New complexes with M-Si-O or M-Si-S linkages (M=Fe or Co)

Article abstract of DOI:10.1016/S0022-328X(01)01403-6

Ph₂XSiFe(CO)₂Cp [X = p-tolylS (1a), MeO (1b)] and Ph[2-MeOC₆H₄]XSiFe(CO)₂Cp [X = Cl (2a), OMe (2b)] have been fully characterised, including X-ray crystal structure determinations for 1a, 1b and 2a. None of the examples showed any tendency for migration of the X groups from silicon to iron, with elimination of silylene. However, very ready loss of the X groups was seen in the electrospray mass spectra, suggesting formation of the cationic silylene-iron complex ions is favoured. This was especially so for 2a and 2b, where intramolecular stabilisation of the silicon centre from the 2-OMe group is possible. The stable siloxane O[SiPh₂{Co(CO)₄}]₂ was also characterised; the X-ray crystal structure analysis shows a Si-O-Si bond angle of 153°.

Full text of DOI:10.1016/S0022-328X(01)01403-6

Journal of Organometallic Chemistry 648 (2002) 237–245
www.elsevier.com/locate/jorganchem
New complexes with MꢀSiꢀO or MꢀSiꢀS linkages (M=Fe or Co)
J. Scott McIndoe, Brian K. Nicholson *
Department of Chemistry, School of Science and Technology, Uni6ersity of Waikato, Pri6ate Bag 3105, Hamilton, New Zealand
Received 20 September 2001; accepted 19 October 2001
Abstract
Ph2XSiFe(CO)2Cp [X=p-tolylS (1a), MeO (1b)] and Ph[2-MeOC6H4]XSiFe(CO)2Cp [X=Cl (2a), OMe (2b)] have been fully
characterised, including X-ray crystal structure determinations for 1a, 1b and 2a. None of the examples showed any tendency for
migration of the X groups from silicon to iron, with elimination of silylene. However, very ready loss of the X groups was seen
in the electrospray mass spectra, suggesting formation of the cationic silylene–iron complex ions is favoured. This was especially
so for 2a and 2b, where intramolecular stabilisation of the silicon centre from the 2-OMe group is possible. The stable siloxane
O[SiPh2{Co(CO)4}]2 was also characterised; the X-ray crystal structure analysis shows a SiꢀOꢀSi bond angle of 153°. © 2002
Elsevier Science B.V. All rights reserved.
Keywords: Cobalt; Iron; Silylene; Siloxane complex; Thia-siloxane
Transfer of ꢀSR is perhaps even more likely, given the
1,2-migration of an H-group described above, since
SiꢀH and SiꢀS bonds have similar strengths (e.g. SiꢀH
320–340 kJ mol−1, Si–S 314 kJ mol−1 [8]).
1. Introduction
The 1,2-migration of a group X from silicon to a
transition metal in compounds containing M–SiR2X
units has been frequently invoked to explain rearrange-
ment processes, especially those where metal-silylenes
are probable intermediates [1–3]. Only recently how-
ever has this been directly observed, by Mitchell and
Tilley in the reversible interconversion of (Et3P)2Pt(H)-
Si(SBut)2OTf and [(Et3P)2Pt(m-SBut)Si(H)(SBut)]+ via
H-migration [4]. More commonly, the systems in which
1,2-migrations are indirectly most strongly implicated
are those involving disilanes, where migration of the
b-silyl group to the metal generates silyl/silylene inter-
mediates [5]. Examples of this type are known for many
transition metals [1].
So far there appear to be no definite examples in
which an ꢀOR or ꢀSR group transfers from silicon to a
transition metal. Migration involving a SiꢀO bond may
be implicated in the metal-catalysed disproportionation
of O(SiMe2H)2 into Me2SiH2 and polysiloxanes, though
other mechanisms can be proposed [6]. A related exam-
ple is the reaction of O(SiH3)2 with Co2(CO)8 to give
products such as Si[Co2(CO)7]2 without SiꢀO bonds [7].
There are now several examples of species with
MꢀSiꢀO linkages described in the literature, with vary-
ing properties. Greene and Curtis tentatively character-
ised the siloxane O[SiMe2{Co(CO)4}]2, but found it to
be extremely unstable [9]. In contrast, species such as
(MeO)3SiCo(CO)4 are apparently readily handled [10].
Braunstein [11] has developed extensive chemistry of
bi-metallic compounds based on (MeO)3Si groups act-
ing as bridging ligands. More recently, stable metallosi-
lanol and metallosiloxane compounds have been
reported, including derivatives (Me5C5)(PMe3)(OC)2-
MSiR2(OH) (M=Cr, Mo, W) from Malisch’s group
[12], and osmium examples from Roper’s laboratory
[13]. Both mono- and octa-Co(CO)4 derivatives of oc-
tasilasesquioxane have been characterised [14], follow-
ing earlier work with smaller cyclosiloxanes [15].
This present paper describes some new complexes
with LnMꢀSiꢀOR or LnMꢀSiꢀSR bonds. The fragment
LnM chosen was mainly Cp(CO)2Fe since this is the
best studied for silyl group migration. The supporting
groups on the silicon were Ph since these stabilise
silylene intermediates better than Me. One example
incorporating Co(CO)4 was also synthesised since this
* Corresponding author. Fax: +64-7-838-4219.
E-mail address: b.nicholson@waikato.ac.nz (B.K. Nicholson).
