Hexacoordinate silacyclobutane dichelate complexes: Structure, properties, and ligand crossover

Article abstract of DOI:10.1021/om101081u

Hexacoordinate dichelate silacyclobutane complexes have been synthesized from dichlorosilacyclobutane and O-trimethylsilylated hydrazides by transsilylation. Like previously reported hexacoordinate silicon complexes, they readily and quantitatively undergo ligand exchange with other silicon compounds (XSiCl₃ and differently substituted O-trimethylsilylated hydrazides), evidence that ionic dissociation does not play a significant role in the exchange mechanism. Germanium tetrachloride causes central-element exchange and formation of analogous hexacoordinate germanium complexes. Likewise, silicon tetrachloride replaces germanium from its hexacoordinate complexes, obeying certain selectivity constraints. When silicon complexes have strongly electron-withdrawing chelate-ring substituents (CF₃ or CH₂CN), GeCl₄ causes, in addition to central-element exchange, also oxidative opening of the four-membered ring and addition of two chlorine atoms. Both chelate exchange and central-element exchange are shown to be dominated by monodentate ligand priorities.

Full text of DOI:10.1021/om101081u

Organometallics 2011, 30, 405–413 405
DOI: 10.1021/om101081u
Hexacoordinate Silacyclobutane Dichelate Complexes: Structure,
Properties, and Ligand Crossover
Shiri Yakubovich,‡ Boris Gostevskii,‡ Inna Kalikhman,‡ Mark Botoshansky,§
Leonid E. Gusel’nikov,†, Vadim A. Pestunovich,z, and Daniel Kost*,‡
‡Department of Chemistry, Ben-Gurion University, Beer Sheva 84105, Israel, §Department of Chemistry,
Technion-Israel Institute of Technology, Haifa 32000, Israel, †Institute of Petrochemical Synthesis,
RAS, Moscow, Russia, and zIrkutsk Institute of Organic Chemistry, RAS,
Siberian Division, Russia. Deceased.
Received November 16, 2010
Hexacoordinate dichelate silacyclobutane complexes have been synthesized from dichlorosilacy-
clobutane and O-trimethylsilylated hydrazides by transsilylation. Like previously reported hexa-
coordinate silicon complexes, they readily and quantitatively undergo ligand exchange with other
silicon compounds (XSiCl3 and differently substituted O-trimethylsilylated hydrazides), evidence
that ionic dissociation does not play a significant role in the exchange mechanism. Germanium
tetrachloride causes central-element exchange and formation of analogous hexacoordinate germa-
nium complexes. Likewise, silicon tetrachloride replaces germanium from its hexacoordinate complexes,
obeying certain selectivity constraints. When silicon complexes have strongly electron-withdrawing
chelate-ring substituents (CF3 or CH2CN), GeCl4 causes, in addition to central-element exchange,
also oxidative opening of the four-membered ring and addition of two chlorine atoms. Both chelate
exchange and central-element exchange are shown to be dominated by monodentate ligand priorities.
Introduction
trigger for the ligand exchange reactions.6 To test this hypo-
thesis, a series of silacyclobutane complexes have been synthe-
sized, in which ionic dissociation is effectively prevented by
the presence of two nonionizable carbon ligands in the four-
membered silacyclobutane ring.
The chemistry of silacyclobutane has attracted consider-
able attention because of its unique reactivity1 and potential
as a reactant in polymer formation.2 Neutral hexacoordinate
silicon dichelates based on hydrazide ligands3 have previ-
ously been reported to readily undergo complete exchange of
ligands with their differently substituted precursors (eq 1),
between two differently substituted complexes (eq 2)4 and
with germanium analogues.5 Since bidentate ligands rapidly
and quantitatively migrate from one silicon atom to another
and to germanium (eq 3), it seemed plausible that the well-
documented reversible ionic dissociation of these complexes
in chloroform and dichloromethane solutions might be the
*To whom correspondence should be addressed. E-mail: kostd@
bgu.ac.il.
(1) For recent studies on silacyclobutanes see: (a) Gusel’nikov, L. E.
Coord. Chem. Rev. 2003, 244, 149–240. (b) Gusel'nikov, L. E.; Avakyan, V. G.;
Guselnikov, S. L. J. Am. Chem. Soc. 2002, 124, 662–671. (c) Pestunovich,
V. A.; Lazareva, N. F.; Albanov, A. I.; Kozyreva, O. B.; Volkova, V. V.;
Gusel'nikov, L. E. ARKIVOC 2006, 116–125. (d) Troegel, D.; Lippert, W. P.;
€
Moller, F.; Burschka, C.; Tacke, R. J. Organomet. Chem. 2010, 695, 1700–
1707.
(2) Gallei, M.; Klein, R.; Rehahn, M. Macromolecules 2010, 43,
1844–1854, and references therein.
(3) For reviews see: (a) Kost, D.; Kalikhman, I. Adv. Organomet.
Chem. 2004, 50, 1–106. (b) Kost, D.; Kalikhman, I. Acc. Chem. Res. 2009,
42, 303–314.
(4) Sergani, S.; Kalikhman, I.; Yakubovich, S.; Kost, D. Organome-
tallics 2007, 26, 5799–5802.
(5) Yakubovich, S.; Kalikhman, I.; Kost, D. J. Chem. Soc., Dalton
Trans. 2010, 39, 9241–9244.
(6) Kost, D.; Kingston, V.; Gostevskii, B.; Ellern, A.; Stalke, D.;
Walfort, B.; Kalikhman, I. Organometallics 2002, 21, 2293–2305, and
other papers in the series “Donor Stabilized Silyl Cations”.
r
2011 American Chemical Society
Published on Web 01/10/2011
pubs.acs.org/Organometallics

406 Organometallics, Vol. 30, No. 3, 2011
Yakubovich et al.
Results and Discussion
Synthesis and Structure. Like other polychlorosilanes,3,7
dichlorosilacyclobutane (1) reacts smoothly with silylated
hydrazides (2) to form neutral hexacoordinate dichelate
complexes (3) in high yields (eq 4). When only one molar equi-
valent of 2 is used, the major isolated product is the penta-
coordinate monochelate (4). The only previously reported
analogues of 3 are dichelates with N-alkylideneimino donor
groups (5, eq 5), which spontaneously rearrange to the tricyclic
complexes (6),8 rather than the present dimethylamino donor
groups.
