me 81 (100)
II. p-fluorotoluene: M+ 110 (32.5); M+- H 109 (71.0); M+- F 91 (100)
m. p-FPhCH2SiMe2F + EtSi(p-FPh)MeF: M+ 186 (21.5); M+- Me 171
(7.60); M+- Et 157 (12.0); M+- CH2(p-FPh) 77 (93.5); p-FPh+ 109
(26.5)*
IV. Et(p-FPhCH2)Si(p-FPh)Me: M+ 276 (2.5); M+- Me 261 (0.50); M+-
Et 247 (2.0); M+- p-FPhCH2 167 (60.0); p-FPhCH2+ 109 (13.5);
M+- p-FPh 181 (1.00); p-FPh+ 95 (0.50)
V. (p-FPhCH2)2SiMe2: M+ 276 (8.5); M+- Me 261 (2.0); M+- p-FPhCH2
167 (100); p-FPhCH2+ 109 (45.0)
primary products unseparable on columns available at the time.
3.2.3d
p-ClPhSiMe2CH2Cl + KF
> p-ClPhCH2SiMe2F + EtSi(p-ClPh)MeF
p-CIPh migration Me migration


48
GC-FID Observations with Respect to Reaction Time of 0.3 fil Supernatant Injections
R/Me primary % completion
Rxn Time (1^ product ratio wrt to S.M. Remarks
1.5 hrs 0.023 11.0 Small amount of Me^SiF2
present. Large quantities of
p-chlorotoluene already
present, in 1.78 x the amount
of the primary products
(assuming equal GC-FID
response factors).
3.5 hrs 0.023 13.7 Approximately double the
quantity ofMe2SiF2asatthe
1.5 hour analysis. Amount of
p-chlorotoluene has increased
by 1.5x the initial analysis.
Ratio of p-chlorotoluene l/2
products now 2.0.
23 hrs 0.023 37.8 Approximately 4x the quantity of
Me2SiF2 as at the 3.5 hour
analysis. Ratio of
p-chlorotoluene l/2 products
now 1.35. Relative quantity of
p-chlorotoluene hasincreased
by a factor of 2.44 since
the 3.5 hour analysis. One side
product and two secondary
products are now present in
near the same quantity as the
primary products. The reaction
was stopped by freezing.
GS-MS Data: Relevant ions and (% abundances) for the main products found were as
follows (listed in elution order):
I. (CH3)2SiF2 M+ 96 (10.1); M+- Me 81 (100)
II. p-chlorotoluene: 35C1M+ 126 (28.0), 37C1M+ 128 (8.12); 35C1
tropylium+ 125 (13.0); 37C1 tropylium+ 127 (3.77); M+- Cl 91 (100)
III. EtSi(p-ClPh)MeF: 35C1M+ 202 (13.0),37Cl M+ 204 (3.77); 35C1M+ -
Me 187 (4.0), 37C1M+ Me 189 (1.16); 35C1M+ Et 173 (10.0),


49
37C1M+ Et 175 (3.0); M+- Cl 91 (100); 35C1 Ph+ 111 (2.0), 37C1
Ph+ 113 (0.6)
IV. p-ClPhCH2SiMe2F: 35ciM+202 (7.0), 37C1M+204 (2.10); 35dM+-
Me 187 (99),37Cl M+ Me 189 (30.0); 35C1 tropylium+ 125 (52.5);
37C1 tropylium+ 127 (15.7); Me2 SiF+ 77 (100)
V. FMe2SiOSi(Et)(p-ClPh)Me: 35C1M+ 276 (2.5), 37C1M+ 278 (0.75);
35d M+ Me 261 (2.5), 37C1M+ Me 263 (0.75); 35C1M+ Et 247
(2.0) , 37C1M+ Et 249 (0.60); M+ p-CIPh 165 (2.0); 35C1M+ F2
MeSiO or Et(Me)Si(p-ClPh)+ (2.5); 37C1M+ F2MeSiO or
Et(Me)Si(p-ClPh)+ 183.7(0.75); F2MeSiOSiMe2+ 151 (100)
VI. (p-ClPhCH2)2SiMe2: 35Cl2M+308 (7.5), 35ci 37CIM+310 (4.5),
37Cl2 M+ 312 (0.70); 35Cl2 M+ Me 293 (2.5), 35C137C1M+ Me
295 (1.25), 37Cl2 M+ Me 297 (0.5); 35C1M+ CH2(p-ClPh) 183
(100), 37C1M+ CH2(p-ClPh) 185 (30.0); 35C1 tropylium+ 125
(22.5); 37C1 tropylium+ 127 (6.75);
VE. EtMeSi(p-ClPh)(p-ClPhCH2): 35C12 M+ 308 (2.0), 35C137C1M+ 310
(1.0) , 37C12 M+ 312 (0.2); 33C12 M+ Me 293 (1.0), 35C137C1M+ -
Me 295 (0.5), 37C12 M+ Me 297 (0.2); 33C12 M+ Et 279 (2.5),
33C137C1M+ Et 281 (1.25), 37C12 M+ Et 283 (0.25);
p-35ClPhSi(Et)Me+ 183 (45), p-37ClPhSi(Et)Me+ 185(13.5);
p-35ClPhCH2Si(Et)Me+ 197 (2.0), p-37ClPhCH2Si(Et)Me+ 199 (0.6);
35C1 tropylium+ 125 (7.5); 37C1 tropylium+ 127 (2.25); 35C1 Cl-Ph
111 (2.0), 37C1 Cl-Ph 113(0.6)
3.2.3e


50
m-CF3PhSiMe2CH2Cl + KF-----> m-CF3PhCH2SiMe2F + EtSi(m-CF3Ph)MeF
m-CF3Ph migration Me migration
GC-FID Observations with Respect to Reaction Time of 0.3|il Supernatant Injections
R/Me primary % completion
RxnTime Q0>) product ratio wrt to S.M. Remarks
1.5 hrs undetermined 5.10
(see GC-MS data)
23 hrs undetermined 48.8
At least 3x the amount of
Me2SiF2 present as contained in
the other reactions at this time.
Relative amount of
m-trifluoromethyltoluene
present in 3.6x and 43x the
amounts of p-chlorotoluene and
p-xylene formed after the same
time in reactions 3.2.3d and
3.2.3a
Relative ratios of the amounts of
Me2SiF2 formed by this time
(this reaction being unity) by
tiie other 3.2.3 reactions
with respect to theX = m-CF3
reaction were: x = p-Me,
(0.08); x = H, (0.13); x = p-F,
(0.23); x = p-Cl, (0.33).
Relative ratios of the amounts of
x substitued toluenes formed by
this time (this reaction being
unity) by the other 3.2.3
reactions (producing
observable ones) with respect
to the x = m-CF3 reaction were:
x = p-Me (0.32); x = p-F
(0.45); x = p-Cl (0.58); Two
secondary products were
formed in at least twice the
relative quantity of primary
products. Halted reaction by
freezing.
GS-MS Data: Relevant ions and (% abundances) for the main products found were as
follows (listed in elution order):
I. Me2SiF2: M+ 96(10.1); M+- Me 81 (100)


51
II. m-CF3-toluene: M+ 160(79.5); m-CF3+tropylium 159 (31.0); M+-
F 151 (17.5); M+ mCF3 91 (100); CF3+ 69 (8.50)
m. *m-CF3PhCH2SiMe2F + EtSi(m-CF3Ph)MeF: M+236 (7.5); M+-Me
221(8.0); M+ Et 207 (5.5); M+-F 217(10.0); M+- CH2Ph(m-CF3)
77 (100); CH2Ph(m-CF3)+ 159 (10.0); M+ Ph(m-CF3) 91 (5.0);
+Ph(m-CF3) 145(1.5)
IV. EtMeSi(m-CF3Ph)(m-CF3PhCH2): M+ 376 (2.0); M+ Me 361 (1.0);
M+ Et 347 (2.0); M+ F 357 (5.0); M+ CH2Ph(xn-CF3) 217
(12.0) ; M+ Ph(m-CF3) 231 (1.5); CH2Ph(m-CF3)+ 159 (9.0);
+Ph(m-CF3) 145 (2.0); H(Me)SiPh(m-CF3)+ (rearranged cation
from Et loss) 189(97.5)
VI. (m-CF3PhCH2)2SiMe2: M+376 (1.0); M+- Me 361 (2.5); M+- F
357 (4.0); M+ CH2Ph(m-CF3) 217 (35.5); CH2Ph(m-CF3)+ 159
(13.0)
Primary products inseparable on columns available at the time.


52
3.3
Competitive Relative Rates of Starting Material Consumption
(for X-substituted arylchloromethyldimethylsilanes + KF 18-Crown-6/mesitylene;
72Q
3.3.1 General Experimental:
The reactions in this study were quantitatively determined competitions for
fluoride anion between the X = p-Me silane and the other four used in section 3.2.
They were conducted at 72C in four 20 ml septum sealed serum cap vials containing
2.5 mmol each of the X = p-Me silane, one of the other four aryl silanes and
hexadecane (Kodak) GC-FID internal standard. The quantities of KF and
18-Crown-6 (Aldrich) added to each of the four vials (along with a miniature magnetic
stirring bar) and 1.5 ml mesitylene solvent (Fischer) were 2.0 mmol and 0.2 mmol
respectively. This stoichiometry was chosen because if both aryl silanes were found
to react with fluoride at the same rate, only 40% of each could be consumed. Since we
were mainly concerned with the rate of primary product formation, the relative lack of
available fluoride should slow the formation of Me2SiF2 and the secondary products.
(This was the effect observed).
The reactions mixtures were prepared on a Cenco balance having a sensitivity
of 0.1 mg and were maintained at 72C in a thermostatically controlled oil
bath/magnetic stirrer manufactured by Sybron, Inc.
The four competitions were sampled after 41.5 hours reaction at 72C.
GC-FID analyses were performed on a 6' x 1/8" OV-l/Chromosorb stainless steel
column in a Perkin elmer 3020B chromatograph. Reference injections of
gravimetrically prepared solutions containing 13.5 mole percent each of the two


53
arylchloromethyldimethylsilanes being studied and the hexadecane internal standard in
mesitylene were made prior to each analysis. ( Same as the intial concentrations of
reactants in the 20 ml serum cap vials).
3.3.2 Results and Discussion
The reaction mixtures in the four competitions studied were found to contain
only small quantities ( < 5 % of total product peak area) of the secondary products
found in reactions 3.2.3 a -> e. This was the expected result in view of the limited
fluoride concentration (approximately calculated^ to be 0.11 mole percent or less)
(
available in solution at any one time. The Hammett correlation of the log(relative rates)
vs cx of the substituents studied is contained below in figure XII
Figure XII
Plot of log Kx / KH vs. Hammett ax Constants for the Relative Rates of Starting
Material Consumption Study (3.3)


54
The 0.99 correlation and 1.58 slope suggest a good degree of anionic
character in the aryl group undergoing migration to form the major primary product
ArCH2SiMe2F. The instability and base strengths of the aryl anions would be
expected to follow the order p-MeAr > Ph > p-FAr > p-CLAr > m-CF3Ar since the cx
values used represent the sum of both resonance (aR) and field (CTj) effects. The p-Me
arylsilane exhibits both the highest observed aryl/naethyl migration ratio (16.76) and
the slowest reaction rate due to anionic instability of [p-MePh]- The m-CF3
arylsilane would probably exhibit (on a 100' or greater capillary column) the lowest
aryl/methyl migration ratio and does exhibit the highest reaction rate due to anionic
stability of [m-CF3Ph]_.
The p-F and 111-CF3 aryl silanes were recently (9-87) reacted in the usual
scale (see reactions 3.2.3 a -> e) with KF/18-Crown-6 in mesitylene at 110C for 4
hours and their supernatants were analyzed on a Hewlett Packard 5890 GC-FID for
the presence of two primary products. The temperature program used on the 25 m x
0.2 mm cross linked methyl silicone coated capillary column near the elution of the
primary products was reduced to 0.5C/min, but only one gaussian peak (no
shoulders, even upon software magnification) was observed. As stated before
(section 3.2.2), GC-MS analysis of this peak showed ions resulting from both
migratory products.


55
.3.3a Detailed Experimental for One of the Four Competitions Studied
Table Xn
Gravimetric Preparation of PhSiMe2CH2Cl vs p-MePhSiMe2CH2Cl
(41.5 hrs, 72C)
Reactant Table MW mmoles mole % f-KF and
PhSiMe2CH2Cl 0.4629 184.80 2.5049 13.5113
p-MePhSiMe?CH9Cl 0.4969 198.77 2.4999 13.4843
KF 0.1171 58.10 2.0155
18-Crown-6 0.0529 264.30 0.2002
Ci6H34(IS) 0.5694 225.45 2.5145 13.5631
Mesitylene 1.3246 120.20 11.0200 balance
total: 18.5393 (- KF and 18-Cr-6)
Table XIII
Gravimetric Preparation of Hexadecane / PhSiMe2CH2Cl, p-Me-PhSiMe2CH2Cl
Reference Solution
Reference Solution:
PhSiMe2CH2Cl
p-MePhSiMe2CH2a
C16H34 aS) 0.5660
Mesitylene
total:
0.4673 184.80 2.5287 13.7253
0.4990 198.77 2.5104 13.6259
225.45 2.4994 13.5662
1.3084 120.20 10.8852 18.4237 balance
GC-FID Parameters for both reference and analysis
Injection size: 0.175 |il
Temperature Program: 70 --> 280C; 8C. min -1
Signal (Attenuation X Range): 128 x 100


Table XIV
Calculation of GC-HD response factors from reference injection:
(via peak heights; chart block units)
reference iniection peak height peak ratio: silane/IS mole ratio: silane/IS RF
Ci6H34(IS) 81.6 0.6141
PhSiMe?CH2Cl 50.7 0.6213 1.0117
p-MePhSiMe2CH2Cl 54.1 0.6630 Table XV 1.0044 0.6601
Example of GC-FID Quantification of Competition Study 3.3.3a Results
Reaction Results peak height equivalent mmoles silane
________________________________________IS peak height* present**_______
Ci6H34 (IS) 92.75
PhSiMe2CH2Cl 32.7 53.2487 1.4436
p-MePhSiMe2CH2Cl 44.65 67.6413 1.8338
* = peak height/RF
** = (eq IS peak height/IS peak height) x mmoles IS
PhSiMeoCHoCl u-MePhSiMeoCHoCl
mmoles silane consumed 1.0613 0.6661
% silane consumed 42.37 26.65
Calculation of relative reaction rate with respect to the phenylchloromethyldimethyl
silane reaction:
[p-MePhSiMe2CH2Cl] (41.5 hrs) = 1.8338/18.5393* = 9.8914 mole percent
A[p-MePhSiMe2CH2Cl] (41.5 hrs) = 9.8914 -13.4843 = -3.5929 mole percent
[PhSiMe2CH2Cl] (41.5 hrs) = 1.4436/18.5393* = 7.7867 mole percent
A[PhSiMe2CH2Cl] (41.5 hrs) = 7.7867- 13.5113 = -5.7246 mole percent
A[p-MePhSiMe2CH2Cl] / A[PhSiMe2CH2Cl] = 0.6276**


57
* Assuming total mmoles reactants-> products is unchanged; no
dimerization.
Reported as 0.63 in reference 14.
3.3.3.b
Table XVI
Results of Competition study starting material disappearance after 41.5 hrs. reaction
at 72C vs p-MePhSiMe2CH2Cl (cx = -0.15)
X comn ArX-PhSiMeoCHoCll Kx/K_ Me KX/KH ax
Arp-MePhSiMeoCHoCll
H -5.7246 mole % -3.5929 mole % 1.5933 0.6276 0.0
p-F -7.6998 mole % -2.8869 mole % 2.6672 1.6739 0.06
p-Cl -6.7595 mole % -1.4505 mole % 4.6601 2.9247 0.23
m-CF3 -5.3414 mole % -0.5842 mole % 9.1431 5.7382 0.47


58
3.4 Appearance rate of x-substituted toluenes (ArCHg)
The acquisition of the rates of formation of substituted toluenes (ArCH/3)
presumably arising from the attack of a second mole of fluoride on the primary aryl
migration product (see fig. XI and eq 15 & 16) was the object of these experiments.
ArCH2SiMe2F + F'------> ArCH2" + Me2SiF2 (eq. 15)
ArCH2- + H+soivent/H20 from KF <-----> ArCH3 (eq. 16)
3.4.1 General Experimental:
Reactions 3.4.3a through 3.4.3e were conducted in 20 ml septum sealed
serum cap vials with mini-magnetic stir bars as in all the previous studies. Mesitylene
was the chosen aromatic solvent because all the substituted toluenes produced in these
reactions (except p-chlorotoluene in 3.4.3d) are easily resolved by non capillary gas
chromotography.
The stoichiometry of reactants chosen was as follows: Arylchloromethyl
dimethylsilanes: 5 mmole; KF (Aldrich): 10 mmole; 18-Crown-6 (Aldrich): 0.5
mmole; C15H34 internal standard (Kodak): 0.27->0.31 mmole; Mesitylene solvent
(Fischer): _1.84 ml, 12.7 -> 14.4 mmoles. Reaction temperature was maintained at
85c by a thermostated oil bath/magnetic stirrer manufactured by Sybron Inc.
GC-FID quantitative analysis for all but the p-Cl substituted toluene was
conducted after 1.5, 3,4.5, 6,10,21.5 -> 28 and 45 -> 47 hours of reaction on a 6' x
1/8" OV-l/chromosorb stainless steel column mounted in a Perkin Elmer 3020B
chromatograph. The reaction (3.4.3d) producing p-chlorotoluene necessitated the use
of a 0.25mm x 10 m OV-101 silicone oil coated open tubular stainless steel capillary
column to effect its resolution from the mesitylene solvent. Reference injections of


59
gravimetrically prepared solutions containing 1.3 -> 1.4 mole percent each of the
x-substituted toluene, C16H34 internal standard, balance mesitylene were made prior
to each analysis in order to obtain FID response factors. (* A Cenco balance having a
sensitivity of 0.1 mg was used to prepare both the reaction mixtures and the
reference solutions**).
sfcsfc
Note: The reference solutions containing 1.3 -> 1.4 mole percent each of
X-Ph-Me and C16H34 were prepared via 10 fold dilutions of solutions containing
~13.5 mole percent each of those components in mesitylene. This was done to ensure
a more accurate weight of the fairly volatile X-Ph-Me compounds present in solution,
and to approximate their concentrations formed in the reactions.
3.4.2 Results and Discussion
The graphical representation of reactions 3.4.3a -> e as displayed in figure
Xm shows two important features: 1) The clustering of the X = p-Me, H and p-F
reaction profiles into one set and the X = p-Cl and m-CF3 profiles into another. 2) A
leveling off or rate decrease in the production of the X-substituted toluenes after 4.5
hours for X = p-Cl and m-CF3, and after 10 hours for the rest.
Observation (1) is not surprising in view of the relative magnitudes [(-0.15,
0, 0.06) vs. (0.23, 0.47)] of the sigma constants for the respective substituents.
Indeed, the GC profiles of reactions 3.2.3 and 3.4.3 showed that the formation of
visible quantities of all primary and secondary products followed the same grouping.
Observation (2) may be explained by the following view of the mechanism
postulated back in Figure XI. Since the sigma constants (ax) used for the aryl


60
Figure XIII
Reaction Profiles from the Rate of ArCH3 Appearance Study (3.4)
substituents represent a sum of resonance and Held effects, it is quite possible that as
each reaction consumes a second mole of fluoride, the resonance stabilization of the
released benzyl anion becomes of more importance than the field (through space)
interaction of the Si bonded aryl group with the growing positive charge on the
chloromethylcarbon about to undergo loss of chloride in determining the type
(primary vs secondary) of product formed. (See Figure XI). If there exists a genuine
equilibrium for X-PhCH2:" + H+------- X-PI1CH3 in the weakly acidic mesitylene
medium, it should be clear that the resonance stabilization of the benzyl anion increases
in the same direction as the sigma constants. When a large enough X-PI1CH3
concentration has formed, the presence of an increased X-benzyl anion concentration
enables the formation of the secondary bi-aromatic silanes from either of the two
pathways in Figure XI. The GC observation of secondary silane formation coincided


61
with the rate decrease in X-PI1CH3 production for all cases. That the relative
magnitude of this decrease was smallest in the X = p-Me case is significant because the
equilibrium concentration of the p-MePhCH2: anion should be the lowest of the five.
The effect of this a relative lack of secondary silane production, even after 45 hours
reaction was observed.
Given the premise that a binary mechanism, which produces and then
consumes benzylic anions in two pathways, is operative, and that the general shapes
of the reaction profiles contained in figure Xm are parabolic, each reaction data set
was fit to the best least squares obtained second order polynomial (see Appendix A).
This was done so that the entire data matrix was treated in a uniform fashion. The
earliest experimentally observed time at which the calculation of a positive quantity of
the x-Ph-Me was possible for all five reactions, was 3.0 hours. [1.5 hours were used
due to an erroneous data set utilized in reference 22]. The relative (to X = H) rates of
X-Ph-Me formation obtained from the quantities of X-Ph-Me compounds calculated at
3 hours reaction were: X = m-CF3 (11.4); X = p-Cl (10.8); X = p-F(2.46); X = H
(1.0), X = p-Me (0.451). The Hammett plot of the log of these rates vs ax is shown
in Figure XIV. Reference 22 erroneously reported the slope (correlation coefficient)
as 4.5 (0.94).
For interests sake, least squares Hammett plots were made of the actual data at
reaction times of 1.5, 3.0, and 4.5 hours. The p values and (correlation coefficients)
obtained respectively, were 3.565 (0.924), 2.760 (0.931), 2.311 (0.917). General
rate equalization of X-Ph-Me formation among the five reactions thereafter would lead
to a lowering of both the slopes and correlation coefficients, providing little (if any)
useful mechanistic information from such Hammett plots.
That the slopes obtained from this study are larger than the 1.58 obtained
from reaction 3.3.3 should come as no surprise in view of the mechanism postulated


62
x
X
Plot of log Kx / Kh (calculated at three hours reaction time) vs. Hammett ax Constants
for the Relative rates of ArCH3 Appearance Study (3.4)
back in figure XI. The migratory ability of a substituted aryl group, although
dependent on anionic resonance stability to a degree, must depend more on its ability
to localize a negative charge through field or solvent-cage effects. Substituted benzyl
anions, on the other hand, should be able to exist as independent entities, capable of
resonance stabilization and possible displacement (of fluoride) reactions at silicon, via
pentacoordination. Such a reaction is necessary to effect the formation of
(X-PhCH2)2SiMe2 and Et(X-Ph)SiCH2PHXMe from the primary migratory
products. Due to the strength of the Si-F bond (154 kcal/mole) however, we believe
that the secondary silanes are formed by the pentacoordinate mechanism forming the
primary products, except that substituted benzyl anion acts in place of fluoride (most
of which has ended up in the primary products and dimethyldifluorosilane by the time
secondary silane formation is observed).


63
3.4.3 Detailed Experimental
3.4.3a p-MePhSiMegCHgCl + KF/18-Crown-6/mesitvlene/85C
Reactant Table gr MW mmoles mole percent KF + 18-Cr-6
p-MePhSiMe9CH?Cl 0.9916 198.77 4.9887 26.6947
KF 0.5820 58.10 10.0172
18-Crown-6 0.1332 264.30 0.5040
Ci6H34 (IS) 0.0604 226.45 0.2667 1.4271
Mesitylene (1.5 ml) 1.6146 120.20 13.4326 Balance
18.6880 - (KF + 18-Cr-6)
Reference Solution er MW mmoles mole oercent
p-Xylene (J.T. Baker) 0.1263 106.16 1.1899 1.3103
Ci6H34 (IS) 0.2712 226.45 1.1976 1.3188
Mesitylene (total) 10.6284 120.20 88.4230 Balance
Total: 90.8105
GC-HD Parameters for both reference and analysis:
Injection size: 0.2 ^1
Temperature program: 70C --> 280C, 8C/min
Signal (Attentuation X range): 32 x 100
Calculation basis: Peak heights (in chart block units)
Reaction Results:
[mmoles IS in reaction = 0.2667]
Rxn Time
1.5 hrs Reference
Reaction
3 hrs Reference
Reaction
4.5 hrs Reference
Reaction
p-xylene
peak height 31.95
mole percent 1.3103
peak height 0.70
mmoles p-xylene present:
peak height 33.70
peak height 0.80
mmoles p-xylene present:
peak height 33.25
peak height 1.75
mmoles p-xylene present:
IS RF
37.05 0.8679
1.3188
55.10
0.0039
39.10 0.8679
28.55
0.0086
37.90 0.8830
30.45
0.0174
(table continued on following page)


64
6 hrs Reference Reaction peak height 33.25 peak height 5.50 mmoles p-xylene present: 37.90 37.10 0.0448 0.8830
lOhrs Reference Reaction peak height 27.6 peak height 29.7 mmoles p-xylene present: 30.55 47.0 0.1853 0.9093
21.5 hrs Reference Reaction peak height 28.2 peak height 28.8 mmoles p-xylene present: 31.5 27.9 0.3055 0.9010
45 hrs Reference Reaction peak height 37.2 peak height 39.85 mmoles p-xylene present: 48.0 20.60 0.6614 0.7800
Table XVn
Results for Rate Studies 3.4.3a -> e
Reaction
mmoles X-ArCH3 produced
Reaction
Time
n-Me H d-F n-Cl m-CF-a
0.5 0.2689
1.0 0.2886
1.5 0.0039 0.0061 0.0193 0.3045
2.0 0.3558
3.0 0.085 0.0227 0.0821 0.2895 0.3726
4.5 0.0174 0.0703 0.1254 0.4559 0.4602
6.0 0.0448 0.0881 0.1357 0.4263 0.4721
10.0 0.1853 0.2557 0.4843 0.4268
21.5 0.3055 0.3664
22.5 0.5645
26.0 0.4802
28.0 0.4846
45.0 0.6614
47.0 0.5164
3.5 Synthesis of Starting Materials
3.5.1 Preparation of (p-CHg-PhlSiMeoCHoCl
I. p-CH3PhBr + Mg ------------> p-CH3PhMgBr
Into a dry 500 ml 3 necked flask, was placed Mg turnings (8.61 gr, 0.354
mole, Fischer) and a magnetic stirring bar. This was covered with 100 ml anhydrous


65
ether. Onto this was attached a N2 equipped reflux condensor and an addition funnel.
Into the funnel, was placed 10 ml Aldrich p-bromotoluene. The system was purged
with N2 and with gentle stirring, was begun the addition of 10 ml portion of
p-bromotoluene. The reaction did not begin until a few drops of 1,2-dibromoethane
were added to the mixture. The balance of the p-bromotoluene was added in a 50/50
solution with ether over a one hour period. Total halide used was 53.7 gr (0.314
mole).
H. p-CH3PhMgBr + ClSiMe2CH2Cl--------> MgBrCl +
(p-CH3-Ph)SiMe2CH2Cl
When Grignard formation was complete (2 hours), 44.9 gr (0.314 mole)
Petrarch ClSiMe2CH2Cl was placed into a clean addition funnel and added the silane
over a one hour period. Precipitation of MgBrCl and ether reflux was observed within
30 minutes. The reaction was magnetically stirred for 24 horns in a darkened hood to
retard the formation of bitolyl from a photochemically caused coupling of the Grignard
reagent.
The unreacted Grignard was destroyed and the MgBrCl cake was dissolved
by adding 400 ml saturated aqueous NH4CI solution to the reaction mixture. The
layers were transfered into a separatory funnel and the aqueous layer was extracted
twice with 200 ml portions of ether. The upper product containing ether layer was
dried over anhydrous sodium sulfate. The sodium sulfate was filtered from the ether
layer in a Buchner funnel and most of the ether was removed under gentle vacuum
through a rotary evaporator. 75 ml of a light yellow fluid was collected.
Vacuum distillation of this fluid through a 30 cm Vigreux column afforded
37.8 gr of a 99.3 % pure (GC-FID) (p-CH3-Ph)SiMe2CH2Cl: MW: 198.77; bp
78.5C/5.5 mm Hg; Yield 60.6 %.