0022-328X/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(01)01403-6

238
J.S. McIndoe, B.K. Nicholson / Journal of Organometallic Chemistry 648 (2002) 237–245
group appears to encourage silyl migration under very
mild conditions [3].
conversion to a new species. The solvent was evapo-
rated and replaced with ether (30 ml). The solution was
filtered, petroleum spirits (10 ml) was added and the
mixture was concentrated to ca. 5 ml, and stored at
−20 °C. Yellow crystals of Ph2(TolS)SiFe(CO)2Cp
were formed (0.400 g, 48%). M.p. 133 °C. Anal. Found:
C, 63.59; H, 4.61. Calc. for C26H22FeO2SSi requires: C,
64.73; H, 4.60%; HRMS: Anal. Found: 482.046, re-
quired 482.0459, IR: w CO (THF) 2005 (s), 1953 (s)
2. Experimental
2.1. General
All manipulations were carried out in an oxygen-free
N2 atmosphere with carefully dried solvents in standard
Schlenk equipment. Ph(2-MeOC6H4)SiCl2 was prepared
by coupling ortholithiated anisole with Ph2SiCl2, and
(Ph2HSi)2O by hydrolysis of Ph2HSiCl. Co2(CO)8 was
freshly sublimed before use while [Fe(CO)2Cp]2 was
used as received. ‘‘TolS’’ refers specifically to p-tolyl-
thio group.
1
cm−1; H-NMR: l 7.67, 7.30 (m, 10H, C6H5), l 6.99
(d, J=7.8 Hz, 2H, C6H4), l 6.79 (d, J=8.1 Hz, 2H,
C6H4), l 4.63 (s, 5H, C5H5), l 2.19 (s, 3H, CH3);
13C-NMR: l 214.3 (CO), 141.4–127.6 (m, C6H5, C6H4),
l 85.0 (C5H5), l 21.0 (CH3); FABMS: m/z: 482 (7%)
[M+H]+, 454 (68%) [M+H−CO]+, 426 (100%)
[M+H−2CO]+, 359 (74%) [M−STol]+, 304 (33%)
[M−STol−2CO]+. Attempted purification by chro-
matography on silica led to extensive decomposition,
2.2. Instrumentation
the
only
product
isolated
being
[Fe2(m-
Infrared spectra were recorded on a Digilab FTS-40
FTIR spectrophotometer. NMR spectroscopy was per-
formed using a Bruker AC300P Multinuclear FT spec-
trometer. Routine electrospray mass spectra (ESMS)
were obtained on a VG Platform II spectrometer oper-
ating under standard conditions. EDESI mass spectra
were collected using a Micromass Quattro LC instru-
ment, in positive-ion mode. Samples were introduced
directly to the source at 4 ml min−1 via a syringe pump
as MeCN–MeOH solutions. Data collection was car-
ried out in continuum mode. The cone voltage was
initially set at 0 V and was increased by increments of
1 V after every scan up to a maximum of 200 V.
Elemental analysis was performed by the Campbell
Microanalytical Laboratory, University of Otago.
Melting points were measured on a Reichart Thermo-
pan melting point apparatus and are uncorrected.
STol)2(CO)2Cp2]+ (identified spectroscopically: ESMS
m/z 544 [M+]; IR w(CO) (THF) 1972 cm−1; ESR: giso
2.004, Dp−p 10G, cf. Ref. [18]) with an unidentified
anion.
2.3.3. Preparation of Ph2(MeO)SiFe(CO)2Cp (1b)
A THF solution (10 ml) of Ph2ClSiFe(CO)2Cp (0.501
g, 1.27 mmol) was treated with a methanolic solution of
NaOMe (0.82 ml, containing 1.25 mmol NaOMe). The
reaction was left to stir overnight. The solvent was
replaced with ether, the solution was filtered and left to
crystallise at 4 °C. Small, thin yellow crystals of
Ph2(MeO)SiFe(CO)2Cp formed (0.399 g, 81%). M.p. 99
°C; Anal. Found: C, 61.57; H, 4.71. Calc. for
C20H18FeO2Si requires C, 61.55; H, 4.65%; IR: w CO
1
(THF) 1998 (s), 1943 (s) cm−1; H-NMR: l 7.60, 7.34
(m, 5H, C6H5), l 4.65 (s, 5H, C5H5), l 3.48 (s, 3H,
OCH3); 13C-NMR: l 214.9 (CO), l 142.6 (ipso), 133.9
(ortho), 128.7 (para), 127.7 (meta) (SiC6H5), l 84.5
(C5H5), l 52.2 (OCH3); 29Si-NMR: l 57.2, ESMS
(MeCN–H2O): Cone 20 V: m/z 408 (100%) [M+
NH4]+, 391 (40%) [M+H]+; Cone 40 V: m/z 400
[M−OMe+MeCN]+. More detailed ESMS informa-
tion about fragmentation processes for this compound
is presented in Section 3.
2.3. Reactions
2.3.1. Preparation of Ph2ClSiFe(CO)2Cp
This known compound was prepared in 60% yield
from Na[Fe(CO)2Cp] and Ph2SiCl2 in THF. M.p. 88–
89 °C (Lit. 89–94 °C [16] or 95–96 °C [17]), IR: w CO
1
(petroleum spirits) 2015 (s), 1965 (s) cm−1; H-NMR: l
7.68, 7.35 (m, 10H, C6H5), l 4.80 (s, 5H, C5H5); 13C-
NMR: l 214.5 (CO), l 142.3 (ipso), 133.6 (ortho), 129.1
(para), 127.8 (meta) (C6H5), l 85.5 (C5H5); ESMS:
(MeCN–H2O), m/z 418 (100%) [M−Cl+MeCN+
H2O]+; 400 (45%) [M−Cl+MeCN]+; 377 (35%)
[M−Cl+H2O]+; 359 (40%) [M−Cl]+.
2.3.4. Preparation of Ph(2-MeOC6H4)ClSiFe(CO)2Cp
(2a)
[Fe(CO)2Cp]2 (0.71 g, 2.00 mmol) in THF (20 ml)
was reduced with sodium amalgam (1%) to
Na[Fe(CO)2Cp]. The resulting solution was transferred
by syringe into a Schlenk flask at 0 °C containing
Ph(2-MeOC6H4)SiCl2 (1.3 g, 4.6 mmol) in ether (10 ml).
The stirred solution was allowed to return to room
temperature and left overnight. The solvent was re-
moved under vacuum and replaced with ether (30 ml).
