Photolysis of (Me2Si)6 in argon matrices doped with high concentrations (ca. 20%) of N2O or C2H4O: Formation of (Me2SiO)6

Article abstract of DOI:10.1016/S0022-328X(01)01184-6

Irradiation using a low pressure mercury lamp (λ=ca. 250 nm) of argon matrices containing ca. 1% (Me₂Si)₆ and ca. 20% ethylene oxide (C₂H₄O) or nitrous oxide (N₂O) for a period of ca. 20 h leads to the formation of the cyclic compound (Me₂SiO)₆. This has a 12-membered ring with alternating Si and O atoms. It is identified by comparison of its infrared spectrum with a spectrum of an authentic sample. The reaction appears to proceed by stepwise insertion of O atoms into Si-Si bonds.

Full text of DOI:10.1016/S0022-328X(01)01184-6

Journal of Organometallic Chemistry 640 (2001) 177–181
www.elsevier.com/locate/jorganchem
Photolysis of (Me2Si)6 in argon matrices doped with high
concentrations (ca. 20%) of N2O or C2H4O: formation of
(Me2SiO)6
Matthew J. Almond a,*, J. Pat Cannady b, Tracey A. Darling a, J. Steven Ogden c,
Robin Walsh a
a Department of Chemistry, Uni6ersity of Reading, Whiteknights, Reading RG6 6AD, UK
b Dow-Corning Corporation, Mail 128, PO Box 995, Midland, MI 48686-0995, USA
c Department of Chemistry, Uni6ersity of Southampton, Highfield, Southampton SO17 1BJ, UK
Received 21 June 2001; accepted 15 August 2001
Abstract
Irradiation using a low pressure mercury lamp (u=ca. 250 nm) of argon matrices containing ca. 1% (Me2Si)6 and ca. 20%
ethylene oxide (C2H4O) or nitrous oxide (N2O) for a period of ca. 20 h leads to the formation of the cyclic compound (Me2SiO)6.
This has a 12-membered ring with alternating Si and O atoms. It is identified by comparison of its infrared spectrum with a
spectrum of an authentic sample. The reaction appears to proceed by stepwise insertion of O atoms into SiꢀSi bonds. © 2001
Published by Elsevier Science B.V.
Keywords: Photolysis; Matrix isolation; (Me2Si)6; Oxidation
In argon matrices doped with low concentrations (ca.
0.5–1%) of the O-donors ethylene oxide (C2H4O) or
1. Introduction
nitrous oxide (N2O) the observed chemistry is some-
There has been considerable interest in the matrix
what different. When (Me2Si)6 is irradiated using a
chemistry of reactive silicon-containing species [1–3].
low-pressure mercury lamp with output centred around
Much of this work has focused on the silylenes—com-
u=254 nm in 0.5–1% N2O-doped argon matrices,
pounds of general formula R2Si, and a convenient route
to dimethylsilylene has been found to be the photolysis
of the cyclic compound (Me2Si)6. This extrudes Me2Si
upon irradiation around u=250 nm to yield (Me2Si)5
[4]. Upon irradiation at longer wavelengths (u=450
nm) Me2Si is slowly isomerised to MeSi(H)ꢁCH2 [5a]; a
Me2Si is generated [6]. Upon annealing to 35 K the
bands of Me2Si decay and new bands are seen at 1186
and 1204 cm−1. It is concluded that the band at 1204
cm−1 is the w(SiꢁO) mode of Me2SiO while that at 1186
cm−1 is w(SiꢁO) of Me2SiO complexed to N2O. When
the experiment is repeated using C2H4O in place of
reaction that can be reversed either by photolysis at
N2O, one band in this region—at 1193 cm−1 is seen—
u=250 nm or by annealing. The thermal reaction only
proceeds at temperatures above 100 K [5b], so it is only
possible to observe it in involatile media such as a
3-methylpentane glass [5a]. If an argon matrix contain-
ing MeSi(H)ꢁCH2 is heated above 50 K, all of the
argon evaporates and the product is the s-bonded
cyclic dimer [5a].
and this is assigned to the complex of Me2SiO with
C2H4O [6]. In both cases the initial product therefore
appears to be the monomeric silanone Me2SiꢁO [6].