66
Structure verification determined by GC-MS. Relevant ions and (%
abundances) found were: 3^C1 M+ 198 (4.0), 37C1 M+ 200 (1.2); ^5q m+ Me
183 (1.0), 37C1 M+ Me 185 (0.3); 35C1 M+ 35ciCH2 and/or 37C1 M+ -
37C1CH2 149(100)
3.5.2 Preparation of ('p-F-Ph'tSiMeoCHoCl
Using a stepwise procedure analogous to synthesis 3.5a, the Grignard reagent
of p-fluorobenzene was prepared from magnesium turnings (8.61 gr, 0.354 mole,
Fischer), 4-fluorobromobenzene (54.96 gr, 0.314 mole, Aldrich) and 150 ml dry ethyl
ether. The Grignard reagent was then reacted with chloromethyldimethylchlorosilane
(44.9 gr, 0.314 mole, Petrarch) to form (p-F-Ph)SiMe2CH2Cl. 50 ml of a greenish
yellow fluid was collected after the hydrolytic work-up of the reaction mixture.
Vacuum distillation of this fluid through a 30 cm Vigreux column afforded
23.8 gr of a 97.3 % pure (via GC-FID) (p-FPh)SiMe2CH2Cl: MW 202.73; bp
62.7C/6.0 mm Hg; Yield 37.4 %.
Structure verification determined by GC-MS. Relevant ions and (%
abundances) found were: 35C1M+ 202 (5.7), 37C1 M+ 204 (2.0); 35C1 M+ Me
187 (1.5), 37C1 M+ Me 189 (0.65); 35ci M+- 35ciCH2 and/or 37C1 M+-
37C1CH2 153(100).
3.5.3 Preparation of ('p-Cl-Ph')SiMe2CH2Cl
Using a stepwise procedure analogous to synthesis 3.5a, the Grignard reagent
of 4-bromochlorobenzene was prepared from magnesium turnings (5.74 gr, 0.236
mole, Fischer), 4-bromochlorobenzene (50 gr, 0.261 mole, Aldrich) and 160 ml dry
ethyl ether. The Grignard reagent was then reacted with


67
chloromethyldimethylchlorosilane (37.35 gr, 0.261 mole, Petrarch) to form
(p-ClPh)SiMe2CH2Cl. 35 ml of a bluish green fluid was collected after the hydrolytic
work-up of the reaction mixture.
Vacuum distillation of this fluid through a 30 cm Vigreux column afforded
18.47 gr of a 97.3 % pure (via GC-FID) (p-ClPh)SiMe2CH2Cl: MW 219.19; bp
89C/5.5 mm Hg; Yield 32.3 %.
Structure verification determined by GC-MS. Relevant ions and (%
abundances) found were: 33Cl2 m+ 218 (6.08), 33C137C1 M+ 220 (4.92); 37Cl2
M+ 222 (0.77), 35ci2 M+ Me 203 (1.17), 35C137C1M+ Me 205 (0.71); 37C12
M+ Me 207 (0.14); 35ci (p-ClPh)SiMe2 169 (100), 37C1 (p-ClPh)SiMe2 171
(35) (loss of CH2CI from the 3 possible molecular ions).
3.5.4 Preparation of (m-CF3-Ph)SiMe2CH2Cl
Using a stepwise procedure analogous to synthesis 3.5a, the Grignard reagent
of m-trifluoromethylbromobenzene was prepared from magnesium turnings (6.06 gr,
0.249 mole, Fischer), m-trifluoromethylbromobenzene (50 gr, 0.222 mole, Aldrich)
and 150 ml dry ethyl ether. The Grignard reagent was then reacted with
chloromethyldimethylchlorosilane (31.8 gr, 0.222 mole, Petrarch) to form
(m-CF3Ph)SiMe2CH2Cl. 50 ml of a clear brown fluid was collected after the
hydrolytic work-up of the reaction mixture.
Vacuum distillation of this fluid through a 30 cm Vigreux column afforded
22.83 gr of a 99.4 % pure (via GC-FID) (m-CF3Ph)SiMe2CH2Cl: MW 252.45; bp
66.5C/5.5 mm Hg; yield 40.7 %.
Structure verification determined by GC-MS. Relevant ions and (%
abundances) found were: 33C1M+ 252 (trace), 37C1M+ 254 (trace); 3^C1 M+ Me


68
237(3.13), 37C1M+-Me 239(1.37); 35C1M+ F 233(5.49), 37C1M+-F 235
(2.04); 33C1M+ 33C1CH2 and/or 37C1M+ 37C1CH2 203 (100).


69
CHAPTER IV
PYROLYTIC REARRANGEMENTS OF SUBSTITUTED
ARYLCHLOROMETHYLMONO (AND DI-) METHYL SILANES
4.1 Introduction:
Between 1975 and 1980, Brook29 and his co-workers investigated the
thermal rearrangements of a-substituted halomethylsilanes, R3SiCHXR', from their
sealed tube pyrolysis at temperatures up to 330C. Generally, they found clean
exchange of X and R groups between carbon and silicon and concluded that the
rearrangement proceeded by the unimolecular formation of an "inverse ylid" wherein
the migration of X from carbon to silicon occurs via a cyclic transition state (eq. 17).
The rate determining step was
R3$iCHXR' -i> R3Si CHR1 -ii->XR2SiCHRR' (eq.17)
Sx/
envisaged to be (i), the formation of the cyclic intermediate, followed by (ii), the rapid
migration of R from silicon to carbon. They considered the kinetic data from the
observed migratory aptitudes of various R groups to be inconsistent with the formation
of a double-bridged transition state (eq. 18) required for simultaneous migration of X
and R in a dyotropic rearrangement.
,K
R3SiCHXR'---- R2Si CHR' (eq.18)
Nx'


70
Radical involvement at 330C was refuted in Brooks' sealed tube studies due
to the lack of any polyhalogenated side products (arising from chain reactions), and in
a recent study conducted by Davidson30 using a strictly gas phase stirred flow (SFR)
pyrolysis apparatus (see appendix C) in which radical trapping reagents added did not
affect reaction rates or product composition (see discussion section 4.5).
The same (sealed tube) rearrangements at lower temperatures (136C) were
found by Jung and Weber31 and by the author of this thesis at 300C (section 4.2)
however, to yield polyhalogenated side products (e.g. XR2SiCHR'X, RR'SiX2 and
free R-H compounds) which are indicative of radical chain processes. Only the true
gas phase SFR pyrolysis as performed by Davidson and Simon on compounds
prepared in this chapter was found to give totally clean rearrangements.
4.2 Isolation of Sealed Tube Pyrolysis Products from
Phenvlchloromethylmethvlsilane
A sealed borosilicate ampoule containing 0.58 gr PhSiHMeCH2Cl (prepared
in section 4.6) under argon was placed in a muffle furnace at 300C for 48 hours. A
startling change (aromatic > pungent) of odor was noted when the ampoule was
opened for the analysis of its contents. The resultant yellowish brown fluid was found
by GC/GC-MS* to contain 71 % of the proton migration product PhMe2SiCl and at
least 5 other products* but absolutely no chloromethyl starting material.
GC-MS anayltical results; relevant ions (abundances) from a Hewlett
Packard 5890 GC with a 25 m x 0.2 mm crosslinked methylsilicone gum capillary
column and a 5970 mass selective detector utilizing electron impact ionization 70 (eV)
and quadrupole ion sorting were as follows:


I.
Starting material:
PhSiHMeCH2Cl: F.W.: 170.71; 35C1 M+ 170(1504), 37C1 M+ 172
(646); 35C1 M+- Me 155 (991), 37C1 M+-Me 157(306); M+
- CH2CI 121 (49184), 35d M+ Ph 93 (3025), 37C1 M+ Ph
95 (2819); Ph+ 77 (5668).
II. Main Products (listed in elution order-):
A. C6H6: M+ 78(50528); M+-H 77(43184); H2C=CH-HC=CH2 53
(1472); [H2C=C=C=CH2]+ 52 (37312); +HC=C=C=CH2 51
(37112).
B. PhSiHMe2: M+ 136(24456); M+-H 135(20904); M+ Me 121
(49984); M+-Ph 59(4951); MeHSi+=CH2 58 (37312); Ph+
77 (4294).
C. PhSiMe3: M+ 150(12634); M+ Me 135(50528); M+-Ph 73
(2307); Ph+ 77 (2627).
D. Main Product: PhSiMe2Cl: 35C1 M+ 170 (11968), 37C1 M+ 172
(4897); 35Q M+-Me 155 (50528), 37C1 M+-Me 157
(24456); 35C1 M+-Ph 93 (9228), 37C1 M+-Ph 95(2462);
Ph+ 77(16384). Lack of M+ CH2C1 at 121; 35C1 +C1(CH2)2
63 (43896), 37C1 +C1(CH2)2 65 (24456).
E. Possible Ph migration product (PhCH2SiHMe) in D due to the
presence of two ions: tropylium cation 91 (19904); M+- CH2Ph
79 (4924) (The relative abundance of these ions increases


72
slightly as spectra are taken towards the end of peak D on the
TIC chromatogram).
F. Mixture of PhMeSiCl2 and PhCH2SiQ2H: 35C12 M+ 190 (19480),
35ci37d M+192 (12164); 37d2 M+ 194 (2307); 35ci2 M+
-Me 175(50528), 35Cl37a M+- Me 177 (49448); 37q2 M+
-Me 179(9848); 35q2 M+-Ph 113(4897), 35Ci37Ci M+-
Ph 115 (3057); 37d2 M+ Ph 117 (744); Ph+ 77 (10452);
tropylium cation 91 (4664); 35q2 M+ CH2Ph 99 (384),
35q37q M+- CH2Ph 101 (622); 37a2 M+-CH2Ph 103
(748); 35q M+. ci 155 (2992), 37q M+ Cl 157 (1141).
G. HPhSi(CH2Cl)2: 35Cl2 M+ 204 (4664), 35Cl37Cl M+ 206 (2773);
37d2 m+ 208 (583); 35a M+ CH2C1 155 (49984), 37q
M+-CH2C1 157 (36912); 35q2 M+-Ph 127(378), 35q37q
M+ Ph 129 (580); 37q2 M+ Ph 131 (368).
4.3 Argon Flow Pyrolysis of ArSiHMeCH2Cl and ArSiMeoCHoCl
4.3.1 General Experimental
Our next plan was to conduct this thermal rearrangement in such a manner as
to let it proceed partway, so that some sort of relative rate data could be obtained when
one varied the aryl substituents on the starting material thus giving insight into the
electronic effects involved in the rearrangement. Two groups of silanes,
ArSiHMeCHCl and ArSiMe2CH2Cl, with the same five aryl substituents used in
Chapter 3 were chosen for the initial pyrolysis study performed by myself. This could
possibly yield relative migration ratios for H/Ar and Ar/Me migration products from
the respective pyrolysis, thus providing additional mechanistic insight. The method


73
chosen was a simple argon-flow drop tube pyrolysis through a calibrated tube furnace.
The pyrolysis tube itself was 22 cm long x 1 cm O.D. (0.75 cm I.D.),
constructed of fused silica with 14/20 fittings at both ends and having a silica frit near
its base so that it could contain packing material. This was chosen to be thirty five 0.6
cm diameter borosilicate beads for obvious reasons, such as surface and packing
uniformity and a relative lack of flow restriction for either gas or liquid.
The furnace itself encased the tube entirely so that only the 14/20 fittings on
either end were out of the hot zone. To the top of the tube was attached a graduated 10
ml addition funnel equipped with an argon inlet. To the bottom, was attached a 10 ml
two neck round bottom flask placed in an ice-salt bath with a small diameter gas exit
tube in one heck.
To establish working pyrolysis parameters, it was then necessary to calibrate
a 120 V variac dedicated to the tube furnace. This was done twice using a
thermocouple pyrometer so that a given reading on the variac corresponded to a certain
temperature in C 5 %. The parameters chosen for the argon flow pyrolyses of
ArSiHMeCH2Cl were as follows: Volume of ArSiHMeCH2Cl = 1 ml; Pyrolysis
temperature = 560 C (below the softening point of borosilicate beads); Hot zone
contact time* per drop of silane = 1.0 sec.; Addition rate = 1 drop/12 seconds over a 2
minute period. Determined from a heat expansion neglected computation of the hot
zones internal volume (via volumetric addition of water into the bead filled tube = 5.04
cc), and adjusting the argon,flow rate through the heated tube to 5.04 cc/sec or 300
cc/min. The argon flow rate was checked just prior to each pyrolysis with a flowmeter
designed by Alltech Inc.
Pyrolysis parameters established for the ArSiMe2CH2Cl compounds were
identical except for a tube furnace temperature of 700C. This necessitated the use of
crushed fused silica packing material in the pyrolysis tube. Pyrolyses of the


m-CF3PhSiMe2CH2Cl compound at 560C and 1000C gave no rearrangement and
destructive carbonization respectively.


74
4.3.2 Results of the argon flow pvrolvses of substituted arvlchloromethvl mono
(and dD methvlsilanes
Approximately 0.5 ml of each pyrolysis effluent was condensed and
immediately submitted for both GC-MS and GC-FID analyses. (Initial FID results
obtained on a 6' x 1/8" SS 10 % OV-1 silicone oil/chromosorb WHP column in a
Perkin Elmer 3020B chromatograph using peak heights; subsequent GC-FID analysis
of the same effluents, after one month refrigerated (-10C) storage in paraffin sealed
vials, were performed on a Hewlett Packard 5890 chromatograph using the same 25 m
x 0.2 mm crosslinked methylsilicone gum capillary column as contained in the HP
GC-MS system). GC results obtained on both HP instruments total ion current
(TIC) from the GC-MS and the FID were reported using software integrated peak
areas. It might be mentioned here that any "quantitative" information derived from the
TIC on GC-MS systems is questionable due to the fact that a given TIC response
factor for any molecule is dependent on its ionization potential and the stability of its
molecular ion to further fragmentation reactions. Thus, the TIC response factors for
the H/PhX and PI1X/CH3 migration products observed from the respective argon flow
pyrolyses of X-PhSiHMeCH2Cl (X-PhSiClMe2 / HSiClMeCH2PhX) and
X-PhSiMe2CH2Cl (ClSiMe2CH2PhX / X-PhSiCIMeEt), would be vastly different
due to significant variance in molecular structure; i.e. the presence of the benzylic
products due to aryl migration would result in a higher TIC response factor than for
migratory products involving protons or methyl groups because of the favorable
fragmentation of the former to substituted tropylium cations (in addition to other
reactions). Table XVm and XIX summarize the observed products and ratios of
interest.


75
Table XVm
560C Argon flow pyrolyses of
X-PhSiHMeCH2Cl--------------> X-PhSiClMe2 + HSiClMeCH2PhX
S.M. volatility X GC-FID % completion TIC H/Ar FID* H/Ar FID** H/Ar
4 p-Me 5.61 11.81 112 OO
2 H 18.4 7.0 743 oo
3 p-F 12.2 52.7 182 OO
5 p-Cl 6.86 333 oo
1 m-CF3 22.7 4.42 138 oo
* 6' column/PE
** 25 m column/HP; one month -10C storage
*** Both proton and aryl migration products were seen to be contained in a
single GC-MS peak. This was evidenced by an increasing abundance of the p-Cl
tropylium cation as one took spectra towards the end of the peak. The volatility of the
aryl migration product was found to be less than the proton migration product in all
cases.
**The complete dissappearance of the aryl migration products after a month's storage
at -10C is probably indicative of the instability of silanes containing three reactive
substituents; H, Cl and CH2Ph.
*The earlier FID ratios obtained suggest little if any correlation with electronic
substituent effects and may only represent an in-situ (column) thermal decomposition
of the aryl migration products into the more stable proton migration products. Indeed,
the major correlation to be observed from the argon flow pyrolysis of
X-PhSiHMeCH2Cl was the percent completion of the rearrangement with respect to
the volatility of the starting material; (i.e., the lower boiling point of the starting
material, the greater the percent completion of the rearrangement at 560Q.
The GC-MS data of the argon flow from the X-PhSiHMeCH2Cl compounds
showed virtually the same analogous product distribution obtained in the


76
sealed-ampoule pyrolysis of the X = H compound.
Table XIX
700C Argon flow pyrolyses of
X-PhSiMe2CH2Cl-----------> X-PhCH2SiClMe2 + EtSiCIMePhX
S.M. GC-FID % TIC FID* FID**
X volatilitv completion Ar/CH3 Ar/CH3 Ar/CH3
p-Me 4 23.32 3.09 5.88 (4.53)
H 2 43.71 2.42 2.93 (4.87)
p-F p-Cl m-CF3 3 5 1 21.43 42.88 22.2 UR*** sfe sfc UR UR UR (4.66) (UR****) (UR****)
UR = unresolved
* 6' column/PE
** 25 m column/HP; one month -10C storage
*** Mass spectra indicated an increasing abundance of fragment ions
common to the methyl migration product (i.e. C2H5+) as one took spectra towards the
end of the pyrolysis product peak. This was taken as evidence of both migrations
occuring.
**** Software magnification of the pyrolysis peak obtained with a slow
(l-2C/min) temperature program showed a shoulder on the tailing side. Modification
of the temperature program and carrier gas flow rate may effect the resolution.
Representative mass spectra of the minor and least volatile migratory products
(X-Ph and Me) from the respective pyrolyses of X-PhSiHMeCH2Cl and
X-PhSiMe2CH2Cl are given below for the X = H compounds (analogous spectra
were obtained from the other compounds):
HSiCIMeCHoPh: 35a M+ 170 (12296), 35a M+ 172(5198); 35aM+-Me
155 (6115), 35q M+- Me 157(2599); 35aM+-H 169 (3057), 35a


77
M+-H 171 (2928); M+-C1 135(1292); 35q M+ CH2Ph 79
(20568), 35C1M+ CH2Ph 81(4845); tropylium cation 91(48912);
toluene cation 92 (48912); Phenyl cation 77 (20680); benzene
cation 78 (12033).
EtSiCIMePh: 35C1 M+ 184 (3961), 33C1M+ 186(1229); 35C1M+-Me 169
(990), 35C1M+ Me 171 (307); 35C1M+ Et 155 (10000), 35C1M+ -
Et 157(9839); M+-C1 149 (157); 35C1M+- Ph 107 (247), 35C1M+
- Ph 109 (37); phenyl cation 77 (1203); benzene cation 78 (985);
relative lack of the X-substituted tropylium cation at 91 and of a
molecular ions loss of the same at 93.
A representative mass spectra of the major (X-Ph) migratory product for the argon
flow pyrolysis of the aryl dimethylchloromethylsilanes is given below for the X = H
compound. (Analogous spectra were obtained from the other compounds).
PhCIP^SiClMeo: 35C1M+ 184 (4892), 37C1M+ 186(1969); 35dM+-Me 169
(979), 35C1 M+- Me 161 (305); M+- Cl 149(621); 33C1 M+ Ph 107(117),
35C1 M+ Ph 109 (7.0); 35C1 M+ CH2Ph 93 (10000), 35C1 M+ CH2Ph 95
(9785); tropylium cation 91 (4838).
Excluding the methyl migration, other minor (and more volatile) products
from the argon flow pyrolysis of the aryldimethylchloromethylsilanes are represented
by the GC-MS data obtained for the p-Me-Ph compound below:
Me?SiCl: 35ciM+ 108 (38), 37C1M+ 110(10); 35dM+-Me 93(1321),


78
35C1M+ Me 95 (509); M+ Cl 73 (382).
Me3SiCh: 35C12M+ 128 (95), 35C137C1M+ 130(71); 35C12M+-Me 113
(1512),35d37ClM+ 115 (1231), 37C12M+-Me 117(189); 35M+-
C1 73 (41). 37C1M+ Cl 75 (31).
p-MePh-SiMe?: M+ 164 (5856); M+ Me 149 (47096); +SiMe3 73 (986);
tropylium cation 91 (1357).
p-MePhSiCIMeo: 35ciM+ 184(5399), 37C1M+ 186(1537); 35dM+-Me
169(24192), 37C1M+ Me 171(10509); 35Cl+SiMe2 93(1292),
37Cl+SiMe2 95(368); tropylium cation 91 (4845).
n-MePhSiCbMe: 35d2M+ 204 (1211), 35d37ClM+ 206(752), 37C12M+
208 (156); 35C12 M+ Me 189 (3987), 35C137C1M+ Me 191 (2819),
37C12 M+ Me 193 (612).
bitolvl: M+ 182(1238);* M+-Me 167(212); MePhCH2+ 105(986); phenyl
cation 77 (3482).
dihvdrobitolvl: M+ 184(45); M+-H 183(188); M+-2H (bitolyl cation)
182(752); M+-3H 181(186); 167(190); MePhCH2+ 105(2320);
phenyl cation 77 (3008).


79
4.4 Stirred flow pvrolvses of X-PhSiHMeCHoCl and X-PhSiMegCHoCl
4.4.1 General Experimental:
After hyper-purification of both sets of 4 compounds via vacuum distillation
through a micro-scale spinning band column (mfd by B/R instruments) to liquid phase
purities of 99.999%, they were pyrolyzed at 2 atm in a nitrogen carrier gas flowing
through a stirred flow pyrolysis apparatus having a flame ionization detector. Injection
of the samples was conducted entirely in the gas phase by pumping the injection
chamber down to a pressure of 0.07 to 0.15 torr* before the introduction of the
nitrogen carrier gas at 2 atm. *The change in the partial pressure of the sample over
this range did not alter the data obtained in this experiment, which was of the kinetic
sort; i.e., providing the Arrhenius parameters (A and Ea) of the rearrangements to
yield the rate constants for the migrations observed in both sets of compounds.
Pyrolysis oven contact time was set at 15 seconds. Computations provided by R.
Simon from these data provided the comparative percent completions (via GC-FID
peak areas) for all but the X = p-Cl compounds presented in tables XX and XXT
below.
4.4.2
Table XX
Results of the Stirred flow pyrolysis of X-PhSiHMeCH2Cl at 560C
calculated % consumption of S.M.
X p-Me chloromethvl starting material 64 Volatility 4
H 10 70 2
p-F 10 64 3
m-CF3 10 62 1


80
4.4.3
Table XXI
Stirred flow pyrolyses of X-PhSiMe2CH2Cl at 700C
approximate calculated % consumption of
X kx Phm{ /kMe chloromethvl starting material* Volatility
p-Me ^10 109 4
H 10 88 2
p-F 10 110 3
m-CF3 10 91 1
* approximate percent completion, 10 % min.
Tables XX and XXI yield the following rates of substituent migration for
chloromethylsilane pyrolyses: Proton/Aryl/Methyl = 100/10/1.
Final note on stirred flow pyrolyses:
Mass spectral analyses of all pyrolysis affluents showed virtually none of the
volatile side products (e.g. PhSiMe3, PhMeSiCl2 from PhSiHMeCH2Cl and
biphenyl; Me3SiCl; Me2SiCl2; PhSiCl2Me from PhSiMe2CH2Cl) obtained in the
argon flow experiments.
4.5 Conclusions from both pyrolytic methods
After noting the lack of starting material volatility correlation with the percent
pyrolysis completion in table XX, compound with the blatant correlation of such in
table XVm (along with the smaller magnitude of percent completions observed in
XVm), it is apparent that 560C may have been too low a temperature for the
obtainance of any mechanistically interesting data from the X-PhSiHMeCH2Cl
compounds. Excluding the X = p-Cl silane however, one obtains a Hammett
correlation coefficient of 0.78 with a p value of 0.79 in a least squares graph of the


81
relative (to X = H) log(percent completions)* vs &x- *(A rough estimate of the
relative rates of pyrolyses with respect to the X = H compound). The presence of the
volatile side products in the argon flow experiments, especially of biaryl compounds is
indicative of the presence of radical coupling chemistry.32 Reference 32 in fact,
details the former large scale production of biphenyl from the pyrolysis of benzene in a
similar apparatus. The presence of electron withdrawing groups (e.g. p-F and
m-CF3) on the aromatic ring in the pyrolysis of the X-PhSiMe2CH2Cl compounds
was seen to lower the overall pyrolysis rate. This is not unexpected because the
unpaired electron involved in a radical mechanism prefers to be in a planar
configuration (with the tc system in this case) with its parent atom. An electron
withdrawing group such as CF3 attached to one of the six carbon atoms of the 7t
system should statistically lower the reactivity of the aryl radical by enticing the
unpaired electron to occupy the tetrahedral sp3 orbitals (in the PI1-CF3 bond) for a
finite percent of its lifetime. This delocalization may also explain the low percent
completion in the p-F case. [That the phenyl (X = H) case was seen to have the
highest percent completion may be explained by the lack of steric problems in the
migration of a more localized radical. The lower percent completion of the p-Me case
(presumably the most localized aryl radical) is probably due only to steric problems.
[The second highest percent completion found for the p-Cl case cannot be explained
from either viewpoint unless one wishes to invoke the possibility of radical
interactions such as Ph-Cl + Cl* > Ph* + CI2*. This may help in the formation of
the dichloro silanes observed as well as in increasing the reactivity of the aryl radical
for migration to a hypothetical Me2SiClCH2* radical, followed by a reversal of to
generate the major p-Cl benzylic product. (Interesting GC-MS results would be found
from the pyrolysis of p-^^ClPhSiMe2CH2^^Cl to either refute or accept such a
hypothesis).