The solution was passed through a filter stick to remove
2.3.2. Preparation of Ph2(TolS)SiFe(CO)2Cp (1a)
To a stirred THF solution (30 ml) of Ph2ClSiFe-
(CO)2Cp (0.683 g, 1.73 mmol) was added Na[STol]
(0.253 g, 1.73 mmol). After 2 days stirring, the IR
spectrum of the reaction solution showed complete

J.S. McIndoe, B.K. Nicholson / Journal of Organometallic Chemistry 648 (2002) 237–245
239
NaCl and the solvent removed once more. Chromatog-
raphy (silica, CH2Cl2–petroleum spirit 1:5) provided a
bright yellow band which was collected and recrys-
tallised from ether–petroleum spirit to provide pale
yellow crystals of Ph(2-MeOC6H4)ClSiFe(CO)2Cp.
(0.54 g, 32%). M.p. 108–110 °C; Anal. Found: C,
56.37; H, 4.02. Calc. for C20H17ClFeO3Si requires C,
56.56; H, 4.03%. IR: w CO (petroleum spirit) 2016 (s),
corrected for absorption by a ƒ scan method. All other
calculations used the SHELX97 programs [20]. The
structures were solved by direct methods, and devel-
oped routinely with refinement based on F2. All non-
hydrogen atoms were assigned anisotropic temperature
factors, and hydrogen atoms were included in calcu-
lated positions. Selected bond parameters are in the
captions to Figs. 1, 3–5.
1967 (s) cm−1 1H-NMR: l 7.9–6.8 (m, 9H, C6H5,
.
C6H4), l 4.72 (s, 5H, C5H5), l 3.68 (s, 3H, OCH3);
13C-NMR: l 214.0, 213.9 (2CO), l 142.5–109.9 (m,
C6H5 and C6H4), l 85.5 (C5H5), l 60.3 (CH3); ESMS:
(MeCN–H2O) m/z 429 [M−Cl+OH+Na]+, 425,
[M−Cl+2H2O], 407 [M−Cl+H2O]+.
2.4.1. Crystal data for Ph2(TolS)SiFe(CO)2Cp (1a)
(
C26H22FeO2SSi, Mr 482.44, triclinic, P1, a=
,
8.5754(2), b=9.9558(2), c=14.7730(3) A, h=
81.066(1), i=75.044(1), k=68.367(1)°, V=1130.15(4)
3
A , Dcalc=1.418 g cm−3, Z=2, F(000)=500, m(Mo–
,
Ka) 0.833 mm−1, Tmax 1.000, Tmin 0.7217, crystal size
0.44×0.37×0.09 mm3. T 148 K.
2.3.5. Preparation of
Ph[2-(MeO)C6H4](MeO)SiFe(CO)2Cp (2b)
A total of 13 606 reflections, 3751 unique (Rint
0.0378) were collected 2BqB25°. Final R1 0.0575
(data with I\2|(I)), 0.0606 (all data), wR2 0.1722,
goodness-of-fit 1.030, final De +1.20/−0.51.
A THF solution (10 ml) of Ph[2-(MeO)C6H5]-
ClSiFe(CO)2Cp (0.174 g, 0.410 mmol) was treated with
a methanolic solution of NaOMe (3.6 ml, containing
0.45 mmol NaOMe). After 14 h the solvent was evapo-
rated and the residue was extracted with ether. After
filtration the solution was stored at −25 °C to give
ragged yellow crystals of Ph[2-(MeO)C6H4](MeO)SiFe-
(CO)2Cp (0.077 g, 45%). M.p. 176–179 °C; IR: w CO
2.4.2. Crystal data for Ph2(MeO)SiFe(CO)2Cp (1b)
C20H18FeO3Si, Mr 390.28, monoclinic, C2/c, a=
,
31.67(3), b=7.609(4), c=17.35(2) A, i=122.85(5)°,
3
V=3512(5) A , Dcalc=1.476 g cm−3, Z=8, F(000)=
,
1
(THF) 2001 (s), 1946 (s) cm−1; H-NMR: l 7.58, 7.27
1616, m(Mo–Ka) 0.91 mm−1, Tmax 0.86, Tmin 0.71,
crystal size 0.82×0.53×0.12 mm3. T 141 K.
A total of 2357 reflections, 2097 unique (Rint 0.0497)
were collected 2BqB22.5°. Final R1 0.0532 (data with
I\2|(I)), wR2 0.1256, goodness-of-fit 1.036, final De
+0.96/−0.37.
(m, 5H, C6H5), l 6.9 (m, 4H, C6H4), l 4.65 (s, 5H,
C5H5), l 3.66 (s, 3H, OCH3), l 3.44 (s, 3H, SiOCH3);
ESMS: (MeCN–H2O); m/z 421 [M+H]+.
2.3.6. Reaction of (Ph2HSi )2O with Co2(CO)8 (3)
To a Schlenk flask was added (PhH2Si)2O (0.562 g,
1.47 mmol), Co2(CO)8 (0.500 g, 1.46 mmol) and
petroleum spirits (25 ml). The mixture was stirred
overnight at room temperature to produce a pale
brown solution with some white precipitate. The sol-
vent was removed under vacuum, and the residue was
washed with small quantities of petroleum spirit to
remove unreacted Co2(CO)8. The crude product was
extracted from small amounts of Co4(CO)12 with Et2O,
and this solution was evaporated slowly to yield large,
colourless crystals of O[SiPh2{Co(CO)4}]2 (0.874 g,
83%). IR w(CO): (petroleum spirits, cm−1) 2097 (m),
2096 (m), 2036 (m), 2035 (m), 2014 (vs), 2005 (s), 1998
(s). The compound was further characterised by an
X-ray crystal structure determination.
2.4.3. Crystal data for Ph(2-MeOC6H4)ClSiFe(CO)2Cp
(2a)
C20H17ClFeO3Si, Mr 424.73, monoclinic, P21/n, a=
,
9.0901(1), b=14.6719(2), c=15.0469(2) A, i=
3
107.38(1), V=1915.20(4) A , Dcalc=1.473 g cm−3
,
,
Z=4, F(000)=872, m(Mo–Ka) 1.006 mm−1, Tmax
0.8496, Tmin 0.7853, crystal size 0.42×0.28×0.22 mm3.