Unfortunately other infrared bands of Me2SiꢁO or of
its complexes could not be identified probably because
the bands are weak and likely to overlap with bands of
starting materials. Intriguingly, the silene MeSi(H)ꢁCH2
yields the same products as Me2Si upon annealing an
N2O-doped argon matrix at 35 K although the silene
does not convert to the silylene at such low tempera-
* Corresponding author. Tel.: +44-1893-16332.
E-mail address: m.j.almond@reading.ac.uk (M.J. Almond).
0022-328X/01/$ - see front matter © 2001 Published by Elsevier Science B.V.
PII: S0022-328X(01)01184-6

178
M.J. Almond et al. / Journal of Organometallic Chemistry 640 (2001) 177–181
tures. It is suggested that the H atom migration is
promoted by the O-donor. In the work reported here
we have studied the photochemistry of (Me2Si)6 in
argon matrices doped with 2–20% of the O-donors
C2H4O or N2O and we find that at these higher O-
donor concentrations very different chemistry is ob-
served from that seen at low O-donor concentrations.
around 254 nm, lines of moderate intensity at 365, 405,
434, 546 and 577 nm, and weak lines at 248, 265, 297,
302 and 334 nm. In order to cut out most of the longer
wavelength emission a filter was employed consisting of
two 25 mm quartz cells filled with saturated solutions
of CoSO4 and NiSO4. This transmitted in the region
220–300 nm only. Infrared spectra were recorded using
a Perkin–Elmer model 983 dispersive spectrometer with
a typical accuracy of 92 cm−1
.
2. Experimental
3. Results and discussion
Experiments were carried out using the matrix isola-
tion apparatus at University of Reading which has been
described elsewhere [7,8]. Gas mixtures were prepared
using standard vacuum line techniques. (Me2Si)6 and
(Me2SiO)6 were deposited by the slow spray-on method
with the sample held at room temperature and the
matrix gas flowed over the sample at a pressure of ca.
7.5×10−5 mmHg. A deposition time of ca. 90 min was
employed. The sample was deposited onto a CsI win-
dow held at ca. 12 K by means of an Air Products
202OSP ‘‘Displex’’ closed-cycle helium refrigerator. The
deposit was irradiated using a low-pressure mercury
lamp (Hanovia), which has one very intense line at
When an argon matrix (at ca. 12 K) doped with ca.
20% ethylene oxide (C2H4O) and containing a low
concentration (ca. 1%) of the cyclic hexamer (Me2Si)6 is
subject to prolonged irradiation using a low-pressure
mercury lamp with output around u=250 nm the
infrared bands of the starting compound gradually
decay in intensity. At the same time new infrared bands
appear and grow. These spectral changes are illustrated
in Fig. 1 while the band positions and their assignments
are listed in Table 1. It may be seen from Fig. 1 that
after about 20 h of photolysis the bands of the starting
compound have decayed to about half of their original
intensity. Several prominent bands are observed which
are assigned to a product A. Most striking is a broad
band around 1070 cm−1 which is likely to arise from a
w(SiꢀOꢀSi) mode of a polymeric silicone species [9,10].
Two conclusions may immediately be drawn from
these spectra. First, A is not the monomeric silanone
(Me2SiꢁO) which is generated when (Me2Si)6 is irradi-
ated in argon matrices doped with 0.5–1% N2O or
C2H4O [6]. Second, it would appear that there is only
one product (A) formed in our experiments. This con-
clusion may be drawn from the fact that all new bands
seen in the infrared spectra (i.e. those not belonging to
starting materials) appear and grow at the same rate
and there is no clear sign of any different bands seen at
shorter photolysis times (but see later).
A likely candidate for A is the cyclic siloxane
(Me2SiO)6—known as D6—which contains a 12-mem-
bered ring with alternating Si and O atoms. An experi-
ment was therefore carried out in which an authentic
sample of D6 was isolated in an argon matrix at ca. 12
K. The observed spectrum is illustrated in Fig. 2, while
the band positions are listed in Table 1. The strong
similarity between this spectrum and the spectrum of A
confirms that A is indeed D6. Our spectra of A are also
similar to those previously recorded for liquid D6 [9].
Further evidence for the nature of A comes from the
result of annealing experiments. No change is seen in
the spectrum of A upon annealing. Indeed if the matrix
is allowed to warm to room temperature over a period
of 4 h so that all of the matrix gas is lost, then the only
changes to be observed are a broadening of the bands
Fig. 1. Infrared transmission spectra in the region 1400–600 cm−1
seen upon photolysis of (Me2Si)6 (at a concentration of ca. 1%) in an
argon matrix doped with 20% ethylene oxide. The matrix support was
a CsI window at 12 K and the photolysis source used was a
low-pressure mercury lamp with output centred at u=254 nm.
