Phosphanes with bulky oligosilyl substituents

Article abstract of DOI:10.1016/j.jorganchem.2008.12.003

By reaction of dichloroheptasilane [(SiMe₃)₂MeSi]₂SiCl₂ with lithiumphosphanides LiPHR, the silylphosphanes [(SiMe₃)₂MeSi]₂SiClPHR with R = 2, 4, 6-tri-tert-butylphenyl ( = super

Full text of DOI:10.1016/j.jorganchem.2008.12.003

Journal of Organometallic Chemistry 694 (2009) 757–762
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Phosphanes with bulky oligosilyl substituents
*
A. Dzambasky, J. Baumgartner, K. Hassler
Institute of Inorganic Chemistry, University of Technology, Stremayrgasse 16, A-8010 Graz, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 31 October 2008
Received in revised form 28 November 2008
Accepted 2 December 2008
Available online 10 December 2008
By reaction of dichloroheptasilane [(SiMe3)2MeSi]2SiCl2 with lithiumphosphanides LiPHR, the silylphos-
phanes [(SiMe3)2MeSi]2SiClPHR with R = 2, 4, 6-tri-tert-butylphenyl ( = supermesityl, Mes ) (1) and
*
Si(SiMe3)3 ( = hypersilyl, Hyp) (2) were prepared. Both compounds were characterized with X-ray diffrac-
tion, multinuclear NMR spectroscopy and elemental analysis. Compound 1 did not react with n-BuLi, but
only with a large excess of tert-BuLi. Phosphasilene [(SiMe3)2MeSi]2Si@PMes* could be identified by a 31P
NMR signal at +346 ppm. All attempts to separate it from the reaction mixture failed due to many by-
products which had formed through SiSi and SiP bond cleavage. Lithiation of 2 was possible with
4.2 equiv. of tert-BuLi, and crystals of the lithiumphosphanide [(SiMe3)2MeSi]2SiClPLiHyp (3) could be
obtained from THF, albeit in a quality not sufficient for X-ray diffraction. All attempts to achieve LiCl elim-
ination and formation of the phosphasilene [(SiMe3)2MeSi]2Si@PSi(SiMe3)3 failed due to the unusual sta-
bility of the lithiumphosphanide. Prolongued refluxing in toluene (110 °C) only led to complete loss of
coordinated THF, and 31P7Li spin spin coupling could be observed in the 31P NMR spectrum (1JPLi = 84 Hz).
Reaction of potassium phosphanide [(SiMe3)3Si]SiMe3PK with SiCl4 led to the formation of [(SiMe3)3-
Si](SiMe3)P(SiCl3) (4), which could be successfully characterized with X-ray diffraction and multinuclear
NMR spectroscopy. SiP bond lengths vary between 218 pm (SiCl3) and 230 pm (hypersilyl). Despite these
differences, 31P29Si coupling constants are nearly identical (92.4 Hz and 85.5 Hz, respectively).
Ó 2008 Elsevier B.V. All rights reserved.
Keywords:
Silylphosphanes
X-ray diffraction
NMR spectroscopy
1. Introduction
the type SiiPr[CH(SiMe3)2]2 [4]. We therefore asked ourselves the
question whether bulky oligosilyl groups containing SiSi single
It is well known that contrary to the classic double bond rule,
multiple bonds between third row elements can be kinetically sta-
bilized with large organic groups. For instance, the synthesis of the
first isolable disilene R2Si = SiR2 bearing four mesityl groups R was
reported in 1981, and a combination of mesityl and supermesityl
groups (supermesityl = Mes* = 2,4,6-tri-tert-butylphenyl) was used
to stabilize the phosphasilene Mes2Si = PMes* [1,2]. Subsequently it
was shown that other substituents such as tert-butyl or CH(SiMe3)2
and even silyl groups may serve the same purpose.
Ab initio calculations for model systems bearing SiH3 substitu-
ents indicate a stabilizing electronic effect of silyl groups, which for
larger SiR3 is superimposed by steric effects. For instance, the SiSi
bond strength in disilenes measured by the Si@Si stretching force
constant increases smoothly in the series H2Si@SiH2, (SiH3)H-
bonds (such as Si(SiMe3)3) might also be capable to stabilize dou-
ble or triple bonds of third row elements. These are substituents
even less electronegative than SiH3 or SiMe3, and interactions with
*
r (SiSi) orbitals supposedly exert a stabilizing effect
antibonding
on the p-electron system. The steric demand of the oligosilyl group
can be increased further by substituting some methyl groups with
i
larger organic substituents such as Pr or tBu. Moreover, as SiSi
bonds are much more reactive than CC or SiC, they might offer
the possibility to perform reactions in the substituent sphere, with-
out affecting the double bond. For instance, SiMe3-groups in oligo-
silyl substituents such as Si(SiMe3)3 can be cleaved with KOtBu at
temperatures as low as ꢀ70 °C, giving an anionic silicon center
SiK(SiMe3)2 and Me3SiOtBu [5].
