The gas-phase reactions of SiCl4 and Si2Cl 6 with CH3OH and C2H5OH: An investigation by mass spectrometry and matrix-isolation infrared spectroscopy

Article abstract of DOI:10.1039/b403586k

The gas phase reactions of SiCl₄ and Si₂Cl ₆ with CH₃OH and C₂H₅OH have been investigated using both mass spectrometry and matrix isolation techniques. SiCl₄ reacts with both CH<

Full text of DOI:10.1039/b403586k

View Article Online / Journal Homepage / Table of Contents for this issue
R E S E A R C H P A P E R
The gas-phase reactions of SiCl4 and Si2Cl6 with CH3OH and
C2H5OH: An investigation by mass spectrometry and
matrix-isolation infrared spectroscopy
Nicola Goldberg,a Matthew J. Almond,*a J. Steven Ogden,b J. Pat Cannady,c Robin Walsha and
Rosa Becerrad
a
School of Chemistry, University of Reading, Whiteknights, Reading, UK RG6 6AD.
E-mail: m.j.almond@rdg.ac.uk
Department of Chemistry, University of Southampton, Highfield, Southampton, UK SO17 1BJ
Dow–Corning Corporation, Mail Stop 128, 2200 W. Salzburg Rd., PO Box 994,
b
c
Midland MI 48686-0994, USA
Instituto de Quimica-Fisica ‘‘Rocasolano’’, CSIC, C/Serrano 119, 28006 Madrid, Spain
d
Received 8th March 2004, Accepted 30th March 2004
F|rst published as an AdvanceArticle on the web 4th May 2004
The gas phase reactions of SiCl4 and Si2Cl6 with CH3OH and C2H5OH have been investigated using both
mass spectrometry and matrix isolation techniques. SiCl4 reacts with both CH3OH and C2H5OH upon
mixing of the vapours for times in excess of 3 h to generate the HCl-elimination products SiCl3OR (R ¼ CH3
or C2H5). The identity of these products is confirmed by deuteration experiments and by ab initio calculations
at the HF/6-31G(d) level. Further products are generated when the mixture is passed through a tube
heated to 750 ꢀC. Si2Cl6 reacts with CH3OH and C2H5OH via a different mechanism in which the Si–Si
bond is cleaved to yield SiCl3OR and HCl. Other products of the type SiCl4ꢁn(OCH3)n are tentatively identified
by a combination of mass spectrometric and matrix isolation measurements. These latter products indicate
further replacement of Cl atoms by OR groups as a result of reaction of CH3OH or C2H5OH with the
initial product.
limited information on the hydrolysis or alcoholysis reactions
Introduction
of silicon tetrahalides. Ault has shown11 that SiF4 forms 1:1
complexes with H2O, CH3OH and CH3OCH3 when the indivi-
dual reagents are condensed by twin-jet deposition onto a cold
CsI window with excess argon. Use of a merged-jet deposition
arrangement did not significantly increase the yield of these
species suggesting that they do not exist at appreciable concen-
tration in the gas phase but rather are formed on the surface of
the matrix during deposition. There was no sign of HF elimi-
nation from these adducts, reflecting most probably the
strength of the Si–F bonds. By contrast codeposition of SiCl4
and CH3OH in twin- or merged-jet mode led to no identifiable
product bands in the condensate.12 HSiCl3 does, however,
form the HCl-elimination product Cl2HSiOCH3 upon co-
deposition with SiCl4 by twin-jet or merged-jet methods.12 In
this case the yield of product is increased significantly on
switching to merged-jet mode suggesting that a gas-phase reac-
tion does occur. The authors of this work were surprised to
find that SiCl4 did not show similar reactivity to that of
HSiCl3 . TiCl4 reacts with CH3OH under these conditions.13
In twin-jet mode the principal product is the 1:1 adduct TiCl4ꢂ
CH3OH but this gives way to the HCl-elimination product
TiCl3OCH3 when merged-jet mode is used to deposit the
matrix. Heating of the merged-jet region leads to destruction
of TiCl3OCH3 and the formation of a secondary product
presumably by elimination of CH3Cl.
The alcoholysis of chlorosilanes has for many years been
an important process in the manufacture of organosilicon
reagents for the chemical industry.1 More than 50 years ago
partial hydrolysis reactions of SiCl4 were carried out in order
to generate a homologous series of silicon oxychlorides of gen-
2,3
eral formula SinOnꢁ1Cl2nþ2
.