Spectrum A, after deposition; spectrum B, after photolysis for 1 h;
spectrum C, after photolysis for 3 h; spectrum D, after photolysis for
7 h; spectrum E, after photolysis for 10 h; spectrum F, after photoly-
sis for 20 h; spectrum G, after allowing the matrix to warm to room
temperature over a period of 4 h.

M.J. Almond et al. / Journal of Organometallic Chemistry 640 (2001) 177–181
179
Table 1
a
Infrared bands (3000–600 cm−1
)
seen to appear upon photolysis of (Me2Si)6 in argon matrices doped with ca. 20% C2H4O or N2O compared
with infrared bands of authentic samples of (Me2Si)6 and (Me2SiO)6
(Me2Si)6
(Me2SiO)6
N2O 20% product b
C2H4O 20% product c
C2H4O 2% product
Approximate description d,e
f
f
f
f
f
f
f
f
f
2945 (m)
2885 (m)
1399 (m)
2945 (m)
2883 (m)
1399 (w)
1262 (s)
w(CH)
w(CH)
l(CH3)
l(CH3)
l(CH3)
l(CH3)
was(SiOSi)
z(CH3)
z(CH3)
z(CH3)
z(CH3)
z(CH3)
z(CH3)
w(SiC)
w(SiC)
w(SiC)
w(SiC)
w(SiC)
w(SiC)
w(SiC)
w(SiC)
w(SiC)
1261 (s)
1260 (s)
1247 (m)
1249 (m)
1238 (m)
1063 (vs)
880 (m)
860 (m)
1060 (vs)
885 (m) g
1062 (vs)
888 (m) g
1038 (s)
h
h
847 (m)
830 (m)
824 (s)
811 (s)
828 (m) g
826 (ms) g
785 (m)
823 (m)
799 (s)
736 (m)
721 (w)
795 (m)
725 (w)
702 (w)
785 (m)
782 (wm)
h
h
h
h
h
705 (vw)
612 (vw)
686 (m)
653 (m)
628 (w)
614 (w)
wsym(SiOSi)
a
92 cm−1
.
b Product formed by irradiation of (Me2Si)6 in 20% N2O-doped Ar matrices.
c Product formed by irradiation of (Me2Si)6 in 20% C2H4O-doped Ar matrices.
d References: (Me2Si)6 [4,6,16]; (Me2SiO)6 [9,17].
e Descriptions are approximate: bands in the fingerprint region results from vibrations which are coupled.
f Product bands not observed: obscured by starting material bands.
g Broad band–lower frequency component seen for pure (Me2SiO)6 is not observed.
h Not observed probably because of lack of intensity.
although our findings do not provide much firm evi-
dence because of the lack of observation of clear bands
due to intermediates. The filtered low pressure lamp
used in our experiments gives a narrow band output
centred around 250 nm. Neither C2H4O nor N2O ab-
sorb strongly in this region and irradiation, under these
of A (Fig. 1). The match of band position shows that A
must be a compound which is stable at room tempera-
ture, not prone to polymerisation and quite involatile
so that it is not pumped off the matrix window over a
period of 4 h. All three of these properties are met by
D6.
We repeated our experiments replacing C2H4O by
N2O as the matrix dopant—again at a concentration of
ca. 20%. Under these conditions we again observed A
as the reaction product. The observed bands of A from
both this experiment and also the experiment using
N2O as the oxidant are listed in Table 1.
As mentioned previously, there is a precedent for our
observed matrix chemistry from previous work carried
out in solution [10]. Alnaimi and Weber have observed
six products upon the oxidation of (Me2Si)6 using m-
chloroperbenzoic acid as a source of O atoms. These
were cyclic siloxanes containing 1–6 O atoms and were
believed to result from the sequential insertion of O
atoms into the SiꢀSi bonds of (Me2Si)6; the final
product being D6.
Fig. 2. Infrared transmission spectra in the region 1400–600 cm−1 of
(Me2SiO)6 isolated in an argon matrix at a concentration of ca. 1%.
