The steric and electronic effects of substituents on pseudorotational activation energies in acyclic pentacoordinate organosilanes

Material Information

The steric and electronic effects of substituents on pseudorotational activation energies in acyclic pentacoordinate organosilanes
O'Connell, Bonnie Kathleen
Publication Date:
Physical Description:
ix, 83 leaves : illustrations ; 29 cm


Subjects / Keywords:
Physical organic chemistry ( lcsh )
Organometallic compounds ( lcsh )
Organosilicon compounds ( lcsh )
Fluosilicates ( lcsh )
Nuclear magnetic resonance ( lcsh )
Fluosilicates ( fast )
Nuclear magnetic resonance ( fast )
Organometallic compounds ( fast )
Organosilicon compounds ( fast )
Physical organic chemistry ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 81-83).
General Note:
Submitted in partial fulfillment of the requirements for the degree of Master of Science, Department of Chemistry
Statement of Responsibility:
by Bonnie Kathleen O'Connell.

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:
18008625 ( OCLC )

Full Text
Bonnie Kathleen O'Connell
B.A., University of Colorado-Denver, 1984
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

This Thesis for the Master of Science degree by
Bonnie Kathleen O'Connell
has been approved for the
Department of
7/?v / Si

O'Connell, Bonnie Kathleen (M.S., Chemistry)
The Steric and Electronic Effects of Substituents on
Pseudorotational Activation Energies in Acyclic
Pentacoordinate Organosilanes
Thesis directed by Professor Robert Damrauer
The activation energies for the
pseudorotational process of a series of nine acyclic,
dicarbon-substituted fluorosiliconates synthesized via
the Damrauer method have been investigated using
variable-temperature F NMR studies and the complete
bandshape method of analysis. The energy values
obtained in this study indicate that substituent
effects on pseudorotation for these compounds are
small. However, trends are observed that may be
attributed to steric and electronic influences of

There are many people who have helped me along
the way, some of whom are listed here. Steve Danahey,
Roger Simon and Vernon Yostfellow students who taught
me much. Jim Crofter, Mike Milash and Louis Aran, who
willingly offered their expertise. Robert Damrauer, for
whom it has been a privilege and a pleasure to work.
Sandy Meyer, who always believed it was possible. My
parents, who let me know they are proud. Bob and Eric,
who gave love and support even when they were tired of
the whole business.

I. INTRODUCTION................................... 1
1.1 Background Information.................... 1
1.2 Pentacoordination in Silicon Compounds 7
1.3 Exchange Occurring at Pentacoordinate
Silicon................................. 9
1.4 Pseudorotation at Pentacoordinate
Silicon.................................. 12
1.5 Our Work................................. 18
II. EXPERIMENTAL RESULTS.......................... 21
2.1 General Instrumental and Experimental
Methods.................................. 21
2.2 List of Compounds Studied................ 23
2.3 References to In-House Preparations.... 24
2.4 Synthesis of o-Tol2SiF2.................. 25
2.5 Synthesis of [o-To^SiF^ ][K+,18-C-6] 26
2.6 Variable Temperature- F NMR Studies... 27
2.7 Complete Bandshape Analyses (CBS)........ 40
2.8 Estimated AG* for o-Xy^SiF^ ............. 53
2.9 Error Analysis for the CBS Method........ 54
III. DISCUSSION.................................... 60
3.1 Results.................................. 60

3.2 A Hammett Study.......................... 62
3.3 Steric Effects on Pseudorotation........ 72
REFERENCES AND NOTES................................ 81

I. F NMR Data: Temperature-Dependent
Behavior of p-Tol2SiF3 .................... 31
II. F NMR Data: Temperature^Dependent
Behavior of (p-N02Ph)2SiF3 ................. 33
III. F NMR Data: Temperature-Dependent
Behavior of o-To^SiF^...................... 34
IV. F NMR Data: Temperature-Dependent
Behavior of a-Nap2SiF3 ..................... 36
V. F NMR Data: Temperature-Dependent
Behavior of PhXylSiF^ ...................... 37
VI. F NMR Data: Temperature-Dependent
Behavior of t-BuPhSiF^ .....1.............. 38
VII. F NMR Data: Temperature-Dependent
Behavior in 50/50 by Volume Acetone-d6 and
Vinyl Chloride of (p-N02Ph)2SiF3 ........... 39
VIII. CBS Analysis for p-Tol2SiF3~.............. 43
IX. CBS Analysis for (p-N02Ph)2SiF3 .......... 44
X. CBS Analysis for o-Tol2SiF3 .............. 45
XI. CBS Analysis for a-Nap2SiF3_.............. 46
XII. CBS Analysis for PhXylSiF3 ............... 47
XIII. CBS Analysis for t-BuPhSiF3 .............. 48
XIV. CBS Analysis for (p-N02Ph)2SiF3 in
VC/Acetone-d6............................. 49
XV. Log vs 1/T................................. 50

, viii
XVI. Log kr/T vs 1/T.............................. 51
XVII. Activation Parameters for Dicarbon
Substituted Siliconates...................... 52
XVIII. Activation Parameters for the
Dicarbon-Substituted Siliconates
Studied by Our Group......................... 61
XIX. Energy_Barriers to Pseudorotation of
X2SiF2~ Type Siliconates..................... 69
XX. Energy Barriers to Inversion of Martin's
Siliconates.................................. 71
XXI. Bond Lengths (A) and Angles (deg) for
Three Fluorosiliconates...................... 75

I. Plot of log kr vs 1/T for p-To^SiF^-........... 58
II. Plot of log k^/T vs 1/T for p-Tol2SiF3".... 59
III. Atom assignments for Table XXI................ 76

1.1 Background Information
Pseudorotation in pentacoordinate silicon
compounds has received considerable attention in the
past twenty years. This introduction will attempt to
demonstrate reasons for this interest as well as
furnish examples of important work in this area.
Organosilicon compounds were first prepared via
a Grignard reactionthe most commonly used method of
preparationin 1904. Using silicon tetrachloride and
ethylmagnesium iodide, EtSiCl-j, Et2SiCl2/ Et^SiCl and
Et^Si were synthesized.1 Since that time a vast
chemistry involving organosilicon molecules has been
Because silicon occurs just below carbon in the
periodic table, there are some close relationships
between the two elements. Equally important and
interesting are some differences. For our work the most
important difference involves silicon's ability to

expand its valence shell and form stable
pentacoordinate species.
There are several differences between silicon
and carbon at the level of atomic structure; these
account for the ability of silicon to expand its
valence shell:
1. The covalent radius of the silicon atom
is one and one half times that of the
carbon atom thus rendering the silicon atom
more sterically accessible.
2. The positive charge of the nucleus in
the silicon atom is shielded with an
additional shell of 8 electrons and the
electronegativity of silicon is lower than
that of carbon.
3. In contrast to a carbon atom, which does
not have vacant d-orbitals, the outer
valence shell of silicon has vacant 3d
atomic orbitals (AO).
4. The polarizability of silicon is greatly
facilitated by a larger (as compared to the
carbon atom) size of valent AO and,
consequently, by a higher diffuseness of
its electron cloud.
5. Valence electrons of silicon are at a
much greater distance from the nucleus than
in the carbon atom due to which the
ionization potential of the silicon atom,
in spite of a greater nuclear charge, is
about 3.1 eV less than that of the carbon
These characteristics combine to give silicon a
considerably greater ability to form penta- and
hexacoordinate compounds than carbon. Also important is
the fact that silicon forms especially strong bonds
with electronegative elements,^ particularly fluorine

(bond energy = 154 kcal/mol), oxygen (bond energy = 128
kcal/mol), chlorine (bond energy = 113 kcal/mol), and
nitrogen (bond energy = 100 kcal/mol). Most
pentacoordinate silicon compounds contain one or more
bonds involving these elements, and fluoride is
particularly useful as a nucleophile in forming
pentacoordinate species.
The ability of silicon to form hypervalent
compounds is more than a curiosity, however.
Pentacoordinate silicon species play an important role
as intermediates in nucleophilic substitution reactions
at silicon. The nucleophilically induced racemization
as well as hydrolysis (or alcoholysis) of halosilanes
has been found to proceed through higher coordinated
silicon intermediates. In a recent study, Stevenson
and Martin5 found strong kinetic and NMR evidence
indicating that compound I undergoes nucleophilically
catalyzed inversion via a pentacoordinate state

(compound II), which may then undergo an intramolecular
rearrangement accounting for the inversion of the
The pentacoordinate state has interesting
symmetry since five ligands cannot be placed
equivalently in space except in a plane. The two
conformations generally favored for pentacoordinate
species are trigonal bipyramidal (TBP), having two
axial and three equatorial bonds, and rectangular
pyramidal (RP), with one apical and four basal
substituents. Transition metal pentacoordinates have
been found to favor the RP geometry, while
pentacoordinates involving nontransition elements tend
toward TBP geometry.
The nonequivalence of ligand positions in
five-coordinate geometry has led to the use of NMR,
particularly F NMR, as an important stereochemical
probe for pentacoordinate compounds. Fluorine is a very

useful nucleus in NMR due to its high sensitivity. It
also exhibits coupling to the Si isotope (4.7%
relative abundance) with large coupling constants in
the 200 Hz range. In main group TBP compounds,
equatorial fluorines are shielded (i.e., upfield)
relative to the averaged F NMR signal, while axial
fluorines are deshielded (i.e. downfield) relative to
the averaged signal.
7 8
Work done by Muetterties et al., on
fluorophosphoranes (e.g., R3PF2 an(^ R2PF3^ usin9 F
NMRtechniques, clearly indicates that such compounds
have TBP geometry with axial sites occupied by the more
electronegative ligands. These workers also observed
fast intramolecular ligand exchange, presumably via
Berry pseudorotaton, in many of the phosphoranes they
studied. Theoretical calculations and diffraction
studies on TBP phosphoranes show axial bonds to be
longer than equatorial bonds with greater s character
present in the equatorial bonds.
The pseudorotation mechanism was proposed by
Berry in 1960 to account for the spectroscopic
equivalence of the fluorines in PF^. This mechanism
consists of a vibrational bending motion involving both

axial and equatorial ligands proceeding through an RP
transition state.
7 8
Muetterties and coworkers, looking at
evidence from vibrational and NMR studies on several
main group pentacoordinate compounds, concluded that
such "positional exchange will be a common low energy
process in five-coordinate structures." The low energy
barrier of this process reflects the small energy
differences between TBP and RP structures. They further
conclude that such a mechanism may be responsible for
the failure to observe isomers in many substituted
Using vibrational, NMR and normal coordinate
analysis studies, Holmes** concluded that the Berry
mechanism defines an acceptable intramolecular exchange
process for nonrigid PF^, PYF^ and P^F^ compounds.