82
Further evidence of radical involvement could also be found by a high
temperature ESR study (i.e., the presence of an ESR absorption would favor radical
processes over anionic mechanisms even though they (the latter) may still be occuring.
When I.N. Jung and W.P. Weber31 conducted the sealed tube pyrolysis of
HSiMe2CH2Cl at 136C, they found that the main initial products were equimolar
amounts of Me3SiH and ClSiMeCH2Cl. Later in the reaction, these products
decreased while the proton "migration" product Me3SiCl became the main product.
Their results were very reasonably interpreted in terms of two consecutive radical
chain sequences represented below by reactions 1 to 5.
1) HSiMe2CH2Cl------> HSiMe2CH2* + Cl-
2) HSiMe2CH2- + HSiMe2CH2Cl -----> -SiMe2CH2Cl + Me3SiH
3) -SiMe2CH2Cl + HSiMe2CH2Cl ------> HSiMe2CH2- + ClSiMe2CH2Cl
4) Me3SiH + Cl------> Me3Si- + HC1
5) ClSiMe2CH2Cl + Me3Si------> Me3SiCl + -O^Sil^SiCl
Simple intramolecular radical shifts between carbon and silicon such as
PhSiMe2CH2C1 70Bp-> Ph$iMe2CH2- + Cl'
SiMe2CH2Ph -SiPhMeEt
ClSiMe2CH2Ph
CISiPhMeEt
Figure XV
Two Step Radical Mechanism for the Pyrolytic Rearrangement of PhSiMe2CH2Cl to
ClSiMe2CH2Ph and CISiPhMeEt


83
PhSiMe2CH2-------> SiM^CE^Ph have been shown not to occur by Wilt33 in
observing them in the liquid phase at 150C. Thus the involvement of a simple two
step mechanism such as shown below for the argon flow pyrolysis of PhSiMe2CH2Cl
is probably non-existent.
The presence of the Me3SiH analogue, PhMe2SiH in our 560C argon flow
and 350C sealed tube pyrolyses of PhSiMe2CH2Cl, indicate a similar radical chain
mechanism in operation, but no ClSiMe2CH2Cl was found. In fact, for both sets of
argon flow pyrolyses, the only side products containing two chlorine atoms were
Figure XVI
Possible Carbene Elimination Mechanism of ArSiCl2Me formation from the Sealed
Tube of Argon Flow Pyrolysis of ArSiMe2CH2Cl
dichlorosilanes of the type X-PhSiMeCl2 and X-PhCH2SiHCl2 from the pyrolysis of
X-PhSiHMeCH2Cl and X-PhSiMeCl2 and Me2SiCl2 from the pyrolysis of
X-PhSiMe2CH2Cl. A possible route for the formation of X-PhSiMeCl2 from the
unobserved ClSiMe2CH2Cl analogue, ClSiMePhCH^Cl, may involve direct
chloromethyl attack at silicon forming a "ylid" pentacoordinate intermediate followed
by carbene elimination (Fig. XVI). Pyrolysis of vinylchloromethyldimethylsilane may
prove carbene involvement if cyclopropylmethyldichlorosilane were to form (Fig.


84
Cl
>
Cl
Cl
+
>'I
1
MeSi
Cl
Figure XVII
Test for a Carbene Elimination Mechanism
XVH).
The formation of such a pentacoordinate "ylid" may only be favorable at
higher temperatures and would thus explain the lack of chloromethyl-chlorosilane
intermediate found in our studies.
The presence of substituted benzenes (in the pyrolysis of X-PhSiHMeCH2Cl)
and of substituted biphenyls (in the pyrolysis of X-PhSiMe2CH2Cl) may be indicative
of the high temperature favored disproportionation of H/MeSiPhMeCH2* radicals into
silenes, RRSi=CH2, with loss of either H* or Ph\ (Loss of Me* may explain the
trimethylarylsilanes observed in both pyrolysis sets).
In observing that the kinetic data obtained from the stirred flow pyrolyses of
both X-PhSiHMeCH2Cl and X-PhSiMe2CH2Cl appear to be independent of the aryl
substituent chosen and that no observable side products possible resulting from
radical processes were found, it is concluded by the author that neither radical nor
simple anionic processes are responsible for the clean reactions provided by gas phase
- stirred flow pyrolysis.


85
The refutation of a gas phase radical pathway was substantiated in a paper
submitted by I.M.T. Davidson30 and S.I. Maghsoodi in 1986. Their work involved
the attempted trapping of the suspected HMe2SiCH2- radical from the stirred flow
pyrolysis of HMe2SiCH2Cl in the same apparatus as used in section 4.4. The
trapping reagents chosen were 10 fold excesses of methanol, methyl chloride, toluene,
butadiene and propene. The kinetics of formation of Me3SiCl were unaffected by any
of these well known gas phase radical forming reagents. No new products were
detected in any of these experiments nor in others with a 30-fold excess of butadiene
or a 200-fold excess of propene.
The authors concluded that the gas phase rearrangement of HMe2SiCH2Cl
> Me3SiCl proceeded by the unimolecular formation of a 3 center "chloronium"
transition state involving pentacoordinate silicon followed by hydride migration to a
probable developing positive charge on the chloromethyl carbon:
Figure XVH[
Unimolecular Mechanism for the Gas Phase Rearrangement of HMe2SiCH2Cl to
Me3SiCl
The degree of charge separation in the proposed three center intermediate
remains to be investigated since the alterations of the aryl substituents in experiments
4.3 and 4.4 did not affect observed product ratios or formation rates to any major
extent. A bit of fine tuning of the pyrolytic parameters (esp. in studies 4.3) to help the
reproducibility of migratory/kinetic data would be of a great asset in the determination
S
+


86
of charge and field effects involved in aryl migration.
That others31 and the author of this thesis have found that sealed-tube or
liquid drop flow pyrolyses of chloromethylsilanes produce radical induced volatile side
products in addition to the predominant migratory (of R or H) monochlorosilanes,
should well confirm the dependence of the reaction mechanism on the molecular
population density change occuring when one pyrolysis a strictly gas phase sample as
compared to the pyrolysis of a two phase system.


Full Text

PAGE 1

SOLUTION AND GAS PHASE REARRANGEMENTS OF CHLOROMETIIYLSll .ANES by Vernon Edward Yost B.A., University of Colorado at Denver, 1981 A Thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fu1fillment of the reqUirements for the degree of Master of Science Department of Chemistry 1988

PAGE 2

This Thesis for the Master of Science Degree by Vernon Edward Yost has approved for the Department of Chemistry by

PAGE 3

ABSTRACT CHAPTER I: The gas phase reactions of R-substituted trimethylsilanes with hydroxide and fluoride ion have been shown to produce the pentacoordinate anions Me3RSiOand Me3RSiFrespectively. Me3RSiOwas found to undergo immediate migratory rearrangement of methyl or R groups to the hydroxyl proton to form Me2RSiOand methane or Me3SiOand R-H. The ratios of RIMe migration were shown to be strongly dependent on the gas phase acidities of the respective R-H hydrocarbon. Me3RSiFwas found to be most stable when R was alkyl, but decomposes to Me3SiF and Rwhen R was benzyl, phenyl, alk:enyl or alkynyl. CHAPTER II: Solutions of R-substituted chloromethyldimethyl silanes in aromatic solvents have been found to react with inorganic fluoride anion via phase transfer catalysts such as 18-crown-6 to effect two primary migrations; that of an R group, to produce RCH2Si(F)Me2, and that of either of the methyl groups, to produce EtSi(F)(Me)R. The ratios of the RIMe migration have been correlated with the gas phase acidities of the respective RH hydrocarbons and with various a type substituent constants in the Hammett fashion. These have made it clear that the cleaved carbon fragment undergoing migration has substantial carbanionic character. CHAPTER III: Solutions of substituted arylchloromethyldimethyl silanes in appropriate aromatic solvents have been found to react with fluoride anion via an

PAGE 4

iv 18-crown-6 phase transfer catalyst to produce two interesting side products (dimethyldifluorosilane and substituted toluenes) in addition to two sets of Ar/Me migration products: 1 ArCH2Si(F)Me2 and EtSi(F)(Me)Ar 2 (ArCH2)2SiMe2 and ArCH2Si(Et)(Me)PhX Hammett studies using O'x were found to correlate very well with a variety of experiments and led to the conclusion of a pentacoordinate (at silicon) reaction mechanism. CHAPTER N: The simple (drop by drop-tube furnace) and vacuum pyrolysis of substituted arylchloromethyldi(and mono)methyl silanes were found to produce chlorosilanes which have undergone RIMe and H/Me migration respectively. The relative migratory ratios obtained in the tube furnace studies were found to give crude estimates of kinetic parameters used in the true gas phase, vacuum pyrolytic method (aka Stirred Flow Pyrolysis). The rearrangement mechanisms were found to be dependent on the technique of pyrolysis used.

PAGE 5

v ACKNOWLEDGEMENTS Much credit for the research contained herein belongs to : Dr. Robert Damrauer, Thesis Supervisor Stephen E. Danahey Bonnie K. O'Connell Roger Simon

PAGE 6

CONTENTS INTRODUcriON........................................................................................... 1 CHAPTER I GROUNDWORK FOR REARRANGEMENT STUDIES 1.1 Gas Phase Acidities of Hydrocarbons ....................... ........... 3 1.1.1 Review................................................................................... 3 1.1.2 Experimental Results ............................................................ 4 1.2 Pentacoordinate Silicon Anions............................................ 6 CHAPTER IT THE RELATIVE :MIGRATORY ABILITIES OF ALKYL, VINYL AND ARYL GROUPS IN SOLUTION PHASE FLUORIDE INDUCED REARRANGEMENTS OF 2.1 2.2 2.3 2.4 2.5 2.5.1 2.5.2 2.5.3 2.5.4 2.6 2.6.1 2.6.2 Introduction ....................................................................... Summary of Optimization Experiments ............................. Results and Discussion ...................................................... Alkoxide Rearrangements of Chloromethylsilanes ......... Synthesis of Starting Materials .......................................... Preparation of EtSiMe2CH2CI.. ........................................ Preparation of n-BuSiMe2CH2CI. ..................................... Preparation of i-PrSiMe2CH2CI. ....................................... Preparation of (CH2)2SiMe2CH2CI.. ................................ Optimized RSiMe2CH2CI + KF Reactions ....................... Detailed Experimental -Reaction ofEtSiMe2CH2Cl with KF ................................ Reaction of n-BuSiMe2CH2Cl with KF ............................ 2.6.3 Large Scale Preparation of n-CsH 11 SiMe2F 8 9 12 17 19 19 20 21 24 26 26 26

PAGE 7

2.6.4 2.6.5 2.6.6 2.6.7 2.6.8 2.6.9 2.6.9a 2.6.9b vii and Et(n-Bu)SiMeF............................................... 27 Reaction of i-PrSiMe2CH2Cl with KF................................ 28 Reaction of cy-PrSiMe2CH2Cl with KF.............................. 29 Reaction of H2C=CHSiMe2CH2Cl with KF....................... 30 Reaction of PhSiMe2CH2Cl with KF ............................. :..... 31 Reaction of 1-( chloromethyl)-1-methyl-silacyclopentane with KF........................................................... 33 Additional Reactions of Mechanistic Interest....................... 33 Reaction of Me3Si(CH2)3Cl with KF.................................. 33 Reaction of (ClCH2)2SiMe2 with KF.................................. 33 CHAPTER ill MECHANISTICKINETIC STUDIES OF ARYL (Clll..OROMETIIYL)-DIME1HYLSILANES 3.1 3.2 3.2.1 3.2.2 3.2.3 3.2.3a 3.2.3b 3.2.3c 3.2.3d 3.2.3e 3.3 3.3.1 3.3.2 3.3.3 Introduction ......................................................................... Aquisition of Primary and Secondary Products from the Reaction of substituted-aryl(chloromethyl) dimethyl silanes with KF .................................................... General Experimental ......................................................... Results and Discussion ....................................................... ,.. Detailed Experimental ....................... ................................ Reaction ofp-Me-PhSiMe2CH2CI with KF ...................... Reaction of PhSiMe2CH2CI with KF .... ............................ Reaction ofp-F-PhSiMe2CH2CI with KF ......................... Reaction of p-Cl-PhSiMe2CH2CI with KF ........................ Reaction of m-CF3-PhSiMe2CH2CI with KF ................... Competitive Relative Rates of Starting Material Consumption .................................................... .................. General Experimental ......................................................... Results and Discussion ........................................................ Detailed Experimental ......................................................... 35 37 37 37 44 44 45 46 47' 49 52 52 53 55

PAGE 8

viii 3.3.3a PhSiMe2CH2Cl and p-Me-PhSiMe2CH2CI....................... 55 3.3.3b->d p-F-PhSiMe2CH2Cl, p-Cl-PhSiMe2CH2Cl, and m-CF3-PhSiMe2CH2Cl vs. p-Me-PhSiMe2CH2CI............ 57 3.4 Appearance Rates of X-sustituted toluenes (ArCH3)........................................................ ........................ 58 3.4.1 General Experimental............................................................ 58 3.4.2 Results and Discussion.......................................................... 59 3.4.3 Detailed Experimental ........................................................... 63 3.4.3a Appearance Rate of p-Xylene from the Reaction of p-Me-PhSiMe2CH2Cl with KF............................................ 63 3.4.3b->e Appearance Rates of p-F, p-Cl, m-CF3 and Normal Toluenes from the Reactions ofp-F, p-Cl, m-CF3 and H-PhSiMe2CH2Cl with KF............................. 64 3.5 Synthesis of Starting Materials............................................ 64 3.5.1 Preparation ofp-Me-PhSiMe2CH2CI................................. 64 3.5.2 Preparation ofp-F-PhSiMe2CH2CI.................................... 66 3.5.3 Preparation of p-Cl-PhSiMe2CH2CI.................................. 66 3.5.4 Preparation of m-CF3-PhSiMe2CH2CI..................... 67 CHAPTER IV PYROL YTIC REARRANGEMENTS OF SUBSTITUTED ARYL-CHLORODI(AND MONO) ME1HYL SILANES 4.1 4.2 4.3 4.3.1 4.3.2 4.4 4.4.1 Introduction and Review .................................................... Isolation of Sealed Tube Pyrolysis Products from Phenylchloromethylmethylsilane ............................... ;.; .... .. Argon Flow Pyrolysis of ArSiHMeCH2Cl and ArS iMe2 CH 2 Cl .......... ............................................... ..... General Experimental ......................................................... Results of the Argon Flow Pyrolysis of Substituted Arylchloro methylmono (and di-) methylsilanes ............... Stirred Flow Pyrolysis of ArSiHMeCH2Cl and ArSiMe2CH2CI .................................................................. General Experimental ......................................................... 4.4.2 Results of Stirred Flow Pyrolysis of 69 70 72 72 74 79 79

PAGE 9

1X ArSilrneCH2CI.................................................................... 79 4.4.3 Results of Stirred Flow Pyrolysis of ArSiMe2CH2Cl...................................................................... 80 4.5 Conclusions from Both Pyrolitic Methods ................. 80 4.6 Synthesis of Arylchloromethylmethylsilanes....................... 87 4.6.1 Preparation of PhSilrneCH2CI.. ...................... 87 4.6.2 Preparation ofp-Me-PhSilrneCH2CI................................. 90 4.6.3 Preparation ofp-F-PhSilrneCH2CI.................................... 91 4.6.4 Preparation ofp-Cl-PhSiHMeCH2CI................................... 92 4.6.5 Preparation of m-CF3-PhSilrneCH2CI.............................. 93 REFERENCES................................................................................................ 94 APPENDIX A Second Order PolynomialLeast Squares Obtained Fit for Reaction 3.4.3a->e Profiles .................................. ,;. 96 B Argon Flow Pyrolysis Apparatus................................................... 102 C Stirred Flow Pyrolysis Apparatus........................................ 103

PAGE 10

TABLE I. II. III. IV. v. VI. VII. VIII. IX. X. XI. XII. XIII. TABLES Relationship of Mlacid and Electron Affmities ofR to the Relative Migratory Abilities of Alkyl, Alkenyl and Aryl Groups in the Gas Phase Reaction ofMe3SiR with OH................................... 4 Relationship of the Relative Migratory Abilities of Alkyl, Alkenyl and Aryl Groups to Taft a* and o1 constants for the Solution Phase Reactions of RSiMe2CH2Cl with KF.............................................. 12 GC-FID of EtSiMe2CH2Cl + KF Reaction Mixture with respect to time.................................................................................. 26 of n-BuSiMe2CH2Cl + KF Reaction Mixture w1th respect to ttme.. .. .. .... .. .. ...... .. .. .......... .. .. ........ .... ..... .. .. .. .... .. .. .. .. .. 27 of scale n-BuSiMe2CH2Cl + CsF Reaction Mixture Wlth respect to time............................................................. 28 of + KF Reaction Mixture Wlth respect to 28 of + KF Reaction Mixture Wlth respect to time............................................................ 29 of H2<;=CHSiMe2CH2Cl + KF Reaction Mixture Wlth respect to time............................................................ 30 of + KF Reaction Mixture w1th respect to time........................................................... 31 Generalized Product Distribution from Reactions of ArSiMe2CH2Cl with KF ............................................................ 38 Grouping and Ratios of Primary and Secondary Products from Reactions of ArSiMe2CH2Cl with KF............................................ 38 Detailed Gravimetric Preparation ofPhSiMe2CH2Cl vs. p-Me-PhSiMe2CH2Cl + KF Competition Reaction Mixture......... 55 Detailed Gravimetric Preparation of Hexadecane/PhSiMe2CH2Cl, p-Me-PhSiMe2CH2Cl Reference Solution..................................... 55 XIV. Example of GC-FID Response Factor Calculation from Reference

PAGE 11

xi Solution Injection........................................................................... 56 XV. Example of GC-FID Quantification of Competition Study 3.3.3a Results ................................................................................ -.............. 56 XVI. Summary of Competition Study 3.3.3a->d Results......................... 57 XVII. Summary of ArCH3 Appearance Rate Study (3.4.3a->e) Results.. 64 x.vm. Results of the 560C Argon Flow Pyrolysis of ArSiHMeCHzCl.. 75 XIX. Results of the 700C Argon Flow Pyrolysis of ArSiMezCHzCl... 76 XX. Results of the 560C Stirred Flow Pyrolysis of ArSiHMeCHzCl.. 79 XXI. Results of the 700C Stirred Flow Pyrolysis of ArSiMezCHzCL 80

PAGE 12

FIGURES FIGURE I. Plot of log ([Me3SiO-] I [Me2RSiO-]) (statistically corrected) vs. AHacid (of R-H) ....................................................... :................... 5 II. Plot oflog ([R migration] I [Me migration]) vs. AHacid (ofR-H) for Solution Phase Reactions (2.6.1-2.6.7) ofRSiMe2CH2Cl with KF ....................... ...................................................... 13 III. Plot oflog ([R migration] I [Me migration]) vs. Taft a* for Solution Phase Reactions (2.6.1-2.6.7) ofRSiMe2CH2Cl with KF........... 13 N. Plot oflog ([R migration] I [Me migration]) vs. Hammett a1 Constants for Solution Phase Reactions (2.6.1-2.6.7) of RSIMe2CH2Cl with KF.................................................................. 14 V. Simple Pentacoordinate Mechanism for Solution Phase Rearrangements of RSiMe2CH2Cl Induced by Fluoride Ion Attack at Silicon .............................................................................. 15 VI. B -elimination Mechanism of the Reaction of (ClCH2)2SiMe2 with KF to form Me2SiF2 and Ethylene ........................ .'........................ 16 VII. illustration of the Proton NMR Splitting of Cyclopropanes........... 25 VIII. Plot of log ([Me2SiF2lx I [Me2SiF2]H) vs. Hammett ax Constants for Solution Phase Reactions (3.2.3a->e) of ArSiMe2CH2Cl with KF ....................................................................... :.................... 39 IX. Plot oflog (I.2/r1 ) Products vs. Hammett ax Constants for Solution Phase Reactions (3.2.3a->e) of ArSiMe2CH2CI with KF.. 40 X. Plot of log ([X-Ph-Me] I [p-F-Ph-Me]) vs. ap Constants for Solution Phase Reactions of ArSiMe2CH2CI with KF. 41 XI. Detailed Mechanism of Primary and Secondary Product Formation for Solution Phase Reactions (3.2.3a->e) of ArSiMe2CH2CI with KF..................................................................................................... 43 XII. Plot oflog Kx I KH vs. Hammett ax Constants for the Relative Rates of Starting Material Consumption Study (3.3)................................ 53 XIII. Reaction Profiles from the Rate of ArCH3 Appearance Study (3.4).. 60

PAGE 13

xiii XIV. Plot of log Kx I KH (calculated at three hours reaction time) vs. Hatmriett crx Constants for the Relative Rates of ArCH3 Appearance Study (3.4) ....................................................................................... 62 XV. Two Step Radical Mechanism for the Pyrolytic Rearrangement of PhSiMe2CH2Cl to ClSiMe2CH2Ph and ClSiPhMeEt................. 82 XVI. Possible Carbene Elimination Mechanism: of ArSiC12Me formation from the Sealed Tube of Argon Flow Pyrolysis of 83 XVIT. Test for a Carbene Elimination Mechanism .......................... 84 :xvrn. Unimolecular Mechanism for the Gas Phase Rearrangement of HMe2SiCH2Cl to Me3SiCl.............................................. 85

PAGE 14

INTRODUCITON In the field of organosilicon chemistry, the determination of a reaction mechanism is a complex task because of silicon's ability to exist as a positive "siliconium ion (A)l;" in neutral, tetracoordinate species (B); as well as in singly and doubly charged penta and hexacoordinate complex anions (C and D). R-4 R-4 R3 R R2 R R2 "si+ "si/ -4' I_ -1, l2/ Rs R / 'R R / 'R /Si-R5 Si R1 I R/I'R 1 2 1 2 R3 1 6 R3 A B c Q In order to arrive at a valid mechanism, one must either perform a very large number of experiments, or more preferably, study the recent literature concerning the reaction of interest and develop a logical series of experiments which examine the reaction under varying conditions. Doing so may elucidate different mechanisms effecting different or identical reactions on the same molecule. This was the case as seen by our research team working under Dr. Robert Damrauer in studying the base-induced solution phase and gas phase rearrangements RR'R"SiCH2Cl + B ----> BSiR'R"CH2R + BSiRR"CH2R' + BSiRR'SiCH2R" + Cl(eq.l) RR'R"SiCH2Cl-=!J....___,> ClSiR'R"CH2R + ClSiRR"CH2R' + ClSiRR'CH2R" (eq.2)

PAGE 15

2 of chloromethyl substituted silanes (eq. 1 and 2). We found that these compounds are unique in that alterations of reaction conditions result in a shift of reactivity between silicon and the chloromethyl carbon. The solution phase rearrangements (eq. 1) induced by base or anionic attack at silicon have been shown by us2 and others3 to occur only in aprotic media aided by phase transfer catalysts. The use of protic media was found4 to shift the reactive site to the chloromethyl carbon and produce only the SN l/SN2 nucleophilic displacement product RR'R"SiCH2B + CI-. This change of reactivity may or may not result in alterations of the product formed (e.g., back in 1965, R.W.Bott, C. Eabom and B.M. Rushton found that Me3SiCH2Cl reacts with A1Cl3a Lewis acid-to give C1SiMe2Ets, the same product which would now be recognized to form from various pyrolytic reactions performed on the same compound). scope of this thesis covers the mechanistic details of rearrangements (to the chloromethyl substituted carbon) of various R groups attached to a chloromethyl substituted silane. These rearrangements were brought about in two ways: (1) by solution phase reaction with fluoride anion, and (2) by thermal or pyrolytic methods.

PAGE 16

CHAPTER I GROUNDWORKFORREARRANGEMENTSTUDffiS 1.1 Gas Phase Acidities of Hydrocarbons 1.1.1 Introduction and Review The work performed by C.H. DePuy, V .M. Bierbaum and R. Damrauer6 on the gas phase reaction of substituted trimethylsilanes ((CH3)3SiR) with hydroxide ion provided data which supported an anionic cleavage mechanism forming the products listed in equations 3 and 4. (CH3)3SiR + OH---> (CH3)3SiO+ R-H (eq. 3) (CH3)3SiR + OH> (CH3)2RSiO+ CH4 (eq. 4) The mechanism was found to involve the formation of a pentacoordinate hydroxy-silane anion, (CH3)3SiOH .. R-, which transfers one of the three methyl groups* or the R group as a carbamon to the hydroxyl proton, forming Cf4 or R'H in addition to the siloxide anions which were observed and quantified. *The true R/CH3 migratory ratios were statistically corrected to be three times the observed siloxide ratios. (= 3 [(CH3)3SiO-] I [(CH3)2SiRO-]) These R/CH3 migration ratios were used in this preliminary study and in our subsequent studies to probe the mechanistic details of carbanionic rearrangements of chloromethylsilanes. Four R groups (phenyl (Ph), H, vinyl (H2C=CH) and methyl (Me)) with

PAGE 17

4 known Miacid values for the gas phase reaction (eq. 5), were used to generate a linear free RH(g) ---.> R(g)+ H(g)+ (eq. 5) energy relationship between the log of the siloxide ratio (eq. 3 I eq. 4) measured (theoretically equivalent, after statistical correction*, to the log of RIMe migration) and This line was used to predict approximate Miacid values for five other RH hydrocarbons (R =ethyl (Et), isopropyl (i-Pr), t-butyl (t-Bu), cyclopropyl (cy-Pr) and methoxymethyl (CH30CH2)). These data are represented in Table I and Figure I. Electron affmities (see eq. 6) ofR-were calculated from Miacid and Mibond-dissoc. R(g) + e------> R-(g) (eq. 6) EA = MI for process The correlation coefficient of the entire data set for log [siloxide(3)/siloxide(4)] vs. Mlacid was seen to be -0.998 with a slope of -0.0828. Table I Relationship of Miacid and Electron Affinity of R to the Relative Migratory Ailities of Alkyl, Alkenyl and Aryl Groups in the Gas Phase Reaction of Me3SiR with OHR [(CH3)3SiO-]* [(CH3)2RSiO-] Miacid(RH) kcal/mole EA(R) kcal/mole Et i-Pr Me t-Bu cyPr CH30CH2CH2CH H 0.44 421 -9 0.61 419 -11 1.0 416.6 1.8 1.7 414 -7 2.4 412 8 6.0 407 0 7.0 406 17 20 400.4 17.4 weakest I I I I I I Ph 24 400.7 25 strongest To investigate further electronic effects, the same log-ratio set was plotted

PAGE 18

5 against Taft a"' and a1 constants.2 Slopes (correlation coefficients) obtained respectively were 1.55 (0.876) and 3.21 (0.601). The correlation of log [siloxide(3)/siloxide(4)] with .6.Hacid clearly indicated the importance of acid strength in determining the relative cleavage aptitude of R groups The fair correlation with a"' suggested that the amount of negative charge localization on R, due to electronic dipole effects, is moderate. Inductive effects, as shown by the rough correlation oflog [siloxide(3)/siloxide(4)] with Hammett derived a1 constants, seemed to outweigh polar effects (slope of 3.21 vs. 1.55) but this was only due to the restricted range of a1 constants (0.1064) compared with that of a"' constants (0.6019). Nevertheless, the positive slope did indicate that electron withdrawing groups lead to more ready cleavage. r---1 r---1 I I 0 0 -(f) Cl) a: ...., N -...., ..., :I: :a: v v -L-.J L-.J 40 20 10 4.0 2.0 1. 0 0.4 0.2 390 400 410 .6H:cid (kcallmole) Figure I error bers Jlh :1: 0.5 X :1:0.0 :1: 2.0 Me :1: 0.8 420 Plot of log ([Me3SiQ-]/[Me2RSiQ-]) (statistically corrected) vs. Miacid (of RH)

PAGE 19

6 1.2 Pentacoordinate Silicon Anions Four years prior to the work summarized above, the same research team aided by five others7, was able to isolate gas phase pentacoordinate silicon anions by the reactions of tetra-methyl, methoxy, vinyl and trimethyl-R-substituted silanes with fluoride ion in a flowing afterglow apparatus (eq. 7). (eq. 7) The purpose of this study was to effect the generation of stable gas phase carbanions. This was possible only when R orR' was aryl, benzyl, alkenyl or alkynyl. The reaction which transpired in these cases was the decomposition of the pentacoordinate anion (eq. 8) into a free gas phase carbanio"n (R-orR'-) and a fluorosilane (R3SiF or RzR'SiF). R R........_ I_ Si-R' R /I R (g) + Mass spectral verification of gas phase pentacoordinate silicon anions led our research team to speculate as to whether or not they were implicated a8 intermediates in the solution phase reactions of chloromethylsilanes with fluoride and alkoxide anions. This question was affirmatively through the work of S.E. DanaheyS with the reactions of various organofluorosilanes with stoichiometric amounts of potassium fluoride and 18-Crown-6 (phase transfer catalyst) in aromatic solvent (eq. 9). The typical product isolated from these reactions was a crytalline 'caged' potassium difluorosiliconate.