We may make some observations regarding the pos-
sible mechanism of the reaction in which A is formed

180
M.J. Almond et al. / Journal of Organometallic Chemistry 640 (2001) 177–181
photosensitisation of the O-donor molecule such as was
seen, for example, in the oxidation of M(CO)6 (M=Cr,
Mo or W) to metal oxides by CO2 or N2O upon
photolysis at wavelengths too long for the O-donor to
absorb [14].
The next reaction steps must involve insertion of O
atoms into the remaining SiꢀSi bonds. Here it is known
from Alnaimi and Weber’s work that SiꢀSi bonds adja-
cent to O atoms are significantly more reactive than
those not bonded to oxygen; the insertion of one O
atom promotes the insertion of other O atoms [10].
Despite this, Alnaimi and Weber were able to observe
and record infrared spectra of all the intermediate
species Me12Si6On (n=1–6). While in most regions
these species show very similar infrared absorptions, for
the strongest band, that arising from wasym(SiꢀOꢀSi),
there are differences in the band position as follows:
n=1, w=1040 cm−1; n=2, w=1030 cm−1; n=3,
w=1070 cm−1; n=4, w=1065–1075 cm−1; n=5, w=
1065–1070 cm−1; n=6, w=1060–1090 cm−1 [10].
Close inspection of the spectra illustrated in Fig. 1
reveals that the exact position of the broad band in this
region does change slightly as photolysis proceeds. Af-
ter 6–9 h of photolysis, there are two peaks centred at
around 1040 and 1070 cm−1; after ca. 20 h of photoly-
sis, the higher wavenumber feature is dominant and the
lower wavenumber feature seems to have disappeared
completely after the window has been warmed to room
temperature. This provides some evidence that there are
species with different numbers of O atoms present in
the matrix, even if only at low concentrations, and that
the species with six O atoms predominates only after
prolonged photolysis and annealing. We observe no
other features which may be assigned to different prod-
ucts, but this is not too surprising given the known
similarity in the infrared spectra of these different oxi-
dation products [10].
In order to search more systematically for intermedi-
ates in this reaction we carried out experiments in
which (Me2Si)6 isolated in a 2% C2H4O-doped argon
matrix was photolysed using the low-pressure mercury
lamp. The purpose of this experiment was to limit the
concentration of the O-donor in order to attempt to
generate less oxygen-rich products. The spectral
changes observed when the matrix was photolysed for
21 h are shown in Fig. 3. New bands are seen but these
are much weaker than in the case of the 20% C2H4O-
doped matrix. The principal new bands seen upon
photolysis are listed in Table 1. These bands are at
similar positions to those seen for the product in 20%
C2H4O- or 20% N2O-doped argon matrices. The most
significant difference is perhaps the position of the
broad band arising from w(SiꢀO). In the 2% doped
matrix this band is centred around 1038 cm−1 which is
close to the reported positions of the most intense
infrared bands of Me12Si6O and Me12Si6O2 [10]. This
Fig. 3. Infrared transmission spectra in the region 1400–600 cm−1
seen upon photolysis of (Me2Si)6 (at a concentration of ca. 1%) in an
argon matrix doped with 2% ethylene oxide. The matrix support was
a CsI window at 12 K and the photolysis source used was a
low-pressure mercury lamp with output centred at u=254 nm.
Spectrum A, after deposition; spectrum B, after photolysis for 3 h;
spectrum C, after photolysis for 21 h.
conditions, of argon matrices containing either 20%
C2H4O or 20% N2O but no (Me2Si)6 led to no observ-
able changes in the infrared spectra. However, (Me2Si)6
is known to absorb at this wavelength resulting in SiꢀSi
bond cleavage [11–13]—probably via s–s* transitions
involving electrons in the SiꢀSi bonds. The first step in
the reaction is probably therefore rupture of a SiꢀSi
bond in (Me2Si)6. In the absence of, or at low concen-
trations of O-donors the reaction then proceeds to
extrude Me2Si and to form the pentamer (Me2Si)5 [4].
When low concentrations of O-donors are present this
may react to give the silanone Me2SiꢁO [6]. At higher
concentrations of O-donors it is not unreasonable to
assume that some form of adduct may be formed
between the O-donor and either the starting material or
the species with the cleaved SiꢀSi bond, leading to
photosensitisation of the O-donor. This, in turn, may
allow insertion of an O atom into the SiꢀSi bond. That
we do not observe any such adduct directly is probably
not surprising given (i) that it would doubtless be much
more reactive than the starting material or the product,
and (ii) that its infrared spectrum would probably be
not much different from that of the starting material.