However, the use of oligosilyl groups poses some severe syn-
thetic difficulties, such as facile migration of silyl groups.
Di(oligosilyl)silylenes which normally form disilenes by dimeriza-
tion undergo very rapid silyl migration, rearranging into disilenes
as shown in Scheme 1.
For this reason, we were unable to obtain [(SiMe3)2MeSi]2-
Si@Si[SiMe(SiMe3)2]2 from silylene [(SiMe3)2MeSi]2Si by dimeriza-
tion. Instead, various cyclotrisilanes were obtained formed through
silyl migration and addition of silylenes to disilenes [6]. So far, only
a handful of disilenes substituted with oligosilyl groups have been
described in the literature, among them a fused bicyclic system as
Si@SiH(SiH3) and (SiH3)2Si@Si(SiH3)2, which can be explained with
*
p-electron delocalisation into antibonding
r
(SiH) orbitals [3].
Moreover, the Si2Si@SiSi2 backbone is planar. Opposite to that, SiSi
stretching force constants in disilanes decrease with increasing
number of silyl substituents.
Recently, the synthesis of the first isolable disilyne has been de-
scribed, which is also stabilized by two bulky silyl substituents of
* Corresponding author. Tel.: +43 316 873 8206.
E-mail address: karl.hassler@tugraz.at (K. Hassler).
0022-328X/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2008.12.003

758
A. Dzambasky et al. / Journal of Organometallic Chemistry 694 (2009) 757–762
RPLi ꢀ SiXR02 (preferably with X = F), a strategy which is also suit-
able for the preparation of arsasilenes. In this publication, we de-
scribe the preparation of the title phospha-halosilanes 1 and 2,
their lithiation with tert-BuLi and the subsequent attempts to
achieve thermal LiX elimination.
2. Results and discussion
Scheme 1. Silylene to disilene rearrangement of (oligosilyl)silylenes.
2.1. Synthesis of 1 and 2
well as disilenes with Si@SiH or Si@SiCl bonds which contain Si-
Me(SitBu3)2 or SiH(SitBu3)2 groups [7–9]. They were prepared by
reduction of a 1,2-dihalodisilane, and not by dimerization of a
silylene.
Previously reported dichloroheptasilane [(SiMe3)2MeSi]2SiCl2
and the lithium phosphanides RHPLi with R = 2,4,6-tri-tert-butyl-
*
phenyl and Si(SiMe3)3 (furtheron denoted as Mes and hypersilyl,
A further problem with the use of oligosilyl groups stems from
the ease of SiSi bond cleavage during reductive dehalogenation,
which normally is used to form silylenes or disilenes from haloge-
nated precursors. The number of by-products may be quite consid-
erable, which certainly does not facilitate isolation of pure
compounds. Noteworthy, silyl migration does not occur with oligo-
silylphosphinidenes such as (SiMe3)3SiP. (SiMe3)3SiP@PSi(SiMe3)3
is formed quickly by dimerization, which is the only di(oligosilyl)
diphosphene described in the literature [10].
A fair number of silicon–phophorus double bonded systems
have been described so far, which also include some P-silylated
derivatives, albeit none with oligosilyl groups [11]. Up to now,
the best method for the synthesis of phosphasilenes is the thermal
abstraction of a lithium halide LiX from phosphanido-halosilanes
respectively) were used to prepare the highly congested chlorosi-
lylphosphanes R02SiCl ꢀ PHR as shown in Scheme 2 (R0 = (SiMe3)2
*
MeSi; R = Mes (1) or hypersilyl (2)).
Both phosphanes could be crystallized successfully from n-
hexane, and X-ray structural data could be obtained. In Fig. 1, the
molecular structures are shown with the help of ORTEP plots (30%
probabilities), and selected interatomic distances and angles are
given.
Due to steric interactions of the substituents, SiSi bonds are on
the average 0.02–0.03 Å longer than considered as normal (2.32–
2.35 Å; for instance 2.338 0.004 Å in dodecamethylcyclohexasi-
lane Si6Me12). With 2.28 Å, the PSi(1)-bond in 1 is 0.035 Å longer
than in 2 (PSi(1) = 2.245 Å), which can be attributed to the in-
*
creased steric demand of Mes compared with hypersilyl. In 2,
the PSi(8) bond is longer than PSi(1) by about 0.04 Å, a difference
1
which is also reflected in differing JPSi coupling constants which
are 80 Hz for PSi(8) and 108 Hz for PSi(1). The steric congestion
is also clearly reflected in some unusual bond angles. For instance,
the angle Si(1)P(1)Si(8) is widened up to almost 121° in 2, and
Si(2)Si(1)Si(5) angles are also near this value (118.7° and 118.3°,
respectively).
