Reactions of the disilane Me3SiSiCl3 with P-Chlorophosphaalkenes: Transient and persistent per-silylated phosphaalkenes

Article abstract of DOI:10.1002/ejic.201000283

Reactions of P-chlorophosphaalkenes (RMe₂Si)2C=PCl (1a: R = Me; 1b: R = Ph) with the disilane Me₃SiSiCl₃ (5) furnish diphosphenes (Cl₃Si)(RMe₂Si)2C-P=P-C(SiCl3)(SiMe ₂R)₂ (4a: R

Full text of DOI:10.1002/ejic.201000283

FULL PAPER
DOI: 10.1002/ejic.201000283
Reactions of the Disilane Me3SiSiCl3 with P-Chlorophosphaalkenes: Transient
and Persistent Per-Silylated Phosphaalkenes
Cristina Mitrofan,[a] Roxana M. Bîrzoi,[a] Delia R. Bugnariu,[a] Jens Mahnke,[a]
Antje Riecke,[a] Emma Dürr (née Seppälä),[a] Wolf-W. du Mont,*[a] Peter G. Jones,[a] and
Heinrich Marsmann*[b]
Keywords: Disilanes / Phosphaalkenes / Diphosphenes / Dichlorosilylene / 29Si-NMR
Reactions of P-chlorophosphaalkenes (RMe2Si)2C=PCl (1a: R
= Me; 1b: R = Ph) with the disilane Me3SiSiCl3 (5) furnish
diphosphenes (Cl3Si)(RMe2Si)2C–P=P–C(SiCl3)(SiMe2R)2 (4a:
R = Me; 4b: R = Ph) by Me3SiCl elimination. The structure of
the new compound 4b was confirmed by X-ray diffraction; it
displays crystallographic inversion symmetry. Monitoring the
reactions with 31P- and 29Si-NMR spectroscopy detected P-
(trichlorosilyl)phosphaalkenes (RMe2Si)2C=PSiCl3 (2a, R =
Me; 2b, R = Ph) as the primary intermediates from reductive
P-silylation of 4a, 4b, and P-[(trichlorosilyl)phosphanyl]phos-
phaalkenes (RMe2Si)2C=P–P(SiCl3)C(SiCl3)(SiMe2R)2 (3a: R
= Me; 3b: R = Ph) as unsymmetric dimerisation products that
rearrange to provide 4a, 4b in step III of the reaction se-
quence. This step (the PǞC 1,3-trichlorosilyl shift reaction)
was mimicked by the synthesis of (Me3Si)2C=P–P(SiCl3)tBu
(7), which rearranges into an unsymmetric diphosphene
tBuP=PC(SiMe3)2SiCl3 (8). The bulkier P-chlorophosphaalk-
ene (iPrMe2Si)2C=PCl (1c) reacts with 5, eliminates Me3SiCl
and thereby provides the first persistent acyclic per-silylated
phosphaalkene (iPrMe2Si)2C=PSiCl3 (2c) in an incomplete
reaction. 2c exhibits an exceptionally large NMR coupling
1J(31P,29Si) = Ϯ249 Hz. Within weeks, the mixtures of 1c and
2c undergo decomposition with loss of the P=C functions.
Introduction
Most of the significant advances in the chemistry of complexes of nucleophilic carbenes with silicon dihalides
stable carbenes and group 14 carbene analogues in the last were isolated,[7] and a carbene adduct of SiCl2 was found
three decades have been based on the use of sterically and to act as source of reactive SiCl2 species.[7b] As a proton-
electronically stabilising substituents.[1] In contrast, syn- free alternative to HSiCl3, we have reported the use of tri-
thetic progress to simple dihalogenosilylenes has been very chlorosilyltrimethylgermane (Me3GeSiCl3)[8] as a reagent
limited since Timms’ high-temperature and matrix stud- for the transfer of SiCl2 moieties to P-phosphanylphos-
ies.[2,3] Oligomeric silanes of empirical formula SiCl2 are phaalkenes that leads under very mild reaction conditions
synthetically available, but they do not behave as sources of either to insertion of two equivalents of SiCl2 into the P=C
[5]
monomeric SiCl2,[4] and the alternative use of Si2Cl6 or bond, or to a diphosphene with remote C–SiCl2–PiPr2 func-
[6]
HSiCl3 as precursors is very limited. Only very recently, tions (Scheme 1).[9]
Scheme 1. Dichlorosilylene transfer from Me3GeSiCl3 to a P-phosphanylphosphaalkene.[9]
[a] Institut für Anorganische und Analytische Chemie der
No intermediates of this reaction were observed spectro-
Technischen Universität, Braunschweig,
scopically. A related diphosphene with remote C–SiCl3
Postfach 3329, 38023 Braunschweig, Germany
Fax: +49-531-3915387
E-mail: w.du-mont@tu-bs.de
[b] Anorganische und Analytische Chemie, Universität Paderborn,
Warburger Strasse 100, 33098 Paderborn, Germany
functions, and also the corresponding germanium com-
pound, were isolated by Zanin, Karnop, et al. in a previous
study on reactions of the P-chlorophosphalkene (Me3Si)2-
View this journal online at
wileyonlinelibrary.com
4462
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2010, 4462–4469

Per-Silylated Phosphaalkenes
Scheme 2. Expected reaction path of the reaction of 1a with a dichlorosilylene source.[10]
C=PCl (1a) with Si2Cl6 and with the GeCl2 dioxane com- rates of steps I–III and especially to enhance the rate of
plex.[10] In the course of these reactions in a heated solvent, step I.
very small 31P-NMR AX patterns were observed, but after 3. Independent efficient access to, and unambiguous char-
complete consumption of 1a, only the NMR signals of the acterisation of, a P-[(trichlorosilyl)phosphanyl]phosphaalk-
diphosphene products remained. The intermediacy of a per- ene related to 3a, which may rearrange to another diphos-
silylated phosphaalkene (Me3Si)2C=PSiCl3 (2a) in the case phene by the postulated 1,3-trichlorosilyl shift reaction
1a/Si2Cl6 was expected by analogy to the reductive tri- (step III).