T 203 K.
A total of 19 366 reflections, 4462 unique (Rint
0.0335) was collected 2BqB28°. Final R1 0.0371 (data
with I\2|(I)), 0.0446(all data), wR2 0.0889, goodness-
of-fit 1.063, final De +1.09/−0.47.
2.4.4. Crystal data for O[SiPh2{Co(CO)4}]2 (3)
(
C32H20Co2O9Si2, Mr 722.52, triclinic, P1, a=
,
2.4. X-ray crystallography
10.362(3), b=10.383(5), c=17.331(5) A, h=74.42(2),
3
,
i=80.87(5), k=61.34(2)°, V=1575.1(10) A , Dcalc
=
For compounds 1a and 2a, unit cell parameters and
intensity data were collected using a Siemens SMART
CCD diffractometer, using standard collection proce-
dures, with monochromatic Mo–Ka X-rays (0.71073
1.523 g cm−3, Z=2, F(000)=732, m(Mo–Ka) 1.183
mm−1, Tmax 0.897, Tmin 0.809, crystal size 0.72×
0.36×0.18 mm3. T 141 K.
A total of 5234 reflections, 4921 unique (Rint 0.0376)
was collected 2BqB24°. Final R1 0.0481 (data with
I\2|(I)), wR2 0.0978, goodness-of-fit 1.029, final De
+0.38/−0.36.
,
A). Corrections for absorption and other effects were
carried out with SADABS [19]. For 1b and 3, data were
collected on a Nicolet R3 four-circle diffractometer and

240
J.S. McIndoe, B.K. Nicholson / Journal of Organometallic Chemistry 648 (2002) 237–245
3. Results and discussion
3.1. Ph2(TolS)Fe(CO)2Cp (1a)
This compound was readily prepared from
ClPh2SiFe(CO)2Cp, on reaction with a stoichiometric
amount of Na[STol] in THF. The yellow crystalline
material is essentially air-stable, but attempted chro-
matography on silica led to rapid decomposition, giving
among other products the known radical cation
[Fe2(CO)2Cp2(STol)2]+, characterised by its ESMS, its
w(CO) spectrum and its characteristic ESR spectrum
[18]. While this formally involves a transfer of the thiol
group from silicon to iron, the reaction was non-spe-
cific, so mechanistic discussion is unwarranted.
Fig. 1. The structure of Ph2(TolS)SiFe(CO)2Cp. Bond parameters
include: Fe(1)ꢀSi(1) 2.296(2), Si(1)ꢀS(1) 2.180(2), Fe(1)ꢀCcp (av)
,
2.097(6) A; Fe(1)ꢀSi(1)ꢀS(1) 113.68, Si(1)ꢀS(1)ꢀC(41) 102.8(2),
Fe(1)ꢀSi(1)ꢀC(21) 118.1(2), Fe(1)ꢀSi(1)ꢀC(31) 112.0(2)°.
These changes are consistent with the formation of an
adduct of the type 5. However all attempts to isolate
the adduct were thwarted. A yellow solid was formed
which contained Hg, I and tolyl groups, and probably
contained (TolS)HgI (or the corresponding symme-
trised equivalents), while the solution showed carbonyl-
region infrared peaks which were consistent with
Ph2ISiFe(CO)2Cp. All this points to the initial forma-
tion of 5, but further reaction exchanges the SiꢀS bond
(possibly via a silylene) for a SiꢀI one. However the
system did not provide anything other than this tenta-
tive information.
The X-ray structure determination of Ph2(TolS)SiFe-
(CO)2Cp was undertaken since the only previous exam-
ples related to this were (TolS)3−n(Tf)nSiRu-
(PMe3)2Cp*, [(TolS)2(MeCN)SiRu(PMe3)2Cp*]+ and
[(TolS)(phen)SiRu(PMe3)2Cp*]2+ from Tilley’s group
[22], and Ph2(PhS)SiRhH2(PMe3)3 [23]. The structure is
illustrated in Fig. 1, which shows the expected arrange-
ment about both the Fe and Si atoms. The orientation
of the tolyl group is towards the metal fragment, which
appears unusual but is presumably the result of crystal
packing interactions on a flexible bond. The closest
intramolecular approach between the two fragments,
In an attempt to induce a migration reaction,
Ph2(TolS)SiFe(CO)2Cp was thermolysed or photolysed,
in the presence of HMPA to trap any silylene interme-
diates. However, only slow decomposition to in-
tractable products was observed, suggesting that the
deoligomerisation reaction observed for disilyl com-
plexes of Fe(CO)2Cp does not have a parallel with SiꢀS
bonds.
,
O(2)···C(41) (3.39 A) is essentially the sum of the van
der Waals radii for the two atoms, so there is no barrier
It is known that Pt(II) thiolate complexes form stable
adducts via a bridging S atom to mercury(II) halides, as
in 4 [21]. It was therefore of interest to see if the S atom
of Ph2(TolS)SiFe(CO)2Cp behaved similarly, to give
complex 5. When Ph2(TolS)SiFe(CO)2Cp and HgI2
were mixed in d6-DMSO in a NMR tube, the 1H
signals from the Cp protons shifted by +0.36 ppm,
and the tolyl-group protons also shifted markedly.