PAGE 20

7 F (0 01 R ,I_ Kr 0 K+ 0 RR'R"SiF + Si-R1 (eq.9) toluene 0_) F/ I 18-Crown-6 '--./ R" Eight of these compounds were characterized by multinuclear cryogenic FI' NMR (lH, 19p and 29Si) as to their geometries and tendencies to undergo psuedorotational processes similar to pentavalent phosphorous.

PAGE 21

CHAPTER IT THE RELATIVE MIGRATORY ABll...ITIES OF ALKYL, VINYL AND ARYL GROUPS IN SOLUTION PHASE FLUORIDE INDUCED REARRANGEMENTS OF CHLOROMETHYLSILANES 2.1 Introduction To find out if the migration of weakly acidic R groups to a chloromethyl carbon in the solution phase reactions of chloromethylsilanes with bases (see eq. 2) takes place through a pentacoordinate silicon anion similar to those discussed in the gas phase groundwork studies, we chose to study the reactions of dimethyl-R-substituted chloromethylsilanes (Me2RSiCH2Cl) with fluoride (eq. 10) so that the ratios of various R to methyl migrations could be obtained and correlated with gas phase (Miaci()) and solution phase ( cr* and cr1 ) parameters. ICI tol1.1ene 18-C:rov.n-6 Rmigration MeRSiF( Et) (eq.lO) Me migration Fluoride anion, a very hard base, was chosen because the strength of the Si-F bond (154 kcal/mole)9 in the products enables their easy identification by GC/MS with electron impact ionization. Phase transfer catalysts such as 18-crown-6 (I), silacrown ethers (ll) or quaternary ammonium salts (lll) were found necessary to release fluoride anion in aprotic solvents, the medium which gave the cleanest reaction mixture.

PAGE 22

9 r-\. (0 (0 (n-Bu)4N+/F0 0 0 0 0_) 0_) + '--/ '--/ ( octyl ) 3 HMe/F I II III The source of the fluoride ion used in the optimized experiments with 18-Crown-6 was anhydrous potassium fluoride. 2.2 Sl.Th1MARY OF OPTIMIZA TIQN EXPERIMENTS Reactions of RSiCCH3l2CH2C1 in the condensed phase with fluoride anion/18-crown-6 system in aromatic solvents Introduction: The optimization of the reaction conditions (temperature, fluoride source, phase transfer catalyst, solvent and apparatus) was conducted for R = Me (chloromethyltrimethylsilane) by S.E. DanaheylO. This material has only one possible migratory product, dimethylethylfluorosilane, resulting from the reaction in equation 11. (eq. 11) An interesting alteration in the relative rate (and yield) of product formation was noted for R = Me by S.E. Danaheyll and with R = Et as the alkali fluoride source changed from potassium to cesium. In both cases the cesium reactions caused a 250 fold increase in the rate of starting material consumption and a 50 % increase in the total product yield. These reactions were conducted in 20 ml serum cap vials at 60C

PAGE 23

1 0 and 74C in 0.5 ml toluene and o-xylenerespectively. The stirring and temperature control of these and further reactions was controlled by a thermostated/magnetically stirred oil bath manufactured by Sybron, Inc. The usual amount of reactants used in these and further reactions of quantitative interest in chapter 2 was: RSi(CH3)2CH2Cl = 5 mmol; KF (oven dried, Aldrich reagent grade) 0.55 g (10 mmol); 18-Crown-6 (Aldrich)= 0.13 g (0.5 mmol) in 0.5 to 2.0 m1 of aromatic solvent, the choice of which depended on the volatilities of starting materials and products. The molar ratio of reactants was held constant throughout the study. Additional Notes: Cesium fluoride reactions were found to proceed almost to completion in the absence of phase transfer catalyst and solvent12, but GC/MS analyses of reaction mixtures were found to contain vast amounts of siloxanes [(RMe2Si)20]*, some of which obscured the migration products of interest *These are due to the hygroscopic nature of even oven dried cesium fluoride. The release of its moisture during reactions of Si-F bonds brings about the formation of Si-0-Si bonds. Its use apart from an additional large scale preparation (7X the usual amounts of reactants) of the n-butyl products (2.5.3) was discontinued. Sodium fluoride was found to give no reaction with Me3SiCH2CP3, whereas, potassium iodide gave the substitution product Me3SiCH2Il4. Additional fluoride/phase transfer systems such as (n-Bu)4N+/F-/THF and (octyl)3NMe+ta+ KF/toluene were tried but gave poor yields of migration products and complex reaction mixtures. IS Pyridine and the non-aromatic solvents dimethylsulfoxide (DMSO), acetonitrile (MeCN) and dimethylformamide (DMF) were also tried, but also resulted

PAGE 24

1 1 in complex product mixtures16. The chloromethyldimethyl-R silanes studied in section 2.6 were ethyl, methyl, isopropyl, cyclopropyl, n-butyl, vinyl, phenyl and silacyclopentano. The source of the R = methyl, vinyl and phenyl compounds was Petrarch Systems, Inc.; 1-(chloromethyl)-1-methyl-silacyclopentane was prepared by R. Damrauer via a di-Grignard reaction: 'si::J ( Cl Syntheses of the ethyl, n-butyl, isopropyl and cyclopropyl are contained herein.

PAGE 25

1 2 2.3 Results and Discussion It here again must be noted that the true (statistically corrected) R/CH3 migration ratio for the reaction of one mole of chloromethyl dimethyl-R-substituted silane is two times the observed product ratio on the GC-FID. The results contained in Table IT were optimized for each reaction before the time (ORT) when secondary product formation was first noted and the peak heights of the primary products were maximized. Table II Relationship of the Relative Migratory Abilities of Alkyl, Alkenyl and Aryl Groups to Taft cr* and 01 Constants for the Solution Phase Reactions of RSiMe2CH2Cl with KF R ORT(hrs) RICH3 fl.* qi-Et 264 1.5 -0.1019 -0.05518 Me 336 1 416.617 0.0019 -0.04618 Cypent. 9.0 00 i-Pr 168 0.67 41917 -0.1919 -0.06418 cyPr 264 0.57 41217 0.1118 n-Bu 168 0.87 42019 0.1319 -0.06018 Vy 48 20 40617 0.5620 0.0519 Ph 1.5 16 39917 0.6019 0.1019 The correlations of log(R/Me) migration with MI0 acid cr* and cr1 constants are given below in Figures II to IV. (It might be mentioned here that the .1H0 acid values contained in reference 6 included t-butyl instead of n-butyl. The synthesis of chloromethyldimethyl(t-butyl)silane was attempted from t-butyl lithium and chloromethyldimethylchlorosilane in pentane, but failed. We were fortunate to find the value listed in reference 19 for n-butane). The cr* constants utilized depict the relative abilities of R group anions to

PAGE 26

1. 5 1.0 c: c: 0 0 --G G 0.5 -"i E a: Q) ...... I: 0 -0.5 400 Vy 405 410 = -0. 0655 carr. = 0. 822 415 cy-Pr .6.H:cid R-H (kcallmole) Figure II 13 i-Pr Plot oflog ([Me3SiO-] I [Me2RSiO-]) vs. Miacid (ofR-H) for Solution Phase Reactions (2.6.1-2.6.7) ofRSiMe2CH2Cl with KF ........ ........ c: c 0 0 G G --E E a: Q) ...... I: --0 1. 5 = 1. 82 co rr. = 0 9 1 5 1.0 0.5 0.6 0.7 Cy-Pr cr* Figure III Plot of log ([R migration] I [Me migration]) vs. Taft cr* for Solution Phase Reactions

PAGE 27

1 4 localize a negative charge via dipole field effects (field effects); a negative a* indicates a highly localized electron pair on a non-polar substituent, and vice versa for c c 0 0 .... Q;l Q;l L. L. 0\ 0\ --E E a: -..:L 0\ 0 Et \-0. 04 i-Pr n-Bu 1.5 Vy 1. 0 = 9. 06 corr. = 0.952 -0.02 0. 02 0. 04 0. 06 0. 08 0. 10 -0.5 Figure IV Plot oflog ([R migration] I [Me migration]) vs. Hammett a1 Constants for Solution Phase Reactions (2.6.1-2.6.7) of RSiMe2CH2Cl with KF a positive a*. The ability of an R group to stabilize a negative charge through inductive processes are represented by the a1 constants; a negative a1 indicates electron flow away from the R group (relative anionic instability) and vice versa for a positive a1 The slopes (correlation coefficients) listed in this thesis for loglO(R/Me) migration vs. the a* and AH0 acid were correctly reported, unlike the 1.81 (0.943) and -0.0674 (0.904) reported in a paper submitted to JACS in 198421, This was a result of an erroneous report of the ratio of cyclopropyVmethyl migration as 7.0 instead of 0.57. The error was reported in a subsequent paper submitted to Organometallics in

PAGE 28

15 198522, but as 0.66. The difference (1.82 vs 9.06) in the slopes of the er*, err correlations is only due to the range of err constants (0.1064) compared with that of er* constants (0.619). Such a comparison serves no purpose because er* constants are methyl based whereas err constants are proton based. The small slope obtained with the .Ml0 acid correlation in both the solution and gas phase experiments is only indicative of the 5.5 % variance between the .6.H0 acid values of the R groups studied. The 0.82-0.95 correlations of all three parameters with the RIMe migration ratio combined with the similar results from the gas phase study mentioned in section 1.1 led us to postulate the mechanistic model below: Rl _____. e 0 8 F -SiCH Cl 1\ t2 Me Me ----------FigureV MeEtSiFR Simple Pentacoordinate Mechanism for Solution Phase Rearrangements of RSiMe2CH2Cl Induced by Fluoride Ion Attack at Silicon This involves the formation of a pentacoordinate anionic intermediate where Me and R migrate competitively to displace chloride ion. The behavior of this solution phase intermediate is analogous to that of real gas phase fluorosilicate anions (section 1.2) in that aryl and vinyl groups are more likely to separate as freer carbanions than are alkyl groups, thus effecting a.higher RIMe migration ratio. The driving force for pentacoordination is the thermodynamic stability of the Si-F bond (20-40 kcal/mol)23, and the migratory efficiency is controlled largely by anion stability. Pentacoordination at silicon is easily accomplished through bonding with fluorine. The thermodynamic

PAGE 29

1 6 stability of the Si-F bond apparently lowers the energy of the unoccupied 3d orbitals sufficiently to cause formation of a resultant trigonal bipyramidal dsp3 hybrid orbital. This is not suprising in view of the fact that the hexacoordinate SiF62anion has been known for years, uses d2sp3 hybrid orbitals and is very stable. The GC-MS identification of Me2Si(CH2Rh in the R =vinyl and phenyl cases and Me2SiF2 in the phenyl case was the basis for the further mechanistic studies of substituted arylchloromethyldimethyl silanes in chapter 3 of this thesis. A mechanism for the H2C=CH2 and Me2SiF2 forming reaction (2.5.10) of bis(chloromethyl)dimethylsilane and KF is postulated below: (\. 1 Me2Sif 2CH2CH2Cl ll elimination Figure VI B-elimination Mechanism of the Reaction of (ClCH2)2SiMe2 with KF to form Me2SiF2 This involves a total of 2 moles of fluoride per mole of starting material, each reacting to form a different pentacoordinate adduct. The first adduct undergoes a chloromethyl migration with loss of chloride and the second undergoes a B elimination of chloride from the chloromethyl migratory product to form ethylene and dimethyldifluorosilane.

PAGE 30

1 7 2.4 Alkoxide Rearrangements of Chloromethylsilanes An interesting parallel to the fluoride reactions contained herein is found in the work of R.L. Kreeger, P.R. Menard, E.A. Sons and H. Shechter with methoxide anion.24 Part of their study involved the reaction of methoxide anion with phenylchloromethyldimethylsilane in either methanol (eq. 12) or dioxane (eq.13), with and without the addition of 18-Crown-6. Analogous phenyl (a) and methyl (b) migration products were formed in somewhat similar (at least in magnitude) ratios when NaOMe/dioxane was used at 30C and 60C. The use of methanol (a polar protic solvent) with sodium, potassium and cesium methoxide resulted in products formed via: chloride displacement by Meo-(c) (in decreasing quantities respectively); chloromethyl displacement by Meo-attack at Si (d) and Meo-attack at Si, displacing benzyl anion from (a) -found in the form of toluene-(e) (both in increasing quantities respectively). No migratory products found in the MeOH system whereas no chloride displacement was found in the aprotic dioxane system. PhSiMe2CH2Cl + MeO-/dioxane -> PhCH2SiOMeMe2 (a)+ EtSiOMePhMe (b) + small quantities of (d) and (e) (eq. 12) PhSiMe2CH2Cl + MeOJMeOH -> PhSiMe2CH20Me (c)+ PhSiMe20Me (d) + CH2C1 (obsetved as CH3Cl) + Me2Si(OMe)2 (e) + cH2Ph (obsetved as MePh) (eq. 13) The formation of Me2Si(OMe)2 and toluene is analogous to that of Me2SiF2 and the substituted toluenes contained in the next chapter. The effect of the crown

PAGE 31

18 ether used in both solvents was to increase and decrease the yields of Me2Si(OMe)2 and (in the dioxane system) respectively. The yield of Me2Si(OMe)2 with respect to time (in the methanol system) was found to increase drastically when CsOMe was used This is entirely analogous to our fmdings with CsF reactions and suggests a higher extent of ionic dissociation with cesium salts in bothprotic and aprotic media. That Me2Si(OMe)2 was found in the methanol system, is suggestive of the instability of benzylic silanes in protic ionic media and perhaps that the pentacoordinate intermediate [PhSiOMeMe2CH2Cl]is rather short lived in such media. This is not unlikely in view of the large proportion of the chloride displacement product (c) formed in the methanol system; hydrogen bonding decreases the "hardness" of the methoxide base. Although no RSiMe2F or RSiMe2CH2F analogues were detected by S.E. Danahey16 in polar solvents, it is clear that H -> F bonding from hygroscopic CsF or polar solvents creates messier reaction mixtures by providing alternate reactions (esp. siloxane formation from trace moisture).

PAGE 32

1 9 2. 5. Synthesis of Starting Materials 2.5.1 Preparation of CH3CH2SiCCH3}2Q!2,Q I. C2HsBr + Mg0 > C2H5MgBr Into a dry, 3 necked 500 m1 flask equipped with an addition funnel, a helium supplied reflux condensor and magnetic stirrer, was placed magnesium turnings (3.06 g, 0.1259 mol, Fischer) and one crystal of iodine. 100 ml anhydrous ether was added, and the system was purged with helium. Into the addition funnel was placed a solution of ethyl bromide (15.6 ml, 0.21mol, Fischer) in 30 ml ether. Excess ethylbromide was used to compensate for its volatility. The halide solution was added slowly at a 0.5 drop/sec rate. Grignard formation began in fifteen minutes and was complete after fifty minutes. The mixture was refluxed gently for an additional hour. II. C2H5MgBr + ClSi(CH3)2CH2Cl -> CH3CH2Si(CH3)2CH2Cl + MgBrCl Using an identical set-up, 50 m1 of Grignard supernatant was syringe transferred into a helium purged (septum sealed) addition funnel. The 500 ml flask contained a solution of ClSi(CH3)2CH2Cl (13.6 g, 0.095 mol, Petrarch) in 40 ml ether. The Grignard was added dropwise over a fifteen minute period, during which time a gentle reflux of ether and MgBrCl precipitation was observed. 70 extra ml ether was added and the mixture was refluxed under gentle heat overnight. The unreacted ClSi(CH3)2CH2Cl and products were extracted and hydrolyzed with 200 ml saturated, aqueous Nl4Cl solution. The salts dissolved as the double ammonium salt (NI4Cl.MgBrCl) in the aqueous layer. Unreacted chlorosilane starting material hydrolyzed to the siloxane((Si(CH3)2CH2Cl)20) and dissolved in the ether layer along with the main product. The flask contents were transferred into a separatory funnel; the lower, aqueous layer was discarded and the

PAGE 33

20 upper ether layer was dried over anhydrous Na2S04. The drying agent was filtered through a Buchner funnel and most of the ether was removed under gentle vacuum through a rotary evaporator. Nine grams was collected (15 ml) of a light yellow fluid, shown by GC-FID to be a 50:50 mix of product and siloxane . Distillation of this fluid at atmospheric pressure through a 20 em Vigreux column afforded 2.83 g, (21.8 % yield) of 99.5 % pure CH3CH2Si(CH3)2CH2Cl: MW: 136.7; bp: 90C/630 mmHg; 1H NMR (CDCl3/CHCl3): B = 0.12 (s, 6H, Si(CH3)2), 0.6-0.76 (t, 3H, CH3CH2Si ), 0.9-1.0 (q, 2H, CH3CH2Si ), 2.8 (s, 2H, SiCH2Cl)*. 2.5.2 Preparation of n-C4H9Si!CH3l2CH2Cl I. In a dry 500 ml 3-necked flask equipped with an addition funnel, a nitrogen supplied reflux condensor and magnetic stirrer, was placed magnesium turnings (10 g, 0.41 mol, Fischer), a couple of iodine crystals and 50 m1 anhydrous diethyl ether. The system was purged with nitrogen. Grignard formation was initiated by adding 5 ml pure n-butyl chloride through the funnel. This was followed by addition of a solution of 40 m1 n-butyl chloride in 60 ml diethyl ether over a 90 minute period. The total amount of n-butyl chloride used was 45 ml (39.88 g, 0.43 mol, Fischer). The reaction was complete within two hours of initial Grignard formation. II. n-C4H9MgCl + ClSi(CH3)2CH2Cl-> n-C4H9Si(CH3)2CH2Cl + MgC12 Using an almost identical set-up, 55 m1 (0.15 mol; 40 % excess) of n-butylmagnesium chloride solution was syringe transferred to a septum sealed addition funnel atop a nitrogen purged 3 necked flask which contained ClSi(CH3)2CH2Cl (15 g, 0.105 mol, Petrarch) dissolved in 50 m1 diethyl ether. The

PAGE 34

21 Grignard was added over a period of one hour. No immediate refluxing without external heat was observed, so gentle heating was used to reflux this reaction over a 48 hr period, during which time most of the ether evaporated. At this time a resultant huge cake of MgC12 was broken up with 800 ml ether (in small portions at a time). The ether/MgCl2 sludge was removed and hydrolyzed with 200 ml saturated aqueo11s Nf4Cl solution to decompose unreacted Grignard and dissolve magnesium salts. The aqueous layer was separated from the organic layer with a separatory funnel. The organic layer was dried over anhydrous sodium sulfate. Mter drying, the sulfate was filtered, and ether evaporated under gentle vacuum with a rotary evaporator. Crude n-C4H9Si(CH3)2CH2Cl (21 ml) was obtained. The crude product was distilled under vacuum through a 20 em Vigreux column. n-C4H9Si(CH3)2CH2Cl: 11.76 g (68.1 %yield; 97% pure by GC-FID); MW: 164.75; bp: 67C/30 nun Hg; 1H NMR (CDCl3/CHCl3): a = 0.11 (s, 6H, Si(CH3)2), 0.65 (t, 3H, CH3CH2CH2CH2Si ), 0.90 (m, 4H, CH3CH2CH2CH2Si ), 1.32 (t, 2H, CH3CH2CH2CH2Si), 2.8 (s, 2H, SiCH2Cl). 2.5.3 Preparation of (CH3)2CHLi + LiCl Lithium dispersion (19.85 g, 30 wt% in mineral oil, 0.858 mol, Aldrich) was washed with 800 ml dry pentane in a three neck, one liter flask equipped with an addition funnel, overhead stirrer, Friedrichs condensor with argon gas inlet and a filter/drain system designed by myself. This consisted of a simple teflon stopcock drain fused to the bottom center of the flask via 1/4 inch glass tubing. Directly above the stopcock was placed 2 inches of fme glass wool filter. Above this was placed a layer of sand on top of which was a layer of coarse, crushed borosilicate glass. This, in conjunction with the overhead stirrer, was used to scrape the surface of lithium

PAGE 35

22 particles to increase the rate of formation and yield of active lithium reagent. Gentle argon pressure was used for draining the mineral oil, pentane wash. Through the funnel, 200 ml pentane was added to the clean lithium along with 10 ml pure (CH3)2CHCI to start reaction. The balance of the chloride (30 ml), in 200 ml pentane under argon, was added over a period of two hours with overhead stirring. Total chloride used was 40 ml (33. 7 g, 0.429 mol, Kodak). Heat from reaction refluxed the pentane for two hours after addition. All but 4 ml of the Li reagent solution (300 ml) was drained under gentle argon pressure through a septum atop into a 500 ml addition funnel attached to an argon purged one liter, three necked flask equipped with a Friedrichs condensor and magnetic stirrer. The remaining 4 ml was titrated under argon with 1.35 ml sec-butyl chloride in a 100 ml three necked, flask containing the lithium reagent, 20 ml toluene, 0.5 ml of a solution of 100 mg 2,2-biquinoline indicator in 50 ml toluene.This indicator changes the solution color to yellow when the lithium reagent is completely reacted. The concentration of the lithium reagent was found to be 0.34 M. ll. ClSi(CH3}zCH2CI + 0.33 SbF3 -> FSi(CH3}zCH 2 Cl + 0.33 SbCl3 A 100 rill single neck flask with a magnetic stirring bar was charged with SbF3 (11.45 g, 0.0641 mol, Aldrich). This was cooled to -78C with a dry ice/acetone bath. A reflux condensor with nitrogen inlet was attached and the flask was purged. Under slow agitation of the fluoride; ClSi(CH3}zCH2Cl (25.0 g, 0.175 mol, Petrarch), via syringe, was added. The solution fumed at frrst, and then solidified within five minutes. Mter removal of the cooling bath, the reaction mixture was then allowed to warm up to ambient temperature over a period of 4 hours. During this period, a gentle reaction occurred with a color change of light brown to

PAGE 36

23 purple and the precipitation of antimony trichloride. A GC of the supernatant fluid showed an almost quantitative conversion to FSi(CH3)2CH2Cl. The crude product was distilled through a 20 em Vigreux column to yield FSi(CH3)2CH2Cl: 19.11 g (19 ml, 86.4 % yield, 98.2 % pure); MW: 126.6; bp: 76C/630 mmHg. +LiF To the purged flask in I., containing FSi(CH3)2CH2Cl (12.7 g, 0.10 mol) in 20 ml pentane, was added the lithium reagent solution (0.34 M, 300 ml, 0.102 mol Li) with magnetic stirring. Gentle reflux began and white clouds of LiF were observed indicating reaction. The reaction mix was stirred for 96 hours, at which time there was no further increase in product or decrease in starting material as observed by GC-FID. The reaction mixture was hydrolyzed with 40 ml cone. HCl in 200 cc crushed ice. When all clouds of LiF had dissolved, the hydrolysis mixture was transferred to a one liter separatory funnel. The layers were separated, and the aqueous layer was then washed with two 100 m1 portions of pentane. The organic phase was dried over anhydrous sodium sulfate. After flltration of the drying agent, pentane was removed from the crude product via distillation through a 40 em Vigreux column. 20 ml of a fluid were obtained. Two major products (close in volatility) in addition to 24.6% of the starting material were obtained. Distillation of this fluid at atmospheric pressure through a 45 em vacuum jacketed teflon annular spinning band column (mfd by Nester/Faust Corporation) afforded fair purification of (CH3)2CHSi(CH3)2CH2Cl; the best fraction collected was shown to be 81.2 % pure by GC. (CH3)2CHSi(CH3)2CH2Cl: 1.28 g (6.49 % yield); bp: 129C/630 mmHg; 1 H NMR (CDCl3/CHCl3): 8 = 0.10 (s, 6H, Si(CH3)2), 1.0 (s, 7H, (CH3)2CHSi ),

PAGE 37

24 2.80 2.5 .4 Pre.paration of Dimethylchloromethylcyclopro.pylsilane I. H2C=CHSi(Me)2CH2Cl + CH2I2 -> (CH2)2CHSi(Me)2CH2CI + Zni2 + Cui Charged into a dry 500 ml three necked flask were zinc dust (18.8 g, 0.288 mol, Mallinkrodt), copper (I) chloride (2.85 g, 0.029, Mallinkrodt) and 50 ml anhydrous diethyl ether. An overhead stirrer, Friedrichs condenser with nitrogen inlet and an addition funnel were then attached to the three necked flask. The system was purged with nitrogen, and H2C=CHSi(CH3)2CH2Cl (15.0 g, 0.111 mol, Petrarch) was added through the funnel, with stirring. This addition was exothermic. Over a fifteen niinute period, diiodomethane (38.8 g, 0.145 mol, Baker) was added to the reaction mixture. Ether reflux, without external heating, was then observed for 2 ... hours . More ether (30 ml) was added, and the solution was maintained at reflux for 24 hours. The reaction mixture was then vacuum filtered through a Celite bedded Buchner funnel into a 500 ml flask. The precipitate was washed with 200 ml ether. This ether mixture was then hydrolyzed with 200 ml saturated, aqueous NH4Cl solution. The mixture was transferred to a separatory funnel, and the aqueous layer discarded. The organic layer was dried over sodium sulfate. After filtration from the drying agent, the ether was removed by rotary evaporation. Fifteen ml of a dark fluid containing four major products was obtained. Vacuum distillation of this fluid through a 20 em Vigreux column afforded (CH2)2CHSi(CH3)2CH2Cl: 2.73 g (16.6 % yield; 96.3 %pure); MW: 148.5; bp: 40C/30 mmHg; 1 H NMR (CDCl3fCHCl3): B = -0.35 (s, 1H, (CH2)2CHSi ), 0.07 (s, 6H, Si(CH3)2), 0.25 (sxt, 2H, (CH2)2CHSi ), 0.60 (m, 2H, (CH2)2CHSi),

PAGE 38

25 2.80 (s, 2H, SiCH2Cl). Note: Cyclopropanes give complex spectra since Ha and Hb are not equivalent and not only have slightly different chemical shifts, but also split each other (fig. 7): Figure7 illustration of the Proton NMR Splitting of Cyclopropanes

PAGE 39

26 2.6 Optimized RSiCCH3)2CH2CIIKF Reactions -Detailed Experimental General Experimental, see section 2.2. 2.6.1 EtSi(CH3)2CH2Cl + KF -> PrSi(CH3)2F + Et(Et)Si(CH3)F + KCl Time of first appearance of migratory products: 48 hrs. Initial Ratio Et/Me migratory products: 0.75 at 168 hrs. Subsequent GC-FID analysis (additive peak height based) performed on a Perkin Elmer 3020B with a 6', 1/8 inch OV-1/chromosorb column. Table ill GC-FID analysis Reaction Time Chrsl % completion wrt S.M. Et/Me product ratios 168 264 456 600 768 <5 7.4 10.0 10.0 48* appearance of secondary products 0.75 0.75 0.60 0.60 0.60 GC/MS analysis: The ethyl migration product was found to elute first, from an identical column in a Finnegan 3200 with a Teknivent data system. This was followed by the adjacent methyl migration product, starting material, solvent, three side products and 18-Crown-6. Relevant ions and (% abundances) for the main products: I. PrSi(CH3)2F M+ 120 (5); M+-Me 105 (12.5); M+-Pr 77 (100) IT. Et(Et)Si(CH3)F M+ 120 (8); M+-Me 105 (4); M+-Et 91 (100) 2.6.2 n-BuSi(Me)2CH2Cl + KF -> PeSi(Me)2F + Et(n-Bu)Si(Me)F + KCl n-Bu migration Me migration Time of first appearance of migratory products: 72 hrs.

