Organometallics 1995, 14, 2868-2879
2868
Unusual Rate Enhancement in the RhCl(PPh3)3=CatalyzedHydrosilylation by Organosilanes Having Two Si-H Groups at Appropriate Distances: Mechanistic Aspects Hideo Nagashima, Kazuo Tatebe, Toshinori Ishibashi, Akihito Nakaoka, Jun Sakakibara, and Kenji Itoh* Department of Materials Science, Toyohashi University of Technology, Toyohashi, Aichi 440,Japan Received October 13, 1994@ Unusual rate enhancement observed in the RhCl(PPh3)s-catalyzed hydrosilylation of carbonyl compounds with certain a,w-bifunctional organosilanes was studied in two series of experiments. First, the reactions with Me2HSi(CHz),SiHMe2 [l (n = l),2 ( n = 2), 3 ( n = 3), and 4 ( n = 4)1, RzHSi(CH2)2SiHPhz [8 (R = Me) and 9 (R = Ph)], and 1,2-[MezHSi(CH2),1[Me2HSi(CH2),,lCsH4 [5 (n = n' = O), 6 (n = 0 , n' = 11, and 7 (n = n' = 1)l were investigated in order to understand the rate acceleration by these two closely spaced Si-H groups. The reactions of five of these bifunctional organosilanes, 2,3,and 5-7, with acetone were unusually rapid and resulted in selective conversion of only one of their Si-H bonds to a Si-OiPr group within several hours at room temperature. The reaction of their remaining Si-H bonds was as slow as the hydrosilylation with monofunctional organosilanes such as EtMezSiH and PhMezSiH; the conversion was below 25% after 1 day at room temperature. The rate of the reaction of acetone with 1 or 4 was similar to that of EtMe2SiH or PhMezSiH. These results suggest that the large enhancement in rate occurred in those bifunctional organosilanes in which two closely spaced Si-H groups were connected by 2-4 carbon units. Similar rate enhancement, compared with monofunctional organosilanes, was observed at 50 "C in the hydrosilylation of 8 or 9 and led to the selective conversion of one Si-H group to a Si-OiPr moiety. Analysis of the products revealed involvement of redistribution of methyl groups in the reaction with 5. Hydrosilylation of unsymmetrical bifunctional organosilanes, 6 or 8, gave a 1:l mixture of two isomers. In the second approach, the stoichiometric reaction of 8 or 9 with RhCl(PPh& was studied by 'H and 31PNMR spectroscopy. The product obtained was dependent on the solvent used; in CDC13, Rh(II1)-
-
oxidative adducts, R12HSi(CH2CH2)R22Si-RhHC1(PPh3)2(R1, R2 = Me or Ph), having a trigonal bipyramidal structure with two PPh3 ligands a t the apical positions were obtained, whereas the spectra of such reaction mixtures obtained in toluene-ds suggested the formation
of Rh(V)-double oxidative adducts, R12Si(CH&H2)R22Si-RhH3(PPh3)2. Since the catalytic hydrosilylation of acetone with 2 or 8 proceeded in toluene-ds, but did not in CDC13, it is likely that the Rh(V)-double oxidative adducts play an important role in the rapid hydrosilylation of ketones with one end of the bifunctional organosilanes. These experimental results allow us to discuss two probable mechanisms involving disilametallacyclic intermediates for the rate enhancement by the two closely spaced Si-H groups. Introduction The hydrosilylation of unsaturated organic molecules in which a Si-H bond adds across a carbon-carbon, carbon-oxygen, or carbon-nitrogen multiple bond with the aid of a transition metal catalyst is an important industrial and laboratory process for the synthesis of organosilicon c0mpounds.l Numerous studies have been carried out using a variety of hydrosilanes, catalysts, and unsaturated substrates from various facets of the @Abstractpublished in Advance ACS Abstracts, May 1, 1995. (1) (a) Speier, J. L. Adu. Organomet. Chem. 1979,17,407.(b) Ojima, I. In The Chemistry of Organic Silicon Compounds; Patai, S., Rapport, Z., Eds.; Willey: New York, 1989.(c) Marciniec, B.;Gulinski, J. J . Organomet. Chem. 1993,446, 15. (dj Comprehensive Handbook on Hydrosilylation; Marciniec, B., Ed.; Pergamon: Oxford, U.K., 1992.