1.2 Pentacoordination in Silicon Compounds
Many studies on pentacoordinate organosilicon
compounds involve chelating ligands or incorporation of
the central silicon atom into a five-membered ring,
since such structural features stabilize the
pentacoordinate geometry. Silatranes, (compound III,
e.g.) have been extensively studied by Voronkov
and other Russian chemists. X-ray diffraction studies
demonstrate the existence of the transannular Si-N bond
and show that the structure is that of a distorted TBP.
Corriu and coworkers1'^ have looked at
pentacoordinate species of the type shown by compounds
IV and V. The F NMR spectrum at -95C for compound IV
displays two lines of equal intensities at -111 ppm and
-154 ppm. The spectrum for compound V shows a triplet
at -128 ppm and a doublet at -148 ppm with relative
intensities of one to two respectively. These spectra
were done at low temperature in nonpolar media and both

are in good agreement with TBP geometry. There also
exists close similarity between the NMR spectra of
these silicon compounds and a germanium analog studied
by these workers. X-ray investigation of the germanium
compound shows its geometry to be that of a distorted
TBP, which suggests TBP geometry in these
pentacoordinated silicon compounds.
Stevenson, Wilson, Martin and Farnham^ have
done extensive work characterizing siliconates of the
type shown by compound VI, where the bidentate ligands
preferentially stabilize the pentacoordinate relative
to four and six coordinate compounds. X-ray work done

on the phenylsiliconate (Y = Ph) shows the silicon atom
to be pentacoordinate, having a somewhat distorted TBP
geometry. The three carbon atoms occupy equatorial
positions and the oxygen termini occupy the axial
positions. The axial Si-0 bond exhibits the lengthening
relative to the equatorial Si-C bonds expected in TBP
structures. The distortion is toward RP geometry along
the Berry pseudorotational pathway postulated for
intramolecular exchange in pentacoordinate compounds.
Accurate structural data on acyclic
fluorosiliconates has been less easy to obtain.
Recently, Schomburg and Krebs were able to obtain
crystal data for [CgHgCH2N(CH3)3]+[SiF5]~ and
[NMe4]+[Ph2SiF3]~. The overall geometry of the SiF5~
anion is a slightly distorted TBP. The small distortion
in the F(eq)-Si-F(eq) angle may be due to packing
effects. The distortion in the observed TBP geometry in
the Ph2SiF3 anion is probably due to intramolecular
repulsions between the axial fluorines and the ortho
hydrogens on the phenyl rings.
1.3 Exchange Occurring at Pentacoordinate silicon
Pseudorotation is a widely postulated mechanism
for intramolecular exchange at pentacoordinated

silicon. However, it is important to note that any
investigation of an exchange process at silicon is
complicated by the multiplicity of processes possible.
Hydrolysis has been shown to be an important
factor in fluorine exchange. Gibson, Ibbott and
Janzen, looking at the role of trace amounts of H20
and HF in fluorine exchange, found conclusive evidence
that intermolecular fluorine exchange in SiF5 is due
to hydrolysis (eq 1). An NMR spectrum of recrystallized
SiF5" + H20 SiF4OH- + HF (1)
[ (CgHyJ^Nj^SiF^] in CH2C12 solvent at 38C showed no
19 29 .
sign of F- Si coupling. Addition to the sample of
hexamethyldisilazane, which scavenges the HF produced
19 29
by H20, leads to the appearance of F- Si satellites.
Subsequent addition of water to the sample results in
the loss of coupling.
Further studies by Marat and Janzen
demonstrate the notable effect of trace amounts of H20,
HF and CH^OH on increasing the rate of intermolecular
fluorine exchange in CH3SiF4. They conclude that the
active catalyst in this exchange process is HF, which
may be formed by the reaction of H20 or CH^OH with

CH^SiF^ Again, the addition of hexamethyldisilazane
inhibits exchange.
Another possible mechanism of intermolecular
exchange may involve the formation of fluorine bridged
intermediates as pathways to exchange (eg 2). After
\/ I
RSiF, + RSlF. PSi FSi-t-F (2)
3 4 ' I / \
carefully eliminating impurities, Marat and Janzen"^
observed rapid intermolecular fluorine exchange in such
systems as SiF^-SiF^ MeSiF^-MeSiF^ and
PhSiF^-PhSiF^ for example. A mixture of SiF^ and
(Pr)2-SiFg resulted in the formation of the
five-coordinate silicon fluoride. No exchange was
observed for SiF^-SiF4 type systems.
There is also the possibility of exchange due
to other types of intramolecular processes. Corriu has
observed rapid ring opening in molecules of the type
shown by compound VII. At higher temperatures an
equilibrium between open and closed forms results in an
averaging of fluorine signals.

The turnstile mechanism has been proposed as an
alternate mechanism for intramolecular ligand exchange.
MO calculations applied to compounds, however,
indicate that such a mechanism would be a higher energy
process than pseudorotation. This mechanism for
intramolecular exchange may become more important in
more highly substituted phosphoranes, however.^
1.4 Pseudorotation at Pentacoordinate silicon
There is evidence from several studies in the
area of pentacoordinated organosilicon compounds in
favor of the Berry pseudorotational mechanism as an
important contributor in intramolecular ligand
In the first work done concerning
pentacoordinate silicon species in solution, Klanberg
and Meutterties,^ prepared and studied siliconates
SiF^ RSiF^ (where R = aryl and alkyl), and Ph^SiF^

as ammonium salts. These workers found little
dependence on the molar ratio of the reactants
usedeven a large excess of fluoride ion did not
result in the formation of hexacoordinate species.
Impurities, however, were difficult to remove in these
substances and markedly affected melting points.
Still, the low temperature NMR spectra of
SiF5, PhSiF4 and Ph2SiFj in nonpolar media are
analogous to those of isoelectronic fluorophosphoranes.
Ph2SiF2 displays at low temperature a spectrum
consisting of two lines with relative intensities of
two to one, as was found for dialkylfluorophosphoranes,
indicative of a "frozen" TBP geometry. As the
temperature was increased, however, the signals
broadened, disappeared and then reappeared as a
gradually sharpening singlet. Although Klanberg and
Muetterties (K-M) interpreted this data as evidence of
Si-F bond breaking, later work by Damrauer and
Danahey indicates this behavior is consistent with
intramolecular exchange throughout the temperature
Low temperature spectroscopic equivalence of
fluorines is seen in both SiF5 and PhSiF^ TBP
geometry is assumed by analogy to the spectra of

isoelectronic phosphoranes. The observation of a single
narrow fluorine signal is consistent with rapid
intramolecular exchange. Again by analogy with
phosphoranes, Berry pseudorotation is the favored
Distinct broadening of the F resonance
observed for MeSiF^ by K-M at room temperature is
evidence for the occurence of intermolecular exchange
phenomena. However, at low temperature,
pentacoordination and pseudorotation are supported by
the spectroscopic equivalence of fluorines in F NMR
spectra (analogous to MePF^) accompanied by Si-F
Farnham and Harlow observed the temperature
dependent behavior of compound VIII between -15C and
70.3C. The spectra, which were concentration
independent as well as generally solvent independent,
show rapid interchange of the diastereotopic CF^ groups

at higher temperatures. The addition of
hexamethyldisilazane or hexamethylphosphoramide to
remove 1^0 or HF did not change the exchange rate.
Visual fit" of observed and calculated spectra allowed
* 29
calculation of the AG = 16.6 kcal/mol. Si NMR
indicated no significant Si-F intermolecular exchange.
The authors consider the Berry pseudorotation process
to be the most reasonable enantiomerization mechanism,
although intramolecular breaking of the Si-0 bond
leading to exchange could not be strictly eliminated.
19 1
Corriu and coworkers saw H NMR evidence
implying pseudorotational behavior in compound IX. Over
the temperature range of -90C to 31C, the spectra for
the NMe2 group show two coalescences. Thefirst, where
four peaks coalesce into two, corresponds to a dynamic
process at silicon with no bond breaking. Ag* for this
process equals 9.4 kcal/mol. The second coalescence
occurs due to the breaking of the Si-N bond accompanied

by rotation around the C-N bond'/ and AG* equals 11.8
kcal/mol. Additional evidence for the intramolecular
19 29
exchange process is that F- Si coupling is observed
at all temperatures. The Si chemical shift is
consistent with the pentacoordinate state and is
temperature dependent. Also, the temperature variable
F studies give a similar Ag equaling 9.3 kcal/mol.
20 21
Holmes, Day, Sau and Holmes have done an
important structural study to characterize the solid-
state distortion coordinate of five-coordinated silicon
compounds. Compounds X and XI were used in this study.
Both should be considerable distorted toward RP
geometry due to the presence of bidentate, cyclic
ligands with identical attachments to silicon involving
the electronegative oxygen atoms. The RP geometry is
also favored by the H-bonding capability of the cation
with the ring oxygens of the anion. In compound XI the
RP geometry is further aided by the presence of

