Heterodehydrocoupling of phosphines and silanes catalyzed by titanocene: A novel route to the formation of Si-P bonds [17]

Full text of DOI:10.1021/ja981570q

12988
J. Am. Chem. Soc. 1998, 120, 12988-12989
Table 1. Heterodehydrocoupling of Phosphines and Silanesa
Heterodehydrocoupling of Phosphines and Silanes
Catalyzed by Titanocene: A Novel Route to the
Formation of Si-P Bonds
reaction
time (h)
phosphine
Ph2PH
silane
product
yield (%)b
p-TolSiH3
2
p-TolH2SiPPh2
p-TolHSi(PPh2)2
p-TolHSi(PPh2)2
CyH2SiPPh2
Ph2HSiPPh2
PhMeHSiPPh2
p-TolH2SiPCy2
p-TolH2SiPCy2
PhH2SiPCy2
CyH2SiPCy2
PhMeHSiPCy2
NR
84
16
Ronghua Shu,† Leijun Hao,† John F. Harrod,*,†
Hee-Gweon Woo,†,§ and Edmond Samuel‡
Ph2PH
Ph2PH
Ph2PH
Ph2PH
Cy2PH
p-TolSiH3
CySiH3
Ph2SiH2
PhMeSiH2
p-TolSiH3
9
2
2
2
3
24
24
24
24
24
24
24
3
100 (75)c
100
100
100
8.2
Chemistry Department
McGill UniVersity
Montreal, Quebec, Canada H3A 2K6
Laboratoire de Chimie
24
21
100
100
Cy2PH
Cy2PH
Cy2PH
PhPH2
PhPH2
PhPH2
CyPH2
PhSiH3
CySiH3
PhMeSiH2
p-TolSiH3
CySiH3
PhMeSiH2
p-TolSiH3
Organome´tallique de l’ENSCP
(URA 403 CNRS) 11 rue P. et M. Curie
75005 Paris, France
NR
NR
ReceiVed May 6, 1998
p-TolH2SiPHCy
p-TolHSi(PHCy)2
c-[-p-TolHSiPCy-]3 (30)
3.3
0.3
CyPH2
p-TolSiH3
650
We report herein the first examples of catalytic heterodehy-
drocoupling between Si-H and P-H (eq 1).
a Reactions were run at room temperature, without solvent, using
3-5 mol % Cp2TiMe2 based on silane. Phosphine/silane ratio ) 1:1.2
unless otherwise noted. All compounds were characterized by 1H, 31P,
and 29Si NMR. NR: no cross-coupling observed; Tol ) tolyl, Cy )
cyclohexyl. b Based on integration of 31P NMR. Figures in parentheses
are isolated yields.. c Phosphine/silane ratio ) 2.1.
R2PH + RSiH3 catalyst8 R2PSiH2R + H2
(1)
Dehydrocoupling and heterodehydrocoupling reactions to form
Sn-Te, Si-N, Si-C, Si-O, and B-N bonds have all been
achieved using group 4 metallocenes as catalysts.1 A notable
absence from the above list is a heterodehydrocoupling reaction
to form Si-P bonds. Catalytic dehydrocoupling has some
advantages over currently available methods for the synthesis of
Si-P bonds2 in that it can be carried out as a one-pot reaction
under mild conditions and it does not produce large amounts of
waste products that may be difficult to dispose of.
Some results for the heterodehydrocoupling of silanes and
phosphines with a Cp2TiMe2 precatalyst are shown in Table 1.
These reactions exemplify the range of reactivities and selectivities
that are achievable. A surprising feature is the much lower
reactivity of the less sterically encumbered phosphines. Such
behavior is counter to that expected for σ-bond metathesis
chemistry where the larger groups should hinder the formation
of the transition state.
Residual tertiary Si-H bonds are inactive toward coupling;
hence, all of the final products contain a Si-H bond. Residual
P-H bonds are more active, and although compounds with P-H
bonds were detected in the slow CyPH2 reactions, they undergo
further coupling to give longer chains. An NMR experiment
showed that, in the reaction of CyPH2 with p-TolSiH3, p-TolH2-
SiPHCy was the initial product, followed by CyP[SiH2p-Tol]2,
1. Subsequently, as the p-TolSiH3 concentration diminished
relative to that of CyPH2, p-TolHSi(PHCy)2, 2, and other higher,
uncharacterized oligomers appeared. After 2 days, white crystals
of 1,3,5-tri(cyclohexylphospha)-2,4,6-tri(p-tolylsila)cyclohexane,
3, began to appear and continued to accumulate, eventually
reaching a yield of 30% after several weeks.3 A similar reaction
of PhSiH3 and CyPH2 gave the previously reported phenylsila
analogue of 3 in an isolated yield of 53%.4 These six-membered
rings could either result from coupling of 1 and 2, or from
redistribution reactions. In either case their isolation is due to
their low solubilities in the reaction medium.