2.2. Lithiation of 1
Compound 1 did not react with n-butyl-lithium even when the
lithiation was tried in
a non-polar solvent and at elevated
temperatures. Just decomposition of n-BuLi was observed. With
Scheme 2. Synthesis of (chloro)oligosilylphosphanes 1 and 2.
MeLi, only substitution of chlorine with methyl occurred, and
*
Fig. 1. ORTEP plots (30% probabilities) of the molecular structures of [(SiMe3)2MeSi]2SiCl–PHMes (1, left) and [(SiMe3)2MeSi]2SiCl–PH[Si(SiMe3)3] (2, right). Selected
*
interatomic distances and angles are as follows (Å and °): [(SiMe3)2MeSi]2SiCl–PHMes : P(1)Si(1): 2.2844(14); P(1)C(1): 1.856(4); Si(1)Cl(1): 2.1245(13); Si(1)Si(2):
2.3696(15); Si(1)Si(5): 2.3839(15); Si(2)Si(3): 2.3640(15); Si(2)Si(4): 2.3516(16); Si(5)Si(6): 2.3532(15); Si(5)Si(7): 2.3589(15); C(1)P(1)Si(1): 113.22(12); Cl(1)Si(1)P(1):
113.22(12); Si(2)Si1)P(1): 97.89(5); Si(5)Si(1)P(1): 103.43(5); Si(2)Si(1)Si(5): 118.74(5); Si(4)Si(3)Si(2): 104.02(13); Si(6)Si(5)Si(7): 107.87(5).[(SiMe3)2MeSi]2SiCl-
PH[Si(SiMe3)3]: P(1)Si)1): 2.245(2); P(1)Si(8): 2.285(2); Cl(1)Si(1): 2.120(2); Si(1)Si(5): 2.346(2); Si(1)Si(2): 2.359(2); Si(2)Si(3): 2.350(2); Si(2)Si(4): 2.357(3); Si(5)Si(6):
2.351(2); Si(5)Si(7): 2.365(2); Si(8)Si(9): 2.351(2); Si(8)Si(10): 2.362(3); Si(8)Si(11): 2.364(2); Si(1)P(1)Si(8): 120.70(9); Cl(1)Si(1)P(1): 111.92(9); P(1)Si(1)Si(2): 101.28(9);
P(1)Si(1)Si(5): 115.16(9); P(1)Si(8)Si(9): 103.68(9): P(1)Si(8)Si(10): 122.38(10); P(1)Si(8)Si(11): 99.82(9); Si(2)Si(1)Si(5): 118.3(9); Si(3)Si(2)Si(4): 110.61(9); Si(6)Si(5)Si(7):
109.94(9); Si(11)Si(8)Si(9): 107.75(9); Si(11)Si(8)Si(10): 108.15(9); Si(9)Si(8)Si(10): 113.57(10).

A. Dzambasky et al. / Journal of Organometallic Chemistry 694 (2009) 757–762
759
*
[(SiMe3)2MeSi]2SiMe-PHMes could be identified as the sole prod-
served. Instead, the extended reflux only led to total loss of coor-
dinated THF, and splitting of the 31P NMR signal into a quartet
due to 31P7Li coupling could clearly be observed (1JPLi = 84 Hz).
NMR data for 3 have been summarized in Section 3. Most cer-
tainly, the unsusual resistance to LiCl elimination is caused by
the less polar (and increased covalent) character of the LiP-bond
due to the electron donating capability of both the hypersilyl
and the Si7 substituent. Treating 3 with MgBr2 resulted in com-
plete decomposition of the phosphanide, no phosphasilene could
be detected spectroscopically.