PAGE 40

27 Initial Ratio n-Bu/Me migratory products: 0.541 at 72 hrs. Table IV GC-FID analysis Reaction Time SiF from n-BuSi
PAGE 41

Observations: Time of first appearance of migratory products: 3.5 hrs. Initial Ratio n-Bu/Me migratory products: 0.463 at 3.5 hrs. Table Y GC-FID analysis Reaction Time Chrs) % completion wrt S.M. 24 42* 48 65.4** 72 90.6*** n-Bu/Me product ratios 0.463 0.453 0.468 initial production of three main low volatile side products ** %of total products being transmuted to side products= 39.5 *** stopped reaction by freezing to -20C at this point 28 The cold reaction mixture was filtered under gentle vacuum with a few extra ml of mesitylene. The flltrate (20 ml) was distilled through a 45 em annular teflon spinning band column (vacuum jacketed). This afforded mesitylene free fractions of PentylSi(CH3)2F and Et(n-Bu)Si(CH3)F (1.6 g mixture; 10.8 mmol; 30.7 % yield; bp l13C/630 mm Hg). Structure verifications were performed, again, by GC/MS. 2.6.4 i-PrSi(Me)2CH2Cl + KF > i-BuSi(Me)2F + Et(i-Pr)Si(Me)F + KCl (81.2% purity) i-Pr migration Me migration Time of first appearance of migratory products: 24 hrs. Initial Ratio i-Pr/Me migratory products: 0.35 at 24 hrs. Table VI GC-FID analysis Reaction Time (hrs) %completion wrt S.M. i-Pr/Me product ratios 72 96 168 240 10 10.5 11.1 76* rapid appearance of three major side products in about 5X the quantity of the main products. Reaction mixture was frozen and stopped at this point. 0.288 0.30 0.30 0.36

PAGE 42

29 GC/MS analysis: As expected, the isopropyl orR migration product was found to elute before the larger adjacent methyl migration product. These were followed by starting material ( + impurity), mesitylene solvent, three major side products and 18-Crown-6. Relevant ions and(% abundances) found for the main products were: I. i-BuSi(CH3)2F M+ 134 (2.00); M+-Me 119 (15.0); M+i-Pr 91 (24); M+i-Bu 77 (100) II. Et(i-Pr)Si(CH3)F M+ 134 (5.00); M+-Me 119 (7 .50); M+-Et 105 (100); M+i-Pr 91 (100) 2.6.5 cy-PrSi(CH3)2CH2Cl + KF -> cy-PrCH2Si(CH3)2F + KCl + Et(cy-Pr)Si(CH3)F cy-Pr migration Me migration Time of first appearance of migratory products: 96 hrs. Initial Ratio cy-Pr/Me migratory products: 0.230 at 168 hrs. Table VII GC-FID analysis Reaction Time (hrsl % completion wrt S.M. 192 5 264 17*3 456 40 600 66.8* 768 -1oo** cy-Pr/Me product ratios 0.250 0.286 0.270 0.286 0.270 appearance of two main side products in 2/3 to 3/4 the relative ** quantity of the major products. Reaction mixture was frozen and stopped at this point. GC/MS analysis: The cyclopropyl (R) migration product once again preceded the methyl migration product in elution. These were followed by starting material,

PAGE 43

mesitylene solvent, two major side products and 18-Crown-6. Relevant ions and (% abundances) found for the main products were: I. cy-PrCH2Si(CH3)2F M+ 132 (8.33); M+-Me 117 (6.50); M+ cy-PrCH2 77 (100) 30 ll. Et(cy-Pr)Si(CH3)F M+ 132 (2.60); M+-Me 117 (10.0); M+Et 103 (100); M+cy-Pr 91 (100); M+(cy-Pr +Me; w/proton rearrangement?) 77 (72) 2.6.6 C2H3Si(CH3)2CH2Cl + KF --> C2H3CH2Si(CH3)2F + Et(C2H3)Si(CH3)F + KCI Me migration Reaction was carried out by S.E. Danahey2S on a scale 20X the usual amount of starting materials Time of first appearance of migratory products: 24 hrs. Initial Ratio C2H3fMe migratory products: 1 0.0 at 24 hrs. Table VIII GC-FID analysis Reaction Time Chrs) % completion wrt S.M. C2H3fMe product ratios 24 48 72 66.7 85.7 too* 10.0 10.0 10.0 one major side products reported at this time; Reaction mixture was frozen and stopped at this point GCIMS analysis: The large vinyl (R) migration product again preceded the methyl migration product in elution. These were followed by a small starting material o-xylene solvent, the lone side product (lll.) and 18-Crown-6. Relevant ions and (% abundances) found for the main products were:

PAGE 44

31 I. C2H3CH2Si(CH3)2F M+ 118 (100); M+Me 103(92); M+C2H3 91 (0.5); M+C2H3CH2 77 (100) II. Et(C2H3)Si(CH3)F M+ 118 (48); M+-Me 103 (100); M+C2H3 91 (100); M+-Et 103 (100); ill. (CH3) 3Si(C2H3CH2)2 M+ 140 (2.5); M+-Me 125 (100); M+ C2H3 113 (0.5); M+C2H3CH2 99 (100); Isolation of products: Distillation of the filtered reaction mixture through a Vigreux column gave 0.5 g of a liquid whose GC indicated that it was a mixture of both C2H3CH2Si(CH3)2F and Et(C2H3)Si(Oi3)F in a tatio of 5:1. The peaks were very close in retention time and were not separated. 2.6.7 PhSi(CH3)2CH2Cl + KF-PhCH2Si(CH3)2F + Et(Ph)Si(CH3)F + KCl Ph migration Me migration Time of first appearance of migratory products: 1.5 hrs. Initial Ratio Ph/Me migratory products: 8.0 at 24 hrs. Table IX GC-FID Analysis Reaction Time (hrs) % completion wn S.M. Ph/Me product ratios 1.5 23 48 18.12 41.4* 58.3** rapid appearance of 2-3 side products ** reaction mixture was frozen and stopped at this point. 8.0 6.37 3.70 GC/MS analysis: The phenyl (R) migration product was the fourth substance

PAGE 45

32 to elute, followed by the adjacent methyl migration product. These were preceded by two highly volatile side products found to be (CH3)2SiF2 and (ClCH2)2Si(CH3)2, and toluene solvent in order, respectively. The main, primary (1 ) products were followed by starting material, the secondary (2) products (PhCH2)2Si(CH3)2, Ph(CH3)2SiOSi(PhCH2)(CH3)2, PhCH2(CH3)2SiOSi(Ph)(Et)(CH3) and 18-Crown-6. Relevant ions and(% abundances) found for the main products were: I. (CH3)2SiF2 M+ 96 (10.1); M+Me 81 (100) II. PhCH2Si(CH3)2F M+ 168 (32.5); :M+Me 153(11.5); M+Ph and/or CH2Ph+ 91 (23); M+CH2Ph and/or Ph+ 77 (100) ill. Et(Ph)Si(CH3)F M+ 168 (26.5); M+Me 153 (9.5); M+Et 139 (100); M+Ph 9i (28.5); Ph+ 77 (100) N. (PhCH2)2Si(CH3)2 M+ 240 (7.0); M+-Me 225 (1.5); M+CH2Ph 91 (23) V. Ph(CH3)2SiOSi(PhCH2)(CH3)2 M+ 300 (0.5); M+Me 285 (1.0); M+CH2Ph 209 (100); M+Me2SiPh 165 (7.5); M+ OSi(PhCH2)(CH3)2 135 (20); Ph(CH3)2SiOSi+(CH3)2 193 (37.0); VI. PhCH2(CH3)2SiOSi(Ph)(Et)(CH3) M+ 314 (0.5); M+Me 299 (1.0); M+Et 285 (4.2); M+Ph 237 (trace); M+CH2Ph 223 (100); M+MeEtSiPh 165 (6.0); M+OSi(PhCH2)(CH3)2 149 (2.0); CH2Ph+ 91 (21.5); (CH3)2Si+SiMeEtPh 207 (28.5)

PAGE 46

33 2.6.8 1-chloromethyl-1-methyl-silacyclopentane + KF > 1-fluoro-1-methylsilacyclohexane + 1-fluoro-1-ethylsilacyclopentane + KCl Ring opening/ R migration -Me migration Ring closed product Time of first appearance of migratory products: 9 hrs. Initial Ratio RIMe migration: oo Short path distillation gave an initial fraction rich in R migration product with mesitylene as the only other detectable component (-41 %). This solution was suitable for proton and carbon-13 NMR analyses, which revealed that only 1-fluoro-1-methyl-silacyclohexane was formed. Additional Reactions of Mechanistic Interest: 2.6.9a Cl(CH2)3Si(CH3)3 + KF --,.--.> BuSi(CH3)2F + KCl Rearrangement within 336 hrs (2 weeks) at this temperature; the only reaction with this system was a possible dimerization of the starting material with loss of chloride. 2.6.9b (ClCH2)2Si(CH3)2 + KF --> (CH3)2SiF2 + C2I4 + KCl This reaction, conducted for three days in a septum-sealed serum cap vial resulted in so much internal pressure that the septum-cap assembly was found to have blown off the vial (along with the reactants). It was conducted again at room temperature for four days after which both gas and solution phase samples of the reaction mixture were analyzed by GC-FID and GC/MS.

PAGE 47

34 The solution phase GC-FID revealed two highly volatile products in addition to a very small amount starting material peak indicative of an almost complete reaction. Subsequent GC/MS revealed the least volatile of the first products to be (CH3)2SiF2: ions and(% abundances): M+ 96 (81); M+-Me 81 (100); M+-F 77 (52). Retention time comparative GC-FID analysis of the gas phase revealed the most volatile of the two products to be ethylene.

PAGE 48

CHAPTER ill :MECHANISTICKINETIC STUDIES OF (ARYLCHLORO:METHYL)DIMETHYLSILANES 3.1 Introduction: 35 The anionic migration mechanism postulated in chapter 2 is further probed in this chapter for its validity. Chloromethyldimethyl-R-silanes with readily migrating R groups were chosen for further study because they (esp. aryl) are used for classic probes of electronic effects. Anionic stability was found to be the predominant parameter responsible for the relative migratory abilities of these and of the alkyl anions in the previous chapter. The presence of secondary products of the Me2Si(CH2R)2 type in the R =Phenyl and Vinyl cases and ofMe2SiF2 (eventually found to be present in all of the solution phase fluoride reactions via GC/MS) led us to speculate about what happens mechanistically when a second* mole of fluoride reacts. [*The fluoride/silane stoichiometry in all reactions contained in study 1 and 2 was kept at 2 : 1, but perhaps only< 0.05 % of the total fluoride was freed by the 18-crown-6 at any given time26 for reaction in the aromatic medium.] The series of reactions chosen for chapter 3 utilized five aryl(chloromethyl)dimethyl silanes rather than vinyl because a great number of Hammett cr constants are known and were derived for substitutions on the benzene ring. These were X= p-Me, H, p-F, p-Cl, and m-CF3. All except the phenyl (X=

PAGE 49

36 H)** were synthesized from a substitutedarylmagnesium halide Grignard reaction with chloromethyldimethylchlorosilane. [section 3.5] [**Obtained from Petrarch Systems, Inc.] The first of these reactions (3.2.3) concentrated on the acquisition of the total primary (1 ) and secondary (2) product distributions (eq. 14). Next, a competition for fluoride between the p-Me-Ph-silane and all the others yielded relative ratios of reaction rates based on starting material disappearance. The last of these solution studies concerned itself with the rate of appearance of X-substituted toluenesthe major secondary product observed in the first series and suspected of forming secondary products of the Me2Si(CH2Rh type.

PAGE 50

37 3.2 Acquisition of primazy (1 ) and secondaty (2) products for reactions of the type: ArSiMe2CH2CI + KF > ArCH2SiMe2F + EtSi(Ar)Me2F (1 ) + Me2SiF2 + ArCH3 + Me2Si(CH2AR)2 + EtMeSi Ar(CH2Ar) (2) (eq. 14) 3 .2.1 General Experimental: These reactions were conducted in 20 ml serum cap vials with miniature magnetic stir bars at 55C maintained by a thermostatted oil bath manufactured by Sybron Inc. Toluene (1.5 ml) was the solvent chosen. The scale of the reactions was cut in half from those in Study 1: 5 mmol silane, 10 mmol KF and 0.5 mmol 18-Crown-6; the 10:20:1 reactant molar ratio was unchanged. The product and migration ratios were determined by GC-FID peak heights on a Perkin Elmer 3020B chromatograph with a 6' x 1/8" OV-1/chromosorb stainless steel column. Fairly reproducible* ( 5 %) 0.3 J.1l injections were used to sample the reaction mixtures which were prepared gravimetrically equivalent on a Cenco balance with a sensitivity of 0.1 mg. Qualitative peak identification was performed on a Finnigan 3200 GC-MS with a Teknivent data System. (* As determined by the 18-crown-6 GC-FID peak heights). 3.2.2 Results and Discussion: The generalized product distribution from reactions of KF and ArSiMe2CH2Cl (based on peak heights relative to the Me migration product (D) EtSiArMeF found in a reactions) after 23 hours reaction at 55C in toluene is shown below in table X. [does not include siloxanes resulting from hydrolysis of Si-F

PAGE 51

38 bonds.] Table X Generalized Product Distribution from Reactions of ArSiMe2CH2CI with KF compoundAr A B c D E F phenyl 1.13 6.8 1.0 0.4 U.D. p-tolyl 0.24 1.7 3.9 1.0 0.06 U.D. pfluorophenyl 0.16 0.95 U.D. 1.0 0.15 0.016 p-chlorophenyl 0.45 1.4 U.D. 1.0 0.58 0.065 m-(CF3)phenyl 1.43 2.5 U.D. 1.0 1.8 0.59 legend: A = Me2SiF2; B = ArCH3; C = ArCH2SiMe2F; D = EtSiArMeF; E = (ArCH2)2SiMe2; F = EtMeSiAr(CH2Ar) one secondary product, toluene, and solvent were identical in this case so it was not included iri calculations, performed below in Table XI. peak ratios of AID were not reported in reference 22 but are included in the calculation of the relative ratio of I,2/L1 products contained below in table XI. U.D. =undetermined presence -even by capillary GC-FID compoundAr p-tolyl pfluorophenyl p-chlorophenyl m-(CF3)phenyl Table XI Grouping and Ratios of 1 and 2 Products L2-A+B +E+F 2.0 1.276 2.495 6.32 5.6 1.0 1.0 1.0 L20fL10 0.357 1.276 2.495 6.32 The solvent phase concentrations of Me2SiF2 (bp 2-3cn60 mm Hg) relative to the X= H reaction (3.2.3b), calculated from GC-FID peak hieghts using 0.3 J.1l 5 % volumetric injections of reaction 3.2.3 supernatants after 23 hours reaction at 55C were: X= p-CH3 (0.62); X= p-H (1.0); X= p-F (1.77); X= p-Cl (2.54); X= m-CF3 (7.69). The concentrations* of substituted toluenes (ArCH3 produced relative to the X= p-F** reaction (3.2.3c), calculated from such volumetric injections after the same reaction time were: X= p-CH3 (0.71); X = p-F (1.0); X= p-Cl (1.29); X = mCF3

PAGE 52

{ r ., f 1-, ; I I I r 39 (2.22). Assuming equal GC-FID repsonse factors. ** The X= p-F reaction was used in place of the X= H reaction as a reference because the toluene produced by reaction 3.2.3b was identical to the solvent. Considering the error inherent in the quantitative method used for reactions 3.2.3 a -> e, Figure VIII shows a remarkable correlation between log([Me2SiF2]X/ [MezSiFz]H) and the Hammett ax constant for the respective substituents. ( O'H defined as 0.0 of course). The O'x values used were28: X= p-Me (-0.15); X= H (0.0); X = p-F (0.06); X = p-Cl (0.23) and X = m-CF3 (0.49). These represent a sum-of both resonance O'R and inductive a1 effects in the classic Hammett expression log (k/ko) = pax. X ::::1: N N LL.. LL.. c;; c;; N N Q) Q) I: I: goo 0 --0.2 p-Me = 1. 748 corr. = 0. 99 0.2 0.3 0.4 0.5 Figure VIII Plot of log ([MezSiFzJx I [Me2SiFz]H) vs. Hammett ax Constants for Solution Phase Reactions 3.2.3a->e of ArSiMe2CHzCl Similar plots were made for log (l:2/Ll 0 ) products vs. ax (figure IX), and

PAGE 53

40 for log ([ArCH3]X/[ArCH3]F) vs. O'p after shifting the proton based O'x values downward by -0.06 to obtain O'p values (figure X) The only RIMe migration ratios obtained to correlate with O"x belong to X = p-Me and X= H. Statistically corrected, they were 16.76:1 and 16.0:1 repsectively. -0 ...... 0 N H -g. 0 --0.2 -0. 1 -0.4 0.2 0.3 0.4 0.5 e of ArSiMe2CH2Cl Data obtained for X= p-F and m-CF3 indicate only one primary product, even on a capillary GC-FID.27 GC-MS data however, reveal this one peak to contain both migratory products (C and D, Table X). The shockingly low ratio for the p-Cl silane (0.046) may indicate a steric problem for p-ClPh migration in the previously postulated pentacoordinate adduct (Figure V).

PAGE 54

-Q) Q) ::E ::EI I .r:.a.. Cl..l IU.. X . c. t:J'I 0 -0.2 Figure X = 0. 791 carr. = 0. 995 0.3 0.4 0.5 41 Plot oflog ([X-Ph-Me] I [p-F-Ph-Me]) vs. cr1 Constants for Solution Phase Reactions (3.2.3a->e) of ArSiMe2CH2Cl with KF The excellent correlations and especially the high slopes (rho values) shown in figures vm and IX suggest that a large degree of negative charge stabilization in some secondary (perhaps another pentacoordinate) transition state is necessary for the formation of the secondary products Me2SiF2, ArCH3, Me2Si(CH2Ar)2, and EtSiMeAr(CH2Ar). The lower slope (0.791), shown in figure X for the approximate rate of ArCH3 formation with respect to p-fluorotoluene vs O'p could be due to the assumption made that all substitUted toluenes possess equal FID response factors. The 0.995 correlation coefficient would then be due to chance. To eliminate this type of uncertainty, all subsequent Hammett studies were performed with an actual unreactive internal standard hexadecane. Nevertheless, a mechanism for 1 and 2 product formation consistent with the observations in section 3.2.3 is presented below in figure XI (see next page). This mechanism explains the lowering of the aryVmethyl migration ratio as the

PAGE 55

42 X-substituent on the aryl group becomes increasingly electron-withdrawing because the attack of a second mole of fluoride on the aryl migration 1 product should form from Me2SiF2 and resonance stabilized benzyl anions which would then react with any available acidic protons to form ArCH3. The secondary migration products Me2Si(CH2Ar)2 and EtSiMeAr(CH2Ar) would then arise from the same rearrangement effected by preacting with the starting material, except that the reacting anion is now ArCH2-, competing with any premaining. As mentioned in figure XI, the alternate route to the secondary migration products involving benzyl anion attack and fluoride displacement is thermodynamically unfavorable in view of the strength of tetracoordinate Si-F bonds (154 kcal/mole) compared with that of Si-C bonds (88 kcal/mole). See Figure XI. The 1Semi1-quantitative information obtained by the 0.3 J.Ll 5 % injections of reaction 3.2.3 supernatants provided the insight necessary to give the go-ahead Me -Si...._ F> ) Me 18-Cr-6 Cl 1 x-@-_CH2 X KCI) Me ...... + 'si -CH2 Me-::. I 1 18-Cr-6 / r@-x Me -Si...._ 1 Me Cl f'" I I I I I I v Me -Si ...._ I Et Cl aryl migration methyl migration .P.roducts

PAGE 56

1KCI Me 's ., 1 I l _,@-x Et -Si M/ '-@-x methyl migration aryl migration secondary_P.roducts 43 Me2SiF2 + x-@--cH3 toluene P.roducts relative amount depends on stability of benzyl anion Possible but not probable route to secondary migratory products: ,r@-x Me -Si-.... I Me + X -@--_cH 2 F r@-x Me -Si + F M/ '-@-x )}-x Me -SI1-.... Et + X -@--_cH2 F ,@-x Et-Si + F M/ '-@-x Figure XI Detailed Mechanism of Primary and Secondary Product Formation for Solution Phase Reactions (3.2.3a->e) of ArSiMezCHzCl decision for performing two actual (quantitatively controlled by a GC-FID internal standard) Hammett studies on these solution phase reactions: 3.3. The relative overall rates of reaction as determined by a competitive starting material consumption between the X = p-Me silane and the other four (two at a tinie of course); 3.4 A kinetic study of the rate of appearance of the X -substituted toluenes chosen over the rate of

PAGE 57

44 Me2SiF2 formation becuase of the volatility of the latter. 3.2.3 Detailed Experimental: 3.2.3a p-CH3-PhSiMe2CH2Cl + KF -> p-CH3-PhCH2SiMe2F + EtSi(p-CH3-Ph)MeF p-MePh migration Me migration GC-FID Observations with Res,pect to Reaction Time of 0.3y.l Supernatant Injections RIMe primary % completion Rxn Time C1 ) product ratio wrt to S.M. 1.5 hrs 8.38 11.5 3.5 hrs 10.0 33.0 23 hrs 3.88 57.0 Remarks Both p-xylene and Me2SiF2 present Relative amounts of Me2SiF2 and p-xylene had doubled and tripled respectively. Slight amounts of secondary products and one other side product had formed. Relative amounts of Me2SiF2 and p-xylene had tripled and octupled respectively from the time of the last observation. Substantial amounts of secondary products had formed. The reaction was halted by freezing. GS-MS Data: Relevent ions and (% abundances) for the main products found were as follows (listed in elution order): I. Me2SiF2: M+ 96 (10.1); M+-Me 81 (100) IT. p-xylene: M+ 106 (33.5); M+ -H 105 (18.5); M+-Me 91 (100) ill. pMePhCH2SiMe2F: M+ 182 (18.0); M+-Me 167 (5.0); M+-CH2Ph(p-Me) 77 (100); p-MePhCH2+ 105 (23.5)

PAGE 58

45 IV. EtSi(p-MePh)MeF: M+ 182 (25.0); M+-Me 167 (8.5); M+-Et 153 (100); M+ -p-MePh and/orp-MePh+ 91 (44.0) V. (pMePhCH2)2SiMe2: M+ 268 (3.5); M+ -Me 253 (8.0); M+ CH2Ph(p-Me) 163 (100); p-MePhCH2+ 105 (23.0) VI. MeJSiOSiMe2(p-MePh): M+-238 (22.0); M+-H 237 (100); M+-H and .0. (Me3SiOSiMe2(Ph-p-CH2+)) 221 (27.5); M+p-MePh 147 (4.0); p-MePh+ 91 (5.2); M+-SiMe3 165 (2.0); Me3Si+ 73 (26.0) 3.2.3b PhSiMe2CH2Cl + KF -->. PhCH2SiMe2F + EtSi(Ph)MeF Ph migration Me migration GC-FID Observations with ReS1Ject to Reaction Time of 0.3y.l Supernatant Injections RIMe primary % completion Rxn Time (1 ) product ratio wrt to S.M. Remarks 1.5 hrs 8.0 23 hrs 6.37 48 hrs 3.70 18.12 41.4 58.3 Trace of Me2SiF2; (any toluene formed was blanketed by the solvent) Slight amounts of 2 product forming. About 6 fold the amount of Me2SiF2 present at initial analysis. Substantial quantitites of secondary and one other side product had formed. Twice the amount of Me2SiF4 present as in the previous analys1s. The reaction was halted by freezing. [GS-MS Data: Relevant ions and(% abundances) for the main products found were as follows (listed in elution order): I. (CH3)2SiF2 M+ 96 (10.1); M+Me 81 (100) IT. (CICH 2 )2Si(CH 3 )2 35a2 M+ -H 155 (100); 35cv 37cl M+-H 157

PAGE 59

(30.0); 37cl2 M+-H 159 (9.0); 35cl2 M+-Me 141 (5.0); 35cv 37cl M+-Me 143 (1.5); 37cl2 M+-Me 145 (0.45) ill. PhCH2Si(CH3)2F M+ 168 (32.5); M+Me 153(11.5); M+Ph and/or CH2Ph+ 91 (23); M+CH2Ph and/or Ph+ 77 (100) N. Et(Ph)Si(CH3)F M+ 168 (26.5); M+Me 153 (9.5); M+Et 139 (100); M+Ph 91 (28.5); Ph+ 77 (100) V. (PhCH2)2Si(CH3)2 M+ 240 (7.0); M+Me 225 (1.5); M+CH2Ph 91 (23) VI. Ph(CH3)2SiOSi(PhCH2)(CH3)2 M+ 300 (0.5); M+Me 285 (1.0); M+-. 46 CH2Ph 209 (100); M+Me2SiPh 165 (7.5); M+OSi(PhCH2)(CH3)2 135 (20); Ph(CH3)2SiOSi+(CH3)2 193 (37.0); Vll. PhCH2(CH3)2SiOSi(Ph)(Et)(CH3) M+ 314 (0.5); M+Me 299 (1.0); 3.2.3c M+Et 285 (4.2); M+Ph 237 (trace); M+CH2Ph 223 (100); M+ MeEtSiPh 165 (6.0); M+OSi(PhCH2)(CH3)2 149 (2.0); CH2Ph+ 91 (21.5); (CH3)2Si+SiMeEtPh 207 (28.5)] p-FPhSiMe2CH2Cl + KF ---> p-FPhCH2SiMe2F + EtSi(p-FPh)MeF p-FPh migration Me migration GC-FID Observations with Respect to Reaction Time of0.3bf.l Supernatant Injections RIMe primary % completion Rxn Time (1 ) product ratio wrt to S.M. Remarks 1.5 hrs undetermined 32 (see GC-MS data) SlightamountofMe2SilF2 present. (It was very hard to determine the presence of p-fluorotoluene at this point because of near identical retention times between it and