Figure 1. ORTEP representation of the molecular structure of
3a in the crystal, depicted at the 30% probability level and
omitting hydrogen atoms. Crystallographic disorder in the CF3
groups is removed for clarity.
The silacyclobutane complexes 3 and 4 were characterized
by their various solution NMR spectra and elemental anal-
yses, and the structures of five of them were also determined
by single-crystal X-ray diffraction analysis. The molecular
structures in the solid state of compounds 3a-d and the
monochelate 4c are depicted in Figures 1-5, respectively,
and selected bond lengths and angles are listed in Table 1.
Compounds 3a-d feature distorted octahedral geometries,
with the nitrogen ligands in trans position relative to each
other and the oxygen atoms cis, as is commonly found in
similar dichelate silicon complexes.3
Figure 2. ORTEP representation of the molecular structure of
3b in the crystal, depicted at the 30% probability level and
omitting hydrogen atoms. Only symmetry-unique atoms have
been labeled.
It may be noteworthy that the N-methyl signals in the 1H
and 13C NMR spectra of compounds 3 appear as one singlet
at ambient temperature in CDCl3 solution (see Experimental
Section), despite the C2 molecular symmetry, which requires
two singlets. Only when cooled to lower temperature (260 K)
do the singlets in 3b and 3d split into two signals in each case.
(7) Kalikhman, I. D.; Gostevskii, B. A.; Bannikova, O. B.; Voronkov,
M. G.; Pestunovich, V. A. J. Organomet. Chem. 1989, 376, 249.
(8) Gostevskii, B.; Kalikhman, I.; Tessier, C. A.; Panzner, M. J.;
Youngs, W. J.; Kost, D. Organometallics 2005, 24, 5786–5788.

Article
Organometallics, Vol. 30, No. 3, 2011 407
Figure 5. ORTEP representation of the molecular structure of
4c in the crystal, depicted at the 30% probability level and
omitting hydrogen atoms.
Figure 3. ORTEP representation of the molecular structure of 3c
in the crystal, depicted at the 30% probability level and omitting
hydrogen atoms. Only symmetry-unique atoms have been labeled.
substituents these complexes have lower barriers for enantio-
merization compared to 3b and 3d, with the electron-releasing
substituents. This is rationalized by noting that electron-
withdrawing substituents weaken the coordination of the
nitrogen donor to silicon.10 As a result, it is easier to extend
this bond to facilitate the twist of the two chelate rings next to
each other in the course of helicity reversal.
The C-Si-C bond angles in the four-membered silacy-
clobutane rings are substantially smaller than 90°, ranging
between 75.6° and 78.1° (see Table 1). In the pentacoordinate
4c, the four-membered ring occupies axial and equatorial
positions about the silicon, presumably to be able to keep
a near 90° bond angle; however, in this case the actual
C-Si-C angle is also in the same range, 77.3°. This small
bond angle allows for a much greater C-C-C angle in the
ring, near 100°. Examination of other, literature reported,11
silacyclobutane crystal structures confirms that, in all cases,
regardless of the silicon coordination number, the C-Si-C
angles are generally near 80°, while the C-C-C bond angles
are near 100°, with various degrees of silacyclobutane ring-
puckering. These angles are geometrically required in order
to accommodate the long Si-C and shorter C-C bonds in
the four-membered ring.
Interestingly, in the pentacoordinate monochelate 4c the
chloro ligand occupies an equatorial position, while carbon
takes the axial position. This uncommon arrangement en-
ables the small C-Si-C angle and avoids excessive ring strain.
Figure 4. ORTEP representation of the molecular structure of
3d in the crystal, depicted at the 30% probability level and
omitting hydrogen atoms.
This indicates a dynamic process, which takes place on the
“NMR time scale” and renders the two diastereotopic N-methyl
groups equivalent. The only reasonable process that can
invert the chirality about the silicon center and cause rapid
equilibration of the N-methyl groups is an interchange of the
oxygen ligands by twisting of the O-Si-O plane through a
“bicapped tetrahedron” intermediate or transition state,9 such
that the chelate rings reverse their helical spatial arrangement,
causing enantiomerization of the complex. The fact that no
splitting of methyl signals is observed down to 260 K in
3a and 3c is evidence that with the electron-withdrawing
(10) (a) Gostevskii, B.; Adear, K.; Sivaramakrishna, A.; Silbert, G.;
Stalke, D.; Kocher, N.; Kalikhman, I.; Kost, D. Chem. Commun. 2004,
1644–1645. (b) Gostevskii, B.; Silbert, G.; Adear, K.; Sivaramakrishna, A.;
Stalke, D.; Deuerlein, S.; Kocher, N.; Voronkov, M. G.; Kalikhman, I.; Kost,
D. Organometallics 2005, 24, 2913–2920.
(11) (a) C-Si-C angle in tetracoordinate silacyclobutane, 79°:
Daiss, J. O.; Penka, M.; Burschka, C.; Tacke, R. Organometallics
2004, 23, 4987–4994. (b) C-Si-C angles in tetracoordinate silacyclobu-
tanes, 79.7°, 80.7°: Peckham, T. J.; Nguyen, P.; Bourke, S. C.; Wang, Q.;
Harrison, D. G.; Zoricak, P.; Russell, C.; Liable-Sands, L. M.; Rheingold,
A. L.; Lough, A. J.; Manners, I. Organometallics 2001, 20, 3035–3043.
(c) C-Si-C angle in tetracoordinate silacyclobutane, 82.6°: Jain, R.;
Brunskill, A. P. J.; Sheridan, J. B.; Lalancette, R. A. J. Organomet. Chem.
2005, 690, 2272–2277. (d) Diequatorial C-Si-C angle in pentacoordinate
silacyclobutane, 82.4°: Gericke, R.; Gerlach, D.; Wagler, J. Organometallics
2009, 28, 6831–6834. (e) Axial, equatorial C-Si-C angles in pentacoordi-
nate silacyclobutanes, 77.3°, 77.8°: Spiniello, M.; White, J. M. Organome-
tallics 2000, 19, 1350–1354.