Some steps in the mechanism of titanocene-catalyzed SiH/PH
cross-dehydrocoupling are suggested by the catalytic and sto-
ichiometric reactions of Cp2Ti(PMe3)2, 4, with RR′PH (R ) Ph,
R′ ) Ph or H; R ) Cy, R′ ) H). Although the presence of the
PMe3 reduces the catalytic activity of 4 relative to that of Cp2-
TiMe2, it greatly facilitates the observation of intermediate
titanocene(III) species by EPR spectroscopy. A series of experi-
ments was performed which demonstrated the occurrence of the
reactions shown in Scheme 1. A reaction of of 4 with Ph2PH
(1:1 molar ratio) at room temperature gave the known phosphido
compound 5a as dark green crystals in 86% isolated yield.5 With
a 2:1 molar ratio of 4 to Ph2PH, EPR spectroscopy showed 5a
and the hydride 6 to be formed in a 1:1 ratio.6 These compounds
could result from the oxidative addition of a P-H bond to 1 (or,
more likely, one of its phosphine dissociation products) to give
the intermediate 7. This intermediate could transfer a hydrogen
atom to a second molecule of 4 to give an equal mixture of 5a
and 6. 7 was not observed, even at low temperatures, presumably
because of its rapid reaction with 4.7
(3) NMR in C6D6 at 25 °C: δ(1Η) ) 0.7-2.2 (m, C6H11), 2.01 (s, p-CH3),
6.0 (m, 3 H, Si-H), 7.04, 7.95 (m, m, 12 H, CH3C6H4); δ(31P) ) -160.0 (br);
δ(29Si) ) -15.0 (br); IR (KBr pellet): υ(SiH) ) 2104 (s) cm-1; FAB-MS
(m-nitrobenzyl alcohol matrix) m/z(%) 703 ((M + 1), 3), 619 ((M - C6H11),
9). Anal. Calcd for C39H57P3Si3 C, 66.63; H, 8.17; P, 13.22. Found C, 66.38;
H, 8.40; P, 13.02. The spectra data are comparable to the reported analogous
compounds.17
(4) (a) Driess, M.; Faulhaber, M.; and Pritzkow, H. Angew. Chem., Int.
Ed. Engl. 1997, 36, 1892. (b) Driess, M.; Pritzkow, H.; Skipinski, M. and
Winkler, U. Organometallics 1997, 16, 5108. (c) Driess, M.; Reisgys, M.;
Prizkow, H. Angew. Chem., Int. Ed. Engl. 1992, 31, 1510 and references
therein. (d) Driess, M.; Angew. Chem., Int. Engl. 1991, 30, 1022. (e) Driess,
M.; Fanda, A. D.; Powell, D. R.; West, R. Angew. Chem., Int. Ed. Engl. 1989,
28, 1038.
(5) Dick, D. G.; Stephan, D. W. Organometallics 1991, 10, 2811.
(6) Hao, L.; Lebuis, A.-M.; Harrod, J. F. J. Chem. Soc.; Chem. Commun.
1998 1089.
(7) The reaction between Ph2PH and 4 occurred at and above -60 °C
(monitored by variable temperature 1H and 31P NMR spectroscopy, which
showed a gradual decrease in the peak intensity of 1 and Ph2PH and an increase
in the peak intensity of the free PMe3 in the 1H and 31P NMR spectra. No
resonances due to 4 were detected.
* Corresponding author. E-mail: harrod@omc.lan.mcgill.ca.
† McGill University.
‡ l’ENSCP.
§ Present address: Department of Chemistry, Chonnam University, Kwangju
50-757, Korea.
(1) For reviews see: (a) Harrod, J. F. In Progress in Catalysis; Smith, K.
J., Sanford, E. C., Eds.; Elsevier: Amsterdam, 1992; p 147. (b) Corey, J. Y.