uct with NMR spectroscopy (d31P = ꢀ110.9 ppm, d, JPH = 226 Hz;
1
d
29Si = ꢀ74.3 ppm (bs,(Si(SiMe3)2Me);
d
29Si = ꢀ10.5 (d, SiMe3,
3JPSi = 4 Hz); d29Si = ꢀ9.7 (d, SiMe3, JPSi = ꢁ1 Hz); d29Si = ꢀ25.1
3
1
(d, SiMe2, JPSi = 69 Hz)). The lithiation of 1 was only possible with
tert-butyl-lithium, but not without serious side reactions. When
one equivalent of tert-BuLi was added, the predominant 31P reso-
nance was from the starting material 1. A small dublet observed
at ꢀ145.8 ppm (2JPH = 31 Hz) probably belonged to [(SiMe3)2Me-
*
Si]2SiH-PLiMes formed through reduction of the SiCl bond with
LiH, which is a decomposition product of lithiumalkyls. A singlet
In a recent publication we have shown that tert-BuOK can be
used to selectively cleave SiMe3-groups from trimethylsilylphos-
phanes under formation of potassium phosphanides and tert-bu-
tyl-trimethylsilyl ether. For instance, HypP(SiMe3)2 reacts
quantitatively with formation of HypP(SiMe3)K, and the hypersilyl
group is not attacked under the reactions conditions (THF at
ꢀ78 °C). We therefore attempted to prepare [(SiMe3)2MeSi]2SiCl-
P(SiMe3)Hyp and to treat it with tert-BuOK to form [(SiMe3)2Me-
Si]2SiCl-PKHyp as shown in Scheme 3.
It turned out that no reaction between [(SiMe3)2MeSi]2SiCl2 and
HypP(SiMe3)K could be achieved even at elevated temperatures of
50–60 °C, certainly due to steric strain the additional trimethylsilyl
group introduces. In contrast, much smaller KP(SiMe3)2 did react
quite readily forming [(SiMe3)2MeSi]2SiCl-P(SiMe3)2 in good yields.
It is also noteworthy that phosphasilene [(SiMe3)2MeSi]2Si @PHyp
could not be prepared by salt elimination from [(SiMe3)2MeSi]2-
SiCl2 and Li2PHyp.
observed downfield at +346.3 ppm (1JPSi = 161 Hz) was from the
*
phosphasilene [(SiMe3)2MeSi]2Si@PMes formed by LiCl elimina-
*
tion from [(SiMe3)2MeSi]2SiCl-PLiMes . More tert-BuLi was added
until the resonance of the starting silylphosphane 1 had disap-
peared completely. After exchange of the ether solvent with hex-
ane followed by decantation from the salts, all attempts to obtain
crystals of the phosphasilene failed due to very low yields. Even
the identification of the phosphasilene by 29Si NMR spectroscopy
was not successful due to the same reasons.
When [(SiMe3)2MeSi]2SiF2 was reacted with one equivalent of
*
LiHPMes under the reaction conditions applied for the synthesis
of 1, only the signals of [(SiMe3)2MeSi]2SiF2 could be detected in
the 29Si NMR spectrum. After replacement of the solvent with
n-hexane and refluxing overnight, the 31P NMR spectrum showed
*
that LiHPMes had completely decomposed and that only a small
part had reacted with the difluorosilane under concomitant LiF
1
elimination (d31P = +350.2 ppm, JPSi = 161 Hz). Due to the very
low yield, neither 29Si resonances of the phosphasilene [(SiMe3)2-
2.4. Synthesis of 4
*
MeSi]2Si@PMes could be observed nor could crystals be obtained.
To explore further reactions of HypP(SiMe3)K with chlorosil-
anes, we reacted it with SiCl4 in an attempt to prepare dichlorodi-
phosphanylsilane Hyp(SiMe3)P–SiCl2–P(SiMe3)Hyp as a potential
precursor for HypKP–SiCl2–PKHyp. It turned out that no disubstitu-
tion of SiCl4 could be achieved due to steric reasons. Instead,
Hyp(SiMe3)P(SiCl3) formed in good yields, a phosphane bearing
three different silyl groups. Crystals of 4 with a quality sufficient
for X-ray analysis could be obtained from n-hexane. The molecular
structure is presented in Fig. 2.
Chiral 4 possesses three different silyl substituents. Conse-
quently, the SiP bonds Si(1)P(1), Si(2)P(1) and Si(3)P(1) (SiCl3-
group, SiMe3-group and hypersilyl group, respectively) differ in
their bond length which increases from 2.185 Å (SiCl3) to 2.289 Å
2.3. Lithiation of 2 to form 3
The lithiation of 2 with tert-BuLi in THF at ꢀ78 °C gave a com-
pletely different outcome. 4.2 equiv. of the lithiating agent were
necessary until the 31P signal of starting 2 had disappeared. Side
reactions involving SiP and SiSi bond cleavage also occurred. After
reducing the volume of the solvent, light-orange crystals of [(Si-
Me3)2MeSi]2SiCl-PLiHyp (3) precipitated, albeit not in a quality
sufficient for X-ray analysis. The crystals were dissolved in THF
and warmed up to 50 °C, but no LiCl elimination could be achieved.