(9) For a discussion of rapid ligand exchange through a bicapped
tetrahedron see: (a) Kost, D.; Kalikhman, I.; Raban, M. J. Am. Chem.
Soc. 1995, 117, 11512–11522. (b) Kost, D.; Kalikhman, I.; Krivonos, S.;
Stalke, D.; Kottke, T. J. Am. Chem. Soc. 1998, 120, 4209–4214.

408 Organometallics, Vol. 30, No. 3, 2011
Yakubovich et al.
The sum of equatorial bond angles amounts to 359.47°, i.e., a
nearly perfect silicon-centered equatorial plane.
Ligand Exchange Reactions. Compounds 3a-e were
allowed to react with ZSiCl3 (7a-e) in CDCl3 solutions in 5
mm NMR tubes (eq 6). With Z = Ph, Cl, and H (7c-e), the
exchange takes place at ambient temperature and is essentially
complete within minutes. With Z = alkyl groups (7a, 7b),
the exchange is slower and generally requires immersion in an
80 °C oil bath for 1 h.
The outcome of the ligand exchange reactions was deter-
mined by their various NMR spectra: the disappearance of
signals due to starting materials and parallel growth of prod-
uct signals, in particular formation of 1. It turns out that
all of the reactants (7) caused complete ligand exchange and
the reactions proceeded smoothly with formation of 1. This is
not too surprising, in view of the previously established “ligand
priority list”:4 since silacyclobutane has two alkyl ligands,
which have the lowest ligand priority, they are readily replaced
by the two incoming ligands of overall higher priority (a chloro
and either a carbon, hydrido, or another chloro ligand).
These results clearly indicate that ionic dissociation is not
an essential feature of the ligand-exchange reactions. In fact,
ionic dissociation is irrelevant, since the present silacyclobu-
tane complexes undergo ligand exchange just as easily and
rapidly as the previously reported chloro complexes.4 In
addition, the results strongly suggest that no Si-C bond
cleavage takes place during ligand exchange: the fact that in
allofthe reactions dichlorosilacyclobutane(1) isformedasthe
only product of the silacyclobutane moiety during ligand
exchange rules out ring-opening, since it would be highly
unlikely to expect that, after opening, the four-membered ring
would spontaneously and quantitatively close again to form1.
A possible mechanistic insight into the exchange described
in eq 6 is gained from the reaction of 3b (R = t-Bu) and 7e
(Z = H), as demonstrated in Figure 6. The 29Si NMR spectra
of the reaction products are shown. 7e was chosen because it
allows immediate recognition of compounds with Si-H
bonds, apparent by their characteristic doublet resulting
from large one-bond Si-H coupling. The reaction was
carried out in two fashions, first (Figure 6A) with excess
(more than 2 molar equiv) of 3b. This caused exhaustive
transfer of chelate rings (i.e., bidentate ligands) to 7e, result-
ing in exclusive appearance of the Si-H doublet in the
hexacoordinate dichelate 8 (R = t-Bu, Z = H). The excess
3b is manifest in the formation of the pentacoordinate mono-
chelate 4b (see eq 4), an intermediate formed from 3b after the
donation of a single chelate ring, which does not react further
to transfer another chelate ring, because there are no more 7e
or 9 (see Figure 6B) molecules available to accept the transfer.
Conversely, when the reaction is carried out applying an
excess of 7e, a different pentacoordinate monochelate inter-
mediate (9, Figure 6B) is observed: the one resulting from
transfer of one chelate ring to 7e. In this case, the hydrogen
atom remains attached to silicon, as is evident from the
doublet at ca. -74 ppm, typical of pentacoordination. In

Article
Organometallics, Vol. 30, No. 3, 2011 409
Figure 6. 29Si NMR spectra of the reaction of 3b with 7e in CDCl3 solution: (A) with an excess of 3b, the pentacoordinate intermediate
4b is formed, along with hexacoordinate 8 (R = t-Bu, Z = H); (B) with an excess of 7e, the pentacoordinate intermediate 9 accumulates;
and (C) same solution after 24 h; 9 has disappeared possibly by disproportionation to 7e and 8.
addition, some of the hexacoordinate exchange product (8)
and a large excess of 7e are found in the spectrum. Interest-
ingly, after 24 h at ambient temperature, the intermediate has
completely disappeared (Figure 6C), probably through dis-
proportionation to 7e and 8, suggesting that the hexacoordi-
nate 8 is thermodynamically more favorable than the inter-
mediate.
These results indicate that exchange occurs by transfer of
whole chelate rings: either one or both rings, from one silicon
atom to another. This must be accompanied by simultaneous
back-transfer of chloro ligands from the accepting silicon
atom to the one donating the chelate ring. In other words, it
may be concluded again that exchange does not involve
silicon-carbon bond cleavage.
A similar ligand-exchange reaction is also found between
hexacoordinate complexes 3a-c and silylated hydrazides 2
(eq 7). The net effect of this exchange is the replacement of
the ring-substituents R in complexes 3. Substituent priorities
in this reaction are similar to those reported previously for
regular complexes 8;4 namely, the more electron-withdrawing
substituents R (R = CF3) are replaced from their complexes
(3a) by less electron-withdrawing groups (t-Bu, in 2b). This
apparently opposite priority of ring substituents in comparison
with monodentate ligands leads to the thermodynamically
most stable hexacoordinate complexes.
The stepwise transfer of bidentate ligands from one silicon
compound to another is demonstrated again in Figure 7 by
the reaction of a silacyclobutane complex, 3a, with the
silylated hydrazide, 2b. Figure 7A shows the 29Si NMR
spectrum of the reaction mixture immediately after dissolu-
tion. The major components are the starting materials 2b and
3a, with traces of the exchange products (2a and 3b) visible.
