Insights into the reactivity and structure of silylene phosphonium ions

Article abstract of DOI:10.1039/b315001a

Experimental evidence for the intermediate occurrence of a so far unknown silylene phosphonium ion is presented and its reactivity and bond situation is discussed on the basis of ab initio calculations.

Full text of DOI:10.1039/b315001a

Insights into the reactivity and structure of silylene phosphonium ions†
Rudolf Pietschnig*
University of Graz, Department of Chemistry, Schubertstrasse 1, A-8010 Graz, Austria.
E-mail: rudolf.pietschnig@uni-graz.at; Fax: +43-(0)316-380-9835; Tel: +43-(0)316-380-5285
Received (in Cambridge, UK) 20th November 2003, Accepted 16th December 2003
First published as an Advance Article on the web 30th January 2004
Experimental evidence for the intermediate occurrence of a so
far unknown silylene phosphonium ion is presented and its
reactivity and bond situation is discussed on the basis of ab initio
calculations.
ppm. The 31P NMR signal appears at 103.9 ppm (4JPH = 12.2
Hz).
The formation of 4 can be explained when the abstraction of a
chloride ion from the tetrachloroaluminate anion of 2 is assumed.
Therefore it seems likely that in an initial step silylene phosphon-
ium ion 3 is actually formed. However, unlike its carbon analogue
3 appears not to be stable under the reaction conditions. Its
electrophilicity is obviously higher than that of AlCl3 which leads
Methylene phosphonium ions are cationic tricoordinate tetravalent
phosphoranes with a formal PC double bond (structure I). Initially
they had been postulated as intermediates,1 but soon after could
even be isolated.2 Some years later the existence of discrete
methylene phosphonium has been proven unambiguously by X-ray
crystallography.3 Moreover, the reactivity of methylene phosphon-
ium ions4 and their unusual bonding situation has been analyzed in
detail.5 In contrast nothing is known about the corresponding
silicon analogues so far.
Formally one might describe an analogous silylene phosphonium
ion (II) as an adduct between a silylene unit and a phosphenium ion.
Both of these fragments should possess singlet ground states6 and
therefore have the tendency to form a twisted non-classical double
bond. According to the CGMT model,7,8 this tendency should be
even more pronounced in silylene phosphonium ions than in the
case of methylene phosphonium ions.
2
to the transfer of a chloride ion from AlCl4 to the silicon
center.
To support this hypothesis we performed quantum chemical
calculations using the program package Gaussian 98.12 In order to
assess the relative electrophilicity of the species involved, we
explored the energy separation between the reactants and products
of the hypothetical reaction depicted in Scheme 2. On Hartree Fock
level using a triple zeta basis (6-311+G(d,p)) the mixture of
chlorosilylphosphane 6 and AlCl3 is about 67.42 Kcal mol21 more
stable than the silylene phosphonium tetrachloroaluminate 5.
Similar results are obtained if electron correlation is taken into
account using DFT methods. Thus with the hybrid functional
B3LYP employing a triple zeta basis set (6-311+G(d,p)) the
phosphinochlorosilane–AlCl3 is favored by 67.87 Kcal mol21
(Table 1). The energy separation for this reaction is even more
pronounced if electron correlation is included using the MP2(fc)
approach (70.87 Kcal mol21).
As a result of our calculations we obtained the geometry of the
silylene phosphonium ion 5, which shows some interesting
features. While the silicon atom shows a planar coordination
environment, the phosphorus atom in contrast adopts a pyramidal
geometry. The P–Si bond length is 2.255 Å, in the typical range for
a single bond,13 and only marginally shorter than in 6 (2.267 Å). On
the basis of a Mulliken population analysis14 both, the phosphorus
In this contribution we report on the formation of an intermediate
silylene phosphonium ion which due to its pronounced electrophilic
behavior is not stable under the reaction conditions. However, the
nature of the subsequent products supports the occurrence of this
species and explains its decomposition. Furthermore, energetic
aspects and the bond situation in a model silylene phosphonium ion
are explored by quantum chemical calculations.
Hypothetical cleavage of a silylene phosphonium ion should
furnish a silylene unit along with a phosphenium ion. Both of these
fragments show nucleophilic and electrophilic behavior at the same
positions and therefore they can be considered as singlet hetero-
carbenes.6 Due to the synthetic accessibility of both of these
components,9 it seems worthwhile to investigate whether a silylene
phosphonium ion can actually be generated from these compo-
nents. To explore this possibility we have chosen the “classical”
stable silylene 1 by West10 which is well known for its nucleophilic
character at the silicon atom11 and a diaminophospheniumtetra-
chloroaluminate 2 which is characterized by a high electrophilicity
as our starting materials.
Due to the electrophilic behavior of 2 the reaction was performed
at a low temperature (280 °C) in solution, where the onset of the
reaction can be observed by a color change to dark red. After the
reaction is finished the products can be identified spectroscopically
directly from the reaction mixture using 31P and 29Si NMR. As it
turns out, silylene phosphonium ion 3 is not the final product but
instead phosphanochlorosilane 4 can be identified (Scheme 1). In
the 29Si NMR spectra 4 shows a doublet (1JSiP: 43 Hz) at 226.5
Scheme 1 Adduct formation between silylene 1 and phosphenium ion 2.
† Dedicated to Professor Edgar Niecke on the occasion of his 65th
Scheme 2 Model reaction to assess the electrophilicity of 5 relative to
birthday.
AlCl3.
546
C h e m . C o m m u n . , 2 0 0 4 , 5 4 6 – 5 4 7
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