chlorosilylations of dialkylchlorophosphanes R2PCl and
bulky alkyldichlorophosphanes RPCl2 with Si2Cl6, provid-
Results and Discussion
ing silylphosphanes R2PSiCl3 and RP(SiCl3)2 by elimi-
nation of silicon tetrachloride,[11,8b] but the existence of 2a
was not unambiguously demonstrated. The weak 31P-NMR
AX pattern was tentatively assigned to a transient dimeric
species, the P-phosphanylphosphaalkene (Me3Si)2C=P–P-
NMR-Study of an Alternative Route to a P-[(Trichlorosilyl)-
phosphanyl]phosphaalkene Related to 3a
In a variation of McDiarmid’s trichlorosilylation of
(SiCl3)C(SiCl3)(SiMe3)2 (3a), which was assumed to re- Me4P2 with Si2Cl6 furnishing two equivalents of Me2P-
arrange by a trichlorosilyl 1,3-PǞC shift reaction (step III, SiCl3,[15] Si2Cl6 was found to act as a reductive silylating
Scheme 2) to (Cl3Si)(Me3Si)2C–P=P–C(SiCl3)(SiMe3)2 (4a), agent towards dialkylchlorophosphanes R2PCl and bulky
an inversion-symmetric diphosphene. A similiar pathway alkyldichlorophosphanes RPCl2, providing silylphosphanes
presumably leads to the C-trichlorogermyl diphosphene R2PSiCl3 and RP(SiCl3)2.[10] This facile conversion of P–Cl
(Cl3Ge)(Me3Si)2C–P=P–C(GeCl3)(SiMe3)2.
to P–SiCl3 functions had prompted the attempt to convert
Using bulkier persistent germylenes and stannylenes, pri- P-chlorophosphaalkene (Me3Si)2C=PCl (1a)[16] into the un-
mary products from carbene-like insertions into the P–Cl known (Me3Si)2C=PSiCl3 (2a); this led, however, experi-
bond of 1a were identified spectroscopically,[12] but the de- mentally via postulated 3a to isolated 4a.[10]
tailed pathway of the Zanin/Karnop silylation and ger-
In our search for an alternative access to P-[(trichloro-
silyl)phosphanyl]phosphaalkenes related to the intermediate
mylation reaction is still an unsolved problem.[10]
Open questions concerning the surprising formation of 3a, we considered exploiting the (trichlorosilyl)phosphane
4a from 1a are: is 2a really an intermediate of the reaction method, which is known to convert the chlorophospha-
(step I in Scheme 2), and, assuming that the role of 2a were alkene 1a into P-phosphanylphosphaalkenes (Me3Si)2C=
–
confirmed experimentally: is 2a formed by Cl–/SiCl3 ex- P–PR2 by reactions with (trichlorosilyl)phosphanes
change or by SiCl2 transfer ?[13] Finally, the proposed con- R2PSiCl3;[17] the selective 1:1 reaction of a bis(trichloro-
stitution of 2a and 3a has yet to be subjected to experimen- silyl)phosphane RP(SiCl3)2 with 1a should provide the cor-
tal evidence by 29Si-NMR including analysis of NMR cou- responding
P-[(trichlorosilyl)phosphanyl]phosphaalkene
plings J(31P,29Si). To obtain access to compounds related to (Me3Si)2C=P–P(R)SiCl3. The availability of tBuP(Si-
the elusive intermediates 2a and 3a from the (Me3Si)2C=P– Cl3)2,[11] either from the Si2Cl6 method or the HSiCl3/NEt3
Cl/Si2Cl6 reaction, we chose the following experimental pathway,[11c] prompted us to choose its the 1:1 reaction with
variations of the system:
1a, which indeed led straightforwardly by elimination of
1. Replacement of Si2Cl6 by the unsymmetric disilane SiCl4 to the desired [tert-butyl(trichlorosilyl)phosphanyl]
Me3SiSiCl3 (5) [analogous to the above-mentioned Me3Ge- phosphaalkene 7. This compound exhibits in 31P-NMR a
SiCl3 (6)] as a more reactive silylating agent,[8] to allow characteristic AX pattern and a large coupling constant
milder reaction conditions for the generation of the short- 1J(31P,29Si) [δ31P = 389.3, P=C; –24.9, P-Si, 1J(P,Si) =
1
lived intermediate 2a.
Ϯ112.6, J(P,P) Ϯ232.4 Hz]. Compound 7 is persistent at
2. (i) Replacement of Me3Si groups in (Me3Si)2C=P–Cl (1a) room temperature, but upon attempted distillation, heating
by bulkier iPrMe2Si groups,[14] which are expected to affect furnishes the unsymmetric diphosphene tBuP=PC(SiMe3)2-
the rates of steps II (dimerisation) and III (PǞC 1,3-silyl SiCl3 (8), which is unambiguously characterized by its 31P-
shift reaction), leading to enhanced lifetimes of intermedi- NMR pattern [δ = 604.0 and 531.3 ppm, 1J(31P,31P) =
ates related to 2a and 3a. (ii) Use of PhMe2Si groups, which Ϯ629.9 Hz]. This result supports the proposed rearrange-
are known to be less efficient than Me3Si in the “hierarchy” ment 3a Ǟ 4a (step III, Scheme 2), establishing a pathway
of silyl groups as protecting agents, in order to vary the to unsymmetric diphosphenes (Scheme 3).[18]
Eur. J. Inorg. Chem. 2010, 4462–4469
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
4463

W.-W. du Mont, H. Marsmann et al.
FULL PAPER
Scheme 3. Formation and rearrangement of a P-[(trichlorosilyl)phosphanyl]phosphaalkene.
Reductive Trichlorosilylation Reactions of Modified
Silylation of 1a with Me3SiSiCl3 (5)
P-Chlorophosphaalkenes
The reaction of 1a with 5 at room temperature proceeds
significantly faster than with 6, allowing the observation in
31P-NMR of small amounts of transient 2a (Scheme 4) and
large amounts of 3a (the main product at room tempera-
ture), accompanied by the diphosphene 4a. After the initial
appearance of some 4a at an early stage of the reaction, its
amount does not increase significantly at room tempera-
ture. Isolation of pure 3a, clearly a persistent species at
room temperature, from such reaction mixtures was not
achieved, but it was spectroscopically analysed by 31P and
29Si NMR spectroscopy. The high concentration of 3a in
the mixture allowed the first determination of its full set of
29Si-NMR parameters (see below).