Consequently, THF was replaced with toluene and the solution was
refluxed for 15 h at 110 °C. Again, no LiCl elimination was ob-
Scheme 3. Attempted reaction between [(SiMe3)2MeSi]2SiCl2 and HypP(SiMe3)K.

760
A. Dzambasky et al. / Journal of Organometallic Chemistry 694 (2009) 757–762
lenghts. Ab initio calculations of coupling constants for the series
H2PSiR3 with R = SiH3, Me, H, Cl and F indicate that 1JPSi is negative
for all substituents, and that the strength of the bond (expressed
with the harmonic SiP stretching force constant) increases with
increasing electronegativity [12]. The calculations also predict
quite similar coupling constants for the groups SiCl3 (calc.:
ꢀ57.2, exp.: 52.2 Hz [13]) and Si(SiH3)3 (calc.: ꢀ50.2 Hz, exp.:
46.7 Hz for H2PSi(SiMe3)3 [14]).
3. Experimental
3.1. General, NMR spectroscopy
As all compounds are air-sensitive, reactions were carried out
either in an inert N2 or Ar atmosphere using standard Schlenk tech-
niques, or in a nitrogen filled glove box MBRAUN Unilab, supplied
by M. Braun Gmbh. All solvents were dried prior to use using a col-
umn solvent purification system and were then distilled under a N2
atmosphere to remove all traces of oxygen. NMR spectra were re-
corded on a Varian INOVA 300 spectrometer (1H: 299.95 MHz, 13C:
75.43 MHz, 29Si: 59.59 MHz, 31P: 121.42 MHz). Shift are reported
Fig. 2. ORTEP plot (30% probabilities) of the molecular structure of (SiMe3)(SiCl3)-
PHyp (4). Hydrogen atoms have been omitted for clarity. Selected bond lengths (Å)
and bond angles (°) are: P(1)Si(1): 2.2890(1); P(1)Si(2): 2.1850(7); P(1)Si(3):
2.3002(8); Si(2)Cl(1): 2.0413(7); Si(2)Cl(2): 2.0533(7); Si(2)Cl(3): 2.0421(9);
Si(3)Si(4): 2.3676(10); Si(3)Si(5): 2.3820(8); Si(3)Si(6): 2.3656(8); Si(1)P(1)Si(3):
117.09(3); Si(1)P(1)Si(2): 100.81(2); Si(2)P(1)Si(3): 105.70(2); P(1)Si(3)Si(4):
121.23(3); P(1)Si(3)Si(5): 102.60(3); P(1)Si(3)Si(6): 103.05(2); P(1)Si(2)Cl(1):
112.15(3); P(1)Si(2)Cl(2): 116.91(3); P(1)Si(2)Cl(3): 111.28(3).
in ppm downfield from TMS (1H, 29Si, 13C) and H3PO4 31P).
(
3.2. X-ray diffraction
For X-ray structure analyses, the crystals were mounted onto
the tip of glass fibers and data collection was performed with a
BRUKER-AXS SMART APEX CCD diffractometer using graphite-
(SiMe3) and to 2.300 Å (hypersilyl), mirroring the decreasing
electronegativity of the substituents on the silicon atom
(Cl ? Me ? SiMe3). This can be understood using the concept of
isovalent hybridisation which states that bonds to electronegative
substituents (such as Cl) have more p-orbital character and that
the remaining bonds therefore are characterized by an increased
s-orbital contribution. Bond strengths usually increase with
growing s-orbital character, and bond lengths decrease.
monochromated Mo K
a radiation (0.71073 Å). The data were re-
duced to F2o and corrected for absorption effects with SAINT [15]
and SADABS [16], respectively. The structures were solved by direct
methods and refined by full-matrix least-squares method
(SHELXL97) [17]. If not noted otherwise, all non-hydrogen atoms
were refined with anisotropic displacement parameters. All hydro-
gen atoms were located in calculated positions to correspond to
standard bond lengths and angles. Diagrams were drawn with
30% probability thermal ellipsoids and all hydrogen atoms were
1
Quite obviously, the corresponding JPSi coupling constants
which are 92.4 Hz (SiCl3), 34.8 Hz (SiMe3) and 85.5 Hz (hypersilyl)
do not correlate with substituent electronegativities and bond
Table 1
Crystal data for 1, 2 and 4.