Table 1 Computed energies for reactants and products of the model reaction in Scheme 2 and computed PSi distances [Å] and atomic charges in 5a
[ > PNSi < ]+
+
> P–SiCl < +
AlCl3 [hartree]
DE/Kcal
AlCl42 [hartree]
mol21
d(P–Si) q(P)
q(Si)
Symmetry
HF//6-311+G(d,p)
MP2//6-311+G(d,p)
B3LYP//6-311+G(d,p)
22932.393388
22933.869408
22938.372453
22932.500824
22933.982339
22938.480609
267.42
270.87
267.87
2.256
2.234
2.255
+0.66
+0.66
+0.55
+1.31
+1.32
+1.10
C1
C1
C1
a Computational details are available as supporting information.
and the silicon atoms are positively charged due to the adjacent
nitrogen atoms that are more electronegative. However the
phosphorus atom in 5 shows only approximately half of the positive
charge of the silicon atom (Table 1). This is in good agreement with
our experimental findings where a chloride ion adds to the silicon
atom in 3 rather than to the phosphorus atom. The electrophilic site
in such ions obviously seems to be at the silicon atom. The
reactivity of 3 and the geometry obtained for 5 differ substantially
from the findings for methylene phosphonium ions, which show a
short but twisted central P–C bond and a positively charged but
planar phosphorus atom concomitant with a negatively charged
planar carbon atom.3,5
Our findings support the formulation of 5 as an adduct in which
the lone pair at the silicon atom of a silylene donates into the vacant
orbital located at the phosphorus atom in a phosphenium ion.
Therefore it is structurally more related to classical base adducts of
phosphenium ions15–17 than to methylene phosphonium ions and
could be described as phosphanyl silyl cation. The bond situation in
5 is represented better by description A rather than B in Scheme
3.
By exploring the chemistry of the chloride adduct of silylene
phosphonium ion, we found that under ambient conditions the PSi
bond in 4 is cleaved hydrolytically and the phosphane fragment is
replaced by a hydrogen atom (Scheme 4). As final products
hydridochlorosilane 7 and bisdicyclohexylaminophosphane oxide
8 can be isolated. Compounds 7 and 8 have been previously
described in the literature18,19 and were identified NMR spec-
troscopically (29Si, 31P, 1H) and by high resolution mass spec-
troscopy.20
An interesting aspect in the formation of 7 is the fact that the
cleavage of the P–Si bond in 6 follows the opposite direction than
would be expected from the electronegativity values of these atoms
(P: 2.1; Si: 1.7).21 For comparison, the hydrolytic bond cleavage in
trisorganosilylphosphanes is in accord with the bond polarity on the
basis of the standard electronegativity values and produces
hydridophosphanes along with silanols and/or their condensation
products.
reactivity of the aromatic silylene 1 is mostly based on the
nucleophilicity of its lone pair, the interaction with the phosphen-
ium ion causes a reversed reactivity at the silicon atom by altering
its character from nucleophilic to strongly electrophilic. Or in other
words, by forming this adduct the electrophilic character of the
phosphenium ion is transferred to the otherwise nucleophilic silicon
atom.
In summary, we could show that silylene phosphonium ions are