to the observed orientation. The FeꢀSi bond length is
,
2.296(2) A, which can be compared to those in the only
other two molecules of the type Ph2XSiFe(CO)2Cp to
,
have been determined, X=F 2.278(1) A [24] and X=
,
OMe 2.292(2) A (see Section 3.2). The trend in the
order STol\OMe\F is that expected on electronega-
tivity grounds. The SiꢀS bond length of 2.180(2), and
the SiꢀSꢀC bond angle of 102.8(2)° are both slightly

J.S. McIndoe, B.K. Nicholson / Journal of Organometallic Chemistry 648 (2002) 237–245
241
to [M+H]+ and [M+NH4]+ arising from chemical
ionisation by attachment of H+ or NH4+, presumably
at the oxygen atom of the methoxy group. More inter-
esting is the behaviour at higher cone voltages, where
the fragmentation patterns can yield some revealing
information. At low cone voltage settings, fragmenta-
tion of the complexes is minimised and the electrospray
ionisation mass spectra consist of ions derived from the
intact parent molecule, sometimes also incorporating a
molecule of the solvent used as a mobile phase. The
appearance of [M+H+solvent]+ ions is strongly de-
pendent on how good a donor the solvent is, so aceto-
nitrile, pyridine and dimethylsulfoxide (DMSO) are
common adducts whereas water, methanol and
dichloromethane are less frequently observed. These
adducts are the first to disappear as the fragmentation
energy (cone voltage) is increased. Further fragmenta-
tion generally consists of the loss of neutral fragments
from the pseudo-molecular ion. These features can all
be observed in the energy-dependent electrospray ioni-
sation mass spectrum (EDESI-MS) [27] of
Ph2(MeO)SiFe(CO)2Cp (Fig. 2).
This data presentation technique provides a two-di-
mensional map of cone voltage versus m/z ratio, and
allows visualisation of the entire fragmentation pattern
in a convenient format. Ph2(MeO)SiFe(CO)2Cp was
analysed in an acetonitrile–methanol solution, with
trace amounts of DMSO present, and the fragmenta-
tion processes are summarised in Scheme 1.
At the lowest cone voltage, only the ion [M+H+
DMSO]+ is observed (A). The DMSO adduct is ob-
served rather than MeCN or MeOH, because DMSO is
a superior donor. As the cone voltage is increased, the
ion [M+H]+ (D) makes an appearance, following loss
of the loosely attached solvent. At ꢀ15 V, three more
ions appear, all silylene species resulting from elimina-
tion of methanol from the parent ion. The three are
[Ph2SiꢁFe(CO)2Cp]+ (E) at 359 m/z, with the base-sta-
bilised species [(MeCN)Ph2SiꢁFe(CO)2Cp]+ (C) at 400
m/z and [(DMSO)Ph2SiꢁFe(CO)2Cp]+ (B) at 437 m/z.
Fig. 2. Positive-ion EDESI-MS map [27] of Ph2(MeO)SiFe(CO)2Cp,
run in a mobile phase of MeCN–MeOH with traces of DMSO
present.
larger than the corresponding parameters in Ph3SiSPh
,
(2.156
A and 99.5°, respectively [25]), but the
differences are not chemically significant.
3.2. Ph2(MeO)SiFe(CO)2Cp (1b)
This compound was readily prepared in good yield
from Na[OMe] and Ph2ClSiFe(CO)2Cp in THF, show-
ing no tendency to undergo base cleavage of the FeꢀSi
bond [26]. The spectroscopic characterisation given in
Section 2 is as expected and only the ESMS merit
further discussion. When run in MeCN–H2O at low
cone voltages, the main peaks observed corresponded
Scheme 1.

242
J.S. McIndoe, B.K. Nicholson / Journal of Organometallic Chemistry 648 (2002) 237–245
positioned so that it can interact with the silicon atom,
stabilising a silylene centre if one is formed. Ogino has
shown that ortho-CH2NMe2 groups can stabilise iron–
silylenes using this approach [23].
The chloro example Ph(2-MeOC6H4)ClSiFe(CO)2Cp
was prepared analogously to Ph2ClSiFe(CO)2Cp. It
could be chromatographed on silica which suggested
that the SiꢀCl bond is relatively unreactive, and this
was confirmed by the lack of any reaction at the SiꢀCl
bond with Na[STol]. However it did react with [OMe]−
to give the corresponding methoxy-silyl species Ph(2-
MeOC6H4)(MeO)SiFe(CO)2Cp.
In the ESMS of Ph(2-MeOC6H4)(OMe)SiFe(CO)2Cp
in MeCN–H2O, the main peak at low cone voltages
was the [M+H]+ ion, formed presumably by attach-
ment at the OMe oxygen atom, together with some
solvated [M+H+H2O]+. However as the cone
voltage was increased, loss of MeOH readily occurred
to generate a strong signal from the silylene cation
[Ph(2-MeOC6H4)SiꢁFe(CO)2Cp]+, together with in-
creasing amounts of [Ph(2-MeOC6H4)SiꢁFe(CO)nCp]+
(n=1, 0) under more forcing conditions. In this case,
EDESI-MS studies of Ph(2-MeOC6H4)(OMe)SiFe-
(CO)2Cp showed that the pendant OMe group prevents
coordination of donors including acetonitrile, pyridine
and DMSO after the elimination of methanol. This
observation is strong evidence that the pendant OMe
group provides a degree of intramolecular stabilisation
by coordination to the silicon centre, certainly enough
to prevent association of other bases.
In an attempt to reproduce the ESMS results on a
preparative scale, Ph(2-MeOC6H4)(MeO)SiFe(CO)2Cp,
was sealed in an NMR tube with [Me3O]BF4 in CDCl3,
in the hope that the MeO group on silicon would be
eliminated as Me2O. However no reaction took place,
even after several days at room temperature.
The X-ray crystal structure of Ph(2-MeOC6H4)-
ClSiFe(CO)2Cp (2a) was undertaken to see if there was
any tendency for the silicon atom to become five-coor-
dinate. Fig. 4 shows that the ortho-OMe group is
orientated so that the oxygen atom is pointing towards
the silicon atom, and there are indications of a weak
bonding intramolecular interaction between these
Fig. 3. The structure of Ph2(MeO)SiFe(CO)2Cp. Bond parameters
include: Fe(1)ꢀSi(1) 2.293(3), Si(1)ꢀO(3) 1.656(5), C(3)ꢀO(3) 1.421(8),
,
Fe(1)ꢀCcp (av) 2.093(7) A; Fe(1)ꢀSi(1)ꢀO(3) 112.3(2), Si(1)ꢀO(3)ꢀC(3)
125.7(4), Fe(1)ꢀSi(1)ꢀC(11) 116.1(2), Fe(1)ꢀSi(1)ꢀC(21) 114.9(2)°.