Empirical formula
Formula weight
T [K]
C32H72ClPSi7
719.95
100(2)
C23H70ClPSi11
722.20
100(2)
C12H36Cl3PSi6
486.27
100(2)
Wavelength [Å]
Crystal system, space group
Unit cell dimension
a (Å)
0.71073
monoclinic, P2(1)/c
0.71073
monoclinic, P2(1)/c
0.71073
monoclinic, P2(1)/c
21.897(4)
17.861(4)
17.076(3)
b (Å)
12.699(3)
9.6358(19)
10.304(2)
c (Å)
16.325(3)
26.830(5)
15.971(3)
a
(°)
90
90
90
b (°)
90.92(3)
95.11(3)
107.62(3)
c
(°)
90
90
90
2678.1(9)
4, 1.206
0.667
V [Å3]
4538.9(16)
4, 1.054
0.324
1576
4599.3(16)
4, 1.043
0.418
1576
Z, DCalc [Mg/m3]
Absorption coefficient [mmꢀ1
F(000)
]
1032
Crystal size [mm]
h Range for data collection [°]
Limiting indices
Reflections collected/unique (Rint
Completeness to h max (%)
Absorption correction
0.44 ꢂ 0.30 ꢂ 0.24
0.40 ꢂ 0.20 ꢂ 0.15
0.52 ꢂ 0.42 ꢂ 0.30
1.85– 25.00
1.52–24.00
2.34–26.37
ꢀ26 < h < 26, ꢀ15 < k < 15, ꢀ19 < l < 19
31651/7963 (0.1055)
99.7
ꢀ20 < h < 20, ꢀ11 < k < 11, ꢀ30 < l < 30
29404/7211 (0.0727)
99.9
ꢀ21 < h < 21,ꢀ12 < k < 12, ꢀ19 < l < 19
20737/5460 (0.0406)
99.7
)
Max. and min. transmission
Refinement method
0.9264 and 0.8707
Full-matrix least-squares on F2
7963/0/397
0.9400 and 0.8506
Full-matrix least-squares on F2
7211/0/352
0.8250 and 0.7231
Full-matrix least-squares on F2
5460/0/211
Data/restraints/parameters
Goodness of Fit on I2
0.969
1.412
1.062
Final R indices [I > 2
R indices (all data)
r
(I)]
R1 = 0.0636, wR2 = 0.1375
R1 = 0.0957, wR2 = 0.1481
0.833 and ꢀ0.395
R1 = 0.0972, wR2 = 0.1669
R1 = 0.1091, wR2 = 0.1714
0.508 and ꢀ0.419
R1 = 0.0295, wR2 = 0.0738
R1 = 0.0358, wR2 = 0.0763
0.604 and ꢀ0.233
Largest difference in peak and hole [e Åꢀ3
]

A. Dzambasky et al. / Journal of Organometallic Chemistry 694 (2009) 757–762
761
omitted for clarity. Unfortunately, the obtained crystal quality for 1
and 2 was poor. This fact is reflected by quite high R and low theta
values. Crystal and structure refinement data for 1–3 are summa-
rized in Table 1.
Crystallographic data (excluding structure factors) for the struc-
tures of compounds 1–3 have been deposited with the Cambridge
Crystallographic Data Center as supplementary publication no.
CCDC 703084 (1), 703085 (2) and 703083 (3). Copies of the data
can be obtained free of charge on application to The Director, CCDC,
12 Union Road, Cambridge CB2 1EZ, UK [Fax: internat. + 44-1223/
336-033; E-mail: deposit@ccdc.cam.ac.uk].
THF at ꢀ78 °C. The colour immediately turned yellow. The reaction
mixture was then allowed to warm to room temperature slowly,
and the completion was checked with 29Si, 31P and 31P{1H} NMR
spectroscopy. Light-orange crystals of 3 precipitated from the
THF solution at room temperature, but the quality of the crystals
was not sufficient for X-ray diffraction. When an equimolar
amount of n-butyl-lithium was used, the formation of the lithium
phosphanide 3 could hardly be observed.
No LiCl elimination was observed upon warming the THF solu-
tion of 3–50 °C over several hours. The solvent therefore was re-
placed by toluene, and the solution was refluxed for 15 h. Again,
no LiCl elimination was observed. The extended reflux only led to
a total removal of coordinated THF, which resulted in the splitting
of the 31P{1H} signal into a 1:1:1:1 quartet due to 31P7Li coupling
(1JPLi) = 84 Hz).