This work was soon extended
to investigate the interaction of alcohols with SiCl4 .4,5 By this
time a wide range of alkyl chlorosilicates of the type
Si(OR)nCl4ꢁn had been prepared and characterised. Such com-
pounds are important synthetically because further hydrolysis
causes replacement of the Cl atoms without simultaneous
replacement of the alkoxy groups to give polymeric species
of the type [SiO(OR)2]n .4
In recent times an impetus has been given to work involving
silicon oxychloride species by the importance of the combus-
tion reaction of SiCl4 in O2 to give SiO2 .6 This reaction is used
in chemical vapour deposition (CVD) and chemical vapour
transport (CVT) processes and there has been interest in the
reaction mechanism. A recent theoretical study has suggested
that the reaction proceeds via a chlorosiloxane ring (Cl2SiO)2 .7
Also connected with the use of SiCl4 in CVD and CVT pro-
cesses is the need to obtain this reagent in a high degree of pur-
ity. As such there is a need to develop a greater understanding
of the reactions of SiCl4 and the role of silicon oxychlorides—
as may be illustrated by the photochemical synthesis and deter-
mination by infrared spectroscopy of trichlorosilanol SiCl3OH
in SiCl4 .8,9
The initial aim of the research presented here was to reinves-
tigate the SiCl4 þ CH3OH reaction using the techniques of
mass spectrometry and matrix-isolation infrared spectroscopy.
It was thought that the use of longer mixing times might allow
products to form via a gas-phase equilibrium, and in order to
obtain a more complete picture, combinations of ethanol,
Matrix isolation is an established method to study reaction
intermediates and mechanism.10 Many silicon compounds
have been investigated using this approach but there is only
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4
3264
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 3 2 6 4 – 3 2 7 0

View Article Online
methanol, Si2Cl6 and SiCl4 were studied. In the case of Si2Cl6 ,
one point of interest was whether Si–Si bond rupture would
compete with HCl elimination in the formation of products.
following ions (m/z values for the most abundþant isotopomer
þ
þ
are given in parentheses): SiClþ4 (170), SiCl3 (135), SiCl2
(98), SiClþ (63), SiCl3OCH3 (166), SiCl2OCH3 (129),
SiClOCH3þ (94) and HClþ (36), and a typical spectrum is illu-
strated in Fig. 1. These observations provide a body of evi-
dence for the formation of the HCl elimination product
SiCl3OCH3 . This reaction is analogous to the formation of
TiCl3OCH3 , which has been reported when TiCl4 and CH3OH
are reacted in a merged-jet system.13
þ
Experimental
Mass spectrometric studies were carried out at Southampton
using a VG mass lab-automated MS quadrupole instrument
operating with an ionisation energy of ca. 70 eV and controlled
by a PC. The matrix-isolation equipment used at Reading14 or
Southampton15 has been described in detail elsewhere. High
purity nitrogen and argon (BOC 99.999%) were employed as
matrix gases in the IR studies and deposition times were typi-
cally 10–30 min. During this period the central CsI window
was maintained at ca. 12 K, and spectra were recorded using
a PE 983G spectrometer interfaced with a data station. A typi-
cal resolution and accuracy of ꢃ2 cmꢁ1 or better was achieved.
Samples of SiCl4 (99% purity), Si2Cl6 (96% purity), CH3OH
(99.9% purity) and C2H5OH (99.9% purity) were supplied by
Aldrich and further purified on a vacuum line by distillation
and by the freeze–pump–thaw process. This ensured the
removal of water and dissolved oxygen from the alcohol sam-
ples. Gas mixtures were made up on a standard vacuum line
using manometric measurements and vapour pressure data
to control the composition of the mixture. Gaseous samples
were allowed to mix before deposition onto the cold window
via a needle valve. Where appropriate the deposition tube
was heated by a spirally-wound resistance wire whose current
was controlled by a variac. Temperatures were controlled to
ꢃ5 ꢀC over a length of 25 cm of tube.