The last two results from collisions with gas-phase
solvent molecules subsequent to methanol elimination.
C is much more intense than B now, because there is no
time for an equilibrium to be set up which the better
donor can dominate. EDESI-MS/MS [27] experiments
show that B does not derive directly from A; if A is
selected and subjected to fragmentation in an argon-
filled collision cell, B does not appear at all. At ꢀ40 V,
both CO ligands are lost simultaneously to form
[Ph2SiꢁFeCp]+ (G). A proportion of these ions react
with MeCN to form [Ph2SiꢁFeCp(MeCN)]+ (F), with
the MeCN possibly attached to the coordinatively un-
saturated iron rather than the silicon atom.
The crystal structure of complex 1b was determined
for comparison with that of Ph2(TolS)SiFe(CO)2Cp
(1a) discussed above. This is shown in Fig. 3. The two
molecules are similar overall, other than the relative
conformation of the STol/OMe groups in the respective
examples. The FeꢀSi bond lengths are not significantly
different despite the more electronegative substituent in
the OMe example, a feature that usually leads to bond
,
,
shortening. The SiꢀO bond in 1b is 1.656(5) A which is
atoms. The Si···O distance of 2.991 A is shorter than
the sum of the van der Waals radii for Si and O at 3.6
shorter than expected by comparison with the SiꢀS
,
,
bond of 1a (2.180(2) A) and the difference in covalent
radii of O and S (0.3–0.4 A), but this is not unusual for
A, though much longer than the sum of the covalent
radii of about 1.84 A. The C(21)ꢀC(26)ꢀO(3) angle is
,
,
bonds to Si where SiꢀO bonds are shortened relative to
SiꢀS by stronger p bonding. The wider SiꢀOꢀC bond of
125.7(4)° in 1b compared to the SiꢀSꢀC bond of
102.8(2)° in 1a is also usual for silyl ethers.
115.5°, suggesting the methoxy group is drawn towards
the silicon to shorten the O···Si distance. The Cl on the
Si atom lies opposite the site where the O···Si interac-
tion is developing. Comparison with the structure of
MeC(O)N(SiMe2Cl)2 which has SiꢀCl bonds to both
four- and five-coordinate silicon atoms [28] shows that
3.3. Ph(2-MeOC6H4)XSiFe(CO)2Cp (X=Cl, OMe)
,
the value of 2.115 A for the SiꢀCl distance in Ph(2-
These compounds were prepared because they have
an ortho-OMe group on one of the aryl rings that is
MeOC6H4)ClSiFe(CO)2Cp is longer than the SiꢀCl
,
bond to the tetrahedral silicon atom (2.050 A) though

J.S. McIndoe, B.K. Nicholson / Journal of Organometallic Chemistry 648 (2002) 237–245
243
methyl and phenyl examples are not obvious, but are
presumably attributable to the steric protection pro-
vided by the phenyl groups. The molecule is very
crowded and perhaps is prevented from adopting a
conformation where a migration reaction is possible, if
this is involved in the decomposition process.
The carbonyl region infrared spectra of O[SiPh2-
{Co(CO)4}]2 showed doubling of the normal pattern
found for terminal Co(CO)4 groups. This is unlikely to
arise from coupling between the two ends of the
molecule and suggests perhaps that two conformers
exist in solution. To determine the solid state structure
an X-ray study was carried out. The crystal contains
discrete molecules, as shown in Fig. 5. As expected, the
backbone of the molecule consists of two silicon atoms,
both tetrahedrally coordinated, joined in a SiꢀOꢀSi
linkage. Each silicon atom is further bonded to two
phenyl rings and to a cobalt atom which has trigonal
bipyramidal coordination, the silicon occupying an ax-
ial site. One parameter of interest is the SiꢀOꢀSi angle
of 153.4(2)°, which is surprisingly bent given the fact
that other bulky disiloxanes (e.g. O(SiPh3)2 [30],
O[SiMeF{Fe(CO)2Cp}]2 [31], O[Si(OBut)3]2 [32] and
O[SiCl2{C(SiMe3)3}]2 [33]) all show linear SiꢀOꢀSi
bonds. It is generally assumed that the linear arrange-
ment is stabilised by Si···O p-bonding, and consistent
with these ideas, the SiꢀO bond length in O[SiPh2-
Fig. 4. The structure of Ph(2-MeOC6H4)ClSiFe(CO)2Cp. Bond
parameters include: Fe(1)ꢀSi(1) 2.2698(2), Si(1)ꢀCl(1) 2.1151(7),
,
,
Fe(1)ꢀCcp (av) 2.093(2) A, Si(1)···O(3) 2.991 A; Fe(1)ꢀSi(1)ꢀCl(1)
108.09(3), C(26)ꢀO(3)ꢀC(3) 118.8(2), C(21)ꢀC(26)ꢀO(3) 115.5(2)°.
much shorter than the SiꢀCl trans to O on the five-co-
,
ordinate Si (2.348 A). This is significant, since a major
factor determining the tendency for five-coordination at
silicon is the capacity for a SiꢀX bond to be stretched,
which is facile for X=Cl [29]. Finally, the sum of the
angles about the silicon atom for the potentially equa-
torial groups [FeꢀSiꢀC(21), FeꢀSiꢀC(31) and
C(21)ꢀSiꢀC(31)] is 342.2°, midway between tetrahedral
(328°) and planar (360°). All of this points to a weak
bonding O···Si interaction, poised to stabilise the silicon
centre once the trans group is lost in the mass
spectrometer.
,
{Co(CO)4}]2 of 1.630(3) A is closer to that in O(SiH3)2
,
(SiꢀOꢀSi 142.2(3)°, SiꢀO 1.634
A [34]) than to
,
O[SiMeF{Fe(CO)2Cp}]2 (SiꢀOꢀSi 180°, SiꢀO 1.603 A
[31]).