In Figure 7B the same reaction is carried out with an excess
of 3a. The major product is the mixed complex 10, with one
CF3 and one t-Bu ring substituent, and only a trace of the
doubly exchanged di-tert-butyl complex 3b. This may be an
indication of the instability of the trifluoromethyl-substituted
3a (as previously reported for trifluoromethyl complexes):4,10
when both 3a and the mixed 10 are available, during the
course of the reaction, for further chelate exchange with 2b,
3a is more reactive and hence exchanges with 2b preferen-
tially over 10, causing accumulation of 10. In other words,
the complex with two CF3 groups wants to get rid of its
substituted chelate ring more strongly than 10, which has
already been partly stabilized by exchange of one of the
chelate rings. However, when the reaction is run under an
excess of 2b (Figure 7C), there is no competition for the
substituted chelate rings, and all of the trifluoromethyl-
substituted material (3a and 10) is quantitatively converted
to the tert-butyl-substituted complex 3b.
The preference of tert-butyl as substituent in hexacoordi-
nate complexes 3 over trifluoromethyl is seen also in the
reverse reaction: when 3b reacts with 2a, even when the latter
is present in great excess, only the mixed complex 10 (R1 =
CF3, R2 = t-Bu) is formed. No trace of the doubly exchanged
complex, 3a, is found.
Similar results were obtained also in the exchange reac-
tions of 3a and 2c: either all of the CF3-substituted complex
(3a) was converted to the phenyl analogue 3c, or, in the
presence of excess 3a, the major product was a mixed com-
plex (10, R1 = CF3, R2 = Ph). In the reverse reaction, 3c
with 2a, the only product observed was the mixed complex
10. CF3 was unable to replace both of the phenyl substituents
in 3c, even after hours of reflux and excess of 2a.

410 Organometallics, Vol. 30, No. 3, 2011
Yakubovich et al.
Figure 7. 29Si NMR spectra of the exchange reaction of 3a with 2b in CDCl3 solution: (A) immediately after dissolution; (B) with an
excess of 3a, featuring accumulation of the mixed complex 10; (C) with an excess of 2b; only the completely exchanged product (3b) is
observed.
Another ligand exchange reaction (eq 8) was attempted
and is described below. Diphenyldichlorosilane (11) has been
known to form relatively weak hexacoordinate complexes.12
When subjected to the exchange reaction with hexacoordi-
nate complexes 3a-c, ligand transfer proceeded only to the
first stage, transfer of one of the chelate rings to 11 to form
12, and modifying the reactant 3, after loss of one of the
chelate rings, to 4. Even in the presence of two molar equi-
valents of 3 relative to 11 and after 2 h at chloroform reflux
temperature, the monochelates 12b and 12c did not take up a
second chelate moiety to form the expected dichelate hex-
acoordinate complexes. No exchange reaction was observed
in a solution of 3a with 11.
than other complexes and tends to have at least one of the
chelate rings replaced with a more favorable chelate. Thus, in
the case of 3a reacting with 3b, the symmetrical reactants
almost completely disappear to form the mixed complex (10).
If one of the reactants is present in excess, the other one
completely disappears, within NMR detection limits, to give
way to the mixed complex and to the excess reactant, pre-
sumably to allow for maximum stabilization of the less stable
complex (i.e., 3a) by formation of the mixed complex 10.
Central-Element Exchange. Compounds 3b-d react read-
ily with GeCl4, with replacement of the central silicon by
germanium to form 13b-d (eq 10). The germanium com-
pounds were characterized by comparison with authentic
samples or by analogy of their 1H and 13C NMR spectra with
those of differently substituted authentic compounds.5 Thus,
when 3b reacts with GeCl4 in CDCl3 solution for 20 min at
ambient temperature, the high-field 29Si NMR signal charac-
teristic of the hexacoordinate reactant 3b disappears, and
instead the signal of dichlorosilacyclobutane is observed. The
structure is further confirmed by the 1H and 13C NMR signals,
which have shifted relative to those in 3b, but remain within the
range expected for hexacoordinate complexes (see Experimen-
tal Section). Similar exchange takes place also with 3c and 3d.
One may safely assume that, like in the other exchange
reactions, the process involves consecutive transfer of chelat-
ing (bidentate) ligand units from silicon to germanium.
The pentacoordinate products 4b, 4c, 12b, and 12c were
identified by their typical 29Si NMR chemical shifts at -41.7,
-43.7, -49.1, and -50.9 ppm, respectively, and by compar-
ison with authentic samples and with the crystallographically
characterized 4c.
Ligand Interchange between Two Complexes. Ligand
exchange takes place also when two hexacoordinate com-
plexes are mixed in solution.4 When two differently sub-
stituted silacyclobutane complexes 3 are mixed in CDCl3
solution, after a short while at ambient temperature a third
signal develops in the 29Si NMR spectrum, corresponding
to the mixed dichelate (eq 9). The product distribution
depends on the nature of the ring substituents R: when
these do not differ substantially in electron-withdrawing
power, such as two alkyl substituents, the product dis-
tribution is roughly the statistical 1:2:1, with 10 being the
major product. However, there appears to be a driving
force favoring the mixed complex even beyond the statis-
tical bias: a symmetrically substituted complex with elec-
tron-withdrawing substituents, such as 3a, is less stable
The reaction of 3a and 3e, with the electron-withdrawing
ring substituents, is quite different. In these two reactions the
central-element exchange is accompanied by a simultaneous
(12) Kalikhman, I.; Gostevskii, B.; Botoshansky, M.; Kaftory, M.;
Tessier, C. A.; Panzner, M. J.; Youngs, W. J.; Kost, D. Organometallics
2006, 25, 1252–1258.

Article
Organometallics, Vol. 30, No. 3, 2011 411
Table 2. Crystallographic Data and Experimental Parameters for the Structure Analyses of 3a-d and 4c
3a 3b 3c 3d
4c
795678
C12H17ClN2OSi
268.82
120(2)
monoclinic
P21/n
7.4164(15)
11.718(2)
15.959(3)
90
99.620(4)
90
1367.4(5)
4
1.306
CCDC no.