AdV. Silicon Chem. 1991, 1, 327. (c) Tilley, T. D. Acc. Chem. Res. 1993, 26,
22 and references therein. (d) Gauvin, F.; Harrod, J. F.; Woo, H. G. AdV.
Organomet. Chem. 1998, 42, 363.
(2) (a) King, R. B. Encyclopedia of Inorganic Chemistry; John Wiley &
Sons: New York, 1994; Vol. 6, p 2989. (b) King, R. B. Encyclopedia of
Inorganic Chemistry; John Wiley & Sons: New York, 1994; Vol. 7, p. 3790.
(c) Fritz, G. Angew. Chem., Int. Ed. Engl. 1966, 5, 53.
10.1021/ja981570q CCC: $15.00 © 1998 American Chemical Society
Published on Web 11/26/1998

Communications to the Editor
J. Am. Chem. Soc., Vol. 120, No. 49, 1998 12989
Scheme 1
Scheme 2
Scheme 3
A similar sequence to that shown in Scheme 1 occurred in the
reactions of 4 with primary phosphines RPH2 (R ) Ph, Cy), to
yield the phosphido compounds Cp2Ti(PHR)(PMe3) (5b, R ) Ph;
5c, R ) Cy), as shown by the EPR spectroscopy.8 5b and 5c
lose PMe3 easily to form the phosphide-bridged dimers [Cp2Ti-
(µ-PHR)]2 (8a, R ) Ph; 8b, R ) Cy). In concentrated solutions
the poor solubility of these dimers pulls the PMe3 dissociation
equilibrium in the direction of 8 by precipitation of the dark purple
crystals in high yield (82% for 8b and 81% for 8c) (see eq 2).
reactions to be complete immediately following mixing at room
temperature, in contrast to the much slower reactions of 5a with
PhSiH3 or that of 9 with phosphine. This accounts for the low
steady-state concentration of 6 and our failure to detect it by EPR,
even at temperatures down to -50 °C.
A set of reactions which conforms to the observed results is
shown in Scheme 2. This scheme involves a sequence of σ-bond
metatheses occurring between substrates and Ti(III)-X species,
where X ) silyl, phosphide, or H.11 It is likely that the σ-bond
metatheses are preceded by PMe3 dissociation from the EPR-
observed titanocene species, although there is no direct evidence
of this.
A contributing factor to the low reactivity of primary relative
to that of secondary phosphines (in addition to the insolubility of
the phosphide dimers, referred to above) is the high stability of
their complexes 10. 5b does react with PhSiH3 (2 equiv) at a
similar rate to 5a to give 9, and 5b is regenerated by treating 9
with PhPH2 (2 equiv). However, under catalytic conditions, the
phosphine substrate is present at a much higher concentration than
the catalyst. Measurements under catalytic conditions clearly show
that the excess substrate phosphine ligand completely converts
5b to 10b as the only EPR-observable product.8 On the other
hand, a mixture of paramagnetic species is observed with Ph2-
PH, or CyPH2, even at phosphine concentrations of >10:1 relative
to the catalyst.
No homodehydrocoupling reaction of PhSiH3 or of RR′PH
occurred in any of the reactions described above. Reaction of 4
with excess PhSiH3 first gives the η-2Si-H complex, Cp2Ti(η2-
HSiH2Ph)PMe3,12,13 which slowly decomposes to give 9 with
evolution of H2. No Si-Si coupling products were detected over
a period of weeks. Thus, the presence of phosphine suppresses
the homodehydrocoupling reactions, either through formation of
9 (silane coupling), or 5 (phosphine coupling). This also leads to
the conclusion that all of the σ-bond metatheses of the catalytic
cycle shown in Scheme 2 require predissociation of the phosphine
ligand.
On the other hand, 5a can be isolated either as a consequence of
tighter binding of the PMe3 or because its more sterically
demanding phosphido ligand destabilizes the dimer.5 Addition
of PMe3 to a toluene solution of 8b or 8c regenerated 5b and 5c,
respectively, proving the reversibility of eq 2.9 No reaction
occurred between 4 and Cy2PH, presumably because of steric
hindrance. The rates of the reactions of 4 with phosphines to give
5 follow the order PhPH2 > Ph2PH ≈ CyPH2 . Cy2PH.