The FeꢀSi bond in Ph(2-MeOC6H4)ClSiFe(CO)2Cp is
,
The CoꢀSi bonds (2.348 A average) are relatively
,
2.2698(6) A, surprisingly shorter than even Ph2FSiFe-
long, with previous values for simple R3SiCo(CO)4
,
(CO)2Cp (2.278(1) A [24]). Based on electronegativity
,
ranging from 2.254 to 2.38 A [35]. This is presumably
grounds alone, a bond length comparable to that of
because of the bulk of the substituents in 3.
,
Ph2(MeO)SiFe(CO)2Cp (2.292(2) A) might have been
expected. This may be related to the small distortion
towards trigonal bipyramidal geometry, with the Fe in
a pseudo-equatorial site, induced by the ortho-OMe
group interaction.
3.4. O[SiPh2{Co(CO)4}]2
The early report of O[SiMe2{Co(CO)4}]2 described it
as extremely unstable, decomposing slowly even at
−78 °C [9]. Since 1,2 silyl-migration reactions of
R5Si2Co(CO)4 have been shown to be extremely facile
[3], it was of interest to synthesise a stable example of a
heteroatom-substituted silyl–cobalt compound. Ac-
cordingly, we reacted O(SiPh2H)2 with Co2(CO)8 and
obtained good yields of O[SiPh2{Co(CO)4}]2. This
showed none of the instability of the methyl analogue,
since it was readily isolated. The solid was even stable
for considerable periods in air, although in solution it
proved to be moderately air-sensitive. The reasons for
the remarkable difference in behaviour between the
Fig. 5. The structure of O[SiPh2(Co(CO)4)]2. Bond parameters in-
clude: Co(1)ꢀSi(1) 2.341(2), Co(2)ꢀSi(2) 2.354(2), Si(1)ꢀO(9) 1.630(3),
,
Si(2)ꢀO(9) 1.629(3) A; Si(1)ꢀO(9)ꢀSi(2) 153.4(2)°.

244
J.S. McIndoe, B.K. Nicholson / Journal of Organometallic Chemistry 648 (2002) 237–245
(m) M. Suginome, Y. Ito, J. Chem. Soc. Dalton Trans. (1998)
In the solid state, molecules of O[SiPh2{Co(CO)4}}2
1925;
adopt a gauche configuration for the two cobalt-con-
taining groups, with interlocking of CO groups and Ph
substituents. A possible explanation for the complex IR
spectrum in solution is that there is also an anti
configuration which interchanges slowly on the IR time
scale, giving rise to doubled peaks. This seems more
likely than coupling of vibrational modes across the
two halves of the molecule.
(n) Y. Zhang, F. Cervantes-Lee, K.H. Pannell, J. Am. Chem.
Soc. 122 (2000) 8327;
(o) S. Sharma, K.H. Pannell, Organometallics 19 (2000) 1225;
(p) S. Sharma, K.H. Pannell, Organometallics 20 (2001) 7.
[6] (a) M.D. Curtis, P.S. Epstein, Adv. Organomet. Chem. 19 (1981)
213;
(b) M.D. Curtis, J. Greene, J. Am. Chem. Soc. 100 (1978) 6362;
(c) M.D. Curtis, J. Greene, W.M. Butler, J. Organomet. Chem.
164 (1979) 371;
(d) L.G. Bell, W.A. Gustavson, S. Thanedar, M.D. Curtis,
Organometallics 2 (1983) 740.
[7] M. Van Tiel, K.M. Mackay, B.K. Nicholson, J. Organomet.
Chem. 462 (1993) 79.
4. Supplementary material
[8] D.W. Smith, Inorganic Substances, Cambridge University Press,
Cambridge, 1990, p. 186.
[9] J. Greene, M.D. Curtis, Inorg. Chem. 17 (1978) 2324.
[10] A.J. Chalk, J.F. Harrod, J. Am. Chem. Soc. 87 (1965) 16.
[11] (a) P. Braunstein, M. Knorr, J. Organomet. Chem. 500 (1995)
21;
(b) P. Braunstein, M. Knorr, Coord. Chem. Rev. 180 (1998) 903;
(c) P. Braunstein, M.X. Morise, Organometallics 17 (1998) 540.
[12] (a) W. Malisch, R. Lankat, O. Fey, J. Reising, S. Schmitzer, J.
Chem. Soc. Chem. Commun. (1995) 1917;
Crystallographic data for the structural analyses have
been deposited with the Cambridge Crystallographic
Data Centre, CCDC nos. 172420–172423 for 1a, 1b, 2a
and 3, respectively. Copies of this information may be
obtained free of charge from the Director, CCDC, 12
Union Road, Cambridge CB2 1EZ, UK (Fax: +44-
1223-336033; e-mail: deposit@ccdc.cam.ac.uk or www:
http://www.ccdc.cam.ac.uk).
(b) W. Malisch, R. Lankat, S. Schmitzer, J. Reising, Inorg.
Chem. 34 (1995) 5701;
(c) W. Malisch, K. Hindahl, H. Kab, J. Reising, W. Adam, F.
Prechtl, Chem. Ber. 128 (1995) 963;
Acknowledgements
(d) W. Malisch, S. Schmitzer, R. Lankat, M. Neumayer, F.
Prechtl, W. Adam, Chem. Ber. 128 (1995) 1251.
[13] (a) C.E.F. Rickard, W.R. Roper, D.M. Salter, L.J. Wright, J.
Am. Chem. Soc. 114 (1992) 9682;
We thank the University of Waikato for financial
support and for a postgraduate scholarship (to J.S.M.).
We thank Associate Professor C.E.F. Rickard and A.
Oliver, University of Auckland, and Professor W.T.
Robinson and Dr J. Wikaira, University of Canterbury,
for X-ray data sets.
(b) K. Hubler, P.A. Hunt, S.M. Maddock, C.E.F. Rickard,
W.R. Roper, D.M. Salter, P. Schwerdtfeger, L.J. Wright,
Organometallics 16 (1997) 5076;
(c) M.T. Attar-Bashi, C.E.F. Rickard, W.R. Roper, L.J. Wright,
S.D. Woodgate, Organometallics 17 (1998) 504.