Silylation of 1a with Me3GeSiCl3 (6)
The reaction of chlorophosphaalkene 1a with (tri-
chlorosilyl)germane 6 at room temperature can be followed
by 31P NMR, exhibiting an AX pattern assigned to inter-
mediate 3a and the singlet signal of the diphosphene 4a,
which slowly increases in intensity. The use of excess 6 leads
to the appearance of further (very weak) AX-patterns in
31P NMR, which are tentatively assigned to products from
SiCl2 addition to the P=C bond of 3b.[9] Compound 2a can
be observed in 31P-NMR at an early stage of the reaction
by its very weak singlet signal, which is just strong enough
to allow the resolution of 29Si satellites, and exhibits a very
large coupling constant 1J(31P,29Si) = Ϯ243.4 Hz, even
The complete conversion of the P-[(trichlorosilyl)phos-
larger in magnitude than couplings found in species with phanyl]phosphaalkene 3a into the known diphosphene 4a
P=Si double bonds.[19] This observation correlates well with requires heating of the reaction mixture. In this respect in-
1
the exceptionally large coupling constants J(31P,117,119Sn) termediate 3a resembles the model compound 7. The obser-
in Niecke’s related stannylene insertion product Ph(Me3Si)- vation that a considerable amount of 4a accompanies 3a at
C=PSn(Cl)(µ-NtBu)2SiMe2, which in 31P-NMR exhibits an early stage of the overall reaction, whereas the complete
1J(31P,117,119Sn) of Ϯ1884.9 Hz and Ϯ1972.5 Hz,[12] i.e., in conversion of 3a into 4a requires heating (i.e., involving a
a similar range to compounds with Sn=P π-bonds and to higher energy of activation for the step III rearrangement)
tin tetrahalide–trialkylphosphane complexes.[20]
needs an explanation. The source of “early” 4a may be the
Scheme 4. Products from the reactions of P-chlorophosphaalkenes 1a–c with Me3SiSiCl3 (5).
Scheme 5. Possible reaction paths from P-(trichlorosilyl)phosphaalkenes to diphosphenes.
4464
www.eurjic.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2010, 4462–4469

Per-Silylated Phosphaalkenes
direct dimerisation of 2a to (symmetric) 4a, competing with equivalent, but magnetically inequivalent CH3 groups (dia-
the dimerisation delivering (unsymmetric) 3a. This means stereomers) appear as X parts of AAЈX systems, δ =
that the proposed mechanism 2a Ǟ 3a Ǟ 4a involves a 0.88 ppm (“t”, CH3), 2.86 ppm (“t”, CH3). Their protons
1
bifurcated pathway with the additional “shortcut” 2a Ǟ 4a. also exhibit diastereomeric behavior, i.e. H-NMR presents
When the supply of 2a is exhausted, 3a remains the only two singlets in the SiCH3 region, δ = 0.7 ppm (s, CH3, 6 H)
source of 4a. The point of bifurcation might be a transient and 0.35 ppm (s, CH3, 6 H). The quaternary carbon atom
phosphanylidene-like species that can dimerise either by at δ = 40.9 ppm was assigned according to the data offered
phosphorus insertion into a P–Si bond of 2a delivering 3a, by H,C-HMBC correlation [2J(H,C) with methyl protons].
or by direct P=P bond formation delivering the diphos- An H,C-HSQC correlated spectrum also allowed the as-
phene 4a (Scheme 5).
signment of the NMR resonances corresponding to phenyl
protons, as follows: 7.45 (d, t, ortho, 2 H), 7.23–7.29 (m,
para, 1 H), 7.15–7.22 (m, meta, 2 H).
Reaction of 1b with Me3SiSiCl3 (5)
This reaction basically follows the reaction sequence cor-
Solid 4b is a inversion-symmetric diphosphene with a
responding to the Zanin–Karnop mechanism[10,13] central P=P bond length of 2.0347(4) Å and C–P=P angles
(Schemes 2 and 4) when we consider the formation of 2b of 106.83(3)°. (Figure 1). Searches of the Cambridge Datab-
and 3b as intermediate products, but it is faster than the ase (Version 1.12)[21] revealed the following results: (i) The
reaction of 1a with 5. High concentrations of intermediate fragment C–P=P–C with two-coordinate phosphorus, no
2b at an early stage of the reaction indicate that step I par- coordination to metals and no annelation appears 23 times
ticularly is accelerated when 1b is used as starting material. with P=P 1.985–2.050, av. 2.024 Å. (ii) There are only three
However, in this case there is also evidence for a competing other structures with Si3C substituents at phosphorus,
second reaction pathway leading initially to an asymmetric namely Zanin’s compound 4a,[10] our derivative with remote
diphosphene that finally delivers the symmetric diphos- C–SiCl2–PiPr2 functions[9] (Scheme 1), and the bis[tris(tri-
phene 4b, the final product and the only isolated and com- methylsilyl)methyl]diphosphene investigated independently
pletely characterized compound from the reaction of 1b by two groups;[22] all these molecules display inversion sym-
with 5.
metry and the C–P=P angles are 106.0–108.3, av. 107.4°.
In the case of 1b + 5 we observed in the reaction mixture
by 31P NMR, apart from 2b, 3b, and 4b, also the AX-
pattern of a further transient species [δ31P = 401.9 ppm (d),
1
332.3 ppm (d), J(P,P) = Ϯ597 Hz]. Both 31P-NMR shifts
are in the “phosphaalkene range”, but the magnitude of
1J(P,P) would fit to a P=P double bond. This species of
unknown structure is apparently a precursor to an asym-
1
metric diphosphene [δ31P = 561.9 (d), 502.5 (d), J(P,P) =
620.5 Hz]. The appearance of these transient species with
31P-NMR AX patterns suggests the existence of a further
reaction path (independent of the Zanin–Karnop mecha-
nism)[10,13] to final 4b. The aspect of an intermediate asym-
metric diphosphene has been neglected so far; traces of an
asymmetric diphosphene were also noticed after the sil-
1
ylation of 1a [δ31P = 569.1 ppm (d), 504.7 ppm (d), J(P,P)
= Ϯ616 Hz (P=P)].
The chemical composition of compound 4b has been
confirmed by a correct elemental analysis and by mass spec-
trometry. The mass spectrum shows a characteristic isotopic
distribution for the molecular ion {[M]+, m/z = 895.9 },
{[M – CH3]+, m/z = 880.9}, {[(M – SiCl3)+], m/z = 761}.