0276-7333/95/2314-2868$09.00/0
reaction since its discovery in the late 1940s, and the hydrosilylation reaction continues to receive much attention. In 1989, we reported an unusual selectivity observed in the RhCl(PPhJ3-catalyzed hydrosilylation of unsaturated molecules with MeaHSiCHzCHzSiHMez, in which only one of the Si-H group in the molecule reacts with the unsaturated substrate (X= Y)while the other one remains intact (Scheme 1).2a This selectivity was caused by the fact that the first step of the reaction, i.e., from MezHSiCHnCHzSiHMez (2) to Me2HSiCH2CH2Si(XYHIMe2, was several to several 10 times faster than the second step, Le., from Me2HSiCH&HzSi(XYH)Mez (2)(a) Nagashima, H.; Tatebe, K.; Ishibashi, I.; Sakakibara, J.;Itoh, K. Organometallics 1989,8,2495. (bj Nagashima, H.; Tatebe, K.; Itoh, K. J . Chem. Sac. Perkin Trans. I 1989,1707.
0 1995 American Chemical Society
RhCl(PPh&-Catalyzed Hydrosilylation
Organometallics, Vol. 14,No. 6,1995 2869
Scheme 1 Me, M e
Si, Me' Me
Scheme 2
Me, ,Me X=Y
Me, .Me
(s:; p!
% H Y (& -Me' 'Me
Me Me
Me, .Me MLn __c
Si
(
>MLn* Si Me' 'Me
X=Y: ketone, olefin, acetylene, catalyst; RhCI(PPh& Me, Me
Me. .Me
Me. .H
Me' 'Me
Me'- 'Me
Me' M e
X=Y; ketone, catalyst; RhCI(PPh&
t o Mez(HYX)SiCHzCHzSi(XYH)Mez.Since the hydrosilylation of unsaturated molecules with monofunctional organosilanes such as EtMezSiH and PhMezSiH was as slow as the second reaction, the first reaction apparently was accelerated for some unrecognized reason. Similar rapid reaction and selective conversion of only one Si-H group were also observed in the hydrosilylation of ketones with 1,2-bis(dimethylsilyl)benzene(51, which was accompanied by a curious redistribution of methyl groups on the silicon atoms to give a mixture of two products as shown in Scheme 1.2bDespite the numerous studies reported in the literature on the catalytic hydrosilylation reaction little is known concerning such reactions of bifunctional organosilanes 2 and 5.l The only example of the catalytic hydrosilylation with bifunctional organosilanes to our knowledge is that of the RhCl(PPh&-catalyzed reactions of nitriles reported by Corriu, which afforded N,N-disilylated amines and enamines by the concomitant reaction of both of the Si-H groups with nitrile^.^ Since nitriles are not hydrosilylated with monofunctional organosilanes under the same conditions, this reaction may suggest a higher reactivity of bifunctional silanes, similar to our observation in the hydrosilylation of ketones, olefins, and acetylenes. However, no detailed study of the mechanism of the reaction has been undertaken. How does the rate enhancement occur? It is wellknown that inductive and steric factors of the substituents on the silicon atom affect the rate of the hydrosilylation, and these are sometimes accompanied by significant differences in selectivities of the reaction.ld As a typical example, the RhCl(PPh3)s-catalyzed hydrosilylation of ketones with dialkyldihydrosilanes proceeded more rapidly than that with trialkylmonohyd r ~ s i l a n e s . The ~ , ~ reaction of a,p-unsaturated ketones with PhzSiHz resulted in the addition of an Si-H bond to the carbonyl group to give the corresponding silyl ether of allylic alcohols, whereas the reaction with Et3SiH proceeded in a 1,4-addition mode to form the corresponding silyl enol ether^.^ Stereoselectivity for the hydrosilylation of 4-tert-butylcyclohexanonewas significantly different between PhzSiHz and Et3SiH.5 These features depend on steric and inductive influence around Si-H bonds between dihydrosilanes and monohydrosilanes. In contrast, the hydrosilylation of mesityl oxide or 4-tert-butylcyclohexanone with bifunctional (3) Corriu, R. J. P.; Moreau, J. J. E.; Pataud-Sat, M. J . Organomet. Chem. 1982,228,301. (4) Ojima, I.; Nihonyanagi, M.; Kogure, T.; Kumagai, M.; Horiuchi, S.;Nakatsugawa, K. J . Organomet. Chem. 1975.94,449. Oiima. I.; Kogure, T. OrganometallicH 1982,1 , 1390. (5) Semmelhack, M. F.; Misra, R. N. J . Org. Chem. 1982,47,2469.