795674
11H18F6N4O2Si
380.38
200(2)
triclinic
P1
795675
795676
795677
empirical formula
form mass, g mol-1
collection T, K
cryst syst
C
C17H35N4O2Si
355.58
293(2)
monoclinic
C2/c
C21H27N4O2Si
395.56
120(2)
monoclinic
C2/c
C13H30N6O2Si
330.52
120(2)
triclinic
P1
space group
˚
a, A
7.7482(7)
9.5333(9)
12.3898(12)
75.284(2)
79.816(2)
72.996(2)
841.25(14)
2
15.667(3)
15.259(3)
8.712(2)
90
90.70(2)
90
2082.6(7)
4
1.134
19.877(6)
8.133(2)
13.351(4)
90
114.675(6)
90
1961.4(10)
4
1.340
8.9359(10)
9.4458(10)
11.0692(12)
80.800(2)
84.202(2)
70.193(2)
866.63(16)
2
˚
b, A
˚
c, A
R, deg
β, deg
γ, deg
3
˚
V, A
Z
F
calcd, Mg/m3
F(000)
1.502
392
1.267
360
780
844
568
θ range, deg
no. of coll reflns
no. of indep reflns
Rint
no. of reflns used
no. of params
Goof
2.29-28.50
8738
4158
0.0198
4158
273
1.86-25.41
7468
1887
0.0610
1887
110
2.26-30.13
8117
2858
0.0185
2858
130
1.87-29.05
6702
4534
0.0167
4534
2.17-27.10
5832
2930
0.0187
2930
156
207
1.008
0.996
1.025
1.009
1.037
R1 wR2 [I > 2σI)]
0.0409
0.0889
0.0519
0.0967
0.388/-0.291
0.0564
0.1568
0.0767
0.1705
0.423/-0.375
0.0397
0.1146
0.0436
0.1181
0.358/-0.387
0.0441
0.1065
0.0527
0.1136
0.0361
0.0996
0.0441
0.1060
0.673/-0.380
R1 wR2 (all data)
max./min. res electron
0.688/-0.348
-3
˚
dens, e A
oxidation and ring-opening reaction, as shown in eq 11. In
this reaction germanium oxidizes the four-membered ring by
addition of two chlorine atoms and ring-opening. The
products 14a and 14e were identified by the various NMR
spectra, although not isolated. In particular, the appearance
of three new signals in the 1H and 13C NMR spectra, shown
(by 1H-decoupling and DEPT experiments) to belong to
three CH2 groups, suggested the structure assigned to 14.
resonances of 15 and 16 are shifted substantially to lower
field, relative to the silacyclobutane compounds (3) and other
hexacoordinate silicon compounds. It is concluded that these
compounds are in dynamic equilibrium in solution with
lower coordination species, as represented in eq 13. This is in
agreement with a recent report by Wagler and co-workers,13
in which similar low-field 29Si NMR resonances had indi-
cated similar equilibria. The low coordination may be attrib-
uted to the low electron-withdrawing power of the cycloalkyl
ligands, which is insufficient to render silicon electrophilic
enough to bind to the donor atoms. The silacyclobutane com-
plexes (3) are exceptional in their relatively high-field 29Si
NMR chemical shifts, probably due to their “ring strain release
Lewis acidity”,11d,14 which prefers the smaller bond angles
associated with higher (five or six) coordination numbers.
To study the generality of this ring-opening reaction, ana-
logous complexes were synthesized with silacyclopentane (15a-
c) and silacyclohexane (16a-c) ring systems. In neither of these
complexes was ring-opening and oxidation observed, and ex-
change of the central element proceeded smoothly (eq 12). It may
be concluded that both ring strain and the presence of electron-
withdrawing susbstituents contribute to the tendency of com-
plexes 3a and 3e to undergo the oxidation reaction.
€
(13) Brendler, E.; Wachtler, E.; Wagler, J. Organometallics 2009, 28,
5459–5465.
(14) (a) Denmark, S. E. J. Org. Chem. 2009, 74, 2915–2927.
(b) Denmark, S. E.; Regens, C. S. Acc. Chem. Res. 2008, 41, 1486–
1499.
Examination of the 29Si NMR spectra of the silacycloalk-
ane compounds (3, 15, and 16, see Table 3) shows that the

412 Organometallics, Vol. 30, No. 3, 2011
Yakubovich et al.
Table 3. Comparison of 29Si NMR Chemical Shifts (ppm) of Silacycloalkane Dichelate Complexes 3a-c, 15a-c, and 16a-c
compd indexa
ring substituenta
silacyclobutane (3)
silacyclopentane (15)
silacyclohexane (16)
a
c
b
CF3
phenyl
t-Bu
-105.5
-111.2
-114.1
-33.3
-44.7
-14.3
-8.7
-16.1
-25.6
a Listed in order of decreasing electron-withdrawing power.
Bis(N-(dimethylamino)benzimidato-N0,O)silacyclobutane (3c).
3c was prepared as described for 3a, from 2c (1.26 g, 5.3 mmol)
and 0.37 g (2.6 mmol) of 1. Yield: 0.99 g (95%), mp 159-161 °C.
1H NMR (CDCl3): δ 1.50 (m, 2H, CH2), 1.90 (m, 4H, CH2), 3.02
(s, 12H, NCH3), 7.28-7.84 (m, 10H, Ph). 13C NMR (CDCl3): δ
12.80, 32.74 (CH2), 49.80 (NCH3), 127.34, 127.91, 130.54, 131.99
(Ph), 164.59 (CdN). 29Si NMR (CDCl3): δ -111.2. Anal. Calcd
for C21H28N4O2Si: C, 63.60; H, 7.12; N, 14.12. Found: C, 63.13; H,
6.99; N, 13.69.
Bis(N-(dimethylamino)N0-(dimethyl)glicineimidato-N00,O)-
silacyclobutane (3d). 3d was prepared as described for 3a, from 2d
(1.31 g, 6.5 mmol) and 0.45 g (3.2 mmol) of 1. Yield: 1.01 g (95%),
mp 120-122 °C. 1H NMR (CDCl3): δ 1.25 (m, 2H, CH2), 1.66 (m,
4H, CH2), 2.66 (s, 12H, NCH3), 2.69 (s, 12H, NCH3). 13C NMR
(CDCl3): δ 12.72, 32.95 (CH2), 36.32 (CNCH3), 50.55 (NNCH3),
162.62 (CdN). 29Si NMR (CDCl3): δ -112.7. Anal. Calcd for
C13H30N6O2Si: C, 47.24; H, 9.15; N, 25.43. Found: C, 47.41; H,
9.43; N, 26.04.