When the experiment was repeated but with the reagents
being flowed together into the mass spectrometer through a
merged-jet system, rather than being premixed, only peaks
due to the reactants were seen, namely thþose arising from the
ions: SiCl4 (170), SiCl3 (135), SiCl2 (98), SiClþ (63),
CH3Oþ Hþ (32), CH3Oþ (31), CH2Oþ (30), CHOþ (29) and
CH3 (15). This shows that no reaction takes place and
demonstrates that SiCl4 is less reactive than TiCl4 under these
conditions in line with the previous observations of Brehm,
Ault and Everhart.13,14
þ
þ
When a premixed (mixing time ¼ 3 h) gaseous sample of
SiCl4 (1 Torr) and CH3OH (1 Torr) was passed through a tube
heated to 630 ꢀC, prior to entering the mass spectrometer, ions
resulting from the product SiCl3OCH3 were not seen, but a
pair of peaks at m/z ¼ 50 and 52 in a 3:1 intensity ratio were
seen, alongside features due to SiCl4 and CH3OH fragmenta-
tion. It is likely that these peaks arise from the molecular ion
of CH3Cl, suggesting that SiCl3OCH3 may decompose ther-
mally to eliminate CH3Cl in a manner similar to the thermal
decomposition of TiCl3OCH3 .13
The reaction was then investigated using matrix-isolation
infrared spectroscopy (N.B. argon replaced dinitrogen in some
matrix isolation experiments but in control experiments we
could find no difference in behaviour when these two matrix
gases were interchanged). As before, blank experiments were
conducted on both individual reagents (SiCl4 (1 Torr) in Ar
(100 Torr) and CH3OH (1 Torr) in Ar (100 Torr)) before
any codeposition experiments were conducted. The spectra
obtained from these reagents were in good agreement with
literature spectra.17,18 When both reagents were codeposited
onto the cold matrix window using merged-jet mode without
premixing only bands of the unreacted reagents were seen.
However, when the reaction was repeated allowing the gaseous
reagents to mix for 3 h before deposition new bands of a reac-
tion product, alongside the band of HCl at 2530 cmꢁ1, were
clearly seen. The position of the HCl stretching frequency sug-
gests that this molecule exists as the CH3OHꢂ ꢂ ꢂClH complex.12
Structural modelling was carried out at the HF/6-31G(d)
level, and vibrational frequency calculations were performed
using the Gaussian 98 software package.16a Calculated vibra-
tional frequencies were scaled by 0.8929 as is standard for
the Hartree–Fock method.16b
Results and discussion
Reaction of SiCl4 with methanol
In the initial experiments, a gas phase mixture was made up
consisting of 1 Torr of SiCl4 and 1 Torr of CH3OH, and the
reagents were allowed to mix at room temperature for a period
of 3 h before being admitted into the mass spectrometer. The
mass spectrum showed prominent peaks arising from the
Fig. 1 Mass spectrum of the products of the gas phase reaction of SiCl4 (1 Torr) with CH3OH (1 Torr). The reagents have been mixed for 3 h at
room temperature prior to being admitted to the mass spectrometer.
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 3 2 6 4 – 3 2 7 0
3265

View Article Online
assigned to n(C–O) which was similarly seen to shift to high
frequency upon perdeuteration of the product.
The bands of these products are listed in Table 1. It is likely
that the product formed alongside HCl is SiCl3OCH3 whose
formation in this reaction was suggested by mass spectrometry.
In order to help confirm this hypothesis the experiment was
repeated using CD3OD in place of CH3OH and the shifts of
the product band positions upon deuteration were noted.
Finally the experimentally observed frequencies for both the
H- and D-containing products were compared with those cal-
culated at the HF/6-31G(d) level for SiCl3OCH3 and
SiCl3OCD3 . The observed and calculated frequencies are listed
in Table 1 where approximate descriptions of the vibrations
giving rise to the bands are given. These assignments are based
on the results of the calculations and on the known regions of
absorption of vibrational modes. The product shows bands
entirely characteristic of the CH3O– moiety together with
strongly-coupled n(Si–O) and n(C–O) modes. It may be noted
that the band assigned to n(C–O) (1100 cmꢁ1 in the
SiCl3OCH3 product) is at a much higher frequency than the
Experiments were then carried out in which a gaseous mix-
ture of SiCl4 , CH3OH and Ar were flowed through a heated
pyrolysis tube at 350 or 750 ꢀC prior to condensation on the
cold window. At 350 ꢀC no changes to the product spectrum
were observed i.e. alongside bands of SiCl4 and CH3OH only
those of SiCl3OCH3 and HCl were seen. The tube was then
heated to 750 ꢀC. In this case weak new features were seen at
1439, 1304, 947, 662, 623, 618 and 612 cmꢁ1. This suggests
the formation of one or more Si–Cl-containing products.
Unfortunately these products could be obtained neither in suf-
ficient quantity nor purity to make identification possible. It is
noteworthy that no CH3Cl was seen as a product in these
matrix experiments.23 In the reaction of TiCl4 and CH3OH
similar pyrolysis experiments were peformed.13 Above 250 ꢀC
TiCl3OCH3 was destroyed and CH3Cl13,23 was formed. It is
clear that SiCl3OCH3 is more thermally stable than is
TiCl3OCH3 and that, while some decomposes to give CH3Cl
(as shown by mass spectrometry) other decomposition path-
ways exist.