Reaction of a toluene solution of 5a with PhSiH3 (2 equiv) for
2 h converted it completely to Cp2Ti(PhSiH2)(PMe3), 9,10 with
the production of Ph2PSiH2Ph and a minor amount of (Ph2P)2-
SiHPh. Addition of Ph2PH (2 equiv) to this purple solution
regenerated 5a fully in 1 h, showing that the conversion of 9 to
5a is about twice as fast as that of 5a to 9. An isolated sample of
910 reacted similarly with Ph2PH to give 5a and Ph2PSiH2Ph.
A solution of 6 was prepared by reaction of titanocene(III)
hydride with PMe310 and converted cleanly to 5a or 9 by adding
Ph2PH or PhSiH3, respectively. EPR spectroscopy showed these
(8) EPR data in toluene at 25 °C for 5b: g ) 1.9902, a(31P) ) 25.3 G,
a′(31P) ) 8.0 G, a(1H) ) 5.0 G.; for the PhPD analogue of 5b: g ) 1.9902,
a(31P) ) 25.3 G, a′(31P) ) 8.0 G, a(47/49Ti) ) 9.8 G.; for 5c: g ) 1.9895,
a(31P) ) 25.3 G, a′(31P) ) 8.4 G, a(1H) ) 2.8 G.; for 10: g ) 1.9728, a(31P)
) 4.66 G, a(47/49Ti) ) 13.20 G.; the same EPR spectrum as 10 was observed
for Cp2Ti(PDPh)(PD2Ph).
(9) The assignments of the structures of 8 are based on this reaction since
8 shows no structurally informative resonances in their EPR, 1H or 31P NMR
spectra. 8b was characterized earlier by a single-crystal X-ray structure analysis
and was shown to undergo the reaction shown in eq 2.1
(10) (a) Bercaw, J. E.; Brintzinger, H. H. J. Am. Chem. Soc. 1969, 91,
7301. (b) Marvich, R. H.; Brintzinger, H. H. J. Am. Chem. Soc. 1971, 93,
2046. (c) Bercaw, J. E.; Marvich, R. H.; Bell, L. G.; Brintzinger, H. H. J.
Am. Chem. Soc. 1971, 94, 1219. (d) Aitken, C.; Harrod, J. F.; Samuel, E. J.
Am. Chem. Soc. 1986, 108, 4059. (d) Kool, L. B.; Rausch, M. D.; Alt, H. G.;
Herberhold, M.; Thewalt, U.; Honold, B. J. Organomet. Chem. 1986, 310,
27 and 1987, 320, 37. (e) Meinhart, J. D.; Anslyn, E. C.; Grubbs, R. H.
Organometallics 1989, 8, 583. (f) Samuel, E.; Mu, Y.; Harrod, J. F.; Dromzee,
Y.; Jeannin, Y. J. Am. Chem. Soc. 1990, 112, 3435. (g) Gauvin, F.; Harrod,
J. F.; Samuel, E.; Britten, J. J. Am. Chem. Soc. 1992, 114, 1489. (h) Xin, S.;
Harrod, J. F.; Samuel, E. J. Am. Chem. Soc. 1994, 116, 11563. (i) Muhoro,
C. N.; Hartwig, J. F. Angew. Chem., Int. Ed. Engl. 1997, 36, 1510. (j) Hao,
L.; Lebuis, A.-M.; Harrod, J. F.; Samuel, E. J. Chem. Soc., Chem. Commun.
1997, 2193.
Acknowledgment. We thank the NSERC of Canada and the Fonds
FCAR du Que´bec for financial support.
Supporting Information Available: 1H, 29Si, and 31P NMR assign-
ments for all of the compounds in Table 1, EPR spectra of 5a-c, 6, 9,
and 10b, and experimental details for syntheses (6 pages, print/PDF).
See any current masthead page for ordering information and Web access
instructions.
JA981570Q
(11) We agree with the suggestion of a reviewer that a reaction of 9 with
phosphine to give 6 directly is less likely than its reaction to give 5 and silane.
(12) Spaltenstein, E.; Palma, P, Kreutzer, K. A.; Willoughby, C. A.; Davis,
W. M.; Buchwald, S. L. J. Am. Chem. Soc. 1994, 116, 10308.
(13) NMR data for Cp2Ti(η2-HSiH2Ph)PMe3 in toluene-d8 at -55 °C: δ
(1H) ) 5.70 (s, br, 2H, SiH2), -4.02 (d, 1H, 2JPH ) 77.7 Hz, TiHSi); δ (31P)
2
) 23.9 (d, 1P, JPH ) 77.7 Hz), -61.5 (free PMe3).