unsaturated ring systems with electron withdrawing
substituents which allow for greater electron
delocalization and thus less repulsion in the RP
Using x-ray crystallography these workers found
that the range of angles around silicon indicates that
the crystal structures of each of these compounds
represent a uniform distribution of solid-state
structures between the TBP and RP geometries. The range
of bond lengths observed also supports this conclusion.
The implication here is that the energy difference
between the TBP and RP geometries in pentacoordinate
organosilicon compounds is small, and thus
pseudorotation, which proceeds through a RP
intermediate, should exist for such molecules.
As noted earlier (see p. 8), Martin and
Stevenson^ have investigated the role of
pentacoordinate intermediates (compound II, e.g.) in
nucleophile-catalyzed inversion. There are several
pieces of evidence for the involvement of
pseudorotation in the inversion of the siliconates
catalyzed by p-(dimethylamino)benzaldehyde or
p-tolualdehyde. The rates of inversion are independent
of solvent polarity, arguing against a mechanism

involving the heterolysis of the Si-0 bond. Also, a
correlation between AG* and Taft inductive constants
for the nucleophiles used indicates that the Ag*'s are
lowered (and hence the rate of reaction increased) by
electron withdrawing groups on the nucleophile. Such an
observation is consistent with maintenance of the
pentacoordinate state throughout the inversion.
Finally, inversion via hexacoordination between the
siliconate and the nucleophile may be ruled out on the
basis of the insensitivity of the process to differing
nucleophile concentrations.
1.5 Our Work
Work done in this lab by Damrauer and Danahey
(D-D) in the area of pentacoordination in organosilicon
compounds has yielded several important related
A new technique for the synthesis of
pentacoordinated organosilicon compounds has been
developed. This synthesis involves the use of equimolar
amounts of the fluorosilane, KF and 18-crown-6 stirred
in toluene at room temperature to form the 18-crown-6
potassium salts of the siliconate. Siliconates
synthesized by this method are evidently much less

hygroscopic than similar siliconates synthesized by
earlier methods, and are stable for long periods
under atmospheric conditions. The presence of
hexamethyldisilazane had no apparent effect on behavior
as observed by NMR, even when the spectra indicated the
presence of small amounts of impurities.
Specific pentacoordinates synthesized and
studied by D-D include Ph2SiF3 MePhSiF^ PhSiF^ and
Ph2SiF3 was previously studied by Klanberg and
Muetterties (K-M), however several new findings were
19 29 .
reported by D-D. F- Si coupling up to -60C was
19 19
noted as well as F- F coupling at -53.7C. The
linearity of the plot of log(k/T) vs 1/T gives good
evidence that a single exchange process is occurring
over the entire temperature range from -77C to 38C.
The observed loss of coupling at higher temperatures,
therefore, is due not to intermolecular exchange as
postulated by K-M, but merely to the signal broadening
which occurs around coalescence.
Similar to Ph2SiF3~ in terms of low temperature
limit structure and dynamic properties, MePhSiF3 is
the first reported acyclic, alkyl, dicarbon-substituted

Ph2SiF2 is the first reported acyclic
tricarbon-substituted siliconate. This compound shows
no temperature dependent behavior with a single F
signal observed over a wide temperature range. This
implies a "frozen" TBP geometry with fluorines axial
and phenyl group in equatorial positions.
Finally, PhSiF^- shows no temperature dependent
behavior, with a single F signal over a wide
temperature range indicating rapid intramolecular
Our current work involves a series of complete
bandshape analyses (CBS) on several more siliconates
synthesized in the above manner. Calculation of the
activation parameters of the pseudorotation process and
the effect of substituents thereon is discussed.

2.1 General Instrumental and Experimental Methods
All commercial reagents used were obtained from
Aldrich Chemical Company or Petrarch Systems. Toluene
was distilled from CaH2 and stored over 4 A molecular
sieves. Anhydrous diethyl ether from previously
unopened containers was used in all Grignard reactions.
Anhydrous reagent grade KF (MCB Chemicals) was stored
in an oven at 120C, cooled and weighed immediately.
The glassware used was oven dried for a minimum of 24
Melting points were determined using a Mel-Temp
(Laboratory Devices) apparatus.
A Perkin-Elmer 3020B chromatograph with a flame
ionization detector was used for all gas
chromatographic analysis. An 1/8 inch by 6 foot 10%
OV-1 silicone oil on Chromosorb WHP packed .column was

GC-MS analysis was performed on a
Hewlet-Packard 5890A Gas Chromatograph interfaced to a
5790 Series Mass Selective Detector. The column was a
-30 meter crosslinked methyl silicone high performance
capillary type.
All NMR experiments were performed on an IBM NR
80/B spectrometer-computer system equipped with 32 K of
available central memory/ a variable-temperature unit
and tuneable multinuclear capability. Chemical shifts
are reported on the 6 scale in parts per million with
reference compounds being tetramethylsilane (TMS) for
and ^C and fluorotrichloromethane (CFCl^) for
NMR. 5 mm o.d. tubes were used for all samples with the
concentrations typically 5-10% silane or siliconate in
acetone d-6.
Calibration of the probe temperature was
accomplished using a Keithley Model 871 Digital
Thermometer with a chromel/alumel thermocouple. The
thermocouple was placed in a 5 mm o.d. tube containing
the standard amount of acetone (no attempt was made to
hold lock during this process), and the tube was placed
in the probe. The probe temperature was altered
gradually, and the tube temperature was allowed to
equilibrate 10 minutes at each probe temperature before

the reading was taken. ^(g) was used to spin the tube
during all' low temperature work to avoid the formation
of ice crystals due to water in the air line.
The chemical shift of CFCl^ was calibrated
prior to each F variable-temperature run using a 50%
by volume solution of CFCl^ in acetone d-6.
All the spectra were recorded in the FT mode
using an internal deuterium lock. The 16K data vector
was used for most spectra taken. Typically, 16
transients were collected for spectra and between
176 and 2400 transients were collected for dynamic F
work. C spectra were obtained under proton
heteronuclear decoupling conditions.
The computer simulation was done with the
EXC2.F77 program using a PRIME 9950 mainframe computer
available through UCD Computing Services. The 9950
system includes 8 megabytes of memory and 1650
megabytes of disk storage. Plotting the simulated
spectra was done via a CALCOMP 1044 36", 8-pen plotter.
2.2 List of Compounds Studied
The following dicarbon-substituted siliconates,
all K+,18-crown-6 salts, were used in this study. The

formulae noted in brackets are used in subsequently
referring to these compounds.
Di-o-tolyltrifluorosiliconate [o-Tol2SiF3~]
Diphenyltrifluorosiliconate [Ph2SiF3 ]
Di-p-tolyltrifluorosiliconate [p-Tol2SiF3~]
Di-o-xylyltrifluorosiliconate [o-Xyl2SiF3 ]
Phenylxylyltrifluorosiliconate [PhXylSiF3~]
22 -
Methylphenyltrifluorosiliconate [MePhSiF3 ]
2.3 References to In-House Preparations
Several compounds used in this study were
synthesized and characterized by other members of our
group. The following is a list of these compounds
accompanied by the appropriate laboratory notebook
p-Tol2SiF2: RS-1-77, RS-1-107
p-Tol2SiF3_: RS-1-111, RS-1-119
(p-N02Ph)2SiF2: RD-18-119

(p-N02Ph)2SiF3 : SD-3-15
a-Nap2SiF2 j RS-1-59
a-Nap2siF3 : RS-1-75
t-BuPhSiF2: V.Y.-II-11/9/85
t-BuPhSiF3 : SD-2-95
PhXylSiF2: RS-1-81
PhXylSiF3: RS-1-85
o-Xyl2SiF2: RD-19-37, SD-2-101
o-Xyl2SiF3~ : SD-2-97, 123, 127
2.4 Synthesis of o-Tol2SiF2
(BO-1-125, BO-1-127)
2-bromotoluene (10.0 g, 58.0 mmol) was placed
in a dropping funnel and added slowly to a round-
bottomed, three-necked flask equipped with a magnetic
stirring bar and reflux condenser. The flask contained
magnesium turnings (1.55 g, 64.0 mmol) covered with
diethyl ether. The flask and funnel were kept under
argon gas atall times. The mixture was stirred
throughout the addition, and stirring continued for
about 1.5 hours after the addition of bromide was
completed. At this time, approximately 3.0 g of SiF^
was bubbled into the reaction mixture, and the mixture
was stirred for about 30 minutes. Work up was done

using saturated NH^Cl solution, and the organic layer
was dried over MgSO^. Analysis of the organic layer by
GC-MS shows predominantly o-TolSiF2. 2.7 ml of
essentially pure silane (determined by GC) was
collected using vacuum distillation between 99-108C at
1.1 mm pressure.
119 13
The silane was characterized by H, F NMR and C NMR
as follows:
NMR: shift(ppm) [multiplicity, area, assignment]
2.404 [s, 3, CH3]
7.33-7.75 [m, 4, Ar]
^F NMR: shift(ppm) [multiplicity]
-137.21 [s (Jpsi = 295 Hz)].
C NMR: shift(ppm) [multiplicity, assignment]
23.13 [s, CH3]
126.87-145.69 [5 peaks easily visible, Ar]
2.5 Synthesis of [o-To^SiF^ ][K+,18-C-6]
1.06 g 18-crown-6 (4.0 mmol) and 0.23 g KF (4.0
mmol) were placed in a small bottle containing 8-10 ml
toluene and equipped with a magnetic stirring bar. The
mixture was stirred briefly, o-Tol2SiF2 (1.0 g, 4.0
mmol) was added and the bottle was capped. The reaction