In order to explore this reaction further a range of mixing
times (5, 20 and 30 min) were employed. As the mixing time
was increased the product concentration also increased but
no new products were seen. The maximum product concentra-
tion was achieved after 3 h mixing. Prolonging the mixing time
to 18 h did not cause any new product to appear. Accordingly,
a mixing time of 3 h was used for all reactions of methanol
with SiCl4 . In all reactions a 1:1 ratio of reactants was used.
Experiments were also performed in which the matrix-isolated
products were irradiated using the quartz-filtered, broad-band
output (l ꢄ 190 nm) of a medium-pressure mercury lamp. No
change to the spectrum was observed under these conditions.
This is expected, given that matrix-cage effects would inhibit
dissociation of the isolated product molecules.
corresponding band of methanol (1034 cmꢁ1 17
and shifts to
)
high frequency upon deuteration. Both observations are indi-
cative of strong coupling not only to n(Si–O) but also to C–
H deformation modes.12 The product shows only one infrared
active n(Si–Cl) stretch at 600 cmꢁ1. Two such bands (A1 þ E),
separated by an appreciable energy gap (nE > nA1) would be
expected from the local C3v symmetry of the SiCl3 moiety,
but comparison with related systems shows that the E mode
is significantly more intense in the infrared than the A1 mode.
Examples of similar XYZ3 molecules include HSiCl3 (573 and
513 cmꢁ1),12 FSiCl3 (640 and 465 cmꢁ1),19 HGeCl3 (709 and
418 cmꢁ1),20 OPCl3 (581 and 486 cmꢁ1 21
and SPCl3 (542
)
and 435 cmꢁ1).22 It is likely, therefore that the band we observe
here arises from the degenerate antisymmetric stretch while the
symmetric stretch occurs at lower frequency, is of lesser inten-
sity, and is not detected. For SiCl3OCH3 nasym(SiCl3) (A1) is
calculated to occur at 573 cmꢁ1. The failure to observe a band
arising from this vibration in our experiments probably results
from lack of intensity. The observed shifts in band position
upon deuteration and the close correlation between observed
and calculated spectra leave little doubt that the product gen-
erated here is SiCl3OCH3 .
Reaction of SiCl4 with ethanol
Mass spectrometric measurements were then made on gaseous
samples of SiCl4 (1 Torr) and C2H5OH (1 Torr) which had
been mixed for 18 h. It was found that much longer mixing
times (18 versus 3 h) were required when ethanol replaced
methanol as a reactant to build up a substantial concentration
of product. The mass spectrum ofþsuch a mixture showed clear
Our observations on SiCl3OCH3 are reminiscent of the find-
ings of Ault and Everhart on the analogous species
TiCl3OCH3 formed by reaction of TiCl4 and CH3OH in a
merged-jet system.13 In this work, just one n(Ti–Cl) mode
was observed (at 486 cmꢁ1), and a band at 1152 cmꢁ1 was
þ
features assigned to SiCl3OC2H5 (180), SiCl3OCH2 (165),
þ
SiCl2CH2 (128), SiCl2OHþ (115) and HClþ (36). The mass
spectrum is illustrated in Fig. 2. This suggests that the HCl-
elimination product has again been formed, in this case
SiCl3OC2H5 . Matrix isolation experiments on gaseous samples
of SiCl4 (1 Torr), C2H5OH (1 Torr) and Ar or N2 (200 Torr)
which had been mixed for 18 h support this assertion. The
observed bands which are assigned to SiCl3OC2H5 are listed
in Table 2, where they are compared with calculated frequen-
cies for the same species. The close correlation between
observed and calculated spectra leaves little doubt that the
product here is SiCl3OC2H5 .
Table 1 Observed and calculated frequencies of infrared absorption
for the products formed by the gas-phase reaction of SiCl4 with
CH3OH or CD3OD on mixing for 3 h at room temperature. The
compounds have been isolated in an argon matrix at 12 K at approxi-
mately a 1:100 dilution
n/cmꢁ1
obs
n/cmꢁ1
n/cmꢁ1
obs
n/cmꢁ1
calca
SiCl3OCH3 SiCl3OCH3 SiCl3OCD3 SiCl3OCD3 description
calca
Approximate
2960
2900
2859
1464
1192
1108
815
2956
2889
—
2244
2077
2143
1068
1076
1128
781
2217
2069
—
n(C–H)
n(C–H)
Reactions of Si2Cl6 with methanol and ethanol
n(H–Cl)b
d(CH3)c
n(Si–O)c
n(Si–O)c
d(Si–O–C)c
The next stage of the work was to investigate the reaction of
Si2Cl6 with methanol and ethanol under similar conditions.