The base peak at m/z = 135 corresponds to the (SiMe2-
Ph)+ ion according to the isotopic distribution. 4b shows in
31P-NMR a singlet at δ = 573.6 ppm. The 29Si-NMR chemi-
cal shifts have been assigned according to an INEPT mea-
surement that allowed a correct calculation of the integrals
for the two distinct 29Si nuclei δ = –4.3 ppm (SiMe2Ph, inte-
gral 2Si) and –7.1 ppm (SiCl3, integral 1Si). Both reso-
nances (expected to be X-parts of AAЈX pattern) appear
as broad signals, ruling out the determination of coupling
Figure 1. Molecular structure of 4b (H atoms omitted). Thermal
ellipsoids are set at 50% probability. Selected bond lengths [Å]: P–
P#1 2.0347(4), P–C(9) 1.8863(8), C(9)–Si(3) 1.9663(8), C(9)–Si(1)
1.8522(9), C(10)–Si(3) 1.8762(9), C(16)–Si(3) 1.8656(9), Cl(1)–Si(1)
2.0553(3). Selected bond angles [°]:C(9)–P–P#1 106.83(3), P–C(9)–
Si(2) 109.26(4), P–C(9)–Si(3) 101.91(4), Si(1)–C(9)–P 112.13(4),
C(9)–Si(1)–Cl(2) 114.73(3), Cl(2)–Si(1)–Cl(3) 105.178(14).
Compound 4b also displays two short contacts: intramo-
lecular P···Cl1#1 3.469 Å and intramolecular H14···Cl1
2.94 Å, connecting the molecules by translation parallel to
the a axis.
Reactions of Phosphaalkene 1c with 5 and with 6
This variation of the system – a starting material with
constants with 31P. The 13C-NMR signals appear as broad- bulkier C-silyl substituents – was intended to inhibit the
ened multiplets, also providing little information about the dimerisation process 2 Ǟ 3, allowing the enrichment and
coupling constants with 31P. The two types of chemically isolation of compound 2c.
Eur. J. Inorg. Chem. 2010, 4462–4469
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
4465

W.-W. du Mont, H. Marsmann et al.
2
FULL PAPER
However, addition of germylsilane 6 to 1c leads to the signs of the two coupling constants J(P,Si) suggested by
formation of only traces of the new species 2c. Enrichment DFT calculations (see above). The 29Si nucleus of the tri-
of 2c from such mixtures was not achieved. The 31P-NMR chlorosilyl group in 2c appears at δ = –0.1 ppm as the center
signals of starting material 1c and product 2c (traces) are of a doublet [1J(P,Si) = Ϯ248.4 Hz] (see below).
accompanied by those of another P-trichlorosilyl species [δ
In mixtures containing 1c, 2c, and 5 almost 60% of the
= –93 ppm, 1J(P,Si) = Ϯ128.8, 1J(P,H) = Ϯ220 Hz] that can starting material 1c remains unconsumed for more than two
be assigned the structure (iPrMe2Si)2C(SiCl3)–PH(SiCl3). weeks, before detectable amounts of a further product start
The amount of this species (31P-NMR intensity Ͻ 5% rela- to appear in solution, indicated by an AX pattern in the
tive to that 1c) does not increase further, even after three 31P-NMR upfield region [31P NMR at 81 MHz, C6D6: δ
weeks.
1
= –117.0 ppm (d), –129 ppm (d), J(P,P) = Ϯ189 Hz]. This
Addition of disilane 5 to 1c leads to the formation of pattern is consistent with a diphosphabicyclobutane struc-
considerable amounts of the new species 2c, and with excess ture, which can be explained by a reaction between 1c and
of 5 the approximate content of 2c in the mixture (from 2c from the reaction mixture with elimination of SiCl4. Re-
31P-NMR intensities) can reach 35–45%. Further en- lated bicyclobutanes were detected by Niecke et al., when
richment of 2c from such mixtures was not achieved. The P-chlorophosphaalkenes were reduced by insertion of 1,3-
above-mentioned species (iPrMe2Si)2C(SiCl3)–PH(SiCl3) di-tert-butyl-2,2-dimethyl-1,3,2,4λ2-diazasilastannetidine, a
also appears in the reaction mixture from 1c and 5, but in cyclic aminostannylene, into the P–Cl bonds.[12] Keeping
even smaller amounts than in the previous case.
the bicyclic product from 1c/2c in solution leads within two
A careful analysis of the 29Si-satellite pattern in the 31P- months to the appearance of another species with three in-
NMR of 2c spectrum allowed us to identify only two of the equivalent phosphorus atoms (see Exp. Section).
1
2
Si,P coupling constants, J(P,Si) = Ϯ249.2 Hz and J(P,Si)
1
= Ϯ32.3 Hz. The exceptional magnitudes of J(P,Si) in 2a–
31P- and 29Si-NMR Spectroscopic Data of Compounds 2–7
2c gave rise to a quantum chemical consideration of com-
pounds related to 2a. For reasons of simplicity, P-(tri-
chlorosilyl)phosphaalkenes with less bulky C-silyl substitu-
ents were studied. The calculations were performed with
Gaussian on B3LYP/6-31G* geometry with GIAO/
BHandHLYP/B2. The calculated coupling constants of
compound (H3Si)2C=PSiCl3 are 1J(31P, 29Si) = –368, 2J(31P,
29Si) [cis-silyl groups] = –28, 2J(31P, 29Si) [trans-silyl groups]
= +32 Hz.[24]
In the three known P-(trichlorosilyl)phosphaalkenes 2a–
c, coupling constants 1J(P,Si) are generally significantly
larger that those of (trichlorosilyl)phosphane functions[8]
(as in 3a–b); they are influenced by the substitution pattern
of neighbouring C-silyl groups in the range iPrMe2Si Ͼ
Me3Si Ͼ PhMe2Si; i.e. bulkier alkyl substituents correlate
with larger 1J(P,Si) couplings. An overview of the NMR
spectroscopic data is given in Table 1.