Bis(N-(dimethylamino)cyanoacetimidato-N0,O)silacyclobutane
(3e). 3e was prepared as described for 3a, from 2e (1.18 g,
5.9 mmol) and 0.42 g (2.9 mmol) of 1. Yield: 0.78 g (82%), mp
128-130 °C. 1H NMR (CDCl3):δ1.31 (m, 2H, CH2), 1.67 (m, 4H,
CH2), 2.72 (s, 12H, NCH3), 3.16 (s, 4H, NCCH2). 13C NMR
(CDCl3): δ 12.20, 32.29 (CH2), 21.55 (NCCH2), 49.27 (NCH3),
114.36 (CtN), 160.17 (CdN). 29Si NMR (CDCl3): δ -109.4.
Anal. Calcd for C13H22N6O2Si: C, 48.42; H, 6.88; N, 26.06. Found:
C, 48.43; H, 7.04; N, 26.19.
Chloro(N-(dimethylamino)pivaloimidato-N0,O)silacyclobutane
(4b). 4b was prepared as described for 3a, from 2b (0.66 g, 3.0
mmol) and 0.43 g (3.1 mmol) of 1. Yield: 0.72 g (96%). 1H NMR
(CDCl3): δ 1.21 (s, 9H, (CH3)3), 1.71 (m, 2H, CH2), 1.90 (m, 4H,
CH2), 2.54 (s, 6H, NCH3). 13C NMR (CDCl3): δ 11.40, 33.51
(CH2), 26.56 (C(CH3)3), 35.00 (C(CH3)3), 46.99 (NCH3), 174.04
(CdN). 29Si NMR (CDCl3): δ -41.2.
Within the families of silacyclobutane (3a-c) and silacy-
clohexane (16a-c) complexes in Table 3 the trend in each
group makes good sense: the stronger the electron-withdraw-
ing power of the ring substituent, the weaker the donor pro-
perty of the corresponding NMe2 group, resulting in weaker
NfSi coordination and hence lower field average chemical
shift. In contrast, the trend fails for the silacyclopentane
compounds, in which 15b stands out with its unexpected
relatively low field resonance. This may be due to some direct
steric repulsion between the bulky tert-butyl groups and the
silacyclopentane ring, which may lack the flexibility attributed to
cyclohexane rings, interfering with proper chelate ring closure.
The change in chemical shift is much greater in 16a-c than
it is in 3a-c, probably because the latter are essentially
hexacoordinate, and hence do not change substantially with
changing substituents. In 16, however, the change of substit-
uents causes much greater changes in the equilibrium position
of eq 13, resulting in greater changes in 29Si chemical shifts.
Experimental Section
The reactions were carried out under dry argon using Schlenk
techniques. Solvents were dried and purified by standard meth-
ods. NMR spectra were recorded on a Bruker Avance DMX-
500 spectrometer operating at 500.13, 125.76, and 99.36 MHz,
respectively, for 1H, 13C, and 29Si spectra. Spectra are reported
in δ (ppm) relative to TMS, as determined from standard resid-
ual solvent proton (or carbon) signals for 1H and 13C and
directly from TMS for 29Si. NMR spectra were measured at
297 K, unless otherwise reported. Melting points were measured
in sealed capillaries using a Buchi melting point instrument and
are uncorrected. Elemental analyses were performed by ME-
DAC Ltd., Chobham, Surrey, UK. Single-crystal X-ray diffrac-
tion measurements were performed on a Bruker Smart Apex on
a D8-goniometer. Crystallographic details are listed in Table 2.
Crystallographic data have been deposited with the Cambridge
Crystallographic Data Centre (CCDC). The CCDC numbers
are listed in Table 2. Product 13a has been reported previously.5
Bis(N-(dimethylamino)trifluoroacetimidato-N0,O)silacyclobutane
(3a). A mixture of 1.32 g (5.8 mmol) of 2a and 0.41 g (0.3 mmol) of
1 in 5 mL of chloroform was stirred at room temperature for 2 h.
The volatiles were removed under reduced pressure (0.1 mmHg),
and the residue was washed with 5 mL of n-hexane. A single crystal
for X-ray analysis was grown from n-hexane. Yield: 1.00 g (92%),
mp 105-106 °C. 1H NMR (CDCl3): δ 1.40 (m, 2H, CH2), 1.79 (m,
4H, CH2), 2.83 (s, 12H, NCH3). 13C NMR (CDCl3): δ 12.23, 32.45
(CH2), 49.22 (NCH3), 117.29 (q, 1J(F-C) = 277 Hz, CF3), 156.99
(q, 2J(F-C) = 37 Hz, CdN). 29Si NMR (CDCl3): δ -105.5. Anal.
Calcd for C11H18F6N4O2Si: C, 34.73; H, 4.77; N, 14.73. Found: C,
34.68; H, 5.21; N, 14.99.
Chloro(N-(dimethylamino)benzimidato-N0,O)silacyclobutane
(4c). 4c was prepared as described for 3a, from 2c (1.27 g, 7.4
mmol) and 1.04 g (7.4 mmol) of 1. A single crystal for X-ray
analysis was grown from diethyl ether. Yield of crystals: 0.72 g
(50%), mp 102-104 °C. 1H NMR (CDCl3): δ 1.39-1.78 (m, 6H,
CH2), 2.62 (s, 6H, NCH3), 7.34-7.96 (m, 5H, Ph). 13C NMR
(CDCl3): δ 11.52, 33.64 (CH2), 47.55 (NCH3), 127.51, 128.65,
129.44, 131.70 (Ph), 163.52 (CdN). 29Si NMR (CDCl3): δ -40.2.
Anal. Calcd for C12H17ClN2OSi: C, 53.62; H, 6.37; N, 10.42.
Found: C, 53.20; H, 6.45; N, 10.36.
Chloro(N-(dimethylamino)pivaloimidato-N0,O)diphenylsilane
(12b). 12b was prepared as described for 3a, from 2b (0.28 g,
1.3 mmol) and 0.54 g (2.1 mmol) of 11. Yield: 0.42 g (90%).