The mass spectrum of Si2Cl6 has been discussed elsewhere.24,25
When a gaseous mixture of Si2Cl6 (1 Torr) and CH3OH
(1 Torr) was mixed for 3 h then flowed into the mass spectro-
meter, peaks were seen in the mass spectrum which werþe
assigned to the ions: Si(OCH3)3Clþ (156), Si(OCH3)Cl2
1463
1182
1108
774
1067
1102
1124
742
600
600
573
597
597
573
n
n
asym(SiCl3)
sym(SiCl3)
d
d
a
Calculated frequencies scaled by 0.8929 (see text). b n(H–Cl) of HCl
(DCl) product. No products other than SiCl3OCH(D)3 and H(D)Cl
(129), Si(OCH3)þ2Clþ (125), Si(OCH3)3 (121), Si(OCH3)Clþ
þ
þ
(94), Si(OCH3)2 (90), SiClþ (63), SiOCH3 (59) and HClþ
(36). This result is indicative of the formation of the products
c
were seen in these experiments. Significant coupling of vibrations
d
occurs in this region. Too weak to be observed (see text).
SiCl4ꢁn(OCH3)n (where n ¼ 1–3). No new peaks were seen at
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4
3266
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 3 2 6 4 – 3 2 7 0

View Article Online
Fig. 2 Mass spectrum of the products of the gas phase reaction of SiCl4 (1 Torr) with C2H5OH (1 Torr). The reagents have been mixed for 18 h at
room temperature prior to being admitted to the mass spectrometer.
higher mass, suggesting that products of formula Si2Cl6ꢁn
-
arise from more substituted products were seen. These features
are listed in Table 3.
(OCH3)n were not formed.
The reaction was then investigated by matrix-isolation infra-
red spectroscopy. Infrared spectra were recorded on condensed
gaseous mixtures of Si2Cl6 (1 Torr), CH3OH (1 Torr) and Ar
(200 Torr) and of Si2Cl6 (1 Torr), CD3OD (1 Torr) and Ar
(200 Torr). The product bands observed in each case are listed
in Table 3 and the spectra observed in the CH3OH reaction are
illustrated in Fig. 3. A comparison of the positions of these
bands with those observed in the earlier experiments utilising
SiCl4 as a reactant demonstrates that the molecules SiCl3OCH3
and SiCl3OCD3 have been formed alongside HCl and DCl
respectively. There is no unequivocal evidence for the forma-
tion of more substituted species in these spectra. However,
bands are clearly seen (when CH3OH is the reactant) at
3005, 2915, 1339 and 631 cmꢁ1 which do not belong to
SiCl3OCH3 and are not seen when SiCl4 was reacted with
CH3OH. When CD3OD was used as the reactant with Si2Cl6
in place of CH3OH a number of weak features which probably
In a final series of experiments premixed gaseous samples
(mixing time ¼ 18 h) of Si2Cl6 (1 Torr), C2H5OH (1 Torr)
and N2 or Ar (200 Torr) were studied by matrix isolation infra-
red spectroscopy. The observed bands are listed in Table 4 and
the spectrum is illustrated in Fig. 4. A comparison of these
bands with those seen when SiCl4 and C2H5OH are reacted
together shows that SiCl3OC2H5 has been formed. A number
of additional bands were seen in the spectrum obtained when
Si2Cl6 was reacted with C2H5OH which were not seen when
SiCl4 was the reactant. The positions of these bands are listed
Table 3 Observed infrared bands of the products of the gas phase
reaction of Si2Cl6 with CH3OH and CD3OD on mixing for 3 h at room
temperature. The products have been isolated in an argon matrix at 12
K at approximately a 1:100 dilution. The positions of the bands are
compared with those of the matrix-isolated compound SiCl3OCH(D)3
formed by the gas phase reaction of SiCl4 with CH3OH or CD3OD
Si2Cl6 þ
CH3OH
n/cmꢁ1
SiCl4 þ
CH3OH
n/cmꢁ1
Si2Cl6 þ
CD3OD
n/cmꢁ1
SiCl4 þ
CD3OD
n/cmꢁ1
Table 2 Observed and calculated frequencies of infrared absorption
for the molecule SiCl3OC2H5 formed by the gas phase reaction of SiCl4
and C2H5OH on mixing for 18 h at room temperature. The compound
has been isolated in an argon matrix at 12 K at approximately a 1:100
dilution
Approximate
description
3005
2958
2915