The 31P-NMR signal of 2c (δ31P = 377.2 ppm) appears
about 33 ppm downfield shifted from the starting material
1c. The high concentration of 2c in the mixture allowed for
the first time the determination of the full set of 29Si-NMR
parameters of a “per-silylated” P-(trichlorosilyl)phospha-
alkene. The three 29Si nuclei appear as doublets as a conse-
quence of the coupling with 31P. The two C-silyl groups cis-
and trans-oriented relative to the P-bonded SiCl3 group at
the C=P double bond in P-trichlorosilyl[bis(isopropyldime-
thylsilyl)methylene]phosphane (2c) give separate doublet
signals in the 29Si-NMR spectrum. In C-bis(trimethylsilyl)-
phosphaalkenes the silyl group with the larger (by magni-
tude) coupling constant 2J(31P, 29Si) is generally assigned to
be oriented cis to the phosphorus lone pair, i.e. trans to the
substituent at phosphorus.[23] In 2c the most deshielded
Table 1. Selected 31P- and 29Si-NMR spectroscopic data of phos-
phaalkenes with P–SiCl3 moieties.
δ = 31P
(σ2,λ3)
[ppm]
δ = 31P
(σ3,λ3)
[ppm]
1J(P,P)
[Hz]
δ = 29Si(S-
iCl3) [ppm] [Hz]
1J(P,Si)
[a]
2a
2b
2c
3a
3b
7
371.6
389.2
377.4
394.1
417.5
389.3
–
–
–
–
–
–
*
243.4
236.0
249.2**
172.0
169.8
112.6
*
–0.1
2.9
*
–26.1
–21.3
–24.9
254.6
260.0
232.4
*
[a] (* = not observed; ** value from the 31P-NMR spectrum).
For the transient compound 3a, the proposed assignment
for the 29Si-NMR signals (for the numbering see Scheme 6)
2
3
2
29Si-NMR doublet signal at δ = 7.9 ppm exhibiting J(P,Si)
is as follows: δ1 = 5.1 ppm (d,d), J(P,Si) = Ϯ47.9, J(P,Si)
2
3
= Ϯ9.5 Hz; δ2 = 2.4 ppm (d,d), J(P,Si) = Ϯ12, J(P,Si) =
= Ϯ32.3 Hz is assigned to the iPrMe2Si group trans to the
SiCl3 group (cis to the phosphorus lone pair), the resonance
at 1.9 ppm [2J(P, Si) = Ϯ11.4 Hz] corresponds to the other
iPrMe2Si group. Compared with the starting material 1c [δ
=
29Si 2.6 ppm, 2J(P, Si) = Ϯ39.9 Hz, trans to Cl and
2
1.9 ppm, J(P, Si) = Ϯ4.3 Hz, cis to Cl], the trans-iPrMe2Si
group in 2c appears less deshielded and with a smaller cou-
pling constant. In contrast, the 29Si chemical shifts of the
cis-iPrMe2Si groups of 1c and 2c are very similar, but in 2c
the coupling constant is larger than that of 1c (in magni-
tude). This virtual “confusion” correlates with the opposite Scheme 6. Numbering of 29Si nuclei in the intermediate 3a.
4466 Eur. J. Inorg. Chem. 2010, 4462–4469
www.eurjic.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Per-Silylated Phosphaalkenes
3
1
2
Ϯ4.7, J(Si,P) Ͻ 1 Hz, C(SiMe3)2]. The mixture was converted to
Ϯ8.7 Hz; δ3 = 2.9 ppm (d,d), J(P,Si) = Ϯ172, J(P,Si) =
2
3
4a after 10 d reflux in toluene. The asymmetric diphosphene is also
consumed to give 4a.
Ϯ10.1 Hz; δ4 = –0.6 ppm (d,d), J(P,Si) = Ϯ40.9, J(P,Si) =
Ϯ24.1 Hz; δ5 = 0.9 ppm (d), 2J(P,Si) = Ϯ4.7 Hz [3J(P,Si)
Ͻ1 Hz].
4b: A mixture of 1b (1.6 g, 4.58 mmol) in 3 mL of C6D6 and 5
(1.04 mg, 5.0 mmol) was stirred at room temperature and 31P-
NMR spectroscopically analyzed. The reaction was spontaneous:
first the compound 2b [δ31P = 389.2 ppm (s), C=P, 1J(P,Si) =
Ϯ236 Hz] was formed, which after 1 d was consumed in favor of
Conclusions
3b [δ31P = 417.5 ppm (d), C=P-P, J(P,P) = Ϯ260 Hz], –21.3 ppm
1
Reactions of P-chlorophosphaalkenes (RMe2Si)2C=PCl
(1a–c; R = Me, Ph, iPr) with the disilane Me3SiSiCl3 (5)
furnish by Me3SiCl elimination the new per-silylated phos-
phaalkenes (RMe2Si)2C=PCSiCl3 (2a–c) as the first spec-
troscopically detectable products (step I). The subsequent
step (II) of the reaction is the dimerisation of 2a, 2b provid-
ing isomeric P-[(trichlorosilyl)phosphanyl]phosphaalkenes
(RMe2Si)2C=P–P(SiCl3)C(SiCl3)(SiMe2R)2 (3a: R = Me;
3b: R = Ph) and diphosphenes (Cl3Si)(RMe2Si)2C–P=P–
C(SiCl3)(SiMe2R)2 (4a: R = Me; 4b: R = Ph). Heating of
these mixtures or the reference compound (Me3Si)2C=P–
P(SiCl3)tBu (7) leads by 1,3(PǞC) SiCl3 group shift reac-
tions (step III) to the corresponding diphosphenes 4a, 4b
and 8. The structure of the new diphosphene 4b was deter-
mined crystallographically. The rates of reaction steps I–III
can be influenced by variation of the C-silyl groups in the
parent chlorophosphaalkenes 1a–c. With the bulkier
iPrMe2Si group, the dimerisation of 2c (step III) is pre-
cluded and solutions containing approximately equivalent
amounts of 2c and 1c can be generated. The remarkable
resistance of 2c to dimerisation allowed for the first time
the 31P- and 29Si-NMR spectroscopic characterisation of a
per-silylated P-(trichlorosilyl)phosphaalkene. On the ag-
enda of open questions in this context are still the mecha-
nisms of (i) the formation of compounds 2a–c [SiCl2 inser-
tion into the P–Cl bonds vs. SN-like Cl/SiCl3 group ex-
change at P(σ2λ3)] and (ii) the dimerisation step II (possible
role of phosphanylidene species).
[d, C=P-P, 1J(P,P) = Ϯ260, 1J(P,Si) = Ϯ169.8 Hz] and finally of 4b.