MLn = e.g. Pt(PPh,),, Fe(CO),, Ru(CO), (ref. 6) Me,-Me
(ref.7)
organosilanes, 2 or 5, is much faster than that with EtMezSiH or PhMezSiH, but the selectivities are similar. Since there is little difference on the substituents of the silicon atom between 2 or 5 and EtMezSiH or PhMezSiH, inductive or steric effects may not be a major reason for the rate enhancement in the reaction with 2 or 5. A probable factor which may result in the special reactivity of 2 or 5 is their bifunctional structure, two Si-H bonds of which can concomitantly interact with the catalyst. Fink and Graham reported that a stoichiometric reaction of 2 or 5 with certain transition metal complexes results in release of molecular hydrogen to form stable disilametallacyclic compounds (Scheme 216 Double oxidative addition of two Si-H groups in 2 or 5 to the metal center followed by reductive elimination of Hz is the probable mechanism for the formation of these disilametallacyclic compounds. Tanaka and coworkers reported a platinum-catalyzed dehydrogenative double silylation of unsaturated molecules with 5 and proposed a platinadisilacyclopentaneintermediate formed by double oxidative addition of two of the Si-H bonds in 5 followed by release of H z . ~The disilametallacyclic intermediate undergoes insertion of unsaturated molecules between Pt-Si bonds to accomplish the dehydrogenative disilylation as shown in Scheme 2. An important requirement for these reactions is the proximity of the two Si-H bonds in the bifunctional structure of 2 or 5, which is crucial in facilitating the double oxidative addition to the metal center. In fact, the above (6) Fink, W. Helu. Chim. Acta. 1974,57,1010; 1976,59,606. Corriu, R. J. P.; Moreau, J.; Pataud-Sat, M. Organometallics 1985,4,623. Vancea, L.; Graham, W. A. G. Inorg. Chem. 1974,13, 511. Recent advances for disilarhodacycles: Osakada, K.; Hataya, K.; Tanaka, M.; Nakamura, Y.; Yamamoto, T. J. Chem. Soc., Chem. Commun. 1993, 576. (7) (a) Uchimaru, Y.; Lautenschlager, H. J.; Wynd, A. J.; Tanaka, M.; Goto, M. organometallics 1992,11, 2639. (b) Tanaka, M.; Uchimaru, Y.; Lautenschlager, H. J. J . Organomet. Chem. 1992,428,l. (c) Reddy, N.P.; Uchimaru, Y.; Lautenschlager, H. J.; Tanaka, M. Chem. Lett. 1992,45. (d) Tanaka, M.; Uchimaru, Y.; Lautenschlager, H. J. Organometallics 1991, 10, 16.
Nagashima et al.