1H NMR (CDCl3): δ 1.34 (s, 9H, (CH3)3), 1.90 (s, 6H, NCH3),
7.42-7.80 (m, 10H, Ph). 13C NMR (CDCl3): δ 27.05 (C(CH3)3),
35.80 (C(CH3)3), 48.44 (NCH3), 171.00 (CdN), 127.75, 128.20,
129.56, 131.62, 131.81, 132.68, 133.92, 136.28 (Ph). 29Si NMR
(CDCl3): δ -49.0.
Chloro(N-(dimethylamino)benzimidato-N0,O)diphenylsilane (12c).
12c was prepared as described for 3a, from 2c (0.46 g, 2.0 mmol)
and 0.52 g (2.1 mmol) of 11. Yield: 0.64 g (89%). H NMR
Bis(N-(dimethylamino)pivaloimidato-N0,O)silacyclobutane (3b).
3b was prepared as described for 3a, from2b (1.23 g, 5.7 mmol) and
0.38 g (2.8 mmol) of 1. Yield: 0.98 g (98%), mp 133-135 °C.
1H NMR (CDCl3): δ 1.04 (s, 18H, (CH3)3), 1.28 (m, 2H, CH2),
1.61 (m, 4H, CH2), 2.74 (s, 12H, NCH3). 13C NMR (CDCl3):
δ 12.64, 32.55 (CH2), 27.43 (C(CH3)3), 35.13 (C(CH3)3),
49.46, 49.64 (NCH3), 174.08 (CdN). 29Si NMR (CDCl3):
δ -114.1. Anal. Calcd for C17H36N4O2Si: C, 57.26; H, 10.18; N,
15.71. Found: C, 56.95; H, 10.45; N, 16.16.
1
(CDCl3): δ 1.93 (s, 6H, NCH3), 7.26-8.02 (m, 15H, Ph).
13C NMR (CDCl3): δ 48.21 (NCH3), 164.61 (CdN), 127.79,
127.94, 128.03, 128.37, 129.85, 131.04, 131.97, 131.64 (Ph). 29Si
NMR (CDCl3): δ -51.0.
Bis(N-(dimethylamino)pivaloimidato-N0,O)dichlorogermanium
(13b). 13b was prepared as described for 3a, from 2b (0.61 g,

Article
Organometallics, Vol. 30, No. 3, 2011 413
2.9 mmol) and 0.32 g (1.5 mmol) of GeCl4. Yield: 0.61 g (94%).
1H NMR (CDCl3): δ 1.08 (s, 18H, (CH3)3), 2.98, 3.07 (2s, 12H,
NCH3). 13C NMR (CDCl3): δ 26.99 (C(CH3)3), 35.98 (C(CH3)3)
50.64, 50.68 (NCH3), 173.64 (CdN).
Bis(N-(dimethylamino)benzimidato-N0,O)dichlorogermanium
(13c). 13c was prepared as described for 3a, from 2c (0.52 g,
2.2 mmol) and 0.24 g (1.1 mmol) of GeCl4. Yield: 0.47 g (91%).
1H NMR (CDCl3): 3.34 (s, 12H, NCH3), 7.37-7.92 (m, 10H,
Ph). 13C NMR (CDCl3): δ 51.06, 51.12 (NCH3), 127.47, 128.20,
130.95, 131.48 (Ph), 163.93 (CdN).
Bis(N-(dimethylamino)N0-(dimethyl)glicineimidato-N00,O)dichloro-
germanium (13d). 13d was prepared as described for 3a, from 2d
(0.41 g, 2.0 mmol) and 0.23 g (1.1 mmol) of GeCl4. Yield: 0.38 g
(89%). 1H NMR (CDCl3): δ 2.72 (s, 12H, CNCH3), 2.93, 3.05
(2s, 12H, NNCH3). 13C NMR (CDCl3): δ 35.64 (CNCH3),
52.07, 52.23 (NNCH3), 161.13 (CdN).
Bis(N-(dimethylamino)cyanoacetimidato-N0,O)dichlorogermanium
(13e). 13e was prepared as described for 3a, from 2e (0.40 g,
2.0 mmol) and 0.24 g (1.1 mmol) of GeCl4. Yield: 0.38 g (87%).
1H NMR (CDCl3): δ 3.04, 3.10 (2s, 12H, NCH3), 3.32 (s, 4H,
NCCH2). 13C NMR (CDCl3): δ 22.42 (NCCH2), 50.61, 50.75
(NCH3), 113.47 (CtN), 160.46 (CdN).
Bis(N-(dimethylamino)pivaloimidato-N0,O)silacyclopentane (15b).
15b was prepared as described for 3a, from 2b (0.67 g, 3.1 mmol)
and 0.24 g (1.6 mmol) of dichlorosilacyclopentane. The volatiles
were removed under reduced pressure (0.1 mmHg), and the oily
colorless residue was washed with 5 mL of n-hexane. Yield:
0.50 g (86%). 1H NMR (CDCl3): δ 0.26-0.37, 1.42-1.55 (2 m,
8H, CH2), 1.06 (s, 18H, (CH3)3), 2.56 (s, 12H, NCH3). 13C NMR
(CDCl3): δ 11.77, 22.83 (CH2), 27.46 ((CH3)3), 35.51 (C(CH3)3),
50.22 (NCH3), 172.67 (CdN). 29Si NMR (CDCl3): δ -14.3.
Bis(N-(dimethylamino)benzimidato-N0,O)silacyclopentane (15c).
15c was prepared as described for 3a, from 2c (0.60 g, 2.5 mmol)
and 0.20 g (1.3 mmol) of dichlorosilacyclopentane. The volatiles
were removed under reduced pressure (0.1 mmHg), and the
white solid residue was washed with 5 mL of n-hexane. Yield:
1
0.45 g (84%), mp 80-83 °C. H NMR (CDCl3): δ 0.61-0.69,
1.53-1.63 (2 m, 8H, CH2), 2.69 (s, 12H, CH3), 7.27-7.92
(m, 10H, Ph). 13C NMR (CDCl3): δ 11.60, 23.36 (CH2), 49.25
(NCH3), 127.21, 127.97, 130.43, 132.59 (Ph), 160.66 (CdN). 29Si
NMR (CDCl3): δ -44.7. Anal. Calcd for C22H30N4O2Si: C,
64.36; H, 7.36; N, 13.65. Found: C, 64.00; H, 7.38; N, 13.66.