2859
2845
1464
1339
1189
1096
1032
811
—
n(C–H)a
n(C–H)b
n(C–H)b
n(H–Cl)c
n(C–H)a
d(CH3)b
2960
2900
2859
—
2245
2078
2143
2244
2077
2143
SiCl3OC2H5
obs n/cmꢁ1
SiCl3OC2H5
calca n/cmꢁ1
Approximate
description
2941
2844
1446
1393
1172
1103
1085
985
2951
—
n(C–H)
1464
—
1062
1096
1068
n(H–Cl)b
d(CH3)c
d(CH3)c
n(C–O)c
n(C–O)c
n(Si–O)c
n(C–C)c
d(C–C–O)c
d(CH3)a
d
1192
1100
—
1076
1128
—
n(Si–O)b e
n(C–O)b e
n(C–O)a e
d(Si–O–C)b e
n(Si–Cl)a
n(Si–Cl)a
n(Si–Cl)b
1453
1380
—
1128
913
780
665
631
600
815
781
—
1101
1085
946
751
595
572
631
600
—
—
779
600
600
600
n
n
asym(SiCl3)
sym(SiCl3)
a
Band of unknown product, probably formed by replacement of Cl
atoms by OCH(D)3 groups. These products are only formed when
Si2Cl6 reacts with CH(D)3OH(D) and not when SiCl4 reacts with the
d
a
b
Calculated frequencies scaled by 0.8929 (see text). Band of HCl
b
c
d
c
product. Significant coupling of vibrations occurs in this region.
d
same alcohols. Band of SiCl3OCH(D)3 . Band of HCl. Band
not observed. e Significant coupling of vibrations occurs in this region.
Too weak to be observed (see text).
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 3 2 6 4 – 3 2 7 0
3267

View Article Online
Fig. 3 Infrared spectrum (200–4000 cmꢁ1) of a low-temperature matrix (12 K) generated by condensing the products of the gas phase reagents
Si2Cl6 (1 Torr), CH3OH (1 Torr) and Ar (200 Torr). The reagents were mixed at room temperature for 3 h prior to deposition. Bands of the species
SiCl3OCH3 , which is also formed when SiCl4 reacts under similar conditions with CH3OH, are identified by the symbol S.
in Table 4. It is likely that these bands arise from further
substitution products.
This is a thermodynamically favourable process. Our HF
6-31G(d) calculations give values of DGorxn ¼ ꢁ43.5 kJ molꢁ1
and Keq ¼ 4.2 ꢅ 107 for the reaction. Our findings are similar
to the observations of Ault and Everhart on the reaction of
TiCl4 and CH3OH.13 The reaction of SiCl4 and CH3OH is,
however, slower, as demonstrated by the observation that,
unlike with TiCl4 and CH3OH, the product is not formed in
a merged-jet system. The differences between these systems
probably reflect the fact that the Ti–Cl bond in TiCl4 is both
Discussion
The mass spectrometric and matrix-isolation infrared spectro-
scopic data show that SiCl4 reacts with CH3OH and C2H5OH
to give the HCl-elimination product SiCl3OR (R ¼ CH3 or
C2H5):
˚
longer (2.185 A) and weaker (dissociation enthalpy,
SiCl4 þ ROH ! SiCl3OR þ HCl
ð1Þ
DH ¼ 345 kJ molꢁ1)26y than the Si–Cl bond in SiCl4 (2.017
A, 462 kJ molꢁ1)27y. The HCl-elimination product is presum-
˚
ably formed via an adduct SiCl4ꢂROH (R ¼ CH3 or
C2H5OH). The analogous adduct SiF4ꢂCH3OH is formed on
deposition of SiF4 with CH3OH, although it does not eliminate
HF.11 The differences between our system and this are under-
standable. Fluorine is well known to stabilise higher coordina-
tion complexes of silicon relative to chlorine and therefore the
lack of direct observation of SiCl4ꢂROH is not surprising. On
the other hand HCl elimination (from a more weakly bound
complex) is more likely than HF elimination due to the weaker
Si–Cl bond relative to the extremely strong Si–F [DH(SiF3–
F) ¼ 697 kJ molꢁ1, DH(SiCl3–Cl) ¼ 462 kJ molꢁ1].27 It is
interesting to note that the principal products formed when
Si2Cl6 reacts with CH3OH or C2H5OH are SiCl3OCH3 and
SiCl3OC2H5 respectively. This suggests that the major
pathway for reaction of Si2Cl6 is via Si–Si bond breaking:-
Table 4 Observed infrared bands of the products of the gas phase
reaction of Si2Cl6 with C2H5OH on mixing for 18 h at room tempera-
ture. The products have been isolated in an argon matrix at 12 K at
approximately a 1:100 dilution. The positions of the bands are com-
pared with those of the matrix-isolated compound SiCl3OC2H5 formed