3.3. Syntheses
3.3.1. Synthesis of 2,4,6-tri-tert-butylphenyl-(3-chloro-1,1,1,2,4,5,5,
5-octamethyl-2,4-bis(tri-methylsilyl)-n-pentasilanyl)phosphane (1)
A diethyl ether solution of 2,4,6-tri-tert-butylphenylphospha-
nide (1.14 g, 4.0 mmol) was added dropwise into a diethyl ether
solutio of 3,3-dichloro-1,1,1,2,4,5,5,5-octamethyl-2,4-bis(trimeth-
ylsilyl)-n-pentasilane (1.91 g, 4.0 mmol) at ꢀ78 °C. The reaction
mixture was then allowed to warm slowly to room temperature
and was stirred overnight to complete the reaction. Diethyl ether
was then removed by evaporation under reduced pressure and
the reddish, oily residue was suspended in n-hexane. After filtra-
tion of the insoluble lithium salts, 1 crystallized at ꢀ80 °C in the
form of colourless crystals suitable for single crystal X-ray analysis.
The yield was 0.31 g (11%).
31P NMR (ppm, THF solution): d = ꢀ339.9.
1
29Si NMR (ppm, THF solution): d = ꢀ6.4 (d, SiCl, JSiP = 114 Hz),
d = ꢀ11.8 (m, SiMe3), d = ꢀ12.5 (s, SiMe3), d = ꢀ14.5 (d, (SiMe3)3Si),
2
2JSiP = 16 Hz), d = ꢀ84.4 (d, Si(SiMe3)2Me, JSiP = 33 Hz), d = ꢀ95.1
1
(d, Si(SiMe3)3, JSiP = 166 Hz).
3.3.4. Synthesis of hypersilyl(trichlorosilyl)trimethylsilylphosphane (4)
A diethyl ether solution of potassium (trimethylsilyl)-[tris(tri-
methylsilyl)silyl]phosphanide 3 (3.91 g, 10.0 mmol) was added
dropwise to a diethyl ether solution of tetrachlorosilane (8.50 g,
5.0 mmol) at ꢀ78 °C. The rection mixture was then allowed to
warm to room temperature slowly and was stirred overnight for
completion. Diethyl ether was then removed by evaporation under
reduced pressure, and the oily residue was suspended in n-hexane.
After filtration of the insoluble potassium salts, 4 crystallized from
n-hexane at ꢀ80 °C. The quality of the colourless crystals was suf-
ficient for a single crystal X-ray analysis. The yield was 2.10 g
(43%). No disubstitution of SiCl4 was observed.
Anal. Calc. for C32H72Si7PCl (719.94 g/mol): C, 53.39, H, 10.08.
Found: C, 53.35, H, 10.59%.
1H NMR (C6D6): d = 0.18 (SiCH3, 6H), d = 0.32 (s, Si(CH3)3, 18H),
d = 0.36 (s, Si(CH3)3, 18H), d = 1.31 (9H, tert-Bu), d = 1.70 (18H, tert-
1
Bu), d = 5.18 (d, PH, 1H, JPH = 226 Hz), d = 7.48 (d, ArH, 2H,
4JPH = 2.4 Hz).
2
29Si NMR (C6D6): d = ꢀ64.4 (s, Si(SiMe3)2Me, JSiP = <1.0 Hz),
Anal. Calc. for C12H36Si6PCl3 (486.26 g/mol): C, 29.64; H, 7.46.
Found: C, 28.74; H, 7.96%.
3
3
d = ꢀ10.7 (d, SiMe3, JSiP = 4 Hz), d = ꢀ10.0 (d, SiMe3, JSiP = 2 Hz),
1
d = 28.4 (d, SiCl, JSiP = 92 Hz).
1H NMR (ppm, C6D6): d = 0.34 (s, Si[Si(CH3)3]3, 27H), d = 0.44 (d,
Si(CH3)3, 9H).
31P NMR (C6D6): d = ꢀ92.5 (d, JPH = 226 Hz).
1
3
13C NMR (ppm, C6D6): d = ꢀ3.4 (d, Si[Si(CH3)3]3, JPC = 2.3 Hz),
2
3.3.2. Synthesis of hypersilyl-(3-chloro-1,1,1,2,4,5,5,5-octamethyl-2,
4-bis(trimethylsilyl)-n-pentasilanyl)phosphane (2)
d = 4.7 (d, Si(CH3)3, JPC = 12.2 Hz).
1
29Si NMR (ppm, C6D6): d = ꢀ89.4 (d, Si(SiMe3)3, JSiP = 85.8 Hz),
2
1
A THF solution of lithium tris(trimethylsilyl)silylphosphanide
(2.06 g, 7.2 mmol) was added dropwise to a solution of 3,3-di-
chloro-1,1,1,2,4,5,5,5-octamethyl-2,4-bis(trimethylsilyl)-n-pent-
asilane in THF at ꢀ78 °C. The reaction mixture was then allowed to
warm to room temperature slowly and was stirred for another
12 h. THF was then removed by evaporation under reduced pres-
sure. The reddish, oily residue was suspended in n-hexane, and
the insoluble lithium salts were separated by decantation. At
ꢀ80 °C, 2 crystallized in form of colourless crystals which, after a
second recrystallization were suitable for X-ray single crystal anal-
ysis. The yield was 1.75 g (37%).
d = ꢀ9.4 (d, Si(SiMe3)3, JSiP = 9.1 Hz), d = 6.7 (d, SiMe3, JSiP
=
1
34.8 Hz), d = 12.7 (d, SiCl3, JSiP = 92.4 Hz).
31P NMR (ppm, C6D6): d = ꢀ206.9.