likely intermediates in the reaction of singlet silylenes with
phosphenium ions. Their reactivity and structure differs quite
substantially from that of their carbon analogues and is charac-
terized by a strong electrophilicity at the silicon atom. In fact there
should be a reasonable chance to isolate such adducts by employing
modern non coordinating counter ions.
Notes and references
1 T. A. van der Knaap and F. Bickelhaupt, Tetrahedron Lett., 1982, 23,
2037.
2 R. Appel and R. Schmitz, Chem. Ber., 1983, 116, 3521.
3 A. Igau, A. Bacareido, H. Grützmacher, H. Pritzkow and G. Bertrand, J.
Am. Chem. Soc., 1989, 111, 6853.
4 J. Thomaier, G. Alcaraz, H. Grützmacher, H. Hillebrecht, C. Marchand
and U. Heim, J. Organomet. Chem., 1997, 535, 91.
5 M. Ehrig, H. Horn, C. Kölmel and R. Ahlrichs, J. Am. Chem. Soc., 1991,
113, 3701.
6 J. F. Harrison, R. C. Liedtke and J. F. Liebman, J. Am. Chem. Soc., 1979,
101, 7162.
7 G. Trinquier and J. P. Malrieu, J. Am. Chem. Soc., 1987, 109, 5303.
8 E. A. Carter and W. A. I. Goddard, J. Chem. Phys., 1986, 90, 998.
9 Synthetic Methods of Organometallic and Inorganic Chemistry, ed. W.
A. Herrmann, Thieme, Stuttgart, 1996.
10 M. Denk, R. Lennon, R. Hayashi and R. West, J. Am. Chem. Soc., 1994,
116, 2691.
11 P. P. Gaspar and R. West, in The chemistry of organic silicon
compounds, Vol.2, ed. Z. Rappoport and Y. Apeloig, 1998, 2463.
12 Gaussian Inc., 1998, www.gaussian.com.
13 P. Rademacher, Strukturen organischer Moleküle, VCH, 1987.
14 R. S. Mulliken, J. Chem. Phys., 1955, 23, 1833.
15 C. W. Schultz and R. W. Parry, Inorg. Chem., 1973, 15, 3046.
16 A. H. Cowley and R. A. Kemp, Chem. Rev., 1985, 85, 367.
17 T. I. Solling, S. B. Wild and L. Radom, Inorg. Chem., 1999, 38,
6049.
Considering the complete reaction sequence, 1 is transformed
into 7, which is formally an oxidative addition of one equivalent of
HCl to the silicon atom of the unsaturated silylene. Since the
18 H. tom Dieck and M. Zettlitzer, Chem. Ber., 1987, 120, 795.
19 H. Wong, M. M. Turnbull, K. D. Hutchinson, C. Valdez, E. J. Gabe, F.
L. Lee and Y. Le Page 1988 110 8422.
20 Chlorosilylphosphane 4: a solution of 0.56 g (1 mmol) of 2 in 10 mL of
dry methylene chloride, obtained in a similar fashion as described in ref.
9, is slowly added to 0.20 g (1 mmol) of silylene 1 dissolved in 10 mL
of diethyl ether at 280 °C. The reaction mixture is concentrated under
vacuum and transferred into an NMR tube with a capillary containing
Scheme 3 Phosphanyl silyl cation vs. silylene phosphonium ion.
1
C6D6 as lock substance. 29Si (49.7 MHz): 226.5 ppm (dt, JSiP = 43
4
Hz); 31P (101.2 MHz): 103.9 ppm, JPH = 12.2 Hz. All attempts to
isolate the product from the reaction mixture resulted in hydrolytic
cleavage of the P–Si bond under formation of 7 and 8. 1H (250.1 MHz):
6.39 (t, 4JHH = 7.0 Hz, 1H, > SiClH), 5.84 (d, 4JHH = 7.0 Hz, 2H, C–
H), 1.32 (s, 18 H, t-Bu). 29Si (49.7 MHz): 242.0 ppm (dt, 1JSiH = 319
Hz, 3JSiH = 319 Hz, MS (EI): 231.1082 (calc.: .1084), M+ 2 H.
21 A. L. Allred and E. G. Rochow, J. Inorg. Nucl. Chem., 1958, 5, 264.
Scheme 4 Hydrolytic cleavage of the Si–P bond in 4.
C h e m . C o m m u n . , 2 0 0 4 , 5 4 6 – 5 4 7
547