by the gas phase reaction of SiCl4 with C2H5OH
Si2Cl6 þ C2H5OH
n/cmꢁ1
SiCl4 þ C2H5OH
n/cmꢁ1
Approximate
description
2941
2844
1480
1445
1394
2951
2844
n(C–H)a
n(H–Cl)b
n(H–Cl)c
d(CH3)a d
d(CH3)a d
n(Si–O)a d
n(C–O)a d
n(C–O)a d
n(C–O)c d
n(C–C)a d
n(C–C)c d
d(C–C–O)a d
n(Si-Cl)c
n(Si–Cl)c
n(Si–Cl)a
n(Si–Cl)c
1446
1393
1172
1103
1085
1171
e
1082
1026
980
887
767
662
631
595
485
Cl3Si SiCl3 þ ROH ! SiCl3OR þ HSiCl3
ð2Þ
985
779
where R ¼ CH3 or C2H5 . The difficulty with this pathway is
that there was no evidence for formation of HSiCl3 . Obviously
some fragmentation peaks from HSiCl3 in the mass spectra
would be difficult to distinguish from those of Si2Cl6 . However
we do not see any features arising from HSiCl3 in any of our
infrared spectra either.12 It is, of course, possible that HSiCl3 ,
if it were produced, would itself react with CH3OH. It is
600
Band of SiCl3OC2H5 . b Band of HCl. c Band of unknown product,
probably formed by replacement of Cl atoms by OC2H5 groups. These
products are only formed when Si2Cl6 reacts with C2H5OH and not
a
y DH
values
obtained
via:
DH(MCl3–Cl) ¼ DHof (MCl3) þ
d
when SiCl4 reacts with the same alcohols. Significant coupling of
e
vibrations occurs in this region. Band not observed.
DHof (Cl) ꢁ DHfo(MCl4). DHof values were taken from ref. 26 (M ¼ Ti)
Ti) and ref. 27 (M ¼ Si).
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4
3268
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 3 2 6 4 – 3 2 7 0

View Article Online
Fig. 4 Infrared spectrum (200–4000 cmꢁ1) of a low-temperature matrix (12 K) generated by condensing the products of the gas phase reagents
Si2Cl6 (1 Torr), C2H5OH (1 Torr) and Ar (200 Torr). The reagents were mixed at room temperature for 18 h prior to deposition. Bands of the
species SiCl3OC2H5 , which is also formed when SiCl4 reacts under similar conditions with CH3OH, are identified by the symbol S.
known that HSiCl3 is much more reactive towards CH3OH
than is SiCl4 .12 However, there is no sign of the expected
Acknowledgements
We thank Dow–Corning for providing a studentship for NG.
product of reaction, namely CH3OSi(H)Cl2 , nor is there any
evidence from any of our reactions of a product containing
an Si–H bond. These observations tend to disfavour the
reaction pathway shown above.
An alternative is that one molecule of Si2Cl6 may react with
two molecules of ROH (R ¼ CH3 or C2H5):
References
1
2
3
4
J. L. Speier and M. S. Tzou, Organometallics, 1993, 12,
1981–1982.
W. C. Schumb and A. J. Stevens, J. Am. Chem. Soc., 1947,
69, 726.
W. C. Schumb and A. J. Holloway, J. Am. Chem. Soc., 1941, 63,
2753.
D. F. Peppard, W. G. Brown and W. C. Johnson, J. Am. Chem.
Soc., 1946, 68, 70–72.
Si2Cl6 þ 2ROH ! 2SiCl3OR þ H2
ð3Þ
H2 is not detectable by infrared spectroscopy and, in any case,
it is non-condensable under our experimental conditions. In
principle it could be seen mass-spectrometrically, but unfortu-
nately our mass spectrometer is not configured to reliably
detect ions at less than ca. 3 amu. It is noteworthy that reac-
tion (3) is predicted to be more exothermic (DHo ¼ ca. ꢁ264
kJ molꢁ1) than reaction (2) (DHo ¼ ca. ꢁ48 kJ molꢁ1).z Thus
there is a considerable driving force in its favour. The
presence of HCl in both mass and infrared spectra demon-
strates that some further reactions also take place. The likeli-
hood is that HCl is generated by further replacement of Cl
atoms by OCH3 groups which generates products such as
SiCl2(OCH3)þ2 and SiCl(OCHþ3)3 . The ions SiCl2(OCH3)þ,
SiCl(OCH3)2 and Si(OCH3)3 seen in the mass spectrum,
provide strong evidence for this. In the infrared spectrum
bands seen occur in the correct regions (n(C–H), d(CH3),
n(C–O or Si–O) and n(Si–Cl), see Table 3) of the spectrum
to belong to one or more such molecules although precise
assignments cannot be made.