We have compared the infrared spectra of (Me2Si)6
isolated in pure argon and in 20% N2O- or C2H4O-
doped argon matrices. These spectra show no signifi-
cant differences. Although there is a slight broadening
of bands in the doped matrices, the band positions are
identical to within 92 cm−1. Thus there is no direct
evidence for the formation of an adduct between
(Me2Si)6 and the O-donor which might account for

M.J. Almond et al. / Journal of Organometallic Chemistry 640 (2001) 177–181
181
Acknowledgements
observation suggests that it is possible preferentially to
generate intermediates in this reaction sequence by
limiting the concentration of the O donor.
We thank EPSRC for providing a studentship for
T.A.D. and Dow Corning for financial support.
This situation, where a complex—clearly multi-
step—oxidation reaction proceeds without clear evi-
dence for intermediates is reminiscent of other matrix
reactions e.g. the photo-oxidation of Mn2(CO)10 to
Mn2O7 in O2-doped argon matrices [15]. It is typical of
reactions where the intermediates are much more reac-
tive than starting material so that their concentrations
are necessarily low. It is exacerbated when the infrared
spectra of starting materials, intermediates and prod-
ucts are all similar.
References
[1] M.J. Almond, A.J. Downs, in: R.J.H. Clark, R.E. Hester (Eds.),
Advances in Spectroscopy. Spectroscopy of Matrix Isolated Spe-
cies, vol. 7, Wiley, Chichester, 1989, pp. 79–84, 117, 166–169.
[2] M.J. Almond, Annu. Rep. R. Soc. Chem. Sec. C 93 (1997) 3.
[3] M.J. Almond, K.S. Wiltshire, Annu. Rep. R. Soc. Chem. Sect. C
97 (2001) 3.
[4] T.J. Drahnak, J. Michl, R. West, J. Am. Chem. Soc. 101 (1979)
5427.
Our experiments demonstrate that in matrices doped
with high concentrations (ca. 20%) of the O donors
C2H4O and N2O, photolysis of (Me2Si)6 proceeds to
form (Me2SiO)6 by O atom insertion into the SiꢀSi
bonds. At lower O-donor concentrations (ca. 2%) it is
possible to form less oxygen-rich products, i.e.
(Me12Si6O) or (Me12Si6O2). These findings may be con-
trasted with the situation where even lower concentra-
tions (0.5–1%) of the same O-donors are present, where
the principal product is the monomer Me2SiꢁO [6]. Our
findings also demonstrate that the same chemistry may
be observed both in low-temperature matrices and in
solution. Previous work had shown quite different oxi-
dation products from (Me2Si)6 in solution [10] or in
matrices [6]. It is the specific reaction conditions in the
matrix that determine the products; as the O-donor
concentration is increased the reaction begins to mirror
that seen in solution.
[5] (a) C.A. Arrington, K.A. Klingensmith, R. West, J. Michl, J. Am.
Chem. Soc. 106 (1984) 525;
(b) It should be noted that this reaction has been observed in a
low-temperature matrix but in the gas phase is predicted only to
occur at elevated temperatures.
[6] C.A. Arrington, R. West, J. Michl, J. Am. Chem. Soc. 105 (1983)
6176.
[7] M.J. Almond, M. Hahne, J. Chem. Soc. Dalton Trans. (1988) 809.
[8] M.J. Almond, R.H. Orrin, J. Chem. Soc. Dalton Trans. (1992)
1229.
[9] R.E. Richards, H.W. Thompson, J. Chem. Soc. (1949) 124.
[10] I.S. Alnaimi, W.P. Weber, Organometallics 2 (1983) 903.
[11] M. Ishikawa, M. Kumada, J. Chem. Soc. Chem. Commun. (1970)
612.
[12] M. Ishikawa, M. Kumada, J. Organomet. Chem. 42 (1972) 325.
[13] L.F. Brough, R. West, J. Am. Chem. Soc. 103 (1981) 3049.
[14] M.J. Almond, A.J. Downs, R.N. Perutz, Inorg. Chem. 24 (1985)
275.
[15] M.J. Almond, J. Mol. Struct. 172 (1988) 157.
[16] G. Raabe, H. Vancik, R. West, J. Michl, J. Am. Chem. Soc. 108
(1986) 671.
[17] N. Wright, M.J. Hunter, J. Am. Chem. Soc. 69 (1947) 803.