2870 Organometallics, Vol. 14,No. 6, 1995
3
2
1
4
Me
Me, Me
Me‘ ‘Me 5
Me
SiHMez (SiHPh,
6
Me
SiHPh, 8
(SiHPh,
9
Figure 1. Bifunctional organosilanes. stoichiometric and catalytic reactions were not accessible with the monofunctional organosilanes. These results are consistent with the fact that the rapid RhC1(PPh&catalyzed hydrosilylation was only accomplished by the bifunctional organosilanes, 2 or 5, but not by EtMezSiH or PhMezSiH, which strongly suggests that concomitant interaction of the two Si-H groups in 2 or 5 with the rhodium center is a clue to understanding the mechanism. In this paper, we wish t o report two experiments performed to understand this unusual rate enhancement of bifunctional organosilanes in the RhCl(PPhd3catalyzed hydrosilylation reaction. First, nine bifunctional organosilanes, shown in Figure 1, which had either a different number of carbon units between the two Si-H groups or different substituents on the silicon atoms, were used in the catalytic hydrosilylation of acetone. The results revealed that the structure of the bifunctional organosilanes appropriate for the concomitant interaction of both of the Si-H groups to the Rh center generally showed a rate enhancement. Second, direct detection of the interaction of two Si-H groups in the bifunctional organosilanes, 2,8, and 9, with RhC1(PPh3)3was examined by lH and 31PNMR spectroscopy, and provided evidence for double oxidative addition, i.e., formation of a Rh(V)-disilametalacyclic intermediate,
-
H~R~(RzS~CHZCHZS~RZ)(PP~~)Z. These results indicate the key role of Rh(V)-disilametallacyclic intermediates in the catalytic hydrosilylation with bifunctional organosilanes. In contrast t o the great success of the hydrosilylation in organic and organometallic synthesis, the reaction mechanism still remains uncertain.l The Chalk-Harrod cycle proposed for the platinum-catalyzed hydrosilylation of olefins offered the first reasonable mechanism, which involves the oxidative addition of a Si-H bond to a platinum-alkene complex, followed by alkene insertion into the F’t-H bond and reductive elimination of the alkyl and silyl groups t o form a new silane.8 The Chalk-Harrod cycle and its variants have been extensively used t o explaih various transition metal-catalyzed hydrosilylations including those mediated by RhC1(PPh3)3.4 However, recent publications claimed the (8)Chalk, A. J.; Harrod, J . F. J . Am. Chem. SOC.1965, 87, 16. Harrod, J . F.; Chalk, A. J. In Organic Synthesis via Metal Carbonyls; Wender, I., Pino, P., Eds.; Willey: New York, 1977;Vol. 2.
involvement of higher oxidation states such as Rh(V),gJO Two Ir(V),ll and F’t(IV)12and metal colloidal ~ata1ysis.l~ reaction pathways involving silylmetalation and hydrometalation of unsaturated molecules also were a subject of controversy.14-17The present report offers a mechanistic discussion on the catalytic hydrosilylation from a fresh perspective.