mixture was allowed to stir for a week, at which time
the resulting precipitate (1.6 g, 70.1% yeild) was
gravity filtered, washed with diethyl ether and air
dried. The material obtained shows softening between
154 and 165C but does not melt sharply. The siliconate
1 19 13
was characterized by H, F and C NMR as follows:
NMR: shift(ppm) [multiplicity, area, assignment]
2.479 [s, 6, CH3]
3.539 [s, 29, 18-C-6 ]
6.986-7.603 [m, 10, Ar]
^F NMR: shift(ppm)
-95.97, broad
C NMR: shift(ppm) [multiplicity, assignment]
23.86 [s, CH3]
70.81 [s, OCH2]
123.96-142.65 [5 peaks easily visible, Ar]
2.6 Variable Temperature F NMR Studies
Acetone d-6 was used as the solvent in all but
one of the variable-temperature NMR studies reported
here. The exception is clearly noted.
Due to the difficulties encountered in
attempting to crystallize these siliconates, most were
studied in the crude state. (Work is now being done by

a member of our group to develop reliable methods for
recrystallization of these compounds. ) Although
impurity peaks were noted in some spectra, linearity in
the complete bandshape analysis was maintained, and
periodic use of hexamethyldisilazane did not alter the
The PhXylSiF^ spectra showed several small
impurity peaks located between the axial and equatorial
siliconate signals which barely integrated at fast or
slow exchange. Although these peaks appeared more
significant around coalescence, they did not
significantly affect the siliconate signal lineshape or
signal strength.
A narrow high field impurity peak in the
t-BuSiF3~ spectra did adversely affect the siliconate
signal strength around coalescence. Compensation for
this effect included increasing the number of
transients collected as well as plotting the spectra so
that the impurity peaks went off scale around
A similar situation existed for the coalescence
region spectra of a-Nap2SiF2_ where two impurity peaks
appear close to the equatorial F signal. Similar

measures to those discussed in relation to t-ButPhSiF^
were taken.
The rest of the siliconates studied appeared
relatively pure by NMR evidence.
Tables I-VII summarize the variable temperature
F data collected for this group of siliconates. The
abbreviation BWHM used in the tables means "band width
at half maximum."
All but one compound studied show similar
variable temperature behavior. At low temperature
(approximately 200 K), two narrow peaks indicating
nonequivalent fluorines are present.. As the temperature
is increased these signals broaden, eventually to the
point where they are too broad to observe. At
coalescence a single F signal appears, the chemical
shift of which is a weighted average of the shifts of
the slow exchange signals. This peak progressively
narrows as the temperature is further increased.
The one exception to this pattern is noted in
the variable temperature behavior of Xy^SiF^- (not
reported in a table). We have taken this compound up to
355 K (82C) without seeing coalescence. Unfortunately,
above this temperature the siliconate rapidly

decomposes, making futher high temperature study
There are several factors which lead us to
believe that these siliconates undergo an
intramolecular exchange process, presumably via
pseudorotation. These compounds exhibit very similar
temperature dependent behavior to that displayed by
Pl^SiF^ which has been discussed in previous work in
17 22
relation to intramolecular exchange processes. The
insensitivity of the NMR spectra of these compounds to
the presence of hexamethyldisilazane and the linearity
of the plots of log kr vs 1/T and log kr/T vs 1/T also
support this idea. Although most of the compounds we
have studied by CBS analysis do not have sufficiently
19 29
narrow peaks at room temperature for F- Si coupling
to be visible, (p-NC^Ph)2SiF3~ spectra do display
coupling around room temperature. This evidence
eliminates the possiblity of intermolecular exchange
processes at higher temperatures in this compound.

Table I. F NMR Data: Temperature-Dependent
Behavior of p-Tol2SiF3 .
-3 o O Apical F's Averaged F's Equatorial F
00 VO -100.00 ppm BWHM=7.7 4Hz J(FSi)=251.OHz Integral=2 -132.14 ppm BWHM=8.60Hz J(FSi)=206.3Hz Integral=l
-58.5 -99.89 ppm BWHM=10.0HZ -133.26 ppm BWHM=15.2Hz
J* 00 a\ -99.74 ppm BWHM=20.6HZ -133.41 ppm BWHM=37.8HZ
-38.7 -99.60 ppm BWHM=57.6HZ -133.55 ppm BWHM=111.0Hz
00 00 (N 1 -99.55 ppm BWHM=139.OHZ -133.63 ppm BWHM=216.OHZ
-23.9 -99.62 ppm Broadening -133.32 ppm Broadening
CO VO -99.7 ppm Broadening -132.8 ppm Broadening
-14.0 -100.6 ppm. Broadening Cannot determine
-9.1 -103.3 ppm Coalescing
-4.1 -108.2 ppm Coalescing
0.9 -109.6 ppm Narrowing
10.8 -110.2 ppm Narrowing

Apical F's_____Averaged F's Equatorial F
-110.53 ppm
-110.48 ppm

Table II. 19f nmr Behavior Data: Temperature of (p-N02Ph)2SiF -Dependent 3 *
T(C) Apical F's Averaged F's Equatorial F
-87.7 -98.59 ppm BWHM=10.3Hz J(FSi)=256.2Hz Integral=2 -134.13 ppm BWHM=18.1Hz J(FSI)=201.2Hz lntegral=l
-77.6 -98.56 ppm BWHM=31.8Hz -134.36 ppm BWHM=55.0Hz
-67.5 -98.66 ppm BWHM=110.0Hz -160.23 ppm BWHM=185.7Hz
-57.3 -98.89 ppm Broadening -135.1 ppm Broadening
-47.2 -103.8 ppm Coalescing
-42.1 -108.5 ppm Coalescing
-37.0 -110.2 ppm Narrowing
-32.0 -110.6 ppm BWHM=462Hz
-27.0 -110.6 ppm BWHM=335Hz
-16.8 -110.7 ppm BWHM=146Hz
-6.7 -110.7 ppm BWHM-67.0Hz
3.4 -110.66 ppm BWHM=33.5Hz
13.6 -110.66 ppm BWHM=17.2Hz J(FSi)=232.1Hz

Table III. F NMR Data: Temperature-Dependent
Behavior of o-Tol2SiF3 .
T( C) Apical F's Averaged F's Equatorial F
r- t~~ 00 1 -78.59 ppm BWHM=8.6HZ J(FSi)=256.2Hz Integral=2 -129.73 ppm BWHM=10.5Hz J(FSi)=214.9Hz Integral=l
-77.6 -78.57 ppm BWHM=11.18HZ J(FSI)=256.2Hz -129.72 ppm BWHM=15.5HZ J(FSi)=218.4HZ
-67.5 -78.80 ppm BWHM=26.6Hz -129.98 ppm BWHM=44.7Hz
-57.3 -78.92 ppm BWHM=69.6Hz -130.06 ppm BWHM=122.1HZ
-47.2 -79.04 ppm BWHM=198HZ -130.25 ppm BWHM=332HZ
-37.1 -80.1 ppm Broadening -127.1 ppm Broadening
-32.0 -83.5 ppm Broadening Cannot determi:
-27.0 -84.5 ppm Coalescing
-21.9 -87.4 ppm Coalescing .
-16.8 -95.0 ppm Coalescing
-6.7 -95.1 ppm BWHM=415HZ
3.4 -95.9 ppm BWHM=295Hz

Table III (continued).
T(C) Apical F's Averaged F's Equatorial F
13.6 -96.2 ppm BWHM=134HZ
23.7 -96.24 ppm BWHM=82.5Hz
33.8 -96.28 ppm BWHM=59.3HZ

Table IV: 19F NMR Behavior Data: Temperature- of Cl-Nap2SiF3 . -Dependent
T(C) Apical F's Averaged F's Equatorial F
-88.2 -70.13 ppm BWHM=12.47HZ J(FSi)=261.3HZ Integral=2 -127.06 ppm BWHM=16.43HZ Integral=l
-78.3 -70.11 ppm BWHM=23.28HZ -127.22 ppm BWHM=40.63HZ
-68.4 -70.25 ppm BWHM=67.0Hz -127.50 ppm BWHM=170Hz
-58.5 -70.5 ppm Broadening -127.6 ppm Broadening
-48.6 -71.1 ppm Broadening Cannot determine
-38.7 -73.8 ppm Broadening Cannot determine
CO 00 CM 1 85.8 ppm Coalescing
-18.9 -89.3 ppm Coalescing
-9.1 -89.1 ppm Narrowing
0.9 -89.53 ppm BWHM=163.0Hz
10.8 -89.95 ppm BWHM=141.7HZ
20.7 -89.79 ppm BWHM=73.0Hz
30.6 -89.91 ppm BWHM=61.0HZ

19 i
Table V. F NMR Data: Temperature-Dependent
Behavior of PhXylSiFg .
T( C) Apical F's Averaged F's Equatorial F
-73.7 -88.83 ppm BWHM=6.8Hz J(FSi)=261.3HZ Integral=2 -127.33 ppm BWHM=8.6HZ J(FSi)=214.9Hz Integral=l
-63.1 -88.78 ppm BWHM=6.8Hz -127.42 ppm BWHM=8.6HZ
-52.5 -88.76 ppm BWHM=10.3Hz -127.56 ppm BWHM=17.2HZ
-42.0 -88.73 ppm BWHM=24.9Hz -127.74 ppm BWHM=47.2HZ
-31.5 -88.70 ppm BWHM=62.8HZ -127.85 ppm BWHM=118.6Hz
-20.9 -88.70 ppm BWHM=162HZ -127.90 ppm BWHM=290HZ
-10.3 -88.9 ppm Broadening -127.2 ppm Broadening
0.2 -90.8 ppm Broadening Cannot determine
5.5 -93.0 ppm Coalescing
10.8 -96.3 ppm Coalescing
16.1 -98.8 ppm Narrowing
26.6 -101.1 ppm BWHM=483Hz
61.5 -101.5 ppm BWHM=105Hz