PAGE 60

3.5 hrs undetermined 41.3 23 hrs undetermined 51.0 49 hrs undertennined 55.7 the solvent) Twice the amount of Me2SiF2 present as at the initial analysis. 47 Enough p-F toluene present to elute just past the solvent peak. Amount ofMe2SiF2 had 10 fold increased since the 3 1/2 hour analysis. Large amounts of secondary products had also formed. Secondary products present in almost equal quantities as the primary products. The reaction was halted by freezing. GS-MS Data: Relevant ions and(% abundances) for the main products found were as follows (listed in elution order): I. (CH3)2SiF2 M+ 96 (10.1); M+Me 81 (100) II. p-fluorotoluene: M+ 110 (32.5); M+H 109 (71.0); M+F 91 (100) ill. p-FPhCH2SiMe2F + EtSi(p-FPh)MeF: M+ 186 (21.5); M+-Me 171 (7.60); M+-Et 157 (12.0); M+CH2(p-FPh) 77 (93.5); p-FPh+ 109 (26.5)* IV. Et(p-FPhCH2)Si(p-FPh)Me: M+ 276 (2.5); M+Me 261 (0.50); M+ Et 247 (2.0); M+p-FPhCH2 167 (60.0); p-FPhCH2+ 109 (13.5); M+p-FPh 181 (1.00); p-FPh+ 95 (0.50) V. (p-FPhCH2)2SiMe2: M+ 276 (8.5); M+Me 261 (2.0); M+p-FPhCH2 167 (100); p-FPhCH2+ 109 (45.0) *primary prOducts unseparable on columns available at the time. 3.2.3d p-CIPhSiMe2CH2C1 + KF --> p-CIPhCH2SiMe2F + EtSi(p-CIPh)MeF p-CIPh migration Me migration

PAGE 61

48 GC-FID Observations with Res,pect to Reaction Time of0.3y.l Supernatant Injections RIMe primary % completion Rxn Time (1 ) product ratio wrt to S.M. Remarks 1.5 hrs 0.023 11.0 3.5 hrs 0.023 13.7 23 hrs 0.023 37.8 SnutilamountofMe4SLF2 present. Large quantities of p-chlorotoluene already present, in 1. 78 x the amount of the primary products (assuming equal GC-FID response factors). Approximately double the quantity of Me2SiF2 as at the 1.5 hour analysis. Amount of p-chlorotoluene has increased by 1.5x the initial analysis. Ratio ofp-chlorotoluene 1/2 products now 2.0. Approximately 4x the quantity of Me2SiF2 as at the 3.5 hour analysis. Ratio of p-chlorotoluene 1/2 products now 1.35. Relative quantity of p-chlorotoluene hasincreased by a factor of 2.44 since the 3.5 hour analysis. One side product and tWo secondary products are now present in near the same quantity as the primary products. The reaction was stopped by freezing. GS-MS Data: Relevant ions and (% abundances) for the main products found were as follows (listed in elution order): I. (CH3)2SLF2 M+ 96 (10.1); M+Me 81 (100) II. p-chlorotoluene: 35cl M+ 126 (28.0), 37cl M+ 128 (8.12); 35cl tropylium+ 125 (13.0); 37a tropylium+ 127 (3.77); M+Cl 91 (100) III. EtSi(p-ClPh)MeF: 35a M+ 202 (13.0), 37cl M+ 204 (3.77); 35cl M+Me 187 (4.0), 37cl M+-Me 189 (1.16); 35cl M+ Et 173 (10.0),

PAGE 62

37ct M+-Et 175 (3.0); M+Cl 91 (100); 35ct Ph+ 111 (2.0), 37ct Ph+ 113 (0.6) 49 N. p-CIPhCH 2 SiMe 2F: 35ct M+ 202 (7.0), 37ct M+ 204 (2.10); 35ct M+ Me 187 (99), 37a M+-Me 189 (30.0); 35ct tropylium+ 125 (52.5); 37 Cl tropylium+ 127 (15.7); Me2 SiP+ 77 (100) V. FMe2SiOSi(Et)(p-CIPh)Me: 35ct M+ 276 (2.5), 37a M+ 278 (0.75); 35ct M+-Me 261 (2.5), 37ct M+-Me 263 (0.75); 35ct M+-Et 247 (2.0), 37ct M+-Et 249 (0.60); M+-p-CIPh 165 (2.0); 35ct M+-F 2 MeSiO or Et(Me)Si(p-CIPh)+ (2.5); 37a M+-F2MeSiO or Et(Me)Si(p-CIPh)+ 183.7 (0.75); F2MeSiOSiMe2+ 151 (100) VI. (p-CIPhCH 2 )2SiMe 2 : 35ct2 M+ 308 (7.5), 35ct37ci M+ 310 (4.5), 37ct2 M+ 312 (0.70); 35ct2 M+-Me 293 (2.5), 35ct37ct M+-Me 295 (1.25), 37 CI2 M+-Me 297 (0.5); 35cl M+ CH 2(p-C1Ph) 183 (100), 37ct M+-CH 2 (p-CIPh) 185 (30.0); 35ct tropylium+ 125 (22.5); 37ct tropylium+ 127 (6.75); Vll. EtMeSi(p-ClPh)(p-CIPhCH 2): 35ct2 M+ 308 (2.0), 35ct37ct M+ 310 (1.0), 37 CI2 M+ 312 (0.2); 35ct2 M+ Me 293 (1.0), 35ct 37 Cl M+ Me 295 (0.5), 37 Cl2 M+ Me 297 (0.2); 35CI2 M+ Et 279 (2.5), 35a37ct M+-Et 281 (1.25), 37ct2 M+-Et 283 (0.25); p-35ciPhSi(Et)Me+ 183 (45), p-37CIPhSi(Et)Me+ 185 (13.5); p-35CIPhCH2Si(Et)Me+ 197 (2.0), p-37CIPhCH2Si(Et)Me+ 199 (0.6); 35a tropylium+ 125 (7 .5); 37 Cl tropylium+ 127 (2.25); 35a 0-Ph 111 (2.0), 37ct Cl-Ph 113 (0.6) 3.2.3e

PAGE 63

50 m-CF3PhSiMezCHzCI + KF --> m-CF3PhCHzSiMezF + EtSi(m-CF3Ph)MeF m-CF3Ph migration Me migration GC-FID Observations with Respect to Reaction Time of 0.3111 Supernatant Injections RIMe primary % completion Rxn Time (1 ) product ratio wrt to S.M. Remarks 1.5 hrs undetermined 5.10 (see GC-MS data) 23 hrs undetermined 48.8 At least 3x the amount of MezSiFz present as contained in the other reactions at this time. Relative amount of m:.trifluoromethyltoluene present in 3.6x and 43x the amounts of p-chlorotoluene and p-xylene formed after the same time in reactions 3.2.3d and 3.2.3a Relative ratios of the amounts of MezSiFz formed by this time (this reaction being unity) by the other 3.2.3 reactions with respect to the X = m-CF3 reaction were: x = pMe, (0.08); X = H, (0.13); X = p-F, (0.23); X = p-Cl, (0.33). Relative ratios of the amounts of x substitued toluenes formed by this time (this reaction being unity) by the other 3.2.3 reactions (producing observable ones) with respect to the x = m-CF3 reaction were: x = p-Me (0.32); x = p-F (0.45); x = p-Cl (0.58); Two secondary products were formed in at least twice the relative quantity of primary products. Halted reaction by freezing. GS..;MS Data: Relevant ions and (% abundances) for the main products found were as follows (listed in elution order): I. Me2SiF2: M+ 96 (10.1); M+-Me 81 (100)

PAGE 64

II. m-CF3-toluene: M+ 160 (79.5); m-CF3+tropylium 159 (31.0); M+ F 151 (17.5); M+-mCF3 91 (100); CF3+ 69 (8.50) 51 ill. *m-CF3PhCH2SiMe2F + EtSi(m-CF3Ph)MeF: M+ 236 (7.5); M+-Me 221 (8.0); M+-Et 207 (5.5); M+-F 217 (10.0); M+-CH2Ph(m-CF3) 77 (100); CH2Ph(m-CF3)+ 159 (10.0); M+-Ph(m-CF3) 91 (5.0); +Ph(m-CF3) 145 (1.5) IV. EtMeSi(m-CF3Ph)(m.-CF3PhCH2): M+ 376 (2.0); M+-Me 361 (1.0); M+-Et 347 (2.0); M+-F 357 (5.0); M+-CH2Ph(m-CF3) 217 (12.0); M+-Ph(m-CF3) 231 (1.5); CH2Ph(m-CF3)+ 159 (9.0); +Ph(m-CF3) 145 (2.0); H(Me)SiPh(m-CF3)+ (rearranged cation from Et loss) 189 (97.5) VI. ( m-CF3 PhCH2)2SiMe2: M+ 376 (1.0); M+-Me 361 (2.5); M+ -F 357 (4.0); M+-CH2Ph(m-CF3) 217 (35.5); CH2Ph(m-CF3)+ 159 (13.0) Primary products unseparable on columns available at the time.

PAGE 65

52 3.3 Competitive Relative Rates of Starting Material Consumption (for X-substituted arylchloromethyldimethylsilanes + KF 18-Crown-6/mesitylene; 72C) 3.3.1 General Experimental: The reactions in this study were quantitatively determined competitions for fluoride anion between the X= p-Me silane and the other four used in section 3.2. They were conducted at 72C in four 20 m1 septum sealed serum cap vials containing 2.5 mmol each of the X = p-Me silane, one of the other four aryl silanes and hexadecane (Kodak) GC-FID internal standard. The quantities of KF and 18-Crown-6 (Aldrich) added to each of the four vials (along with a miniature magnetic stirring bar) and 1.5 ml mesitylene solvent (Fischer) were 2.0 mmol and 0.2 mmol respectively. This stoichiometry was chosen because if both aryl silanes were found to react with fluoride at the same rate, only 40% of each could be consumed. Since we were mainly concerned with the rate of primary product formation, the relative lack of available fluoride should slow the formation of Me2SiF2 and the secondary products. (This was the effect observed). The reactions mixtures were prepared on a Cenco balance having a sensitivity of 0.1 mg and were maintained at 72C in a thermostatically controlled oil bath/magnetic stirrer manufactured by Sybron, Inc. The four competitions were sampled after 41.5 hours reaction at 72 C. GC-FID analyses were performed on a 6' x 1/8" OV-1/Chromosorb stainless steel column in a Perkin elmer 3020B chromatograph. Reference injections of gravimetrically prepared solutions containing 13.5 mole* percent each of the two

PAGE 66

53 arylchloromethyldimethylsilanes being studied and the hexadecane internal standard in mesitylene were made prior to each analysis. <*Same as the intial concentrations of reactants in the 20 m1 serum cap vials). 3.3.2 Results and Discussion The mixtures in the four competitions studied were found to contain only small quantities ( < 5 % of total product peak area) of the secondary products found in reactions 3.2.3 a-> e. This was the expected result in view of the limited fluoride concentration (approximately calculated26 to be 0.11 mole percent or less) available in solution at any one time. The Hammett correlation of the log(relative rates) vs O'x of the substituents studied is contained below in figure Xll go 0 p-Me 0.2 0.3 0.4 0.5 e = 1. sa* corr. = 0. 99 *slope listed in reference 14 as 1.60 FigureXll Plot oflog Kx I KH vs. Hammett O'x Constants for the Relative Rates of Starting Material Consumption Study (3.3)

PAGE 67

54 The 0.99 correlation and 1.58 slope suggest a good degree of anionic character in the aryl group undergoing migration to form the major primary product ArCH2SiMe2F. The instability and base strengths of the aryl anions would be expected to follow the order p-MeAr >Ph> p-FAr > p-CIAr > m-CF3Ar since the ax values used represent the sum of both resonance ( aR) and field ( a 1 ) effects. The pMe arylsilane exhibits both the highest observed ary]/methyl migration ratio (16.76) and the slowest reaction rate due to anionic instability of [p-MePh]. The m-CF3 arylsilane would probably exhibit (on a 100' or greater capillary column)* the lowest ary]/methyl migration ratio and does exhibit the highest reaction rate due to anionic stability of [m-CF3Ph]. *The p-F and m-CF3 aryl silanes were recently (9-87) reacted in the usual scale (see reactions 3.2.3 a -> e) with KF/18-Crown-6 in mesitylene at ll0C for 4 hours and their supernatants were analyzed on a Hewlett Packard 5890 GC-FID for the presence of two primary products. The temperature program used on the 25 m x 0.2 mm cross linked methyl silicone coated capillary column near the elution of the primary products was reduced to 0.5C/min, but only one gaussian peak (no shoulders, even upon software magnification) was observed. As stated before (section 3.2.2), GC-MS analysis of this peak showed ions resulting from both migratory products.

PAGE 68

.3.3a Detailed Experimental for One of the Four Competitions Studied Table XII Gravimetric Preparation ofPhSiMe2CH2Cl vs p-MePhSiMe2CH2Cl (41.5 hrs, 72C) 55 Reactant Table gr mmoles mole % (-KF and 18-Cr-6) PhSiMe2CH2Cl 0.4629 184.80 2.5049 13.5113 p-MePhSiMe2CH2Cl 0.4969 .198.77 2.4999 13.4843 KF 0.1171 58.10 2.0155 18-Crown-6 0.0529 264.30 0.2002 C16H34 (IS) 0.5694 225.45 2.5145 13.5631 Mesitylene 1.3246 120.20 11.0200 balance total: 18.5393 (-KF and 18-Cr-6) Table XIII Gravimetric Preparation of Hexadecane I PhSiMe2CH2Cl, p-Me-PhSiMe2CH2Cl Reference Solution Reference Solution: PhSiMe2CH2Cl 0.4673 184.80 p-MePhSiMe2CH2Cl 0.4990 198.77 C16H34 (IS) 0.5660 225.45 2.4994 Mesitylene 1.3084 120.20 total: 2.5287 2.5104 13.5662 10.8852 18.4237 GCFID Parameters for both reference and analysis Injection size: 0.175 J.1l Temperature Program: 70 --> 280C; 8C. Inin -1 Signal (Attenuation X Range): 128 x 100 13.7253 13.6259 balance

PAGE 69

reference injection Table XIV Calculation of GC-FID response factors from reference injection: (via peak heights; chan block units) peak height peak ratio: mole ratio: silaneas silaneas 81.6 RF C16H34 (IS) PhSiMe2CH2Cl 50.7 p-MePhSiMe2CH20 54.1 0.6213 0.6630 1.0117 1.0044 0.6141 0.6601 T!lbleXV Example ofGC-FID Quantification of Competition Study 3.3.3a Results Reaction Results peak height equivalent mmoles silane IS peak height* present**_ C16H34 (IS) 92.75 PhSiMe2CH2Cl 32.7 53.2487 1.4436 p-MePhSiMe2CH20 44.65 67.6413 1.8338 =peak height!RF ** = (eq IS peak height/IS peak height) x mmoles IS PhSiMe2!J!2.Cl p-MePhSiMe2CH2,Q mmoles silane consumed 1.0613 0.6661 % silane consumed 42.37 26.65 56 Calculation of relative reaction rate with respect to the phenylchloromethyldimethyl silane reaction: [p-MePhSiMe2CH2Cl] (41.5 hrs) = 1.8338/18.5393* = 9.8914 mole percent (41.5 hrs) = 9.8914-13.4843 = -3.5929 mole percent [PhSiMe2CH20l (41.5 hrs) = 1.4436/18.5393* = 7.7867 mole percent (41.5 hrs) = 7.786713.5113 = -5.7246 mole percent I = 0.6276**

PAGE 70

* Assuming total mmoles reactants ->products is unchanged; no dimerization. ** Reported as 0.63 in reference 14. 3.3.3.b Table XVI 57 Results of Competition study starting material disappearance after 41.5 hrs. reaction at 72C vs p-MePhSiMe2CH2Cl (crx = -0.15) Xcomn L\fX-PhSiMe2CH2Cll ....Kxl.Kp_Me crx_ L\fn-MePhSiMe2CH2Cll H -5.7246 mole % -3.5929 mole % 1.5933 0.6276 0.0 p-F -7.6998 mole% -2.8869 mole % 2.6672 1.6739 0.06 p-Cl -6.7595 mole% -1.4505 mole% 4.6601 2.9247 0.23 m-CF3 -5.3414 mole % -0.5842 mole % 9.1431 5.7382 0.47

PAGE 71

58 3.4 Appearance rate of x-substituted toluenes (ArCH3). The acquisition of the rates of formation of substituted toluenes (ArCH3) presumably arising from the attack of a second mole of fluoride on the primary aryl migration product (see fig. XI arid eq 15 & 16) was the object of these experiments. ArCH2SiMe2F + p-> ArCH2+ Me2SiF2 (eq. 15) ArCH2+ H+ solvent/H20 from KF <-> ArCH3 (eq. 16) 3 .4.1 General Experimental: Reactions 3.4.3a through 3.4.3e were conducted in 20 m1 septum sealed serum cap vials with mini-magnetic stir bars as in all the previous studies. Mesitylene was the chosen aromatic solvent because all the substituted toluenes produced in these reactions (except p-chlorotoluene in 3.4.3d) are easily resolved by non capillary gas chromatography. The stoichiometry of reactants chosen was as follows: Arylchloromethyl dimethylsilanes: 5 mmole; KF (Aldrich): 10 mmole; 18-Crown-6 (Aldrich): 0.5 mmole; C16fi34 internal standard (Kodak): 0.27 -> 0.31 mmole; Mesitylene solvent (Fischer): _1.84 ml, 12.7 -> 14.4 mmoles. Reaction temperature was maintained at 85c by a thermostated oil bath/magnetic stirrer manufactured by Sybron Inc. GC-FID quantitative analysis for all but the p-Cl substituted toluene was conducted after 1.5, 3, 4.5, 6, 10, 21.5 -> 28 and 45 -> 47 hours of reaction on a 6' x 1/8" OV -1/chromosorb stainless steel column mounted in a Perkin Elmer 3020B chromatograph. The reaction ( 3.4.3d) producing p-chlorotoluene necessitated the use of a 0.25mm x 10m OV-101 silicone oil coated open tubular stainless steel capillary column to effect its resolution from the mesitylene solvent. Reference injections of

PAGE 72

59 gravimetrically prepared solutions containing 1.3 -> 1.4 mole percent each of the x-substituted toluene, C1@f34 internal standard, balance mesitylene were made prior to each analysis in order to obtain FID response factors. <* ACenco balance having a sensitivity of 0.1 mg was used to prepare both the reaction mixtures and the reference solutions**). ** Note: The reference solutions containing 1.3 -> 1.4 mole percent each of X-Ph-Me and C16H34 were prepared via 10 fold dilutions of solutions containing _13.5 mole percent each of those components in mesitylene. This was done to ensure a more accurate weight of the fairly volatile X-Ph-Me compounds present in solution, \ and to approximate their concentrations formed in the reactions. 3.4.2 Results and Discussion The graphical representation of reactions 3.4.3a -> e as displayed in figure xrn shows two important features: 1) The clustering of the X= p-Me, Hand p-F reaction profiles into one set and the X = p-Cl and m-CF3 profiles into another. 2) A leveling off or rate decrease in the production of the X-substituted toluenes after 4.5 hours for X = p-Cl and m-CF3, and after 10 hours for the rest. Observation (1) is not surprising in view of the relative magnitudes [(-0.15, 0, 0.06) vs. (0.23, 0.47)] of the sigma constants for the respective substituents. Indeed, the GC profiles of reactions 3.2.3 and 3.4.3 showed that the formation of visible quantities of all primary and secondary products followed the same grouping. Observation (2) may be explained by the following view. of the mechanism postulated back in Figure XI. Since the sigma constants ( crx) used for the aryl

PAGE 73

0.7 0.6 .... c 4) o.s 4) lo... Q. 4) 0 .... l: .c Q. I X 0.3 4) 0 E 0.2 E 0.1 s 10 15 20 25 30 35 Reation time ( hours ) at 85C FigureXlll so Reaction Profiles from the Rate of ArCH3 Appearance Study (3.4) 60 substituents represent a sum of resonance and field effects, it is quite possible that as each reaction consumes a second mole of fluoride, the resonance stabilization of the released benzyl anion becomes of more importance than the field (through space) interaction of the Si bonded aryl group with the growing positive charge on the chloromethy lcarbon about to undergo loss of chloride in determining the type (primary vs secondary) of product formed. (See Figure XI). If there exists a genuine equilibrium for X-PhCH2:+ H+ --X-PhCH3 in the weakly acidic mesitylene medium, it should be clear that the resonance stabilization of the benzyl anion increases in the same direction as the sigma constants. When a large enough X-PhCH3 concentration has formed, the presence of an increased X-benzyl anion concentration enables the formation of the secondary hi-aromatic silanes from either of the two pathways in Figure XI. The GC observation of secondary silane formation coincided

PAGE 74

61 with the rate decrease in X-PhCH3 production for all cases. That the relative magnitude of this decrease was smallest in the X = pMe case is significant because the equilibrium concentration of the p-MePhCH2:anion should be the lowest of the five. The effect of this -a relative lack of secondary silane production, even after 45 hours reaction was observed. Given the premise that a binary mechanism, which produces and then consumes benzylic anions in two pathways, is operative, and that the general shapes of the reaction prof:U.es contained in figure xm are parabolic, each reaction data set was fit to the best least squares obtained second order polynomial (see Appendix A). This was done so that the entire data matrix was treated in a uniform fashion. The earliest experimentally observed time at which the calculation of a positive quantity of the x-Ph-Me was possible for all five reactions, was 3.0 hours. [1.5 hours were used due to an erroneous data set utilized in reference 22]. The relative (to X= H) rates of X-Ph-Me formation obtained from the quantities of X-Ph-Me compounds calculated at 3 hours reaction were: X = m-:CF3 (11.4); X = p-Cl (10.8); X = p-F (2.46); X = H (1.0), X = p-Me (0.451). The Hammett plot of the log of these rates vs crx is shown in Figure XIV. Reference 22 erroneously reported the slope (correlation coefficient) as (0.94). For interests sake, least squares Hammett plots were made of the actual data at reaction times of 1.5, 3.0, and 4.5 hours. The p values and (correlation coefficients) obtained respectively, were 3.565 (0.924), 2.760 (0.931), 2.311 (0.917). General rate equalization ofX-Ph-Me formation among the five reactions thereafter would lead to a lowering of both the slopes and correlation coefficients, providing little (if any) useful mechanistic information from such Hammett plots. That the slopes obtained from this study are larger than the 1.58 obtained from reaction 3.3.3 should come as no surprise in view of the mechanism postulated

PAGE 75

62 c G) )( I: G) :1... p-c1. c m-CF 3 I G) .., .&:. ::II :1... Q. .... ::II I 0 0 X .&:. .., I") .., G) G) .... -.... 0 0 E < s E "CCI E Cl) p = 2. 435 < '--" .... corr. = 0. 931 ::II u 0 .... < II u )(I :I: 0.2 0.3 0.4 0.5
PAGE 76

3.4.3 Detailed Experimental 3.4.3a p-MePhSiMe2CH2Cl + KF/18-Crown-6/mesitylene/85C Reactant Table gr MW mmoles mole percent -KF + 18-Cr-6 p-MePhSiMe2CH20 0.9916 198.77 4.9887 26.6947 KF 0.5820 58.10 10.0172 18-Crown-6 0.1332 264.30 0.5040 C16H34 (IS) 0.0604 226.45 0.2667 1.4271 Mesitylene (1.5 ml) 1.6146 120.20 13.4326 Balance 18.6880 -(KF + 18-Cr-6) Reference Solution gr MW mmoles mole percent p-Xylene (J.T. Baker) 0.1263 106.16 1.1899 C16H34 (IS) 0.2712 226.45 1.1976 Mesitylene (total) 120.20 88.4230 Total: 90.8105 GC-FID Parameters for both reference and analysis: Injection size: 0.2 J.Ll Temperature program: 70C --> 280C, 8C/min Signal (Attentuation X range): 32 x 100 Calculation basis: Peak heights (in chart block units) Reaction Results: [mmoles IS in reaction= 0.2667] RxnTime p-xylene IS 1.5 hrs 3 hrs 4.5 hrs Reference peak height 31.95 mole percent 1.3103 Reaction peak height 0. 70 mmoles p-xylene present: Reference peak height 33.70 Reaction peakheight 0.80 mmoles p-xylene present: peak height 33.25 Reaction peak height 1. 7 5 mmoles p-xylene present: (table continued on following page) 37.05 1.3188 55.10 0.0039 39.10 28.55 0.0086 37.90 30.45 0.0174 1.3103 1.3188 Balance RF 0.8679 0.8679 0.8830.

PAGE 77

6 hrs 10hrs 21.5 hrs 45 hrs Reaction Time 0.5 1.0 1.5 2.0 3.0 4.5 6.0 10.0 21.5 22.5 26.0 28.0 45.0 47.0 64 Reference peak height 33.25 37.90 37.10 0.0448 0.8830 Reaction peak height 5.50 mmoles p-xylene present: Reference peak height 27.6 Reaction peak height 29.7 mmoles p-xylene present: Reference peak height 28.2 Reaction peak height 28.8 mmoles p-xylene present: Reference peak height 37.2 30.55 47.0 0.1853 31.5 27.9 0.3055 0.9093 0.9010 0.7800 Reaction peak height 39.85 mmoles p-xylene present: 48.0 20.60 0.6614 p-Me 0.0039 0.085 0.0174 0.0448 0.1853 0.3055 0.6614 TableXVll Results for Rate Studies 3.4.3a -> e Reaction mmoles produced H 0.0061 0.0227 0.0703 0.0881 0.4802 p-F 0.0193 0.0821 0.1254 0.1357 0.2557 0.3664 0.5164 p-Cl 0.3045 0.2895 0.4559 0.4263 0.4843 0.4846 m-CF30.2689 0.2886 0.3558 0.3726 0.4602 0.4721 0.4268 0.5645 3.5 Synthesis of Starting Materials 3.5.1 Preparation of (p-CH3-Ph)SiMe2CH2Cl I. p-CH3PhBr + Mg0 ----.> Into a dry 500 ml 3 necked flask, was placed Mg turnings (8.61 gr, 0.354 mole, Fischer) and a magnetic stirring bar. This was covered with 100 ml anhydrous

PAGE 78

65 ether. Onto this was attached a N2 equipped reflux condensor and an addition funnel. Into the funnel, was placed 10 m1 Aldrich p-bromotoluene. The system was purged with N2 and with gentle stirring, was begun the addition of 10 m1 portion of p-bromotoluene. The reaction did not begin until a drops of 1,2-dibromoethane were added to the mixture. The balance of the p-bromotoluene was added in a 50/50 solution with ether over a one hour period. Total halide used was 53.7 gr (0.314 mole). IT. p-CH3PhMgBr + ClSiMe2CH20 --> MgBrCl + (p-CH3-Ph)Si.Me2CH2Cl .When Grignard formation was complete (2 hours), 44.9 gr (0.314 mole) Petrarch ClSiMe2CH2Cl was placed into a clean addition funnel and added the silane over a one hour period. Precipitation of MgBrCl and ether reflux was observed within 30 minutes .. The reaction was magnetically stirred for 24 hours in a darkened hood to retard the formation of bitolyl from a photochemically caused coupling of the Grignard reagent. The unreacted Grignard was destroyed and the MgBrCl cake was dissolved by adding 400 m1 saturated aqueous NF4Cl solution to the reaction mixture. The layers were transfered into a separatory funnel and the aqueous layer was extracted twice with 200 m1 portions of ether. The upper product containing ether layer was dried over anhydrous sodium sulfate. The sodium sulfate was filtered from the ether layer in a BUchner funnel and most of the ether was removed under gentle vacuum through a rotary evaporator. 75 m1 of a light yellow fluid was collected. Vacuum distillation of this fluid through a 30 em Vigreux column afforded 37.8 gr of a 99.3 %pure (GC-FID) (p-CH3-Ph)SiMe2CH2Cl: MW: 198.77; bp 78.5C/5.5 mm Hg; Yield 60.6 %.