Results and Discussion I. Catalytic Hydrosilylation of Acetone with Bifunctional Organosilanes 1-9. Organosilanes of general formula HMezSi(CHz),SiMezH (1-4 ( n = 1-41] were synthesized and used for the catalytic hydrosilylation of acetone (an equimolar amount with respect to the silane) in the presence of RhCl(PPh& (1mol %) in C6D6 a t 30 “C under a nitrogen atmosphere. Reaction profiles following the disappearance of acetone as determined by lH NMR spectroscopy are shown in Figure 2. The rate apparently was dependent on the number of carbon units ( n )between the two MezSiH groups; the reactions with silanes 2 and 3 were complete within 1 h, whereas the conversion of acetone in the reactions of 1 and 4 was less than a few percent after 1 h. Similar reactions of EtMezSiH and PhMezSiH were as slow as those of 1 and 4. It was estimated from the initial rate analysis that the reactions of 2 and 3 were approximately 50 and 120 times faster, respectively, than those of 1, 4, EtMezSiH, and PhMezSiH.18 That no rate acceleration occurred in the hydrosilylation with 1 and 4 indicates that the distance between the two Si-H groups in the bifunctional organosilane is particularly important; that in 1 is too short, whereas that in 4 is too long. (9)(a) Millan, A.;Fernandez, M. J.;Bentz, P.; Maitlis, P. M. J . Mol. Catal. 1984,26,89. (b) Ruiz, J.;Bentz, P. 0.;Mann, B. E.; Spencer, C. M.; Taylor, B. F.; Maitlis, P. M. J. Chem. SOC.,Dalton Trans. 1987, 2709. (10)(a)Duckett, S. B.; Perutz, R. N. Organometallics 1992,1I,90. (b) Duckett, S. B.; Haddleton, D. M.; Jackson, S. A,; Perutz, R. N.; Poliakoff, M.; Upmacis, R. K. Organometallics 1988,7,1526. (11)Tanke, R. S.;Crabtree, R. H. J . Chem. SOC.,Chem. Commun. 1990,1056; J. Am. Chem. SOC. 1990,112,7984;Organometallics 1991, 10,415. (12)Chu, H.K.;Frye, C. L. J . Organomet. Chem. 1993,446,183. (13)Lewis, L. N. J . Am. Chem. SOC.1990,112,5998.Lewis, L.N.; 1986,108,7228. Lewis, N. J . Am. Chem. SOC. (14)Ojima, I ’ Clos, N.; Donovan, R. J.; Ingallina, P. Organometallics 1990,9,3127. Ojima, I.; Donovan, R. J.;Clos, N. Organometallics 1991, 10, 2606. (15)Brookhart, M.; Grant, B. E. J . Am. Chem. SOC. 1993,115,2151. (16)(a) Tamao, K.;Nakagawa, Y.; Ito, Y. Organometallics 1993,12, 2297.(b) Bergens, S. H.; Noheda, P.; Whelan, J.; Bosnich, B. J. Am. Chem. SOC. 1992,114,2121,2128. (17)Seitz, F.; Wrighton, M. S. Angew. Chem., Int. Ed. Engl. 1988, 27,289.Mitchener, J . C.; Wrighton, M. S. J. Am. Chem. SOC.1981, 103,975.Schroeder, M. A.;Wrighton, M. S. J . Organomet. Chem. 1977, 128,345. (18)One may consider that a further kinetic study could contribute to a better understanding of the reaction profiles. However, the large experimental errors observed in each run made us hesitate to do kinetic experiments in detail. Three kinetic runs with 2 under the same conditions as above (under a nitrogen atmosphere) gave a rate constant with +20% experimental error, whereas four experiments conducted in an NMR tube sealed in vacuum gave a smaller rate constant with f 5 0 % experimental error. The origin of the large experimental errors was attributed to oxygen contamination of the reaction mixture, because it is known that a trace amount of oxygen accelerates the RhCKPPh&catalyzed hydrosilylation of olefins.1g As long as the reaction was carried out under a nitrogen atmosphere, at least three repeated runs with each hydrosilane (1-9)proved reproducibility of the initial rate constant with maximum uncertainty 130%, which is enough for qualitative comparison of the reaction rate as described in the text. However, we consider that attempted quantitative analysis using these data would be unproductive. (19)Dickers, H. M.; Haszeldine, R. N.; Malkin, L. S.; Mather, A. P.; Parish, R. V. J . Chem. SOC.,Dalton Trans. 1980,308.Faltynek, R.A. Inorg. Chem. 1981,20, 1357.