4. Supplementary material
CCDC 703084, 703085 and 703083 contain the supplementary
crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crytallographic Data Centre
via www.ccdc.cam.ac.uk/data_reques/cif.
Anal. Calc. for C23H70Si11PCl (722.21 g/mol): C, 38.25; H, 9.77.
Found: C, 38.27; H, 9.76%.
Acknowledgement
1H NMR (C6D6): d = 0.36 (s, 36H), d = 0.41 (s, 27H), d = 0.49 (s,
1
6H), d = 2.05 (d, 1H, JPH = 198 Hz).
The authors gratefully acknowledge financial support by the
‘Fonds zur Förderung der wissenschaftlichen Forschung’, FWF,
Vienna (Project P-18176-N11).
1
2
29Si NMR (C6D6): d = 29.8 (d, SiCl, JSiP = 108 Hz, JSiH = 6 Hz),
2
d = ꢀ9.5 (d, Si(Si(CH3)3)3, JSiP = 10.0 Hz,), d = ꢀ11.2 (d, Si(CH3)3,
3JSiP = 1.0 Hz), d = ꢀ11.5 (d, Si(CH3), JSiP = 6.0 Hz), d = ꢀ67.3 (d, Si-
3
2
1
(SiMe3)2Me, JSiP = 12 Hz), d = ꢀ90.2 (d, Si(SiMe3)3Me JSiP
=
References
80.0 Hz).
31P NMR (C6D6): d = ꢀ210.1 (d, JPH = 198 Hz).
1
[1] R. West, M. Fink, J. Michl, Science 214 (1981) 1343.
[2] C.N. Smit, F.M. Lock, F. Bickelhaupt, Tetrahedron Lett. 25 (1984) 3011.
[3] M. Flock, private communication.
[4] R. Kinjo, M. Ichinohe, A. Sekiguchi, J. Am. Chem. Soc. 129 (2007) 26.
[5] C. Marschner, Eur. J. Inorg. Chem. 2 (1998) 221.
[6] A. Dzambaski, J. Baumgartner, K. Hassler, Silicon Chem. 3 (2008) 271.
[7] H. Kobayashi, T. Iwamoto, M. Kira, J. Am. Chem. Soc. 127 (2005) 15376.
3.3.3. Synthesis of lithium hypersilyl-(3-chloro-1,1,1,2,4,5,5,
5-octamethyl-2,4-bis(trimethylsilyl)-n-pentasilanyl)phosphanide (3)
Five milliliters of a 1.0 M n-pentane solution of n-butyl-lithium
(0.50 mmol) was added to a solution of 0.09 g (0.12 mmol) of 2 in

762
A. Dzambasky et al. / Journal of Organometallic Chemistry 694 (2009) 757–762
[8] N. Wiberg, W. Niedermayer, H. Nöth, M. Warchhold, Z. Anorg. Allg. Chem. 627
(2001) 1717.
[14] V. Cappello, J. Baumgartner, A. Dransfeld, K. Hassler, Eur. J. Inorg. Chem. (2006)
4589.
[9] N. Wiberg, W. Niedermayer, G. Fischer, H. Nöth, M. Suter, Eur. J. Inorg. Chem.
(2002) 1066.
[15] SAINTPLUS: Software Reference Manual, Version 6.45; Bruker-AXS, Madison, WI,
1997–2003.
[10] V. Cappello, J. Baumgartner, M. Flock, A. Dransfeld, K. Hassler, Eur. J. Inorg.
Chem. (2006) 2393.
[11] For a recent review see: M. Driess, Coord. Chem. Rev. 145 (1995) 1.
[12] U. Laky, M. Flock, K. Hassler, in preparation.
[16] R.H. Blessing, Acta Crystallogr. A 51 (1995) 33;
SADABS, Version 2.1; Bruker AXS: Madison, WI, 1998.
[17] G.M. Sheldrick, SHELX97, Programs for Crystal Structure Analysis (Release 97-2),
Universität Göttingen, Göttingen, Germany, 1998.
[13] M. Driess, C. Monse, K. Merz, Z. Anorg. Allg. Chem. 627 (2001) 1225.