PAGE 79

66 Structure verification determined by GC-MS. Relevant ions and (% abundances) found were: 35cl M+ 198 (4.0), 37cl M+ 200 (1.2); 35cl M+-Me 183 (1.0), 37 Cl M+ Me 185 (0.3); 35ct M+ -35clCH2 and/or 37 Cl M+ -37clCH2 149 (100) 3.5.2 Preparation of (p-F-PhlSiMe2Qi2Cl Using a stepwise procedure analogous to synthesis 3.5a, the Grignard reagent of p-fluorobenzene was prepared from magnesium turnings (8.61 gr, 0.354 mole, Fischer), 4-fluorobromobenzene (54.96 gr, 0.314 mole, Aldrich) and 150 ml dry ethyl ether. The Grignard reagent was then reacted with chloromethyldirnethylchlorosilane (44.9 gr, 0.314 mole, Petrarch) to form (p-F-Ph)SiMe2CH2Cl. 50 ml of a greenish yellow fluid was collected after the hydrolytic work-up of the reaction Inixture. Vacuum distillation of this fluid through a 30 em Vigreux column afforded 23.8 gr of a 97.3 %pure (via GC-FID) (p-FPh)SiMe2CH2Cl: MW 202.73; bp 62.7C/6.0 mm Hg; Yield 37.4 %. Structure verification determined by GC-MS. Relevant ions and (% abundances) found were: 35cl M+ 202 (5.7), 37cl M+ 204 (2.0); 35cl M+-Me 187 (1.5), 37 Cl M+ Me 189 (0.65); 35cl M+ 35ClCH2 and/or 37 Cl M+ -37ciCH2 153 (100). 3.5.3 Preparation of (p-Cl-Ph)SiMcc2Q:I20. Using a stepwise procedure analogous to synthesis 3.5a, the Grignard reagent of 4-bromochlorobenzene was prepared from magnesium turnings (5.74 gr, 0.236 mole, Fischer), 4-bromochlorobenzene (50 gr, 0.261 mole, Aldrich) and 160 ml dry ethyl ether. The Grignard reagent was then reacted with

PAGE 80

67 chloromethyldimethylchlorosilane (37 .35 gr, 0.261 mole, Petrarch) to form (p-ClPh)SiMe2CH2CL 35 m1 of a bluish green fluid was collected after the hydrolytic work-up of the reaction mixture. Vacuum distillation of this fluid through a 30 em Vigreux column afforded 18.47 gr of a 97.3% pure (via GC-FID) (p-ClPh)SiMe2CH2Cl: MW 219.19; bp 89C/5.5 mm Hg; Yield 32.3 %. Structure verification determined by GC-MS. Relevant ions and (% abundances) found were: 35cl2 M+ 218 (6.08), 35cl37cl M+ 220 (4.92); 37cl2 M+ 222 (0.77), 35cl2 M+-Me 203 (1.17), 35cl37cl M+-Me 205 (0.71); 37cl2 M+-Me 207 (0 14); 35cl (p-ClPh)SiMe2 169 (100), 37cl (p-ClPh)SiMe2 171 (35) (loss of CH2Cl from the 3 possible molecular ions). 3.5.4 Preparation of Cm-CF3-PhlSiMc2CH2Q Using a stepwise procedure analogous to synthesis 3.5a, the Grignard reagent of m-trifluoromethylbromobenzene was prepared from magnesium turnings (6.06 gr, 0.249 mole, Fischer), m-trifluoromethylbromobenzene (50 gr, 0.222 mole, Aldrich) and 150 ml dry ethyl ether. The Grignard reagent was then reacted with chloromethyldimethylchlorosilane (31.8 gr, 0.227 mole, Petrarch) to form (m-CF3Ph)SiMe2CH2CL 50 m1 of a clear brown fluid was collected after the hydrolytic work-up of the reaction mixture. Vacuum distillation of this fluid through a 30 em Vigreux column afforded 22.83 gr of a 99.4 % pure (via GC-FID) (m-CF3Ph)SiMe2CH2Cl: MW 252.45; bp 66.5C/5.5 mm Hg; yield 40.7 %. Structure verification determined by GC-MS. Relevant ions and (% abundances) found were: 35cl M+ 252 (trace), 37CIM+ 254 (trace); 35ci M+-Me

PAGE 81

68 237 (3.13), 37ci M+-Me 239 (1.37); 35ci M+-F 233 (5.49), 37ci M+-F 235 (2.04); 35ci M+-35CICH2 and/or 37ci M+-37CICH2 203 (100).

PAGE 82

CHAPTERN PYROL YTIC REARRANGEMENTS OF SUBSTITUTED ARYLCHLOROMETIIYLMONO (AND DI-) METHYL SILANES 4.1 Introduction: 69 Between 1975 and 1980, Brook29 and his co-workers, investigated the thermal rearrangements of a-substituted halomethylsilanes, R3SiCHXR', from their sealed tube pyrolysis at temperatures up to 330C. Generally, they found clean exchange of X and R groups between carbon and silicon and concluded that the rearrangement proceeded by the unimolecular formation of an ylid" wherein the migration of X from carbon to silicon occurs via a cyclic transition state ( eq. 17). The rate determining step was (eq.17) envisaged to be (i), the formation of the cyclic intermediate, followed by (ii), the rapid migration of R from silicon to carbon. They considered the kinetic data from the observed migratory aptitudes of various R groups to be inconsistent with the formation of a double-bridged transition state (eq. 18) required for simultaneous migration of X and R in a dyotropic rearrangement. ....R' R3SiCHXR'-----+ R Si CHR' 2 .... (eq.l8) X

PAGE 83

70 Radical involvement at 330C was refuted in Brooks' sealed tube studies due to the lack of any polyhalogenated side products (arising from chain reactions), and in a recent study conducted by Davidson30 using a strictly gas phase stirred flow (SFR) pyrolysis apparatus (see appendix C) in which radical trapping reagents added did not affect reaction rates or product composition (see discussion section 4.5). The same (sealed tube) rearrangements at lower temperatures (136C) were fotmd by Jung and Weber31 and by the author of this thesis at 300C (section 4.2) however, to yield polyhalogenated side products (e.g. XR2SiCHR'X, RR'SiX2 and free R-H compounds) which are indicative of radical chain processes. Only the true gas phase SFR pyrolysis as performed by Davidson and Simon on compounds prepared in this chapter was found to give totally clean rearrangements. 4.2 Isolation of Sealed Tube Pyrolysis Products from Phenylchloromethylmethylsilane A sealed borosilicate ampoule containing 0.58 gr PhSiHMeCH2Cl (prepared in section 4.6) under argon was placed in a muffle furnace at 300C for 48 hours. A startling change (aromatic--> pungent) of odor was noted when the ampoule was opened for the arialysis of its contents. The resultant yellowish brown fluid was found by GC/GC-MS* to contain 71 %of the proton migration product PhMe2SiCl and at least 5 other but absolutely no chloromethyl starting material. GC-MS anayltical results;-relevant ions (abundances)from a Hewlett Packard 5890 GC with a 25m x 0.2 mm crosslinked methylsilicone gum capillary column and a 5970 mass selective detector utilizing electron impact ionization 70 ( e V) and quadrupole ion sorting were as follows:

PAGE 84

71 I. Starting material: PhSiHMeCH 2CI: F.W.: 170.71; 35ci M+ 170 (1504), 37ci M+ 172 (646); 35ci M+-Me 155 (991), 37ci M+-Me 157 (306); M+ CH 2 CI.121 (49184), 35ci M+Ph 93 (3025), 37ci M+Ph 95 (2819); Ph+ 77 (5668). II. Main Products (listed in elution oider): A. C6H6: M+ 78 (50528); M+-H 77 (43184); H2C=CH-HC=CH2 53 (1472); [H2C=C=C=CH2]+ 52 (37312); +HC=C=C=CH2 51 (37112). B. PhSiHMe2: M+ 136 (24456); M+-H 135 (20904); M+-Me 121 (49984); M+Ph 59 (4951); MeHSi+=CH2 58 (37312); Ph+ 77 (4294). C. PhSiMe3: M+ 150 (12634); M+-Me 135 (50528); M+Ph 73 (2307); Ph+ 77 (2627). D. Main Product: PhSiMe2CI: 35a M+ 170 (11968), 37 Cl M+ 172 (4897); 35a M+-Me 155 (50528), 37ci M+-Me 157 (24456); 35ci M+ Ph 93 (9228), 37 Cl M+ Ph 95 (2462); Ph+ 77 (16384). Lack of M+-CH2CI at 121; 35ci +CI(CH2)2 63 (43896), 37ci +ct(CH 2 )2 65 (24456). E. Possible Ph migration product (PhCH2SiHMe) in D. due to the presence of two ions: tropylium cation 91 (19904); M+CH2Ph 79 (4924) (Tile relative abundance of these ions increases

PAGE 85

slightly as spectra are taken towards the end of peak D on the TIC chromatogram). 72 F. Mixture ofPhMeSiCl2 and PhCH2SiC12H: 35Cl2 M+ 190(19480), 35cl37cl M+ 192 (12164); 37cl2 M+ 194 (2307); 35cl2 M+ -Me 175 (50528), 35cl37cl M+-Me 177 (49448); 37cl2 M+ -Me 179 (9848); 35a2 M+Ph 113 (4897), 35cl37cl M+Ph 115 (3057); 37cl2 M+Ph 117 (744); Ph+ 77 (10452); tropylium cation 91 (4664); 35cl2 M+-CH2Ph 99 (384), 35cl37cl M+-CH 2 Ph 101 (622); 37cl2 M+-CH 2 Ph 103 (748); 35cl M+-Cl 155 (2992), 37cl M+-Cl 157 (1141). G. HPhSi(CH 2 Cl)2: 35a2 M+ 204 (4664), 35cl37cl M+ 206 (2773); 37cl2 M+ 208 (583); 35cl M+-CH 2 Cl 155 (49984), 37cl M+ CH 2 Cl 157 (36912); 35cl2 M+ Ph 127 (378), 35cl37 Cl M+Ph 129 (580); 37ct2 M+Ph 131 (368). 4.3 Argon Flow Pyrolysis of ArSiHMeCH2CI and ArSiMe2CH20 4.3.1 General Experimental Our next plan was to conduct this thermal rearrangement in such a manner as to let it proceed partway, so that some sort of relative rate data could be obtained when one varied the aryl substituents on the starting material thus giving insight into the electronic effects involved in the rearrangement. Two groups of silanes, ArSiHMeCHCl and ArSiMe2CH2Cl, with the same five aryl substituents used in Chapter 3 were chosen for the initial pyrolysis study performed by myself. This could possibly yield relative migration ratios for HI Ar and Ar/Me migration products from the respective pyrolysis, thus providing additional mechanistic insight. The method

PAGE 86

73 chosen was a simple argon-flow drop tube pyrolysis through a calibrated tube furnace. The pyrolysis tube itself was 22 em long x 1 em O.D. (0.75 em I.D.), constructed of fused silica with 14/20 fittings at both ends and having a silica frit near its base so that it could contain packing material. This was chosen to be thirty five 0.6 em diameter borosilicate beads for obvious reasons, such as surface and packing uniformity and a relative lack of flow restriction for either gas or liquid. The furnace itself encased the tube entirely so that only the 14/20 fittings on either end were out of the hot zone .. To the top of the tube was attached a graduated 10 m1 addition funnel equipped with an argon inlet. To the bottom, was attached a 10 ml two neck round bottom flask placed in an ice-salt bath with a small diameter gas exit tube in one neck. To establish working pyrolysis parameters, it was then necessary to calibrate a 120 V variac dedicated to the tube furnace. This was done twice using a thermocouple pyrometer so that a given reading on the variac corresponded to a certain temperature in C 5 %. The parameters chosen for the argon flow pyrolyses of ArSiHMeCH2Cl were as follows: Volume of ArSiHMeCH2Cl = 1 ml; Pyrolysis temperature= 560 C (below the softening point of borosilicate beads); Hot zone contact time* per drop of silane= 1.0 sec.; Addition rate= 1 d.rop/12 seconds over a 2 minute period. *Determined from a heat expansion neglected computation of the hot zones internal volume (via volumetric addition of water into the bead filled tube = 5.04 cc), and adjusting the argon.flow rate through the heated tube to 5.04 cc/sec or 300 cc/min. The argon flow rate was checked just prior to each pyrolysis with a flowmeter designed by Alltech Inc. Pyrolysis parameters established** for the ArSiMe2CH2Cl compounds were identical except for a tube furnace temperature of 700C. This necessitated the use of crushed fused silica packing material in the pyrolysis tube. **Pyrolyses of the

PAGE 87

74 m-CF3PhSiMe2CH2CI compound at 560C and 1000C gave no rearrangement and destructive carbonization respectively.

PAGE 88

74 4.3.2 Results of the argon flow I!yrolyses of substituted ar.ylchloromethyl mono (and dil methylsilanes Approximately 0.5 m1 of each pyrolysis effluent was condensed and immediately submitted for both GC-MS and GC-FID analyses. (Initial FID results obtained on a 6' x 1/8" SS 10% OV-1 silicone oil/chromosorb WHP column in a Perkin Elmer 3020B chromatograph using peak heights; subsequent GC-FID analysis of the same effluents, after one month refrigerated ( -1 0C) storage in paraffm sealed vials, were performed on a Hewlett Packard 5890 chromatograph using the same 25 m x 0.2 mm crosslinked methylsilicone gum capillary column as contained in the HP GCMS system). GC results obtained on both HP instruments total ion current (TIC) from the GC-MS and the FID-were reponed using software integrated peak areas. It might be mentioned here that any "quantitative" information derived from the TIC on GC-MS systems is questionable due to the fact that a given TIC response factor for any molecule is dependent on its ionization potential and the stability of its molecular ion to further fragmentation reactions. Thus, the TIC response factors for the H/PhX and PhX/CH3 migration products observed from the respective argon flow pyrolyses of X-PhSiHMeCH2Cl (X-PhSiClMe2 I HSiCIMeCH2PhX) and X-PhSiMe2CH2Cl (ClSiMe2CH2PhX I X-PhSiCIMeEt), would be vastly different due to significant variance in molecular structure; i.e. the presence of the benzylic products due to aryl migration would result in a higher TIC response factor than for migratory products involving protons or methyl groups because of the favorable fragmentation of the former to substituted tropylium cations (in addition to other reactions). Table XVIIT and XIX summarize the observed products and ratios of interest.

PAGE 89

75 TableXVIll 560C Argon flow pyrolyses of X-PhSiHMeCH2Cl ----,> X-PhSiCIMe2 + HSiC1MeCH2PhX S.M. GC-FID% TIC FID* FID** volatility X completion H/Ar H/Ar H/Ar 4 p-Me 5.61 11.81 112 00 2 H 18.4 7.0 743 00 3 p-F 12.2 52.7 182 00 5 p-Cl 6.86 *** 333 00 1 m-CF3 22.7 4.42 138 00 :*6' column/PE 25 m column/HP; one month -10C storage *** Both proton and aryl migration products were seen to be contained in a single GC-MS peak. This was evidenced by an increasing abundance of the p-Cl tropylium cation as one took spectra towards the end of the peak. The volatility of the aryl migration product was found to be less than the proton migration product in all cases. **The complete dissappearance of the aryl migration products after a month's storage at -10C is probably indicative of the instability of silanes containing three reactive substituents; H, C1 and CH2Ph. *The earlier FID ratios obtained suggest little if any correlation with electronic substituent effects and may only represent an in-situ (column) thermal decomposition of the aryl migration products into the more stable proton migration products. Indeed, the major correlation to be observed from the argon flow pyrolysis of X-PhSiHMeCH2Cl was the percent completion of the rearrangement with respect to the volatility of the starting material; (i.e., the lower boiling point of the starting material, the greater the percent completion of the rearrangement at 560C). The GC-MS data of the argon flow from the X-PhSiHMeCH2Cl compounds showed virtually the same analogous product distribution obtained in the

PAGE 90

76 sealed-ampoule pyrolysis of the X = H compound. Table XIX 700C Argon flow pyrolyses of S.M. GC-FID% TIC FID* FID** X volatilitv completion Ar/CH3 Ar/CH3 Ar/CH3 p-Me 4 23.32 3.09 5.88 (4.53) H 2 43.71 2.42 2.93 (4.87) p-F 3 21.43 UR*** UR (4.66} p-Cl 5 42.88 UR*** UR (UR****) m-CF3 1 22.2 UR*** UR CUR****) 6' column/PE ** 25 m column/HP; one month -10C storage *** Mass spectra indicated an increasing abundance of fragment ions common to the methyl migration product (i.e. C2H5+) as one took spectra towards the end of the pyrolysis product peak. This was taken as evidence of both migrations occuring. **** Software magnification of the pyrolysis peak obtained with a slow (1-2C/min) temperature program showed a shoulder on the tailing side. Modification of the temperature program and carrier gas flow rate may effect the resolution. Representative mass spectra of the ininor and least volatile migratory products (X-Ph and Me) from the respective pyrolyses of X-PhSiHMeCH2 Cl and X-PhSiMe2CH20 are given below for the X= H compounds (analogous spectra were obtained from the other compounds): HSiCIMeCH 2 Ph: 35cl M+ 170 (12296), 35cl M+ 172 (5198); 35cl M+-Me 155 (6115), 35cl M+-Me 157 (2599); 35ci M+-H 169 (3057), 35cl

PAGE 91

M+-H 171 (2928); M+-Cl 135 (1292); 35cl M+CH 2 Ph 79 (20568), 35cl M+CH2Ph 81 (4845); tropylium cation 91 (48912); toluene cation 92 (48912); Phenyl cation 77 (20680); benzene cation 78 (12033). 77 EtSiClMePh: 35Cl M+ 184 (3961), 35ci M+ 186 (1229); 35cl M+-Me 169 (990), 35cl M+-Me 171 (307); 35cl M+-Et 155 (10000), 35cl M+Et 157 (9839); M+-Cl 149 (157); 35cl M+Ph 107 (247), 35cl M+ Ph 109 (37); phenyl cation 77 (1203); benzene cation 78 (985); relative lack of the X-substituted tropylium cation at 91 and of a molecular ions loss of the same at 93. A representative mass spectra of the major (X-Ph) migratory product for the argon flow pyrolysis of the aryl dimethylchloromethylsiianes is given below for the X = H compound. (Analogous spectra were obtained from the other compounds). PhCH 2SiC1Me2 : 35cl M+ 184 (4892}, 37cl M+ 186 (1969); 35cl M+-Me 169 (979), 35cl M+ Me 161 (305); M+ Cl 149 (621); 35cl M+ Ph 107 (117), 35cl M+ -Ph 109 (7.0); 35cl M+ CH 2 Ph 93 (10000), 35cl M+ CH 2 Ph 95 (9785); tropylium cation 91 (4838). Excluding the methyl migration, other minor (and more volatile) products from the argon flow pyrolysis of the aryldimethylchloromethylsilanes are represented by the GC-MS data obtained for the p-Me-Ph compound below: Me 3 SiCl: 35cl M+ 108 (38), 37cl M+ 110 (10); 35cl M+-Me 93 (1321),

PAGE 92

78 35cl M+ Me 95 (509); M+ Cl 73 (382). Me 3SiC12 : 35cl2 M+ 128 (95), 35cl37cl M+ 130 (71); 35cl2 M+-Me 113 (1512), 35cl37cl M+ 115 (1231), 37cl2 M+-Me 117 (189); 35M+ Cl 73 (41). 37cl M+-Cl 75 (31). p-MePh-SiMe3: M+ 164 (5856); M+-Me 149 (47096); +siMe3 73 (986); tropylium cation 91 (1357). p-MePhSiCIMe 2 : 35cl M+ 184 (5399), 37cl M+ 186 (1537); 35ct M+-Me 169 (24192), 37a M+-Me 171 (10509); 35ct+siMe2 93 (1292), 95 (368); tropylium cation 91 (4845). 35ci2 M+ 204 (1211), 35cl37cl M+ 206 (752), 37cl2 M+ 208 (156); 35cl2 M+-Me 189 (3987), 35cl37cl M+-Me 191 (2819), 37cl2 M+-Me 193 (612). bitolyl: M+ 182 (1238);* M+-Me 167 (212); MePhCH2+ 105 (986); phenyl cation 77 (3482). dihydrobitolyl: M+ 184 (45); M+-H 183 (188); M+-2H (bitolyl cation) 182 (752); M+-3H 181 (186); 167 (190); MePhCH2+ 105 (2320); phenyl cation 77 (3008).

PAGE 93

79 4.4 Stirred flow pyrolyses ofX-PhSiHMeCH2Cl and X-PhSiMe2CH2,Q 4.4.1 General EXP&imenta1: After hyper-purification of both sets of 4 compounds via vacuum distillation through a micro-scale spinning band column (mfd by B/R instruments) to liquid phase purities of 99.999%, they were pyrolyzed at 2 atm in a nitrogen carrier gas flowing through a stirred flow pyrolysis apparatus having a flame ionization detector. Injection of the samples was conducted entirely in the gas phase by pumping the injection chamber down to a pressure of 0.07 to 0.15 torr* before the introduction of the nitrogen carrier gas at 2 atm. *The change in the partial pressure of the sample over this range did not alter the data obtained in this experiment, which was of the kinetic sort; i.e., providing the Arrhenius parameters (A and Ea) of the rearrangements to yield the rate constants for the migrations observed in both sets of compounds. Pyrolysis oven contact time was set at 15 seconds. Computations provided by R. Simon from these data provided the comparative percent completions (via GC-FID peak areas) for all but the X= p-Cl compounds presented in tables XX and XXI below. 4.4.2 Table XX Results of the Stirred flow pyrolysis of at 560C k approximate mi.sr.. Lkx-Phmigr: p-Me 1U H 10 p-F 10 m-CF3 10 calculated % consumption of chloromethyl starting material* 64 70 64 62 S.M. Volatility 4 2 3 1

PAGE 94

4.4.3 Table XXI Stirred flow pyrolyses of X-PhSiMezCHzCI at 700C approximate X migr fkMe migr: p-Me 10 H 10 p-F 10 m-CF3 10 calculated % consumption of chloromethyl starting material* 109 88 110 91 approximate percent completion, 10 % min. Volatility 4 2 3 1 80 Tables XX and XXI yield the following rates of substituent migration for chloromethylsilane pyrolyses: Proton/AryVMethyl = 100/10/1. Final note on stirred flow pyrolyses: Mass spectral analyses of all pyrolysis affiuents showed virtually none of the volatile side products (e.g. PhSiMe3, PhMeSiClz from PhSiHMeCHzCl and biphenyl; Me3SiCl; MezSiCiz; PhSiCizMe from PhSiMezCHzCl) obtained in the argon flow experiments. 4.5 Conclusions from both nyrolytic methods After noting the lack of starting material volatility correlation with the percent pyrolysis completion in table XX, compound with the blatant correlation of such in table XVIII (along with the smaller magnitude of percent completions observed in XVIII), it is apparent that 560C may have been too low a temperature for the obtainance of any mechanistically interesting data from the X-PhSiHMeCHzCl compounds. Excluding the X = p-Cl silane however, one obtains a Hammett correlation coefficient of 0. 78 with a p value of 0. 79 in a least squares graph of the

PAGE 95

81 relative (to X =H) log(percent completions)* vs ax. *(A rough estimate of the relative rates of pyrolyses with respect to the X = H compound). The presence of the volatile side products in the argon flow experiments, especially of biaryl compounds is indicative of the presence of radical-coupling chemistry.32 Reference 32 in fact, details the former large scale production of biphenyl from the pyrolysis of benzene in a similar apparatus. The presence of electron withdrawing groups (e.g. p-F and m-CF3) on the aromatic ring in the pyrolysis of the X-PhSiMe2CH2Cl compounds was seen to lower the overall pyrolysis rate. This is not unexpected because the unpaired electron involved in a radical mechanism prefers to be in a planar configuration (with the 1t system in this case) with its parent atom. An electron withdrawing group such as CF3 attached to one of the six carbon atoms of the 1t system should statistically lower the reactivity of the aryl radical by enticing the unpaired electron to occupy the tetrahedral sp3 orbitals (in the Ph-CF3 bond) for a finite percent of its lifetime. This delocalization may also explain the low percent completion in the p-F case. [That the phenyl (X= H) case was seen to have the highest percent completion may be explained by the lack of steric problems in the migration of a more localized radical. The lower percent completion of the pMe case (presumably the most localized aryl radical) is probably due only to steric problems. [The second highest percent completion found for the case cannot be explained from either viewpoint unless one wishes to invoke the possibility of radical interactions such as Ph-Cl + Cl -> Ph + C12 *. This may help in the formation of the dichloro silanes obsetved as well as in increasing the reactivity of the aryl radical for migration to a hypothetical Me2SiClCH2 radical, followed by a reversal of* to generate the major p-Cl benzylic product. (Interesting GC-MS results would be found from the pyrolysis of p-35ClPhSiMe2CH237c1 to either refute or accept such a hypothesis).