VI. F NMR Data: Temperature
Behavior of t-BuPhSiF^ .
Apical F1 s____Averaged F's
-101.63 ppm
-101.64 ppm
-101.68 ppm
-101.65 ppm
-102.7 ppm
Very broad
-113.2 ppm
-113.5 ppm
-114.1 ppm
-114.5 ppm
Equatorial F
-135.68 ppm
-135.80 ppm
-135.99 ppm
-135.9 ppm
Cannot determine

VII. F NMR Data: Temperature-Dependent
Behavior in 50/50 by Volume Acetone-d6 and
Vinyl Chloride of (p-NC^Ph)2Sip3
Apical F's_____Averaged F's Equatorial F
-98.09 ppm
-98.15 ppm
-98.29 ppm
-98.4 ppm
-99.0 ppm
-102 ppm
-106 ppm
-110.1 ppm
-110.4 ppm
-110.5 ppm
-134.36 ppm
-134.48 ppm
-134.64 ppm
-134.6 ppm
Cannot determine
-110.52 ppm

2.7 Complete Bandshape Analyses (CBS)
Our CBS studies involved the visual fitting of
computer simulated spectra with actual spectra taken
over a wide temperature range for each of the compounds
noted in Tables I-VII. Spin-spin coupling was ignored
in the fit process due to its small value compared with
the large chemical shift differences found in our F
spectra. The variable in the simulation program is T,
the mean lifetime value, which is the inverse of the
rate constant for intramolecular exchange. The
determination of a series of rate constants over a wide
range of temperatures for a particular siliconate
allows the calculation of the thermodynamic quantities
for the intramolecular exchange process of that
compound. The equations used for these calculations are
discussed below.25,26
The fundamental equation used in determining
the free energy of activation is the Arrhenius equation
(eq 3a,b):
k = Aexp(-E /RT) (3a)
L 3
log k = (-E/2.303R)(1/T) + log A
J. 3

where: R = 1.987 cal/K-mol = the gas constant.
Plotting log kr vs 1/T for a series of rate constants
determined at various temperatures allows the
determination of the activation energy (Ea) for the
exchange process. In solution Eg = AH* + RT.
The Eyring equation, derived from statistical
mechanics, is used to determine AH* and As* (eq 4a,b):
k = (kaT/h)exp(-AH*/RT)exp(-As*/R) (4a)
r d
log kr/T = (-Ah*/2.303R)(1/T) + (4b)
(AS*/2.303R) + log (k0/h) +
log K
where: R = 1.987 cal/K-mol = the gas constant
kn = 1.381E-16 erg/K = Boltzman's constant
h = 6.626E-27 erg-sec = Plank's constant
K = the transmission coefficient, here taken
as 1.
Thus, a plot of log k^/T vs 1/T for a series of rate
constants at given temperatures yields the values for
the enthalpy and entropy of activation. These values

are then used (with a specific temperature value) to
calculate Ag* where Ag* = AH* TAS*.
Tables VIII-XIV display the values obtained for
log kr/ log k^/T and 1/T from the computer simulation
studies done on this group of siliconates. The slope
(AX/AY), intercept (X=0), correlation coefficient (r),
and standard deviations for the scatter of points
around the best fit line (a(X) and O(Y)) for the log kf
vs 1/T and log kr/T vs 1/T plots appear in Tables XV
and XVI. Table XVII lists the activation parameters
obtained for the siliconates studied.

Table VIII. CBS Analysis for p-To^SiFg .
T(C) T(K) 1/TE3
-48.6 224.5 4.459
-38.7 234.4 4.266
-28.8 244.3 4.093
-23.9 249.3 4.011
-18.9 254.2 3.934
o ii i 259.2 3.858
-9.1 264.1 3.786
-4.1 269.1 3.716
0.9 274.0 3.650
10.8 283.9 3.522
20.7 293.8 3.404
30.6 303.7 3.293
T E5 kE-3 log k
1.30E3 0.0769 1.886
320 0.312 2.494
110 0.909 2.958
55.0 1.82 3.260
32.0 3.12 3.494
20.0 5.00 3.699
15.0 6.67 3.824
10.0 10.0 4.000
7.50 13.3 4.124
3.30 30.3 4.481
1.70 58.8 4.769
0.900 111 5.045

log k/T

Table IX. CBS Analysis for (p-NC^Ph^SiF^ .
T ( C) T(K) 1/TE3
-77.6 195.5 5.115
-67.5 205.7 4.861
-57.3 215.8 4.634
-47.2 225.9 4.427
-42.1 231.0 4.329
-37.0 236.1 4.235
-32.0 241.1 4.148
-27.0 246.2 4.062
-16.8 256.3 3.902
-6.7 266.5 3.752
3.4 276.6 3.617
TE5 kE-3 log k
800 0.125 2.097
150 0.667 2.824
45.0 2.22 3.346
15.0 6.67 3.824
11.0 9.09 3.958
6.00 16.7 4.223
3.90 25.6 4.408
2.30 43.5 4.638
1.10 90.9 4.958
0.410 244 5.387
0.180 555 5.744
log k/T

X. CBS Analysis for o-Tol^iF^ .
T(K) 1/TE3 TE5 kE-3 log k log k/T
195.5 5.115 5.30E3
205.7 4.861 990
215.8 4.634 260
225.9 4.427 82.0
236.1 4.235 28.5
241.1 4.148 16.0
246.2 4.062 10.0
251.3 3.979 8.30
266.5 3.752 2.50
276.6 3.617 1.20
286.7 3.488- 0.530
296.8 3.369 0.280
307.0 3.257 0.200
1.89E-3 0.276 -2.015
0.101 2.004 -0.309
0.385 2.585 0.251
1.22 3.086 0.732
3.51 3.545 1.172
6.25 3.796 1.414
10.0 4.000 1.609
12.0 4.079 1.679
40.0 4.602 2.176
83.3 4.921 2.479
189 5.276 2.819
357 5.553 3.080
500 5.699 3.212

Table XI. CBS Analysis for a-Nap2SiF3~.
T( C) T(K) 1/TE3 T E5 kE-3 lag k log k/T
-78.3 194.8 5.133 1.80E3 0.0556 1.745 .-0.544
-68.4 204.7 4.885 300 0.333 2.522 0.211
in 00 in l 214.6 4.660 80.0 1.25 3.097 0.765
-48.6 224.5 4.454 33.0 3.03 3.481 1.130
i u> 00 ~~J 234.4 4.266 13.0 7.69 3.886 1.516
-28.8 244.3 4.093 4.00 25.0 4.398 2.010
-18.9 254.2 3.934 2.50 40.0 4.602 2.197
-9.1 264.1 3.786 1.30 76.9 4.886 2.464
0.9 274.0 3.650 0.600 167 5.223 2.785
10.8 283.9 3.522 0.310 322 5.509 3.055
20.7 293.8 3.404 0.190 526 5.721 3.253

Table XII. ' CBS Analysis for PhXylSiF3
T( C) T(K) 1/TE3 T E5 kE-3 log k loq k/T
-31.5 241.7 4.138 270 0.370 2.568 -0.185
-20.9 252.2 3.964 95.0 1.05 3.021 0.619
-10.3 262.8 3.805 40.0 2.50 3.398 0.978
0.222 273.4 3.671 20.0 5.00 3.699 1.262
5.50 278.6 3.589 13.5 7.69 3.886 1.441
10.8 283.9 3.522 10.0 10.0 4.000 1.547
16.1 289.2 3.458 6.80 14.7 4.167 1.706
26.6 299.8 3.336 3.90 25.6 4.409 1.931
61.5 334.6 2.988 0.680 147 5.167 2.643

Table XIII. CBS Analysis for t-BuPhSiF^ .
T(C) T(K) 1/TE3 T E5 kE-3 log k log k/T
l o h-1 203.1 4.924 950 0.105 2.021 -0.286
-59.9 213.2 4.690 250 0.400 2.602 0.273
-19.1 254.1 3.935 8.00 12.5 4.097 1.692
-8.9 264.3 3.784 4.70 21.3 4.328 1.906
1.3 274.5 3.643 2.50 40.0 4.602 2.164
11.5 284.6 3.514 1.50 66.7 4.824 2.370

Table XIV. CBS Analysis for (p-NO~Ph)-SiF., in
VC/Ace-d6. ^ z j
T(C) T(K) 1/TE3 T E5 kE-3 log k loq k/T
-84.2 188.9 5.293 1.60E3 0.0625 1.796 -0.480
-73.7 199.5 5.014 280 0.357 2.553 0.253
-63.1 210.0 4.761 87.0 1.15 3.060 0.738
-52.6 220.6 4.534 28.0 3.57 3.553 1.209
-47.3 225.8 4.428 15.0 6.67 3.824 1.470
-42.0 231.1 4.326 11.0 9.09 3.959 1.595
-31.5 241.7 4.138 4.50 22.2 4.347 1.963
(Ti O CM 1 252.2 3.964 1.90 52.6 4.721 2.319
-10.3 262.8 3.805 1.00 100 5.000 2.580
0.22 273.4 3.671 0.450 222 5.347 2.910

Table XV. Log kr vs 1/T.
Siliconate Ay/Ax o ll X! r CT(X) a(Y)
p-To^SiF^- -2.666E3 13.89 0.998 2.3E-5 0.054
(p-N02Ph)2SiF3" -2.360E3 14.23 0.999 1.9E-5 0.032
o-Tol2SiF3 -2.210E3 13.30 0.999 2.1E-5 0.049
a-Nap2siF3~ -2.233E3 13.39 0.998 3.8E-5 0.060
PhXylSiF3 -2.239E3 11.89 0.999 1.8E-5 0.028
t-Bu2SiF3 -1.972E3 11.79 0.999 2.4E-5 0.046
In VC/Ace-d6: (p-N09Ph)-SiF, -2.124E3 13.15 0.999 2.4E-5 0.056