A competing reaction takes place, leading to small amounts of an
asymmetric diphosphene δ31P
= =
561.9 ppm [d, 1J (P,P)
1
Ϯ620.5 Hz], 502.5 ppm [d, J (P,P) = Ϯ620.5 Hz]. A transient pre-
cursor of this compound was observed 30 min after the reaction
started: δ = 31P = 401.9 [d, J(P,P) = Ϯ597 Hz], 332.3 [d, J(P,P) =
Ϯ597 Hz]. The asymmetric diphosphene was consumed after about
10–14 d. After 30 d NMR spectra of the reaction mixture showed
compounds 4b and 3b in the ratio 95:5 (high crude yield). All vola-
tiles were removed in vacuo and 0.5 mL of CDCl3 was added to
the residual brown oil. At 0 °C a few colorless crystals of 4b, appro-
priate for X-ray analysis, were obtained.
1
1
4b: 1H-NMR: δ = 0.7 (s, CH3, 6 H), 0.35 (s, CH3, 6 H) dia-
stereotopes (Me2Si), 7.45 (d,“t”, ortho, 2 H), 7.23–7.29 (m, para, 1
H), 7.15–7.22 (m, meta, 2 H). 13C-NMR: δ = 0.88 (“t”, CH3), 2.86
(“t”, CH3), 40.9 [m, C(SiCl3)(SiMe2Ph)2], 127.7 (s, m-C), 129.9 (s,
p-C), 135.6 (s, o-C), 136.7 (s, ipso-C); 29Si-NMR: δ = –4.3 (br. s,
SiMe2Ph, Integr. 2Si), –7.1 (br. s, SiCl3, integr. 1 Si); 31P-NMR: δ
= 573.6 (s, P=P). MS(EI): m/z (%) = 895.9 [M+, 4], 880.9 [(M –
CH3)+, 1], 761 [(M – SiCl3)+, 1], 743.8 [(M – SiCl3–CH3)+, 1], 723.9
[(M – SiCl4)+, 4], 478.9 [M – (C(SiCl3)(SiMe2Ph)2)+, 3], 135 [(Si-
Me2Ph)+, 100], 62.9 [(P=P)+, 22]. Elemental analysis (%):
C34H44Cl6P2Si6 (895.89) calcd.: C 45.58, H 4.95; found C 43.55, H
5.02.
Reaction of 1c with 6: 153 mg (0.6 mmol) 6 were added in one por-
tion to a solution of 170 mg (0.6 mmol) 1c and 0.5 mL of C6D6.
After 10 min, a small peak corresponding to the short-lived species
2c was observed with 31P NMR spectroscopy; δ = 377.4 ppm,
1J(P,Si) = 248.4 Hz. After 5 d, another compound is identified as
1
[(Me2iPr)Si]2C(SiCl3)–P(H)SiCl3 [δ = –93.4 ppm, s, J(P,H) = 220,
1J(P,Si) = 127.1 Hz]. Even 27 d later the reaction does not proceed
further, and the two reaction products mentioned so far are still
clearly distinguished in 31P NMR spectrum together with 1c.
Experimental Section
General Methods: All experiments were carried out under oxygen-
free nitrogen by using standard Schlenk techniques. NMR spectra
were recorded using Bruker spectrometers AC 200, Avance 200, Av-
ance 400 and AMX 300, with 85% H3PO4, and SiMe4 as external
or internal standards.
Reaction of 1c with 5: To (0.5 g, 1.78 mmol) of 1c in 0.3 mL of
C6D6, 2 equiv. of 5 (0.70 g, 3.5 mmol) was added at room tempera-
ture. From the mixture, an NMR sample was prepared and sealed
by melting. The signals assigned to 2c are stable for several weeks,
but their intensity shows a conversion of only 35–40%, while 1c
remains unconsumed.
4a: A mixture of (224 mg, 1.0 mmol) 1a and (414 mg, 2.0 mmol) 5
in 2 mL of toluene was stirred at room temperature. The new spe-
cies 2a [δ = +371.6 ppm C=P, J(P,Si) = Ϯ43.4 Hz] was observed
1
2c: 29Si NMR (C6D6): δ = 7.9 ppm 2J(P,Si) = 32.3 Hz [d, Si-
about 20 min after the reaction began. This is consumed to give 3a
2
(CH3)2iPr]; 1.9 ppm J(P,Si) = 11.4 Hz [d, Si(CH3)2iPr]; –0.1 ppm,
[δ = +394.1 ppm (d) C=P-P, 1J(P,P) = Ϯ254.6 Hz, –26.1 ppm (d)
1J(P,Si) = 248.4 Hz (d, SiCl3). 31P NMR (81 MHz, C6D6): δ =
1
C=P–P, J(P,P) = Ϯ254.6 Hz]. The reaction is very slow. After 5 d
1
2
+377.2 ppm, J(P,Si) = 249.2, J(P,Si) = 32.3 Hz.
the reaction shows three products in 31P NMR: 3a as major prod-
uct, 4a beginning to be formed, and traces of an asymmetric di-
After 2–3 weeks, in 31P-NMR an AX pattern and subsequently (2
months) also an AMX-pattern appear in at the expense of both 1c
and 2c. AX pattern: 31P NMR (81 MHz, C6D6): δ = –117.0 ppm
(d), –129 ppm (d), 1J(P,P) = 189 Hz. AMX pattern: δA = 184.7 ppm
1
1
phosphene [δ = 569.1 ppm, J(P,P) = Ϯ616 Hz, 504.7 ppm, J(P,P)
= Ϯ616 Hz (P=P)]. The concentration of 3a in this case is sufficient
for 29Si NMR spectroscopy (C6D6): δ = 5.1 pm [dd, 2J(Si,P) =
Ϯ47.9, 3J(Si,P) = Ϯ9.5 Hz, P=CSiMe3], 2.4 ppm [dd, 2J(Si,P) =
Ϯ12, 3J(Si,P) = Ϯ8.7 Hz, P=CSiMe3], 2.9 ppm [dd, 1J(Si,P) =
(dd), 1J(PA,PM)
= = =
Ϯ149, 2J(PA,PX) Ϯ111.9, 1J(PA,H)
Ϯ14.7 Hz, δM = 28.3 ppm (dd), 1J(PA,PM) = Ϯ149, J(PM,PX) =
Ϯ141.5 Hz, δX = –55.9 ppm (dd), 2J(PA,PX) = Ϯ111.9, J(PM,PX) =
Ϯ141.5 Hz.