RhCKPPh&-Catalyzed Hydrosilylation
Organometallics, Vol. 14, No. 6, 1995 2871
100
0
40
P ."E
8 2 u
20
0 20
40
60
80
100
120
time (min)
20
40
60
80
100
120
time (min)
Figure 2. Reaction profiles for the catalytic hydrosilylation of bifunctional organosilanes 1-7. All reactions were carried out in CsDe at 30 "C in the presence of 1mol % of RhCl(PPh& as the catalyst. The reaction was monitored by following the decrease of acetone against toluene as the internal standard. In Figure 2 are also illustrated reaction profiles of the hydrosilylation with the silanes 1,2-[MezHSi(CH~),l[MezHSi(CHz),,]CsH4 [5 ( n = n' = O), 6 ( n = 0, n' = l),and 7 ( n = n' = 113, under the same conditions as above. The reactions of these silanes were slower than those of 2 or 3 but were significantly faster than those of 1, 4, EtMezSiH, and PhMezSiH. The relative rate of 5 to PhMezSiH was 20:l. A significant induction period was observed in the reactions of 6 and 7, but after the induction period, rapid hydrosilylation occurred (260 and 70 times faster than PhMezSiH for 6 and 7, respectively). The additional two experiments apparently showed that the rate of the reaction is sensitive to the structure of the bifunctional organosilanes; an isomer of 5, 1,4-(HMezSi)zCsH4,reacted with acetone as slowly as PhMezSiH, whereas 1,8-bis(dimethylsilyl)naphthalene did not react at all. Similar to 2 and 3, silanes 5 and 6, in which two of the Si-H groups are connected by two or three carbon units, showed higher reactivity than 1 and 4. Although there are four carbon units between them, the rate enhancement occurred in the reaction with 7 (unlike 4). This can be attributed to the presence of the benzene ring in 7,which restricts
the distance between the two Si-H moieties. Thus, an adequate distance between the two Si-H groups is important for the rate acceleration. The distance between the two silicon groups is too long in 1,4-(HMezSi)CsH4 but is too short in 1,8-bis(dimethylsilyl)naphthalene. Diphenylsilyl derivatives also were the subjects of a similar study. The hydrosilylation of acetone with diphenylsubstituted organosilanes, 8 and 9, was substantially slower than that of their methyl analogue 2. In reactions at 50 "C, the relative rates between 8 and EtMezSiE-I were estimated to be 8:l. Although 9 reacted with acetone slowly (the rate was approximately 0.2 times that of EtMezSiH), EtPhzSiH did not react with acetone at all under the same conditions. The order of the reaction rate, 8 > EtMezSiH and 9 >> EtPhzSiH, clearly demonstrates that the bifunctional structure of 8 and 9 enhances the reactivity of both the MezSiH and the PhzSiH moieties. In Table 1 are summarized the products and the yields of the enhanced hydrosilylation of bifunctional organosilanes. All of the reactions were carried out in benzene in the presence of RhCl(PPh& (1 mol %) a t room temperature under a nitrogen atmosphere. In all cases, only one Si-H group of the silane was converted to a Si-OiPr moiety. The results provided two interesting problems with respect to the products. One is that redistribution of the substituents on the silicon atoms was only observed for 5. The fact that no methyl group migration was observed in the reactions of 6 or 7 as was seen in those of 5 suggests that both of the dimethylsilyl groups must be bonded directly t o an aromatic ring for the redistribution of the methyl groups t o occur.2o Since methyl group migration is not a general reaction of the bifunctional organosilanes, we will not discuss this further.21 The second problem was raised in the reactions of the unsymmetrical organosilanes 6 and 8, which gave a 1:l mixture of isomers. The result obtained with 8 is especially curious, because the experiments with EtMezSiH or EtPhzSiH described above showed that the MezSiH group is much more reactive than the PhzSiH moiety. If the two Si-H groups in 8 were independently activated by the catalyst, this large difference in reactivity would result in selective conversion of the MezSiH bond in 8 to MezSiOiPr group with the PhzSiH moiety remaining intact. Thus, the lack of chemoselectivity implicates the activation pathways of both of MezSiH and PhzSiH moieties in 8 during the reaction. In summary, the above experiments revealed that bifunctional organosilanes with two Si-H groups connected by two or three carbon units generally showed accelerated hydrosilylation of acetone. It should be pointed out that all of the bifunctional organosilanes showing the rate enhancement also have a structure conducive to the concomitant activation of both Si-H (20) The methyl group migration also occurred in the hydrosilylation of acetone with 3,4-bis(dimethylsilyl)toluene.A mixture of four isomers, two regioisomers of rearranged and unrearranged compounds, was obtained. (21) Redistribution of alkyl groups of organosilanes was extensively observed in the transition metal-catalyzed reactions of hydrosilanes.22 It was ponted out that oxidative addition of low-valent metal species between Si-alkyl bonds might induce the redistribution.22A possible mechanism for the 1,4-methyl group migration in the hydrosilylation with 5 is shown in Scheme 8. (22) Curtis, M. D.; Epstein, P. S. Adu. Organomet. Chem. 1981,19, 220. For a recent publication for the alkyl group redistribution in the platinum-catalyzid process with 5, see ref7b. ~~
2872 Organometallics, Vol. 14, No. 6, 1995
Nagashima et al.