Table XVI. Log kr/T vs 1/T.
Siliconate Ay/Ax X=0 r CT(X) a(Y)
p-To^SiF^ -2.553E3 11.04 0.997 3.1E-5 0.062
(p-N02Ph)2SiF3 -2.269E3 11.43 0.999 2.0E-5 0.032
0-Tol2SiF3 -2.201E3 10.46 0.999 2.2E-5 0.051
a-Nap2SiF3 -2.219E3 10.57 0.977 4.1E-5 0.088
PhXylSiF3" -2.326E3 9.72 0.987 4.9E-5 0.12
t-Bu2SiP3" -1.868E3 8.98 0.999 2.6E-5 0.050
In VC/Ace-d6: (p-NO-Ph)-SiF^- -2.038E3 10.39 0.998 2.9E-5 0.057

Table XVII. Activation Parameters3 for Dicarbon-
Substituted Siliconates.
Siliconate________Ea_______AH*______As*_____Ag* (298)
p-Tol2SiF3 12.2 U.7 3.3 10.7
(p-N02Ph)2SiF3 10.8 10.3 5.1 8.8
o-Tol2SiF3 10.6 10.1 0.65 9.9
a-Nap2SiF3 10.2 9.7 1.3 9.3
PhXylSiF3 10.2 10.6 -2.7 11.4
t-BuPhSiF3 9.0 8.6 -6.1 10.4
In VC/Ace-d6: (p-N02Ph)2SiF3 9.7 9.3 0.15 9.2
aThe units for the activation parameters are kcal/mol for the activation energy, free energy and enthalpy of activation and cal/K-mol for the entropy of activation

2.8 Estimated Ag* for o-Tol2SiF.j~
Due to steric hinderance in o-Xy^SiF^" and to
the tendency of this siliconate to decompose at higher
temperatures, we have been able to make only an
estimate of the minimum activation energy required for
the pseudorotational process in this compound using the
coalescence temperature (T ) method (equations 5f 6).
AG = RT ln(knT /hk ) = cal/mol (5)
C C D c c
k = nAv/21/2 = sec"1 (6)
where: Av = distance betwen the two peaks at slow
exchange = Hz
R =1.987 cal/K-mol
T = coalescence temperature (K)
kg = 1.381E-16 erg/K
h = 6.626E-27 erg-sec
The application of this method to o-Xyl2SiFg
using a minimum Tc of 335 K has been thoroughly covered
20 *
in previous work. The AG resulting from that
calculation was 13.7 kcal/mol. We were subsequently
able to heat a solution of this siliconate up to 355 K

in the NMR instrument without seeing coalescence. Thus,
using a higher minimum T of 355 K, we obtain a Ag* of
14.5 kcal/mol, which is nearly a kcal higher than the
previous calculation. Even so, we see that fairly large
coalescence temperature differences have relatively
little effect on the activation energy obtained.
2.9 Error Analysis for the CBS Method
The standard deviations in the scatter of
points around the best fit line, as noted in Tables XV
and XVI (vide supra), give an indication of the random
error present in our analyses.
In addition to the possibility for random error
in the complete bandshape method of analysis, there are
several components of this method subject to systematic
error which may adversely affect the accuracy of the
activation parameters obtained. Hence, it is important
2 8
to deal quantitatively with such errors.
Error in the determination of temperature for a
particular NMR run is a major contributor to overall
error in the analysis. We calibrated the variable
temperature control unit using a thermocouple placed in
an NMR tube filled with acetone and equilibrated in the

NMR for ten minutes. We determined the accuracy of our
temperature calibrations to be within +/-2C.
Error present in the determination of the rate
constants for the intramolecular exchange process being
studied is also an important consideration. Here there
are two major areas of concern: in the band shape or
fit of the computer simulated peaks with the actual
spectra; and in the effect of the temperature
dependence of the chemical shift on the process.
In the discussion which follows, the
p-To^SiF^ system will be used as a representative
example of an error analysis based on these factors.
General application of the conclusions reached to the
other dicarbon-substituted siliconate systems studied
is reasonable due to the similarities in these systems.
Chemical shifts depend on temperature to
varying degrees in different systems. In doing CBS
analysis, it is important to determine the extent of
such temperature dependence, since a large temperature
dependence introduces substantial error into the
By extrapolating the temperature dependence
displayed by the chemical shift of the F signals at
low temperature (slow exchange), one can estimate the

hypothetical chemical shift difference at high
temperature as if there were no intramolecular exchange
process occurring to average the signals. For
p-To^SiF^ this hypothetical shift difference at 303.7
K would be 2698 Hz. The actual chemical shift
differences between the axial and equatorial signals at
184.9 K equals 2452 Hz. Thus, the change in chemical
shift difference due to temperature over a 119
temperature range is 246 Hz, or 10% of the total shift
difference at low temperature. Hence, if the
extrapolation of the chemical shift at higher
temperatures is used in the complete bandshape
analysis, error due to the temperature dependence of
chemical shift is not a significant source of error in
determining k^ in these systems.
One way of determining the outside limits of
error due to temperature calibration as well as
bandshape errors is via a-graphical method. This
process is illustrated for p-To^SiF^- in' Figures I and
II. Rectangles formed from error bars are derived from
the assumption of +/-2C possible error in temperature
and from the closeness of fit in the determination of
the rate constant. Lines drawn from the corner of one
rectangle on one end of the scale to an opposite corner

of a rectangle at the other end of the scale show the
outside limits in variations of the best fit line used
to determine the activation parameters. The ranges for
the activation parameters determined by this method run
from 11.1 to 14.2 kcal/mol for E 10.1 to 12.4
kcal/mol for AH*, and -3.0 to 9.5 cal/K mol for As*.
It can thus be seen that the outside error
limits on the activation parameters for p-To^SiF^
show close to the same range of values as the entire
range determined for all the siliconates studied.
Although there appears to be a fairly large possibility
for error in the absolute values of activation
parameters calculated by the CBS method, it is
important to remember that all these studies were done
in a systematic way and the values obtained should be
valid in a relative sense.

Figure I. Plot of log vs 1/T for p-To^SiF^ The
solid line is the best fit line with slope = -2.666E3.
The dotted lines estimate the outside limits for error
in the best fit line with slopes = -2.4E3 and -3.1E3.

3.2 3.4 3.6 3.8 4.0 4.2 4.4
1/T E3
I. Plot of log kr/T vs 1/T for p-Tol2SiF3 The
le is the best fit line with slope = -2.553E3
rcept = 11.04. The dotted lines estimate the
limits for error in the best fit line with
-2.2E3 and -2.9E3 and intercepts = 9.7 and

3.1 Results
Table XVIII lists the activation parameters for
all the dicarbon-substituted siliconates studied via
CBS methods by our group, as well as the approximate
Ag* for Xy^SiF^". It is evident from the small range
of values displayed that the substituent effects
exerted on the pseudorotation process are not very
large in this series of siliconates.
Another factor apparent from the information in
Table XVIII is the importance of using a standard
solvent system for CBS analysis in these compounds if
the activation parameters obtained are to be directly
comparable. The variable-temperature behavior of
(p-NC^Ph)2SiF3 was studied both in pure acetone d-6,
our standard solvent system, and in a 50/50 (by volume)
mixture of vinyl chloride and acetone d-6. The
activation energy determined in pure acetone d-6 was
10.8 kcal/mol, while the value obtained in the 50/50

Table XVIII. Activation Parameters3 for the
Dicarbon-Substituted Siliconates
Studied by Our Group.
Siliconate Ea Ah* As* Ag*(298)
p-Tol2SiF3 12.2 11.7 3.3 10.7
bPh2siP3- 11.7 10.0 2.5 9.2
(p-N02Ph)2SiF3_ 10.8 10.3 5.1 8.8
o-Xyl2SiF3 >14
o-Tol2SiF3~ 10.6 10.1 0.65 9.9
a-Nap2SiF3~ 10.2 9.7 1.3 9.3
PhXylSiF3 10.2 10.6 -2.7 11.4
bMePhSiF3 9.9 9.4 -4.4 10.7
t-BuPhSiF3 9.0 8.6 -6.1 10.4
In VC/Ace-d6 (p-N02Ph)2SiF3 9.7 9.3 0.15 9.2
aThe units for the activation parameters are kcal/mol
for the activation energy, the free energy and enthalpy
of activation and cal/K-mol for the entropy of
activation. bCBS analysis done by Stephen Danahey and
reported in his thesis (University of Colorado, Denver;
1.986) .

solvent mixture was 9.7 kcal/mol.
Previously, the behavior of MePhSiF^ was
investigated in this solvent mixture at intermediate
temperatures. No attempt was made to quantify the
effect, but a narrowing of bandwidths relative to those
noted in pure acetone-d6 was observed. It is unlikely
that a solvent polarity effect is operating, since
intramolecular exchange should involve little charge
separation. However, some type of specific weak
solute-solvent interaction is probably responsible for
the effect.38
It is important in looking at these data to
keep in mind the fact that the estimated maximum error
calculated for a representative siliconate (vide supra)
covers about the same range of values as those reported
for all the compounds in the table. As noted earlier,
however, the relative values should be valid, and there
are some interesting trends exhibited by these
activation parameters which are worth exploring.
3.2 A Hammett Study
Several of the substituents used in this study
were chosen in order to allow a Hammett study to be
done. A Hammett study is used to determine the

electronic effect of substituents on the rate or course
of a reaction.
The standard reference reaction chosen by
Hammett was the ionization of para- or meta-
substituted benzoic acid in aqueous solution at 25C.
Benzene rings transmit electronic substituent effects,
while steric interference by the para or meta
substituents is minimized. Hammett measured the effects
of para and meta substituents on the observed pK of the
ionizations and defined a substituent constant, ax (eq
^KX^KH^benzoic acid CTX
where: Kx = the equilibrium constant for the
substituted benzoic acid
Kr = the equilibrium constant for the
unsubstituted benzioc acid.
Thus, by definition, ax for unsubstituted benzoic acid
is zero. Electron withdrawing groups (EWG) favor the
ionization reaction and have positive Ox values, while
electron donating groups (EDG) disfavor the reaction
and have negative a values.