PAGE 96

82 Further evidence of radical involvement could also be found by a high temperature ESR study (i.e., the presence of an ESR absorption would favor radical processes over anionic mechanisms even though they (the latter) may still be occuring. When I.N. Jung and W.P. Weber31 conducted the sealed tube pyrolysis of HSiMe2CH2Cl at 136C, they found that the main initial products were equimolar amounts of Me3SiH and CISiMeCH2CL Later in the reaction, these products decreased while the proton "migration" product Me3SiCl became the main product. Their results were very reasonably interpreted in terms of two consecutive radical chain sequences represented below by reactions 1 to 5. 1) HSiMe2CH2Cl --> HSiMe2CH2 + Cl 2) HSiMe2CH2 + HSiMe2CH2Cl --> SiMe2CH2Cl + Me3SiH 3) SiMe2CH2Cl + HSiMe2CH2Cl --> HSiMe2CH2 + ClSiMe2CH2Cl 4) Me3SiH + Cl --> Me3Si + HCl 5) ClSiMe2CH2Cl + Me3Si --> Me3SiCl + CH2SiMe2SiCl Simple intramolecular radical shifts between carbon and silicon such as PhSiMezCHzCl ) PhSiMe2CH2 + Cl SiMe2CHzPh SiPhMeEt !cr lcr C1SiMe2cH2Ph ClSiPhMeEt Figure XV Two Step Radical Mechanism for the Pyrolytic Rearrangement of PhSiMe2CH2Cl to ClSiMe2CH2Ph and ClSiPhMeEt

PAGE 97

83 PhSiMe2CH2 --> SiMe2CH2Ph have been shown not to occur by Wilt3 3 in observing them in the liquid phase at 150C. Thus the involvement of a simple two step mechanism such as shown below for the argon flow pyrolysis of PhSiMe2CH2Cl is probably non-existent. The presence of the Me3SiH analogue, PhMe2SiH in our 560C argon flow and 350C sealed tube pyrolyses of PhSiMe2CH2Cl, indicate a similar radical chain mechanism in operation, but no ClSiMe2CH2Cl was found. In fact, for both sets of argon flow pyrolyses, the only side products containing two chlorine atoms were )) Me -si--() Cl Cl ll ) Me-si /\"' I 'ct' Ct + Figure XVI )) ----+> Me -Si -1 Ct + :cH 2 Ct Possible Carbene Elimination Mechanism of ArSiCl2Me formation from the Sealed Tube of Argon Flow Pyrolysis of ArSiMe2CH2Cl dichlorosilanes of the type X-PhSiMeCl2 and X-PhCH2SiHCl2 from the pyrolysis of X-PhSiHMeCH2Cl and X-PhSiMeCl2 and Me2SiCl2 from the pyrolysis of X-PhSiMe2CH2Cl. A possible route for the formation of X-PhSiMeCl2 from the unobserved ClSiMe2CH2Cl analogue, ClSiMePhCH2Cl, may involve direct chloromethyl attack at silicon forming a "ylid" pentacoordinate intermediate followed by carbene elimination (Fig. XVI). Pyrolysis ofvinylchloromethyldimethylsilane may prove carbene involvement if cyclopropylmethyldichlorosilane were to form (Fig.

PAGE 98

) Me -Si-... () Cl Cl XVIT). 6 ) l_ rcH2 Me Si '/ \ ""\ I 'Cl" Cl + Figure XVII 84 ) Me -Sii -...cl + :cHz Cl 1 7 Me -Si-... I Cl Cl Test for a Carbene Elimination Mechanism The formation of such a pentacoordinate "ylid" may only be favorable at higher temperatures and would thus explain the lack of chloromethyl-chlorosilane intermediate found in our studies. The presence of substituted benzenes (in the pyrolysis of X-PhSiHM:eCH2Cl) and of substituted biphenyls (in the pyrolysis of X-PhSiMe2CH2Cl) may be indicative of the high temperature favored disproportionation of H/MeSiPhMeCH2 radicals into silenes, RR'Si=CH2, with loss of either H or Ph. (Loss of Me may explain the trimethylarylsilanes observed in both pyrolysis sets). In observing that the kinetic data obtained from the stirred flow pyrolyses of both X-PhSiHMeCH2Cl and X-PhSiMe2CH2Cl appear to be independent of the aryl substituent chosen and that no observable side products possible resulting from radical processes were found, it is concluded by the author that neither radical nor simple anionic processes are responsible for the clean reactions provided by gas phase stirred flow pyrolysis.

PAGE 99

85 The refutation of a gas phase radical pathway was substantiated in a paper submitted by I.M.T. Davidson30 and S.I. Maghsoodi in 1986. Their work involved the attempted trapping of the suspected HM:e2SiCH2. radical from the stirred flow pyrolysis of HMe2SiCH2Cl in the same apparatus as used in section 4.4. The trapping reagents chosen were 10 fold excesses of methanol, methyl chloride, toluene, butadiene and propene. The kinetics of formation of Me3SiCl were unaffected by any of these well known gas phase radical forming reagents. No new products were detected in any of these experiments nor in others with a 30fold excess of butadiene or a 200-fold excess of propene. The authors concluded that the gas phase rearrangement of HM:e2SiCH2Cl -> Me3SiCl proceeded by the unimolecular formation of a 3 center "chloronium" transition state involving pentacoordi.riate silicon followed by hydride migration to a probable developing positive charge on the chloromethyl carbon: FigureXVlll Unimolecular Mechanism for the Gas Phase Rearrangement of HMe2SiCH2Cl to Me3Sia The degree of charge separation in the proposed three center intermediate remains to be investigated since the alterations of the aryl substituents in experiments 4.3 and 4.4 did not affect observed product ratios or formation rates to any major extent. A bit of fme tuning of the pyrolytic parameters (esp. in studies 4.3) to help the reproducibility of migratory/kinetic data would be of a great asset in the determination

PAGE 100

86 of charge and field effects involved in aryl migration. That others31 and the author of this thesis have found that sealed-tube or liquid drop flow pyrolyses of"chloromethylsilanes produce radical induced volatile side products in addition to the predominant migratory (of R or H) monochlorosilanes, should well confirm the dependence of the reaction mechanism on the molecular population density change occuring when one pyrolysis a strictly gas phase sample as compared to the pyrolysis of a two phase system.

PAGE 101

87 4.6 Syntheses of five substituted ru:ylchloromethylmethylsilanes Before beginning the pyrolyses, it was of course necessary to have samples of each of the substituted arylchloromethylsilanes on hand. The same five aryl substituents were chosen as in Chapter 3. The preparation of each compound involved two steps. (I) A Grignard addition of 0.5 equivalents of the appropriately substituted phenylmagnesium halide to a dilute solution containing one equivalent of chloromethylmethyldimethyl silane (Petrarch Systems, Inc.) in ether. (This should effect a monoarylation, but a certain amount of diarylation occurred in all reactions). (ll) The reduction of the remaining silicon chlorine bond by means of one fourth an equivalent of LiAll4. I. ArBr + Mg -> ArMgBr + Cl2SiMeCH2C1 -> ArSiC1MeCH2Cl + MgBtCl IT. 4 ArSiC1MeCH2Cl + LiAlf4 -> 4 ArSiHMeCH2Cl + LWCl4 4.6.1 Preparation of PhSiHMeCH:zQ I (a) PhBr + Mg -> PhMgBr Approximately 0.2 moles of phenylmagnesium bromide was prepared by the dropwise addition of 0.2 mole (31.4 gr) of bromobenzene (Fischer) to an argon purged 500 ml -3 necked round bottom flask containing a slight excess 0.21 mole (5.1 gr)-of Fischer magnesium turnings under 75 ml of dry ether. It was necessary to add a couple of I2 crystals for the Grignard formation to begin. With gentle magnetic stirring, the bromobenzene was added over a two hour period. The reflux of ether without external heating was observed for an additional hour, after which the Grignard formation was deemed complete.

PAGE 102

88 I (b) PhMgBr + Cl2SiMeCH2Cl -> PhSiC1MeCH2Cl + MgBrCl Approximately 76.4 % of the Grignard solution from I(a) was transferred (via a cooled 50 cc syringe) into a 250 m1 septum sealed addition funnel atop an argon purged 500 m1 -3 necked round bottom flask containing 0.153 mole (25 gr) Petrarch chloromethylmethyldichlorosilane dissolved in 100 m1 of dry ether and a 1.5 inch magnetic stir bar. To the center neck of this flask was attached the argon supplied reflux condensor. The addition of the Grignard was carried out slowly over a 1.5 hour period, during which time, ether reflux and MgBt;Cl precipitation was noted. After 24 hours additional reaction, a small sample of the supernatant above the salts was then submitted for GC-FID analysis which determined that the monoarylation was about 86 % complete with respect to the dichlorosilane starting material. Diarylation occurred, but only to about 11.5 %. II. 4 PhSiC1MeCH2Cl + LiAll4 ->4 PhSiHMeCH2Cl + LiAlCl4 The reaction mixture obtained from I(b) was filtered, under gentle argon pressure, through a dry 24/40 glass fritted tube. The MgBrCl was washed with two 100 ml portions of ether and approximately 1/2 of the ether was removed from the filtrate through a rotary evaporator. The resulting solution was then cooled to -20C as to effect the crystallization of any unreacted Grignard which would iself be reduced with the monochlorosilane by LiAll4. A few Grignard crystals were removed by cold filtration under gentle argon pressure through another 24/40 glass fritted tube. The purified 150 m1 ofmonochlorosilane (approximately 0.132 mole) in ether solution was filtered directly into an argon purged 250 m1 addition funnel atop a 1 litre -3 necked round bottom flask containing a 28 % excess of 1/4 and equivalent of

PAGE 103

89 Aldrich Li.All4 (0.042 mole; 1.6 gr) under 100 ml dry ether and a 2 inch magnetic stir bar. An argon supplied Friedrichs condenser was used to contain the ether reflux under an inert atmosphere. The addition of the monochlorosilane solution was carried out with magnetic stirring over a 2 hour period. Each drop resulted in vigorous ether reflux. The reaction was allowed to continue for 24 additional hours. A small sample taken of the supernatant for GC-FID analysis showed the reaction to be almost quantitative. The excess LiAIH4 and LiAlC4 precipitate were decomposed and/or dissolved from the reaction mixture via an in situ (under argon) dropwise addition of 200 ml saturated aqueous NI4 Cl solution with gentle stirring. The resulting layers were then transferred into a 2 liter separatory funnel and the aqueous lithium salt layer was discarded. The product containing ether layer was dried over anhydrous sodium sulfate and the drying agent was removed by argon pressure filtration through a 24/40 glass fritted tube. The ether was then removed from the filtrate through a rotary evaporator and 25 ml of a yellowish fluid shown by GC-FID to be a 66.6% pure product was collected. Vacuum distillation of this fluid at 3 mm Hg through a 10 em Vigreux column afforded 8.8 gr of a 98.5% pure (via GC-FID) PhSiHMeCH2Cl: MW 170.71; bp 54C I 3.0 mm Hg; yield from Cl2SiMeCH2Cl 33.7 %. Structure verification determined by GC-MS. Relevent ions and (abundances) found were: 35a M+ 170 (2270), 37ci M+ 172 (769); 35ct M+ -Me 155 (1224), 37cl M+-Me 157 (376); M+-CH2CI 121 (49184); M+-Cl 135 (153); 35ct M+ Ph 93 (2976), 37CI M+Ph 95 (2422).

PAGE 104

90 4.6.2 Preparation ofp-MePhSiHMeCH2Q p-MePhBr + Mg -> p-MePhMgBr p-MePhMgBr + C12SiMeCH2Cl -> p-MePhSiCIMeCH2Cl + MgBrCl 4 p-MePhSiC1MeCH2Cl + LiAlH4 -> 4 p-MePliSiHMeCH2Cl + Using a stepwise procedure analogous to that of synthesis 2.A.1, the Grignard reagent of p-bromotoluene was prepared from magnesium turnings (5.1 g, 0.21 mol, Fischer), p-bromotoluene (34.2 g, 0.2 mol, Aldrich) and 100 ml dry ethyl ether. The Grignard reagent was then reacted with chloromethylmethyldichlorosilane (32.7 g, 0.2 mol, Petrarch) in 200 ml dry ethyl ether to form p-MePhSiC1MeCH2Cl. GC-FID analysis at the end of reaction revealed approximately 76 % complete reaction with respect to the starting dichlorosilane. Diarylation occurred only to a meager 1.6 %. The was then reduced with lithium aluminum hydride (2.1 g, 0.055 mol, Aldrich). Vacuum distilled through a 30 em Vigreux column gave p-MePhSiHMeCH2Cl: 15.6 g (42.2% yield, 99% pure (via GC-FID)); bp 60-62C /2.0mmHg. Structure verification determined by GC-MS. Relevent ions and (abundances) found were: 35cl M+ 184 (4897), 37a M+ 186 (1537); 35ci M+ -Me 169 (354), 37ci M+-Me 171 (109); M+-CH2Cl 135 (49720); M+-Cl 149 (194); 35cl M+ p-MePh 93 (9031), 37ci M+ p-MePh 95 (2270); p-MePh+ 91 (7078); toluene cation 92 (5984).

PAGE 105

91 4.6.3 Preparation of p-FPhSiHMeCH2Q p-FPhBr + Mg -> p-FPhMgBr p-FPhMgBr + C12SiMeCH2CI -> p-FPhSiCIMeCH2Cl + MgBrCl 4 p-FPhSiCIMeCH2Cl + Li.All4 -> 4 p-FPhSiHMeCH2CI + LiAlC4 Using a stepwise _procedure analogous to that of synthesis 2.A.1, the Grignard reagent of p-bromofluorobenzene was prepared from magnesium turnings (5.1 g, 0.21 mol, Fischer), p-bromofluorobenzene (35.0 g, 0.2 mol, Aldrich) and 100 ml dry ethyl ether. The Grignard reagent was then reacted with chloromethylmethyldichlorosilane (32.7 g, 0.2 mol, Petrarch) in 200 ml dry ethyl ether to form p-FPhSiCIMeCH2CI. GC-FID analysis at the end of reaction revealed approximately 93 % complete reaction with respect to the starting dichlorosilane. Diarylation occurred to an extent of 10.3 percent The monochlorosilane was then reduced with lithium aluminum hydride (2.1 g,"0.055 mol, Aldrich). Vacuum distillation through a 30 em Vigreux column gave p-FPhSiHMeCH2CI: 16.36 g (43.3 % yield, 99.6% pure (via GC-FIO)); bp 41.6C /1.5 mm Hg. Structure verification determined by GC-MS. Relevent ions and (abunchnces) found were: 35a M+ 188 (9329), 37cl M+ 190 (3091); 35cl M+ -Me 173 (1013), 37ci M+-Me 175 (354); M+-CH2Cl 139 (49984); M+-Cl 153 (790); 35cl M+-p-FPh 93 (2685), 37cl M+-p-FPh 95 (5370)*; 35cl M+ F 169 (110), 37ci M+-F 171 (378); p-F-benzene+ 96 (6215); and/or p-FPh+; phenyl cation 77 (22920).

PAGE 106

92 4.6.4 Preparation of p-ClPhSiHMeCHlCl LiAIC4 p-CIPhBr + Mg -> p-CIPhMgBr p-CIPhMgBr + Cl2SiMeCH2Cl -> p-CIPhSiC1MeCH2Cl + MgBrCl 4 p-ClPhSiC1MeCH2Cl + LiAlH4 -> 4 p-ClPhSiHl\1eCH2Cl + Using a stepwise procedure analogous to that of synthesis 2.A.1, the Grignard reagent of p-bromochlorobenzene was prepared from magnesium turnings (5.1 g, 0.21 mol, Fischer), p-bromochlorobenzene (38;29 g, 0.2 mol, Aldrich) and 100 ml dry ethyl ether. The Grignard reagent was then reacted with chloromethylmethyldichlorosilane (32.7 g, 0.2 mol, Petrarch) in 200 ml dry ethyl ether to form p-ClPhSiC1MeCH2Cl. GC-FID analysis at the end of reaction revealed approximately 81 % complete reaction with respect to the starting dichlorosilane. Diarylation occmred to an extent of 5.1 percent The monochlorosilane was then reduced with lithium aluminum hydride (2.1 g, 0.055 mol, Aldrich). Vacuum distillation through a 30 em Vigreux column gave p-CIPhSiHMeCH2Cl: 12.02 g (29.3% yield, 99.87% pure (via GC-FID)); bp 64C /2.0mmHg. Structure verification determined by GC-MS. Relevent ions and (abundances) found were: 35cl2 M+ 204 (2896), 35cl37 Cl M+ 206 (1554), 37cl2 M+ 208 (358); 35a2 M+-Me 189 (253), 35a37cl M+-Me 191 (147), 37cl2 M+-Me 193 (28); 35ct M+-CH 2 Cl 155 (49720), 37cl M+-CH 2CI 157 (21240); 35cl M+-Cl 169 (693), 37ct M+-Cl 171 (197); 35cl M+-p-CIPh 93 (1512), 37 Cl M+ p-CIPh 95 (343); 35CI p-CIPh+ 111 (686), 37 Cl p-ClPh+ 113 (1448); 35cl p-Cl-benzene+ 112 (769), 37ci p-Cl-benzene+ 114 (388); pheriyl cation 77 (6182).

PAGE 107

4.6.5 Preparation ofm-CF:lPhSiHMeCH20 m-CF3PhBr + Mg -> m-CF3PhMgBr m-CF3PhMgBr + Cl2SiMeCH2C1 -> m-CF3PhSiClMeCH2Cl + MgBrCl 93 4 m-CF3PhSiC1MeCH2Cl + Li.All4 -> 4 m-CF3PhSiHMeCH2Cl + LiA1Cl4 Using a stepwise procedure analogous to that of synthesis 2.A.1, the Grignard reagent of p-bromo-meta-trifluorobenzene was prepared from magnesium turnings (5.1 g, 0.21 mol, Fischer), p-bromometa-trifluorobenzene (45.0 g, 0.2 mol, Aldrich) and 100 m1 dry ethyl ether. The Grignard reagent was then reacted with chloromethylmethyldichlorosilane (32.7 g, 0.2 mol, Petrarch) in 200m1 dry ethyl ether to form m-CF3PhSiClMeCH2Cl. GC-FID analysis at the end of reaction revealed approximately 82 % complete reaction with respect to the starting dichlorosilane. Diarylation occurred to an extent of 29.3 percent. The monochlorosilane was then reduced with lithium aluminum hydride (2.1 g, 0.055 mol, Aldrich). Vacuum distillation through a 30 em Vigreux column gave m-CF3PhSiHMeCH2Cl: 15.27 g (32% yield, almost 100% pure (via GC-FID)); bp 41C I 2.0 mm Hg. Structure verification determined by GC-MS. Relevent ions and (abundances) found were: 35a M+ 238 (193), 37cl M+ 240 (85); 35cl M+Me 223 (609), 37ci M+-Me 225 (184); 35cl M+-F 219 (2960), 37ci M+-F 221 (1135); 35cl M+-CH 2CI 189 (49184), 37ci M+-CH 2 Cl 191 (2928); M+Cl 203 (125); 35a M+ m-CF3Ph 93 (646), 37 Cl M+ m-CF3Ph 95 (1496); CF2-benzene+ 127 (12033); CF3-benzene+ 146 (191); CF3-phenyl cation 145 (606).

PAGE 108

94 REFERENCES 1. Bott, R.W.; Eabom, C.; Rushton, B.M. J. Chern., 1965, J., 458. 2. Damrauer, R.; Danahey, S.E.; Yost, V.E. 1984, lfi6, 7633-7634. 3. Kreeger, R.L.; Menard, P.R.; Sans, E.A.; Shechter, H. Tett. Lett., 1985, .2,6, 1116. 4. Ibid., p.1115. 5. Bott, p. 455. 6. DePuy, C.H.; Bierbaum, V.M.; Damrauer, R. JACS. 1984, 106. 4051-4053. 7. DePuy, C.H.; Bierbaum, V.M.; Flippin, L.A.; Grabowski, J.J.; King, G.K.; Scmidtt, R.J.; Sullivan, S.A. lACS., 1980, 102, 5012-5015. 8. Danahey, S.E. Master's Thesis. Univeristy of Colorado at Denver, 1986, Chapter4. 9. Murphy, M.K.; Beauchamp, J.L. 1977,22.2085-2089. 10. Danahey, Chapter 2. 11. !bid., pgs. 10-12. 12. !bid., pgs. 11-12. 13. Thid . pg.11. 14. IJ:llil., pg. 10. 15. .IJ:llii., pgs. 14-15. 16. l.b.kl., pg. 14. 17. DePuy, C.H.; Bierbaum, V.M.; Damrauer, R., p. 4052. 18. Shorter, S. "Correlation Analysis of Organic Chemistry." Research Studies Press: New York, 1982. 19. Hine, J. "Structural Effects on Equilibrium in Organic Chemistry." Wiley Interscience: New York, 1975.

PAGE 109

95 20. Barlin, G.B.; Perrin, D.O. 0. Rey. Chern. Soc., 1966,1Q, 77-101. 21. Damrauer, R.; Danahey, S.E.; Yost, V.E., p. 7364. 22. Damrauer, R.; Yost, V.E.; Danahey, S.E.; O'Connell, B.K. Organometallics, 1985, 1779. 23. Larson, J.W.; McMahon, J.B. 1985,.1DZ, 766. 24. Kreeger, pgs. 1115-1118. 25. Damrauer, R.; Yost, V.E.; Danahey, S.E.; O'Connell, B.K., p. 1782. 26. Danahey, pgs. 32-33. 27. Yost, V.E. Master's Thesis. University of Colorado at Denyer, 1988, Chapter 3, p. 49. 28. Barlin, G.B.; Perrin, D.O. 0. Rey. Chern. Soc., 1966,l!l, 247. 29. Brook, A.G.; Bassingdale, A.R.; Jones, P.F.; Lennon, J.M. Can. J. Chern .. 1975, 332-337. 30. Davidson, I.M.T.; Ijadi-Maghsoodi, S. 1986, S.. 2087. 31. Jung, I.N.; Weber, W.P. J. Chern., 1976, !1. 946. 32. Brown, R.F.C. "Pyrolytic Methods in Organic Chemistry." New York: Academic Press, 1980, pg. 2. 33. Wilt, J.W.; Kolewe, 0.; Kraemer, J.F. lACS.. 1969, .21.2624.

PAGE 110

APPENDIX A Polynomial Curve Fitting of Order 2 on Column 1 vs 2 Data file: vem.1 Raw Data (p-Me-PhSiMe2CH2Cl-> p-xylene) Point X Chrs) Obs. Y (mmoles) 1 1.5 0.0039 2 3 0.0086 3 4.5 0.0174 4 6 0.0448 5 10 0.1853 6 21.5 0.3055 7 45 0.6614 Statistical Analysis Order of Equation 2 Number of Points 7 Total Sum of Squares 0.3524748 Sum of Squares due to regression Sum of Squares due to Deviation Goodness of Fit 0.9883352 Correlation Coefficient 0.9941505 Equation Coefficients: X"O c -3.888165E-02 X" 1 b 1. 795293E-02 X"2 a -5.409995E-05 Calc. Y (mmoles) -1.207398E-02 1.449025E-02 4.081103E-02 6.688836E-02 0.1352377 0.3220987 0.659448 0.3483632 4.111592E-03 y = -5.40x1o5 x2 + 0.01795 x -0.03888 t(y0 ) = 2.18 hrs (2.176) Diff. 1.597398E-02 -5.89025E-03 -2.341103E-02 -2.208836E-02 5.006232E-02 -1.659873E-02 1.952052E-03 96

PAGE 111

Polynomial Curve Fitting of Order 2 on Column 1 vs 2 Data file: vern.2 Raw Data (PhSiMe2CH2Cl -> toluene) Point X
PAGE 112

Polynomial Cwve Fitting of Order 2 on Column 1vs 2 Data file: vern.3 Raw Data (p-F-PhSiMe2CH2Cl-> p-fluorotoluene) Point X Chrs Obs. Y (mmoles) 1 1.5 0.0193 2 3 0.0821 3 4.5 0.1254 4 6 0.1357 5 10 0.2557 6 21.5 0.3664 7 47 0.5164 Statistical Analysis Order of Equation 2 Number of Points 7 Total Sum of Squares 0.1856957 Sum of Squares due to regression Sum of Squares due to Deviation Goodness of Fit 0.98444 Correlation Coefficient 0.9921895 Equation Coefficients: XJ\0 1.098074E-02 XJ\1 2.353333E-02 XJ\2 -2.734414E-04 t(y0 ) = 0.469 hrs Calc. Y (mmoles) 0.0456655 7 .911977E-02 0.1113436 0.1423369 0.2189699 0.3905491 0.5130154 0.1828063 2.889425E-03 Diff. -0.0263655 2.980225E-03 1.405646E-02 -6.636843E-03 3.673007E-02 -2.414912E-02 3.38459E-03 98

PAGE 113

Polynomial Curve Fitting of Order 2 on Column 1vs 2 Data file: vern.4 Raw Data (p-Cl-PhSiMe2CH2Cl-> p-chlorotoluene) Point X Chrs Obs. Y (mmoles) I Calc. Y (mmoles) Diff. 1 1.5 0.3045 0.3004556 4.044473E-03 2 3 0.2895 0.3460809 -5.658093E-02 3 4.5 0.4559 0.387474 6.842601E-02 4 6 0.4263 0.4246349 1.665145E-03 5 10 0.4843 0.503039 -1.873902E-02 6 28 0.4846 0.4834157 1.184285E-03 Statistical Analysis Order of Equation 2 Number of Points 6 Total Sum of Squares 3.907174E-02 Sum of Squares due to regression 0.0308165 Sum of Squares due to Deviation 8.255243E-03 Goodness of Fit 0.7887158 Correlation Coefficient 0.8880967 Equation Coefficients: X "0 0.2505987 X"1 3.464923E-02 X"2 -9.40511E-04 t(y0 ) = -6.192 hrs 99

PAGE 114

Polynomial Curve Fitting of Order 2 on Column 1vs 2 Data file: vern.5 100 Raw Data Cm-CF3-PhSiMe2CH2Cl -> m-trifluorotoluene) Point X Chrs Obs. Y (mmoles) Calc. Y Cmmoles) Diff. 1 0.5 0.2689 0.3013619 -3.246191-02 (1.5) (0.3283) 2 1 0.2886 0.3150197 -2.641973E-02 3 2 0.3558 0.3412365 0.0145635 4 3 0.3726 0.3659884 6.611616E-03 5 4.5 0.4602 0.4003693 5.983073E-02 6 6 0.4721 0.4314541 4.064593E-02 7 10 0.4268 0.4982321 -7 .143203E-03 8 22.5 0.5645 0.555838 8.662045E-03 Statistical Analysis Order of Equation 2 Number of Points 8 Total Sum of Squares 6.889141E-02 Sum of Squares due to regression 5.647421E-02 Sum of Squares due to Deviation 0.0124172 Goodness of Fit 0.819757 Correlation Coefficient 0.9054043 Equation Coefficients: X "0 0.2873379 X"1 2.841427E-02 X"2 -7.32486E-04 t(y0 ) = 8.326 hrs

PAGE 115

1 01 earance of X-substitut -Ph SiMezQ!2C1 + KF + 18-CrQwn-6 in at85C (mmoles with respect to time) X eaction ime -Me -H -Cl m-CF3_ 0.5 0.2689 1.0 0.2886 1.5 0.0039 0.0061 0.0193 0.3045 2.0 0.3558 3.0 0.0086 0.0227 0.0821 0.2895 0.3726 4.5 0.0174 0.0703 0.1254 0.4559 0.4602 6.0 0.0448 0.0881 0.1357 0.4263 0.4721 0.0 0.1853 0.4843 0.4268 1.5 0.3055 0.3664 I 2.5 0.5645 I 6.0 0.4802 '8.0 0.4846 5 0.6614 7 0.5164

PAGE 116

APPENDIXB ARGON FLOW PYROLYSIS APPARA1US 1 0 ml graduated addition ---+funnel sed silica tube with ate beads or crushed 14/20 tube furnace ----t---20C ice-salt bath ml silane added dropwise over 2 minutes at constant rate argon out, __ ,.. to flow meter variac condensate submitted for immediate GC-F I DIMS analysis 102

PAGE 117

0 :< ..... A z ll.l p. ('t) 0 fill p. fll ..... fll ...:I 0 p:: p. i3= 0 ...:I IX.. A ll.l p:: p:: I=! fll flo\o/ controller Berytron (pressure guege) thermocouple port N 2 inlet -+---oven 1 stirred flo\o/ reactor ( SFR) to vacuum Gas Chromatograph valve sample loop sample Chert Recorder Computer