[14] (a) G. Calzaferri, R. Imhof, K.W. Tornroos, J. Chem. Soc.
Dalton Trans. (1993) 3741;
References
(b) M. Rattay, D. Fenske, P. Jutzi, Organometallics 17 (1998)
2930.
[1] H.K. Sharma, K.H. Pannell, Chem. Rev. 95 (1995) 1351.
[2] T.D. Tilley, in: S. Patai, Z. Rappoport (Eds.), The Chemistry of
Organic Silicon Compound, Wiley, Chichester, 1989 Chapter 24.
[3] J.S. McIndoe, B.K. Nicholson, J. Organomet. Chem. 577 (1999)
181.
[4] G.P. Mitchell, T.D. Tilley, J. Am. Chem. Soc. 120 (1998) 7635.
[5] (a) K.H. Pannell, J.R. Rice, J. Organomet. Chem. 78 (1974) C35;
(b) H. Tobita, K. Ueno, H. Ogino, Chem. Lett. (1990) 369;
(c) K.H. Pannell, J. Cervantes, C. Hernandez, J. Cassias, S.
Vincenti, Organometallics 5 (1986) 1056;
[15] J.F. Harrod, E. Pelletier, Organometallics 3 (1984) 1064.
[16] H.C. Clark, A.T. Rake, J. Organomet. Chem. 74 (1974) 29.
[17] T.J. Marks, A.F. Seyam, J. Organomet. Chem. 31 (1971) C63.
[18] R.E. Dessy, R. Kornmann, C. Smith, R.G. Hayter, J. Am.
Chem. Soc. 90 (1968) 2001.
[19] R.H. Blessing, Acta Crystallogr. A51 (1995) 33.
[20] G.M. Sheldrick, SHELX97 Programs for X-ray Crystallography,
University of Go¨ttingen, Go¨ttingen, Germany, 1997.
[21] W. Henderson, L.J. McCaffrey, B.K. Nicholson, Polyhedron 17
(1998) 221.
(d) H. Tobita, K. Ueno, H. Ogino, Chem. Lett. (1986) 1777;
(e) H. Tobita, K. Ueno, H. Ogino, J. Am. Chem. Soc. 112 (1990)
3415;
(f) K. Ueno, H. Tobita, H. Ogino, J. Organomet. Chem. 430
(1992) 93;
[22] (a) S.D. Grumbine, R.K. Chadha, T.D. Tilley, J. Am. Chem.
Soc. 114 (1992) 1518;
(b) S.D. Grumbine, D.A. Straus, T.D. Tilley, Polyhedron 14
(1995) 127;
(c) See also: S.K. Grumbine, G.P. Mitchell, D.A. Straus, T.D.
Tilley, A.L. Rheingold, Organometallics 17 (1998) 5607.
[23] H. Kobayashi, K. Ueno, H. Ogino, Organometallics 14 (1995)
5490.
(g) K. Ueno, H. Ogino, Bull. Chem. Soc. Jpn. 68 (1995) 1955;
(h) K. Ueno, K. Nakano, H. Ogino, Chem. Lett. (1996) 459;
(i) K.H. Pannell, M.C. Brun, H. Sharma, K. Jones, S. Sharma,
Organometallics 13 (1994) 1075;
[24] U. Schubert, G. Kraft, E. Walther, Z. Anorg. Allg. Chem. 519
(1984) 96.
(j) G.P. Mitchell, T.D. Tilley, G.P.A. Yap, A.L. Rheingold,
Organometallics 14 (1995) 5472;
[25] H. Schumann, S. Jurgis, E. Hahn, J. Pickhardt, J. Blum, M.
Eisen, Chem. Ber. 118 (1985) 2738.
[26] M. Hofler, J. Scheuren, J. Organomet. Chem. 55 (1973) 177.
(k) G.P. Mitchell, T.D. Tilley, Organometallics 15 (1996) 3477;
(l) K.H. Pannell, H.K. Sharma, R.N. Kapoor, F. Cervantes-Lee,
J. Am. Chem. Soc. 119 (1997) 9315;

J.S. McIndoe, B.K. Nicholson / Journal of Organometallic Chemistry 648 (2002) 237–245
245
[27] (a) P.J. Dyson, B.F.G. Johnson, J.S. McIndoe, P.R.R. Lan-
gridge-Smith, Rapid Commun. Mass Spectrom. 14 (2000) 311;
(b) P.J. Dyson, A.K. Hearley, B.F.G. Johnson, J.S. McIndoe,
P.R.R. Langridge-Smith, C. Whyte, Rapid Commun. Mass
Spectrom. 15 (2000) 895.
[28] K.D. Onan, A.T. McPhail, C.H. Yoder, R.W. Hillyard, J. Chem.
Soc. Chem. Commun. (1978) 209.
[29] R.J.P. Corriu, J.C. Young, in: S. Patai, Z. Rappoport (Eds.),
The Chemistry of Organic Silicon Compounds, Wiley,
Chichester, 1989, p. 5.
[31] K. Suwinska, G.J. Palenik, R. Gerdil, Acta Crystallogr. 42C
(1986) 615.
[32] W. Ries, T. Albright, J. Silvestre, I. Bernal, W. Malisch, C.
Burska, Inorg. Chim. Acta 111 (1986) 119.
[33] C. Eaborn, P.B. Hitchcock, P.D. Lickiss, A.D. Taylor, J.
Organomet. Chem. 398 (1990) 59.
[34] A. Almenningen, O. Bastienson, V. Ewing, K. Hedberg, M.
Traetteberg, Acta Chem. Scand. 17 (1963) 17.
[35] K.M. Mackay, B.K. Nicholson, in: G. Wilkinson, F.G.A. Stone,
E.W. Abel (Eds.), Comprehensive Organometallic Chemistry,
vol. 6, Pergamon, Oxford, 1982, pp. 1043–1114.
[30] H. Burger, Angew. Chem. Int. Ed. Engl. 12 (1973) 474.