Ϯ172, J(Si,P) = Ϯ10.1 Hz, C–P–SiCl3], –0.6 ppm [dd, 2J(Si,P) =
2
Ϯ40.9, 3J(Si,P) = Ϯ24.1 Hz, P–C–SiCl3], 0.9 ppm, [d, 2J(Si,P) =
Eur. J. Inorg. Chem. 2010, 4462–4469
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
4467

W.-W. du Mont, H. Marsmann et al.
FULL PAPER
[2]
[3]
Stable dihalides of the heavier elements Ge, Sn, Pb are access-
ible as starting materials, and the chemistry of transient dihalo-
genocarbenes is well established: W. Kirmse, Carbene, Carbeno-
ide und Carbenanaloge, Verlag Chemie, Weinheim 1969.
P. L. Timms, Inorg. Chem. 1968, 7, 387–389; W. H. Atwell,
D. R. Weyenberg, Angew. Chem. 1969, 81, 485–493; Angew.
Chem. Int. Ed. Engl. 1969, 8, 469; O. M. Nefedov, M. N. Man-
akov, Angew. Chem. 1966, 78, 1039–1056; Angew. Chem. Int.
Ed. Engl. 1966, 5, 1021; O. M. Nefedov, S. P. Kolesnikov, A. I.
Ioffe, Organomet. Chem. Rev. 1977, 5, 181–219.
Reaction of 1a with tBuP(SiCl3)2: A mixture of 2.33 g (10.4 mmol)
of 1a and 3.7 g (10.4 mmol) tert-butylbis(trichlorsilyl)phosphane
was refluxed for 5 h in 20 mL of dichloromethane. The consump-
tion of 1a and tBuP(SiCl3)2 in favour of compound 7 was con-
firmed by 31P-NMR spectroscopy. The solvent and silicon tetra-
chloride were removed in vacuo; attempts to distil the orange reside
led to thermal decomposition. In the residue from the attempted
distillation, compounds 7 and 8 could be identified by 31P NMR
spectroscopy. Separation of 8 from this mixture was not achieved.
7: 1H NMR (C6D6): δ = 0.2 ppm [d, 9 H, 4J(H,P) = 2.4 Hz Si-
(CH3)3], 0.4 ppm [s, 9 H, Si(CH3)3], 1.3 ppm [d, 9 H, 3J(H,P) =
13.1 Hz C(CH3)3]. 13C NMR (C6D6): δ = 2.0 ppm [d, 3J(C,P) =
[4]
[5]
E. Hengge, D. Kovar, Z. Anorg. Allg. Chem. 1979, 458, 163; H.
Stüger, E. Hengge, Monatsh. Chem. 1988, 119, 873; J. R. Koe,
D. R. Powell, J. J. Buffy, S. Hayase, R. West, Angew. Chem.
1998, 110, 1514–1515; Angew. Chem. Int. Ed. 1998, 37, 1441–
1442.
D. Kummer, H. Köster, M. Speck, Angew. Chem. 1969, 81,
574–575; Angew. Chem. Int. Ed. Engl. 1969, 8, 599; D. Kummer,
H. Köster, Angew. Chem. 1969, 81, 897; Angew. Chem. Int. Ed.
Engl. 1969, 8, 878.
H. H. Karsch, F. Bienlein, A. Sladek, M. Heckel, K. Burger, J.
Am. Chem. Soc. 1995, 117, 5160–5161.
a) R. S. Ghadwal, H. W. Roesky, S. Merkel, J. Henn, D. Stalke,
Angew. Chem. 2009, 121, 5793–5796; b) A. C. Filippou, O.
Chernov, G. Schnakenburg, Angew. Chem. 2009, 121, 5797–
5800.
a) L. Müller, W.-W. du Mont, F. Ruthe, P. G. Jones, H. C.
Marsmann, J. Organomet. Chem. 1999, 579, 156–163; b) W.-
W. du Mont, L. Müller, R. Martens, P. M. Papathomas, B. A.
Smart, H. E. Robertson, D. W. H. Rankin, Eur. J. Inorg. Chem.
1999, 1381–1392; c) S. L. Hinchley, L. J. McLachlan, H. E. Ro-
bertson, D. W. H. Rankin, E. Seppälä, W.-W. du Mont, Inorg.
Chim. Acta 2007, 360, 1323–1331.
W. W. du Mont, T. Gust, E. Seppälä, C. Wismach, P. G. Jones,
L. Ernst, J. Grunenberg, H. C. Marsmann, Angew. Chem. 2002,
114, 3977–3979; Angew. Chem. Int. Ed. 2002, 41, 3829–3832.
3
4
14.7 Hz, Si(CH3)3], 4.3 ppm [dd, J(C,P) = 7.8, J(C,P) = 2.9 Hz,
Si(CH3)3], 32.2 ppm [dd, 2J(C,P)
C(CH3)3], 35.5 ppm [dd, 1J(C,P)
=
=
12.4, 3J(C,P)
23.7, 2J(C,P)
=
=
3.6 Hz,
3.8 Hz,
C(CH3)3], 236.5 ppm [dd, 1J(C,P) = 99.7, 2J(C,P) = 11.2 Hz, C=P],
29Si NMR (C6D6): δ = –4.1 ppm [dd, 2J(Si,P) = 11.7, 3J(Si,P) =
12.8 Hz, Si(CH3)3], 0.1 ppm [dd, 2J(Si,P) = 39.2, 3J(Si,P) =
[6]
[7]
22.1 Hz, Si(CH3)3], 31P NMR (C6D6): δ = –24.9 ppm [d, J(P,P) =
1
232.4 Hz, 29Si satellites: 1J(Si,P) = 112.6, J(Si,P) = 22.1, J(Si,P)
3
3
= 12.8 Hz, P(SiCl3)tBu], 389.3 ppm [d, 1J(P,P) = 232.4 Hz, 29Si sat-
2
2
ellites: J(Si,P) = 39.2, J(Si,P) = 11.7 Hz, C=P]. MS(EI) m/z (%)
= 444 (1) [not assigned], 412 (2) [M]+, 387 (2) [?]+, 355 (3) [M –
tBu]+, 325 (4) [M – SiMe3–CH2]+, 209 (14) [P(SiCl3) Bu]+, 147 (40)
[Me3SiCPP]+, 93(100) [tBuH+Cl]+, 73 (82) [Me3Si]+, 57(56) [tBu]+.