Extending this treatment to other systems
implies that substituent effects in these systems will
be proportional to the substituent effects in the
standard reaction. The proportionality constant is p
(eq 8,9).
I09 = P1o 'Wbenzoic acid <8)
log (K/K) = pav (Hammett equation) (9)
A positive p value indicates that the reaction in
question responds to substitution in the same way as
the benzoic acid system, i.e. that the reaction is
favored by EWG's. A negative value implies that EDG's
favor the reaction. The magnitude of the value is a
measure of the susceptibility of the reaction to the
polar effects of the substituent on the benzene ring.
Due to the relationship between equilibrium
constants and free energy, the Hammett equation may be
written as a linear free energy relationship (eq 10).
log (Kx/Kh) = "(AGX -,Agh)/2.3RT

If the free energy relationship is written in
terms of AG and Ag *, which are directly related to
A ri
rate, the Hammett equation may be expressed in terms of
rate constant's (eq 11).
log (kx/kH) = pax (11)
A method for extending linear free energy
relationships to systems where steric factors are
important was developed by Taft in the 1950's.^ He
noted that the p value for the base-catalyzed
hydrolysis of para- and meta-substituted benzoate
esters is positive and large (2.51), while acid
catalyzed hydrolysis of the same compounds shows almost
no substituent effect (p = 0.03). Extending his
consideration of base and acid catalyzed ester
hydrolysis to more general systems, including aliphatic
ones (RCC^Et), Taft noted the close similarity between
the transition states (T.S.) for the rate limiting
steps of base-catalyzed hydrolysis (compound XII) and
of acid-catalyzed hydrolysis (compound XIII). Both
transition states are tetrahedral and differ only by
the presence of two protons. Protons are small and

exert little steric effect, so it is reasonable to
assume that any steric effect exerted is due to group
R. Hence, Taft proposed that subtracting the Hammett
log term for acid-catalyzed hydrolysis from that for
base-catalyzed hydrolysis should eliminate the steric
terms, leaving only the desired electronic terms in the
resulting equation (equation 12).
19 *
p in equation 12 was given a value of 2.48 so
that a and O values would be of the same magnitude.
Taft took for his reference compound MeCC^Et rather
than HC02Et, i.e. the reference substituent is R ='Me. '
The general Taft equation is given in equation 13:

We have only three compounds with para or meta
substitution as required for a Hammett study:
p-To^SiF^-, Pl^SiF^, and (p-NC^Ph^SiF^-. We made
several attempts to synthesize para-substituted
dianiline or di-N,N,dimethylaniline fluorosiliconate,
which would have added a siliconate with very
electron-donating groups to the study. We were unable
to synthesize either of these compounds, however, and
chose not to synthesize more siliconates having
substituents with a values in the middle range. The
overall effect of substituents on the pseudorotation
process was obviously small, and including such
compounds would have added little to our understanding.
It was found that a+ constants^ gave the best
correlation for the three compounds. a+ constants were
derived to account for the effects of through
conjugation between suitable electron-donating para-
substituents and a reaction center at which a positive
charge is developing. Although such a situation does
not exist for intramolecular exchange in these
siliconates, the correlation is best using these
constants, a situation often found in Hammett
studies. A plot of log kv/k (determined at 298 K) vs
X ri
Ea+ for these compounds yields a p+ value of 0.640 with

a correlation coefficient of 0.9931 (Table XIX).^ The
low magnitude of this p+ value signifies a small
electronic substituent effect on the pseudorotational
process in these siliconates. It is interesting that
the p+ value is positive, however, indicating that
EWG's favor the process.
It is important in interpreting such results to
keep in mind that the pseudorotational intramolecular
exchange process proceeds from a ground state TBP
configuration through an RP transition state. Any
factor which lowers the energy of the RP transition
state or which moves the ground state configuration
along the pseudorotational transition coordinate will
favor the pseudorotational process. In the case of the
para-nitro substituent, the electron delocalization
possible in that system may lower the electron density
in the C-Si bonds and thus lessen the bond repulsion
occurring in the RP transition state. This would in
turn lower the energy of the transition 'state. The
opposite effect may occur for the somewhat electron-
donating para-methyl substituent, leading to more
electron density in the bond and greater bond repulsion
in the transition state.

Table XIX.
Energy_Barriers'to Pseudorotation of
X2SiF3 Type Siliconates.
Ea____________log k(X)_______g+
11.7 0.00 0.00
10.8 1.135 0.79

An attempt was made to correlate the full range
* 31
of our compounds using a values. A scatter plot
resulted, however, indicating that these compounds are
too different to correlate using a single simple
parameter. Also, steric effects appear to be important
factors in some of these compounds as will be discussed
A similar study was done by Martin and
coworkers^ on the inversion of arylsiliconates of the
type shown by compound VI (vide supra). This ligand
permutation is postulated to occur through a series of
pseudorotations. The activation energies for inversion
for the full range of substituents listed in Table XX
were plotted against a constants. A linear correlation
resulted with a slope of -3.37 and a correlation
coefficient of 0.991, despite the use of varying
counter ions, solvents and methods of measurement. It
was also noted that inversion rates of four
arylsilconates (Y = Ph, 3-CF3Ph, 3,5-CF3Ph, CgF5)
correlated best with ct+ constants resulting in a p+
equaling 0.33 and a correlation coefficient of 0.9819.
Martin, postulate that the highest
energy intermediate state in the inversion process will

Table XX. Energy Barriers (AG ,424 R) to
Inversion of Martin's Siliconates
Compound VI).11
3, 5-CF,Ph
aValues determined by magnetization transfer
experiments. values calulated from rates of
interconversion of diastereomersaverage of forward
and reverse reactions. cValues calculated from results
of NMR line-shape analyses.

resemble compound XIV/ having the five-membered ring in
a strained diequatorial placement and the replacement
of a highly electronegative axial oxygen by a more
electropositive carbon ligand. (It is not possible for
the five-membered ring to span both axial positions.)
These workers have, assumed that this conformation is
very close in geometry to the highest energy transition
state for the process. Hence/ the more apicophilic the
monodentate ligand, the greater the driving force
toward inversion and the lower the energy of the
transition state relative to the ground state. The
positive correlation shown by the activation energies
with a values is evidence for the proposed pathway for
3.3 Steric Effects on Pseudorotation
Four of the compounds studiedXyl2SiF3,
o-To^SiF^-, a-Nap2SiF3_, and PhXylSiF^-have steric

hinderance due to the presence of ortho substituents.
As was previously noted we have been unable to
accurately determine the energy of activation of
o-Xy^SiF^- due to its apparently less mobile nature.
One can easily imagine significant interference in the
pseudorotational process due to the presence of four
ortho-methyl groups. Interestingly/ however, the other
three compounds containing ortho substituents have very
similar activation energies (see Table XX) which are
all at least a kilocalorie lower than the activation
energy determined for Pt^SiF^-.
It is likely that steric effects are the key to
this trend since these ortho substituents should exert
little electronic effect. Ortho substituents may
distort the ground state TBP geometry toward the RP
geometry leading to a lowering of pseudorotational
activation energies. In TBP geometry the axial bonds
are longer than the equatorial bonds of identical
ligands. The angle between axial and equatorial bonds
is 90, and between equatorial ligands the angle is
120. For RP geometry the angle between the apical and
a basal'ligand is 105, and the angle between basal
ligands is 86.

We have been able to obtain a crystal structure
34 -
for o-Xyl2SiF3 as detailed in Table XXI. Table XXI
also lists crystal structure data for Ph2SiF3 reported
by Schomburg and Krebs. It is important to note that
the cation involved in these systems is different.
NMe^+ serves as counter ion for Ph2SiF3, whereas the
K+,18-C-6 complex acts as counter ion for o-Xyl2SiF3 .
There is a strong interaction between the K+ ion and
one of the axial fluorines in o-XylSiF^ Considerable
lengthening of that axial Si-F bond involved results.
0-Xyl2SiF3 shows.lenghtening of all bonds to
silicon relative to Ph2SiF3_. There is distortion of
the Fl-Si-F3 bond from the ideal 180 to 175.7. The
CAl-Si-CBl bond angle is 127.5, whereas the ideal
angle between equatorial ligands is 120. The
equivalent angle in Ph2SiF3~ is 118.9. Hence, the
crystal structure gives significant evidence for
distortion due to steric hinderance of the TBP geometry
of o-Xyl2SiF3 along the transition coordinate. If
these same effects are present in the other
ortho-substituted siliconates, such distortion may play
a role in lowering activation energies for
intramolecular exchange.

Table XXI.
asee Figure
Bond Lengths (A) and Angles (deg) for
Three Fluorosiliconates.
1.688 (1)
1.648 (2)
1.893 (2)
86.2 (1)
92.0 (1)
120.6 (1)
118.9 (1)
1.682 (4)
1.639 (4)
1.781 (4)
1.897 (6)
1.902 (6)
87.3 (2)
93.4 (2)
92.2 (2)
171.9 (2)
117.5 (3)
124.3 (2)
84.8 (2)
92.0 (3)
90.8 (2)
118.2 (2)
1.703 (6)
1.652 (6)
1.725 (5)
1.919 (9)
1.925 (8)
88.9 (3)
90.9 (3)
92.7 (3)
175.7 (3)
117.1 (4)
115.3 (3)
86.8 (3)
91.2 (3)
89.0 (3)
127.5 (4)
1.701 (4)
1.621 (4)
1.689 (4)
1.882 (7)
1.845 (7)
87.4 (2)
90.9 (3)
90.6 (3)
175.6 (2.)
118.8 (2)
118.4 (3)
88.3 (2)
91.4 (3)
91.3 (3)
122.8 (3)
III (vide infra) for atom assignments.