Table 1. Isolation of Monohydrosilylation Products of Bidentate Organosilanes with Acetone Entry
Substrate
Temp (.c)
1
2
r.t.
3
Me2HSiCH2CH2Si(OiPr)Me2(10)
75
2
3
r.t.
3
Me2HSiCH2CH2CH2Si(OiPr)Me2(11)
71
Time(h)
Products
Me, 3
5
r.t.
19
Yield(o/o)
Me
Me,
0 ' : f P r
(5
6
r.t.
74
Si*H Me' 'O'PrlSb
Me' Me 1 2 a
5
.Me
a s i * M e
:
4)
3
88 Me
13a(i:i)
Meia
Me 6
7
r.t.
73
48
Me
a
7
a
50
5
a
9
50
10
14
Ph2HSiCH2CH2Si(OiPr)Me2(15a) Ph2('PrO)SiCH2CH2SiHMe2(15b) (1 : 1) Ph2HSiCH2CH2Si(OiPr)Ph2 (16)
All reactions were carried out with 1 mol % of RhCl(PPh& in benzene.
64
2Ob
It was difficult to isolate 16. Conversion determined by
NMR is reported.
groups as in the double oxidative addition to form disilametallacyclic intermediates. As noted above, the lack of chemoselectivity in the hydrosilylation with unsymmetrical bifunctional organosilanes, 6 and 8, supports the concomitant activation of both of the Si-H groups in 6 and 8. To examine the interaction of bifunctional organosilanes with the catalyst, we carried out stoichiometric reaction of RhCl(PPh& with some bifunctional organosilanes as described below. 11. Double Oxidative Addition of 2, 8, and 9 to RhCl(PPh3)s: An NMR Study. The RhCl(PPh&catalyzed hydrosilylation of unsaturated molecules usually is explained by variants of the Chalk-Harrod cycle, in which the oxidative addition of R3SiH to RhCl(PPh& to form (R3Si)RhCl(H)(PPh3)2is the probable initial step of the catalytic c y ~ l e . ~Studies .~ of the reaction of monohydrosilanes R3-,ClnSiH ( n = 0-3) with RhC1(PPhd3 were actively carried out in the 1960s by WilkinsonZ3and H a ~ z e l d i n e and , ~ ~ several oxidative adducts, generally formulated as (R3-,ClnSi)RhHCl(PPhdz, were isolated and characterized. The structure of a stable oxidative adduct (C13Si)RhHCl(PPh& was determined by X-ray analysis to be that of a trigonal bipyramid with trans-phosphines at the apices and H, (23) de Charentenay, F.; Osborn, J. A,; Wilkinson, G. J. Chem. SOC. A 1968. 787. (24) Haszeldine, R. N.; Parish, R. V.; Parry, D. J. J. Chem. SOC.(A) 1969,683. Hazeldine, R. N.; Parish, R. V.; Taylor, R. J. J . Chem. SOC. A 1974, 2311.