[8]
8: 31P NMR (C6D6): δ = 531.3 ppm [d, 1J(P,P) = 628.9 Hz,
1
P=PtBu], 604.0 [d, J(P,P) = 628.9 Hz, P=PtBu].
X-ray Structure Determination of Compound 4b: Crystal data:
¯
C34H44Cl6P2Si6, Mr = 895.87, triclinic, space group P1, a =
[9]
9.4125(4), b = 10.9299(4), c = 11.8668(4) Å, α = 66.618(3), β =
71.536(3), γ = 88.715(3)°, V = 1055.4 Å3, Z = 1, ρcalc = 1.410 Mg/
m3, µ(Mo-Kα) = 0.68 mm–1, F(000) = 464, T = 100 K; yellow block
0.45ϫ0.3ϫ0.2 mm3. Of 37527 reflections collected to 2θ 63°, 6912
were independent (Rint = 0.021). Final R1 = 0.0209 [IϾ2σ(I)], wR2
= 0.0601 (all data) for 221 parameters; S = 1.05, max. ∆ρ 0.5 eÅ–3.
[10]
[11]
A. Zanin, M. Karnop, J. Jeske, P. G. Jones, W.-W. du Mont, J.
Organomet. Chem. 1994, 475, 95.
a) R. Martens, W.-W. du Mont, L. Lange, Z. Naturforsch., Teil
B 1991, 46, 1609; b) R. Martens, W.-W. du Mont, Chem. Ber.
1992, 125, 657; c) L.-P. Müller, A. Zanin, W.-W. du Mont, J.
Jeske, R. Martens, P. G. Jones, Chem. Ber./Recueil 1997, 130,
377; d) L. Müller, W. W. du Mont, F. Ruthe, P. G. Jones, H. C.
Marsmann, J. Organomet. Chem. 1999, 579, 156–163.
E. Niecke, H. J. Metternich, R. Streubel, Chem. Ber. 1990, 123,
67–69.
E. Seppälä, W.-W. du Mont, T. Gust, C. Wismach, J. Or-
ganomet. Chem. 2004, 689, 1331.
D. Bugnariu, Dissertation, Technical University of
Braunschweig, 2007.
T. A. Banford, A. G. Mc Diarmid, Inorg. Nucl. Chem. Lett.
1972, 8, 733.
R. Appel, W. Westerhaus, Angew. Chem. 1980, 92, 578; Angew.
Chem. Int. Ed. Engl. 1980, 19, 556.
J. Mahnke, A. Zanin, W.-W. du Mont, F. Ruthe, P. G. Jones, Z.
Anorg. Allg. Chem. 1998, 624, 1447–1454.
M. Yoshifuji, in: M. Regitz, O. J. Scherer, Multiple Bonds and
Low Coordination in Phosphorus Chemistry, Georg-Thieme Ver-
lag, Stuttgart, 1990, chapter D9, p. 321–337.
Coupling constants for Si–P π-Bonds: C. N. Smit, F. Bickel-
haupt, Organometallics 1987, 6, 1156.
a) Coupling constants for Sn–P π-bonds: C. Couret, J. Escudie,
J. Satgé, J. Am. Chem. Soc. 1985, 107, 8280; b) For Sn–P cou-
pling constants for phosphane SnCl4 complexes, see: J. F. Ma-
lone, B. E. Mann, Inorg. Nucl. Chem. Lett. 1972, 8, 819; A.
Yamasaki, F. Fluck, Z. Anorg. Allg. Chem. 1973, 396, 819.
Data were recorded using Mo-Kα radiation (λ = 0.71073 Å) on an
Oxford Diffraction Xcalibur S diffractometer. An absorption cor-
rection was based on multi-scans. The structure was refined using
the program SHELXL-97 [25]. Hydrogen atoms were included using
rigid methyl groups or a riding model.
[12]
[13]
[14]
[15]
[16]
[17]
[18]
CCDC-777799 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
Acknowledgments
We thank Dr. Alk Dransfeld and Prof. Dr. Manuela Flock (TU
Graz, Austria) for their help with orienting DFT calculations on
compounds 2, and we acknowledge financial support from the
Deutsche Forschungsgemeinschaft (DFG) at an early stage of this
work.
[19]
[20]
[1] A. J. Arduengo III, Acc. Chem. Res. 1999, 32, 913–921; D. Bou-
rissou, O. Guerret, F. P. Gabbai, G. Bertrand, Chem. Rev. 2000,
100, 39–91; M. Driess, H. Grützmacher, Angew. Chem. 1996,
108, 900–929; Angew. Chem. Int. Ed. Engl. 1996, 35, 828–856;
N. Tokitoh, R. Okazaki, Coord. Chem. Rev. 2000, 210, 251–
272; R. West, Acc. Chem. Res. 2000, 33, 704.
[21]
[22]
F. H. Allen, Acta Crystallogr., Sect. B 2002, 58, 380–388.
a) J. Escudié, C. Couret, H. Ranaivonjatavo, J. Satgé, J. Jaud,
Phosphorus Sulfur 1983, 17, 221; b) A. H. Cowley, J. E. Kilduff,
4468
www.eurjic.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2010, 4462–4469

Per-Silylated Phosphaalkenes
J. G. Lasch, S. K. Mehrotra, N. C. Norman, M. Palulski, B. R.
Whittlesey, J. L. Atwood, W. E. Hunter, Inorg. Chem. 1984, 23,
2582–2593.
[24] C. Mitrofan, R. M. Bîrzoi, E. Seppälä, D. Bugnariu, T. Gust,
W.-W. du Mont, H. Marsmann, A. Dransfeld, M. Flock,
4th European Silicon Days (Abstract), Bath (England),
2007.
[23] a) G. Becker, G. Gresser, W. Uhl, Z. Anorg. Allg. Chem. 1980,
463, 144–148; b) K. Issleib, H. Schmidt, C. Wirkner, Z. Anorg.
Allg. Chem. 1981, 473, 86–90.
[25] G. M. Sheldrick, Acta Crystallogr., Sect. A 2008, 64, 112–122.
Received: March 12, 2010
Published Online: August 5, 2010
Eur. J. Inorg. Chem. 2010, 4462–4469
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
4469