Figure III. Atom assignments, for Table XXI. For
Pl^SiF^ CA1 and CBl are phenyl calrbons. For
o-Xyl2siF3", CA1 and CBl are o-xylyl carbons. For
t-BuPhSiF^-f CA1 is a t-butyl carbon and CBl is a
phenyl carbon. For MePhSiF3, CA1 is a phenyl carbon
and CBl is a methyl carbon.

The alkyl/phenyl-substituted siliconates also
exhibit a lowering of activation energies for
pseudorotation relative to Pl^SiF^-. Here again steric
effects may play a role, since methyl and t-butyl
groups are more spherically symmetric and thus
sterically bulky than planar phenyl groups. The trend
of lower activation energies for intramolecular
exchange going from the methyl to the bulkier t-butyl
group supports the notion of steric hinderance as a
factor in decreasing activation energies.
Table XXI also lists the bond lengths and
angles from the crystal structure of t-BuPhSiF^ The
counter ion is the K+,18-C-6 complex. Again, a strong
interaction between K+ and an axial fluorine
contributes to the lengthening of one axial Si-F bond
in the crystal.
The bond lengths of t-BuPhSiF^ are similar to
those reported for Pt^SiF^ The Fl-Si-F3 bond in
t-BuPhSiF3 is 171.9, which is considerably distorted
from the ideal 180 and even smaller than the 175.7
reported for the equivalent bond in o-Xyl2SiF3. It is
reasonable to assume that a sterically bulky ligand in
an equatorial position would most strongly affect axial
ligands. There is also considerable distortion in the

equatorial F2-Si-CB1 bond angle to 124.3.
Surprisingly, however, both equatorial bond angles
involving CA1 (the t-butyl carbon) are smaller than the
ideal of 120. The effect on the distortion of ground -
state TBP geometry due exclusively to steric hinderance
from the alkyl group is thus not clear cut.
The crystal structure of MePhSiF^ has recently
been reported by Holmes, and these parameters are
also Table XXI. In this structure there is
slight bond shortening of most bonds to silicon the bond lengths reported for Pl^SiF^ The
Fl-Si-F3 bond angle is 175.6 and the CAl-Si-CBl angle
is 122.8. The cation involved in this system is
It is important to remember that extrapolating
crystal state data to liquid state phenomena is risky
in that solid state structures may not reflect
accurately liquid state conformations. Packing forces
may play a large role in determining the' eryst-a 1 :
structure. As mentioned earlier the K+ ion interacts
strongly with one axial fluorine in both the
t-BuPhSiF^ and o-xy^SiFj crystal structures. The
distances between the atoms involved are 2.544 A in
t-BuPhSiF^ and 2.703 A in o-Xy^SiF^ These distances

are comparable to effective ionic radii between K+ and
F ions. If this strong interaction exists in
solution, one would expect to see it reflected in the
chemical shifts of the involved atoms. The presence of
such an interaction in solution is not evident from F
NMR studies, however. Unfortunately, it is difficult to
know how much of the observed distortion in the crystal
structure is due to substituent effect and how much is
due to this ligand-cation interaction.
It seems clear that there is distortion in the
TBP ground state geometry in these compounds due to
steric factors. However, it is difficult to directly
correlate the observed distortions and the activation
parameters obtained.
Calculations on the lowest energy pathway for
pseudorotation in these compounds would be very helpful
in determining how structural variations affect the
energy of intramolecular exchange. These calculations
turn out to be very difficult to perform, however, due
to the complexities added by the aryl and alkyl ligands
in the fluorosiliconates.
Undoubtedly there are also electronic factors
involved in the differences between the
pseudorotational activation energies for

alkyl/aryl-sustituted and aryl/aryl-substituted
siliconates due to the very different natures of alkyl
and aryl systems. A n system is not present in alkyl
groups as for aryl groups. Also, alkyl groups tend to
push electrons whereas aryl groups act to delocalize
electrons and are thus relatively electronegative.
In conclusion, investigation by CBS analysis
indicates that-there is a small effect by substituents
on the process of intramolecular exchange in this group
of fluorosiliconates. It is possible to see trends in
energy values for pseudorotation which can be
rationalized by invoking a combination of electronic
and steric substituent effects. Obtaining crystal
structures for more of these compounds would be useful
in determining the validity of arguments presented
here. Although very difficult, calculations would be
most helpful in understanding the factors which affect
the pseudorotational process in these compounds.

References and Notes
1. Eaborn,C. "Organosilicon Compounds"; Academic
Press: New York, 1960; pp 1, 10.
2. Tadura, S. N.; Voronkov, M. G.; Alekseev, N. V.,
Top. Curr. Chem., 1986, 131, 99.
3. Walsh, R., Acc. Chem. Res., 1981, 14, 246.
4. Corriu, R. J. P.; Guerin, C.; Moreau, J. J. E.,
Top. Stereochem.,1984, 15, 158.
5. Stevenson, W. H. Ill; Martin, J. C., J. Am. Chem.
Soc., 1985, 107, 6352.
6. Holmes, R. R., Acc. Chem. Res., 1972, 5, 296.
7. Muetterties, E. L.; Mahler, W.; Packer, K. J.;
Schmutzler, R., Inorg. Chem., 1964, 3, 1298.
8. Muetterties, E. L.; Mahler, W.; Schmutzler, R.,
Inorg. Chem., 1963, 2, 613.
9. Musher, J. I., J. Chem. Ed., 1974, 51, 94.
10. Breliere, C.; Carre, F.; Corriu, R. J. P.;
De Saxce, A.; Poirier, M.; Royo, G., J. Organomet.
Chem,, 1981, 205, Cl.
11. Stevenson, W. H. Ill; Wilson, S.; Martin, J. C.;
Farnham, W. B., J. Am. Chem. Soc., 1985, 107,
12. Schomburg, D.; Krebs, R.; Inorq. Chem., 1984, 23,
13. Gibson, J. A.; Ibbott, D. G.; Janzen, A. F.,
Can. J. Chem., 1973, 51, 3203.
14. Marat, R. K.;
55, 116.
Janzen, A. F., Can. J. Chem., 1977,

15. Marat, R. K.; Janzen, A. F., Can. J. Chem., 1977,
55, 3845.
16. Klanberg, F.; Muetterties, E. L., Inorg. Chem.,
1968, 7, 155.
17. Damrauer, R.; Danahey, S. E., Organomet., 1986, 5,
18. Farnham, W. B.; Harlow, R. L., J. Am. Chem. Soc.,
1981, 103, 4608.
19. Corriu, R. J. P.; Kpoton, A.; Poirier, M.; Royo,
G.; Corey, J., J. Organomet., 1984, C25.
20. Holmes, R. R.; Day, R. 0.; Harland, J. J.; Sau,
A. R.; Holmes, J. M., Organomet., 1984, 3, 342.
21. Holmes, R. R.; Day, R. 0.; Harland, J. J.;
Holmes, J. M., Organomet., 1984, 3, 347.
22. Danahey, S. E., M.S. Thesis, University of
Colorado, Denver, 1986,.
23. Simon R., M.S. Thesis, University of Colorado,
Denver, in progress.
24. EXC2 program and much assistance supplied by
Professor S. S. Eaton, University of Colorado-
25. Carey, F. A.; Sundberg, R. J. "Advanced Organic
Chemistry"; Plenum Press: New York; 1984; Section
26. Harris, R. K. "Nuclear Magnetic Resonance
SpectroscopyA Physiochemical View"; Pitman:
London; 1983; Section 5.6.
27. Lambert, J. B.; Shurvell, H. F.; Verbit, L.;
Cooks, R. G.; Stout, G. H. "Organic structural
Analysis"; Macmillan: New York; 1976; Chapter 6.
28. Professor S. S. Eaton supplied valuable
assistance in doing this error analysis.
29. Harris, J. M.; Wamser, C. C. "Fundamentals of

Organic Reaction Mechanisms"; John Wiley and
Sons, Inc: New York; 1976; Section 3.4.
30. Sykes, P. "A Guidebook to Mechanism in Organic
Chemistry"; Longman: New York; 1984; Section 13.5.
31. Values of substituent constants used obtained from
Hansch, C; Leo, A. "Substituent Constants for
Correlation Analysis in Chemistry and Biology";
Wiley: New York; 1979.
32. A discussion of this phenomenon can be found in
any good physical organic chemistry book, e.g.,
Lowry, T. H.; Richardson, K. S. "Mechanism and
Theory in Organic Chemistry", Harper and Row: New
York; 1981.
33. The values of 0+ used were doubled due to the
presence of two phenyl rings with substituents on
these three siliconates. The "additivity of the
effect of multiple substitution" is noted in
Hammett, L. P. "Physical Organic Chemistry";
McGraw-Hill: New York; 1970; Section 11.20.
34. The crystal structure data was kindly supplied by
Roberta Day> University of Massachusetts, Amherst.
35. Huheey, J. E. "Inorganic Chemistry", Harper and
Row; New York; 1983.
36. The abbreviations in the notebook references
refer to the following workers: Robert Damrauer
(RD), Roger Simon (RS), Stephen Danahey (SD), and
Vernon Yost (VY).
37. Harland, J. J.; Payne, J. S.; Day, R. 0.; Holmes,
R. R., Inorg. Chem., 1987, 26, 760.
38. Nekipelov, V. M.; Zamaraev, K. I., Coord. Chem.
Rev., 1985, 61, 1985.