C1, and Sic13 in the trigonal plane.25 The NMR spectra of the oxidative adducts suggested that the structure in solution would be analogous to that formed in the crystal structure; the five-coordinate complexes ( R s - ~ C ~ , S ~ ) R ~ H C generally ~ ( P P ~ ~ showed )~ a Rh-H peak split into a doublet of triplets due to the coupling with lo3Rh(&-a = 21-27 Hz) and two equivalent PPh3 ligands (JH-Rh-P = 13-15 Hz).23,24We confirmed that the reaction mixture of RhCl(PPh& and HSiMezPh in CDCl3 gave a characteristic Rh-H peak split into a doublet Of triplets at -15.33 ppm (&-Rh = 22.7 Hz and J H - R h - P = 15.4 Hz) in the 'H NMR spectrum, whereas a doublet was observed at 35.7 ppm ( J R h - P = 124.5 Hz) in 31P(1H}NMR.26 Similar NMR spectra were obtained in toluene-&; a Rh-H peak appeared at -14.60 ppm as a doublet of triplets in the lH NMR spectrum, and a Rh-P signal at 36.0 ppm as a doublet. These spectral features are consistent with HRhCl(SiMezPh)(PPh& (17)having the structure shown in Figure 3. The reaction was reversible, and the equilibrium favored the reactant. Appropriate spectra, in which only small peaks derived from the starting materials appeared, were obtained in the reaction of 2 molar equiv of PhMezSiH with 1 equiv of RhCl(PPh&. Two ill-resolved Rh-H signals also appeared at -10.2 ppm as a broad doublet and at -17.5 ppm as a broad singlet in the lH (25) Muir, K. W.; Ibers, J. A. Inorg. Chem. 1970, 9, 440. (26)31PNMR of several rhodium-phosphine complexes: Brown, T. H.; Green, P. J. J . Am. Chem. SOC.1970, 92, 2359.
Organometallics, Vol. 14, No. 6,1995 2873
RhCl(PPh&-Catalyzed Hydrosilylation
Table 2. NMR Data for the Oxidative Adductsa-d BRh-H
17 19a 2Oa 21b 226
6Rh-P
-15.33" (dt, J = 15.4, 22.7 Hz), -14.60b(dt, J = 15.4, 22.0 Hz) -15.54 (dt, J = 15.4, 23.5 Hz) -15.18 (dt, J = 15.4,22.0 Hz) -10.8 (brd, J = 125 Hz, l H ) , -7.3 (brs, l H ) , -6.7 (brs, 1H) -10.9 (brd, J = 130 Hz, l H ) , -6.3 (brs, 2H)
3E1.7~(d, J = 124.5 Hz), 36.0b (d, J = 124.5 Hz) 37.4 (d, J = 128.5 Hz) 31.6 (d, J = 128.5 Hz) 27.7 (dd, J = 16.1, 84.4, lP), 36.6 (dd, J = 16.1, 104.4 Hz, 1P) 32.1 (dd, J = 16.1,86.3Hz, 1P)
a I n CDC13. b I n toluene-& a t -60.0 "C. 'H NMR spectral data for other parts are a s follows. 17: (CDC13) 6 0.01 (s, Rh-SiMe), 6.8-7.7 (m, Ph); (toluene-&) b 0.4 (rRh-SiMe), 6.7-7.9 (m, Ph). 19: 6 -0.25 (s, RhSiMe), 0.4-0.55 and 1.0-1.15 (m, SiCHz), 4.44 (br-s, Ph2SiH), 6.8-7.8 (m, Ph). 20: 6 1.15 (s, SiCH2), 4.32 (brs, PhzSiH), 6.8-8.7 (m, Ph). 21: 6 -0.1 to 1.3 (m, MeSiCHZ), 6.5-8.1 (m, Ph). 22: b 1.1-1.5 (m, SiCH2), 6.5-8.0 (m, Ph). Spectra in CDC13 were measured in the presence of cyclohexane a s the internal standard. RhCl(PPh&: bRh-p 29.6 (dd, J = 38, 145 Hz, 2P), 46.8 (dt, J = 38, 190 Hz, 1P) in CDC13; 28.7 (dd, J = 38, 145 Hz, 2P), 46.0 (dt, J = 38, 190 1P) in toluene-&. RhHzCl(PPh3)z (18): 6Rh-H -17.5 (brs, l H ) , -10.2 (brd, J = 125 Hz, 1H) i n CDC13 and -16.6 (brs, l H ) , -9.3 (brd, J = 125 Hz, 1H) in toluene-&; bRh-p 35.2 (d, J = 116.4) in CDC13 and 32.8 (d, J = 116.4 Hz) in toluene-&.
H
17
Me,ye
20
19
PPh,