Bis(N-(dimethylamino)trifluoroacetimidato-N0,O)silacyclo-
hexane (16a). 16a was prepared as described for 3a, from2a (0.87g,
3.8 mmol) and 0.32 g (1.9 mmol) of dichlorosilacyclohexane. The
volatiles were removed under reduced pressure (0.1 mmHg), and
the oily colorless residue was washed with 5 mL of n-hexane. Yield:
Bis(N-(dimethylamino)trifluoroacetimidato-N0,O)-3-(chloro)-
propylchlorosilicon(IV) (14a). A mixture of 0.32 g (0.8 mmol) of
3a and 0.28 g (1.3 mmol) of GeCl4 in 5 mL of chloroform was
stirred at room temperature for 1 h. The volatiles were removed
under reduced pressure (0.1 mmHg), and the residue was washed
with 5 mL of n-hexane. Two products, 14a and 13a, were
obtained and identified by their NMR spectra (13a by analogy
with an authentic sample) in the mixture. An excess of GeCl4
was added, resulting in quantitative conversion of 14a to 13a.
14a: 1H NMR (CDCl3): δ 1.19-1.28 (m, 2H, CH2), 1.90
1
0.66 g (84%). H NMR (CDCl3): δ 1.02-1.80 (m, 10H, CH2),
2.60 (s, 12H, CH3). 13C NMR (CDCl3): δ 15.31, 24.47, 29.47
1
(CH2), 46.35 (CH3), 117.46 (q, J(F-C) = 277 Hz, CF3), 140.81
(q, 2J(F-C) = 37.4 Hz, CdN). 29Si NMR (CDCl3): δ -8.7.
Bis(N-(dimethylamino)pivaloimidato-N0,O)silacyclohexane
(16b). 16b was prepared as described for 3a, from 2b (0.67 g,
3.1 mmol) and 0.24 g (1.6 mmol) of dichlorosilacyclohexane.
The volatiles were removed under reduced pressure (0.1 mmHg),
and the oily colorless residue was washed with 5 mL of n-hexane.
Yield: 0.28 g (82%). 1H NMR (CDCl3):δ 1.33-1.76 (m, 10H, CH2),
1.10 (s, 18H, (CH3)3), 2.30 (s, 12H, NCH3). 13C NMR (CDCl3):
δ 14.78, 25.11, 29.10 (CH2), 27.70 ((CH3)3), 36.35 (C(CH3)3),
47.19 (NCH3), 166.86 (CdN). 29Si NMR (CDCl3): δ -25.6.
Bis(N-(dimethylamino)benzimidato-N0,O)silacyclohexane (16c).
16c was prepared as described for 3a, from 2c (0.52 g, 2.2 mmol)
and 0.20 g (1.2 mmol) of dichlorosilacyclohexane. The volatiles
were removed under reduced pressure (0.1 mmHg), and the oily
colorless residue was washed with 5 mL of n-hexane. Yield: 0.43
g (86%). 1H NMR (CDCl3): δ 1.01-1.88 (m, 10H, CH2), 2.52
(s, 12H, CH3), 7.28-8.02 (m, 10H, Ph). 13C NMR (CDCl3):
δ 14.47, 24.84, 29.08 (CH2), 46.99 (CH3), 127.09, 127.96, 130.05,
133.42 (Ph), 155.47 (CdN). 29Si NMR (CDCl3): δ -16.1.
3
3
(t, J(C-H) = 10 Hz, 2H, CH2), 2.04 (t, J(C-H) = 10 Hz, 2H,
CH2), 2.99 (b, 12H, NCH3). 13C NMR (CDCl3): δ 22.20, 29.87,
35.62 (CH2), 51.03 (NCH3), 116.93 (q, 1J(F-C) = 277 Hz, CF3),
156.42 (q, 2J(F-C) = 37 Hz, CdN). 29Si NMR: δ -126.7.
Bis(N-(dimethylamino)cyanoacetimidato-N0,O)-3-(chloro)pro-
pylchlorosilicon(IV) (14e). 14e was prepared as described for
14a, from 3e (0.51 g, 1.6 mmol) and 1.70 g (7.9 mmol) of GeCl4.
14e: 1H NMR (CDCl3): δ 1.19-1.29 (m, 2H, CH2), 1.89
3
3
(t, J(C-H) = 10 Hz, 2H, CH2), 2.04 (t, J(C-H) = 10 Hz, 2H,
CH2), 2.91 (b, 12H, CH3), 3.26 (s, 4H, NCCH2). 13C NMR
(CDCl3): δ 22.31, 30.05, 35.78 (CH2), 21.33 (NCCH2), 52.40
(NCH3), 113.42 (CtN), 159.88 (CdN). 29Si NMR (CDCl3):
δ -128.16.
Bis(N-(dimethylamino)trifluoroacetimidato-N0,O)silacyclo-
pentane (15a). 15a was prepared as described for 3a, from 2a
(0.74 g, 3.3 mmol) and 0.25 g (1.6 mmol) of dichlorosilacyclo-
pentane. The volatiles were removed under reduced pressure
(0.1 mmHg), and the oily gray residue was washed with 5 mL of
n-hexane. Yield: 0.59 g (91%). 1H NMR (CDCl3): δ 0.58-0.62,
1.55-1.58 (2 m, 8H, CH2), 2.60 (s, 12H, NCH3). 13C NMR
(CDCl3): δ 11.28, 23.17 (CH2), 48.40 (NCH3), 117.35 (q, 1J(F-C)
= 277 Hz, CF3), 150.09 (q, 2J(F-C) = 37.4 Hz, CdN). 29Si NMR
(CDCl3): δ -33.3. Anal. Calcd for C12H20F6N4O2Si: C, 36.54;
H, 5.11; N, 14.21. Found: C, 36.06; H, 5.50; N, 14.12.
Acknowledgment. Financial support by the Israel Science
Foundation, grant No. ISF-242/09, is gratefully acknowl-
edged.
Supporting Information Available: Crystallographic data for
3a-d and 4c in the form of CIF files are available free of charge
via the Internet at http://pubs.acs.org.