5
6
W. Gerrard and A. H. Woodhead, J. Chem. Soc., 1951, 519–522.
M. Binnewies and K. Jug, Eur. J. Inorg. Chem., 2000,
1127–1138.
7
A. Kumar, T. Homann and K. Jug, J. Phys. Chem. A, 2002, 106,
6802–6809.
8
9
R. Gooden, Inorg. Chem., 1983, 22, 2272–2275.
R. Gooden and J. W. Mitchell, J. Electrochem. Soc., 1982, 129,
1619–1622.
10 M. J. Almond and K. S. Wiltshire, Annu. Rep. Prog. Chem., Sect.
C: Phys. Chem., 2001, 97, 3–60.
11 B. S. Ault, J. Am. Chem. Soc., 1983, 105, 5742–5746.
12 M. Brehm and B. S. Ault, J. Mol. Struct., 2003, 649, 95–103.
13 B. S. Ault and J. B. Everhart, J. Phys. Chem., 1996, 100, 15 726–
15 730.
14 M. J. Almond and M. Hahne, J. Chem. Soc., Dalton Trans., 1988,
2255–2259.
15 J. S. Ogden and R. S. Wyatt, J. Chem. Soc., Dalton Trans., 1987,
859–865.
16 (a) M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A.
Montgomery Jr., R. E. Stratmann, J. C. Burant, S. Dapprich,
J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas,
J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci,
C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson,
P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. Ciolowski, J. V. Ortiz,
z These enthalpy change estimates are based on bond dissociation
enthalpies listed in refs. 27 and 28.
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 3 2 6 4 – 3 2 7 0
3269

View Article Online
21 M. L. Delwaulle and F. Francois, Compt. Rend., 1945, 220, 817.
22 H. Gerding andWestrik, Rec. Trav. Chim., 1942, 61, 842.
23 A. J. Barnes, H. E. Hallam, J. D. R. Howells and G. F.
Scrimshaw, J. Chem. Soc., Faraday Trans. 2, 1973, 69, 738–741.
24 N. Goldberg, J. S. Ogden, M. J. Almond, R. Walsh, J. P.
Cannady, R. Becerra and J. A. Lee, Phys. Chem. Chem. Phys.,
2003, 5, 5371–5377.
25 N. Goldberg, PhD Thesis, University of Reading, 2003.
26 M. W. Chase, Jr., C. A. Davies, J. R. Downey, Jr., D. J. Frurip,
R. A. McDonald and A. N. Syverud, J. Phys. Chem. Ref. Data,
1985, 14(Suppl. No. 1).
27 R. Becerra and R. Walsh, in The Chemistry of Organic Silicon
Compounds, ed. Z. Rappoport and Y. Apeloig, Wiley, Chichester,
1998, vol. 2, ch. 4, p. 153.
A. G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe,
P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres,
C. Gonzalez, M. Head-Gordon, E. S. Replogle and J. A. Pople,
Gaussian 98, Revision A9, Gaussian Inc., Pittsburgh, 1998.(b)
J. B. Foresman and A. Frisch, Exploring chemistry with Electronic
Structure Methods: A Guide to Using Gaussian, 1st edn., Gaussian
Inc., USA, 1993.
17 A. Lee-Smith, J. Chem. Phys., 1953, 21, 1997–2004.
18 S. Woo Han and K. Kim, J. Phys. Chem., 1996, 100, 17 124–
17 132.
19 J. Goubeau, F. Ha¨nschke and A. Ruoff, Z. Anorg. Allg. Chem.,
1969, 366, 113–115.
20 A. Ruoff, H. Burger, S. Biedermann and J. Cichon, Spectrochim.
28 S. W. Benson, Thermochemical Kinetics, 2nd edn., Wiley, New
York, 1976.
¨
Acta, 1974, 30A, 1647–1651.
T h i s j o u r n a l i s Q T h e O w n e r S o c i e t i e s 2 0 0 4
3270
P h y s . C h e m . C h e m . P h y s . , 2 0 0 4 , 6 , 3 2 6 4 – 3 2 7 0