of the products were made, but this should readily be feasible by
using the same protocol that was applicable for the isolation of
5b.{ Preliminary studies indicate that the catalytic reaction is also
suitable for the alkylation of other silyl phosphines.
The spectroscopically determined yields of products in the
individual transformations are listed in Table 1 and reveal that the
reactions studied may roughly be divided into three different
categories. Highly reactive alkyl chlorides such as chloroacetyl
acetate and allyl chloride (entries 1 and 2) were found to give
quantitative conversion of 6 into the corresponding tertiary
phosphines within 24 h at ambient temperature even without
catalyst. Addition of a catalytic amount of 1b had no visible effect
in these cases. Moderately reactive alkyl halides such as benzyl
chloride and a-chloroacetonitrile (entries 3 and 4) required longer
reaction times or increased reaction temperatures. Formation of
minor amounts of the corresponding tertiary phosphines was in
this case observed even without catalyst but the yields improved
substantially in the presence of a catalytic amount of 1b. Finally,
unreactive alkyl halides such as dichloromethane and chlorobutane
(entries 5 and 6) were completely unreactive, even at 90 uC, in the
absence of catalyst but afforded 90% or higher yields of the
alkylation product in the presence of 1b. The coupling of 6 with
CH2Cl2 yielded exclusively the monosubstituted a-chloromethyl-
phosphine 3b; phosphination of the second chlorine atom to afford
diphenylphosphinomethane (dppm) was not observable, even
when an excess of the silyl phosphine was employed.
31P NMR studies of the catalytic reactions confirmed that the
diphosphine 1b is in fact formed under the reaction conditions and
is finally converted back into the P-chloro-diazaphospholene 2b
when the silyl phosphine 6 is completely consumed. Apart from 1b,
no further phosphorus-containing reaction intermediates were
detected. These observations confirm that the diazaphospholene
participates actively in the metathesis reaction and brings about an
acceleration of the phosphorus–carbon bond formation process.
The data are further in accord with the anticipated mechanism
outlined in Scheme 2 and suggest that alkylation of 1b is
presumably the rate determining step. Considering that the overall
process relies on the recycling of a Lewis acidic phosphenium
fragment [R2P+] between the diphosphine 1b and the P-chloro-
diazaphospholene 2b, the crucial factor for the operation of the
catalytic cycle is obviously the P–X bond polarisation (X = Cl,
PPh2) in both intermediates, which controls the weakening of the
P–X bonds and, at the same time, enhances the nucleophilicity of
X;5,7 the interplay of both effects makes the intermediates behave
as ‘‘phosphenium-phosphide’’ complexes and allows easy turnover
of the anionic fragments X. As the bond polarisation is intimately
connected with the unique stability of diazaphospholenium
cations5,7 it appears that the decisive element for the functioning
of the catalysis lies indeed in the characteristics of the electrophilic
[R2P+] fragment in the intermediates 1b and 2b rather than their
nucleophilicity.
Scheme 2 Diazaphospholene-catalysed cross-coupling of Me3SiPPh2 (6)
with alkyl chlorides XCH2Cl.
take place under very mild conditions, it is conceivable that a
stepwise transformation 6 A 1b A 3b may actually proceed faster
than direct alkylation of the silyl phosphine 6. As the P-chloro-
diazaphospholene 2b that is required for the formation of 1b is
regenerated in the alkylation step, the whole process can, under
these conditions, be formulated as a cyclic sequence (Scheme 2)
which would permit the accomplishment of the acceleration of the
phosphorus–carbon bond formation in the presence of a catalytic,
rather than a stoichiometric, quantity of a diazaphospholene.
In order to verify this hypothesis we studied the diazapho-
spholene-catalysed alkylation of silyl phosphine 6 with a range of
alkyl chlorides. The reactions were conducted by stirring a toluene
solution containing 6, the appropriate alkyl chloride (2 molar
equiv.) and a catalytic amount of the P-chloro-diazaphospholene
2b (20 mol% with respect to 6), at temperatures between 25 uC and
90 uC, for periods between 24 and 170 h (see Table 1). In each case
a control reaction was carried out by performing a separate
reaction under identical conditions but without catalyst. All
reaction mixtures were analysed by 31P NMR and 1H,31P HMQC
NMR spectroscopy. The products were identified by the
evaluation of characteristic correlation signals in the 2D NMR
spectra and by comparison of the observed chemical shifts with
those of authentic samples.8 The yields of products and unreacted
starting materials were determined by integration of the appro-
priate NMR signals. In all cases, the yield of product exceeded the
amount of 2b added, thus proving that the latter acts as a catalyst
rather than a stoichiometric reagent. No attempts toward isolation
Table 1 Reaction conditions and yields for cross-coupling reactions
of Ph2PSiMe3 (6) with alkyl chlorides RCl
Reaction
T/uC time/h
Yield of
Catalyst/mol%a Ph2PR (%)b
Entry R
1
2
3
4
5
6
a
–CH2CO2Et
–CH2CO2Et
–CH2CHLCH2 25
–CH2CHLCH2 25
–CH2Ph
–CH2Ph
–CH2CN
–CH2CN
–CH2Cl
–CH2Cl
n-Bu
25
25
24
24
120
120
48
48
72
72
24
20
—
20
—
20
—
20
—
20
—
20
—
.99
.99
.99
.99
90
10
80
40
.99
0
90
0
50
50
25
25
90
90
90
90
In summary, it has been demonstrated that the P-phosphinyl-
diazaphospholene 1b undergoes not only stoichiometric metathesis
with alkyl halides but may also act as an organocatalyst that
promotes the formation of a phosphorus–carbon bond in the
condensation of a silyl phosphine with alkyl chlorides. The use of a
diazaphospholene as catalyst allows such reactions to be
conducted at lower temperatures than usual and permits the
coupling of a silyl phosphine with unactivated alkyl chlorides
24
170
170
n-Bu
Given values denote the molar ratios of the catalyst (2b) relative to
the substrate (6). As determined by integration of 31P{1H} NMR
spectra.
b
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 2810–2812 | 2811