Introduction

The previous chapter described the studies on the structure and reactivity of alkyl and hydride yttrium complexes containing the [DADMB]^2- ligand (DADMB = 2,2'-bis(tert-butyldimethylsilylamido)-6,6'-dimethylbiphenyl).¹ Early transition metal complexes, including chelating diamide complexes of this type, have been extensively investigated as catalysts for olefin polymerizations,^2-17 and we have also been interested in the potential of d⁰ systems to function as silane dehydropolymerization catalysts.^18-22 While the yttrium-DADMB complexes were not active as olefin or silane polymerization catalysts, we have found that they can be used as olefin hydrosilylation catalysts.²³ The hydrosilylation of olefins by early transition metal catalysts is a well known process, and recently many studies have been devoted to investigating the reactivity and selectivity of such catalysts towards various unsaturated organic substrates.^24-32 The mechanism and energetics of the hydrosilylation catalytic cycle has also been investigated extensively.²⁴ As is usually the case in early transition metal chemistry, the ancillary ligands employed in these studies have been based on cyclopentadienide and its derivatives. The potential of non-cyclopentadienyl ligand sets in hydrosilylation catalysis by d⁰ metals has yet to be explored. This chapter describes the hydrosilylation chemistry of [DADMB]Y complexes, and the initial results on the use of a resolved chiral catalyst, (S)-[DADMB]YMe(THF)₂, in the enantioselective hydrosilylation of olefins.

Results and Discussion

The yttrium complexes 1 and 2 were prepared according to previously described methods.¹ While these complexes do not appear to be promising as olefin polymerization catalysts, the yttrium hydride 2 was found to react cleanly with olefins to give single-insertion alkyl products.^1,23 Further reactions with olefin are very slow, and measurable amounts of polyolefins were not observed. Complexes 1 and 2 were also found to be inactive as dehydropolymerization catalysts, and whereas 2 is inert toward silanes such as PhSiH₃, the yttrium alkyls 1 and [DADMB]Y[CH(SiMe₃)₂](THF)(OEt₂) react with hydrosilanes in a metathesis-type reaction to form 2 and an alkylsilane. The latter observation of Si-C bond formation prompted us to investigate the possibility of catalytic hydrosilylation, since it was envisioned that the yttrium hydride 2 could react with an olefin via insertion, and then the resulting yttrium alkyl group might be transferred to a silane with regeneration of the yttrium hydride catalyst (Scheme 1). In fact, similar hydrosilylations are known to occur with other group 3 and lanthanide hydrides (e.g., [Cp*YH]₂) or their precursors.^24-27,29

Initial experiments were designed to test the activity of 2 as a hydrosilylation catalyst. Addition of a large excess of PhSiH₃ to a THF-d₈ solution of 2 did not result in reaction, but when the mixture was pressurized with ethylene (5 psi, room temperature), the clean formation of PhEtSiH₂ was immediately observed, and within 12 h all the PhSiH₃ had been consumed. 1-Hexene was also found to react with PhSiH₃ in the presence of 2, to give PhSiH₂(CH₂)₅CH₃ as the major product. Under similar conditions cyclohexene did not undergo hydrosilylation, consistent with its lack of reactivity towards the hydride 2.

The complete insolubility of the hydride dimer in non-coordinating solvents required the use of THF for the hydrosilylation reactions. However, the yttrium methyl complex 1 can also be used as an active hydrosilylation catalyst, presumably because it first reacts with the silane to give a small concentration of reactive, monomeric hydride. Thus, hydrosilylations may be carried out in a nonpolar solvent such as benzene using 1 as the catalyst precursor. The active hydride species under these catalytic conditions has not been observed, as addition of a hydrosilane to 1 led to rapid disappearance of the resonances for 1 (by ¹H NMR spectroscopy) with formation of the methyl silane, but no yttrium hydride resonances could be detected (apparently due to the very low concentration of the active species).

To probe the selectivity of this catalytic system, a range of olefins was tested, using PhSiH₃ and PhMeSiH₂ as representative primary and secondary silanes. The results are presented in Table 1. The steric bulk of the ligand apparently limits the reactivity of the catalyst such that only terminal olefins react at a measurable rate, and norbornene is the only disubstituted olefin observed to react. With both PhSiH₃ and PhMeSiH₂, no reaction was observed with cyclohexene, 1-phenyl-1-methylethene (alpha-methylstyrene), and trans-1,2-diphenylethene, and PhMeSiH₂ did not react with norbornene. As has been noted for other lanthanide and yttrium hydrosilylation catalysts,^24-26,29 both 1,2- and 2,1-additions of the silane to the double bond are observed, the ratio of the two isomers being controlled by steric and electronic effects. Aliphatic olefins give predominantly the terminal addition product, while styrene preferably gives benzylsilane derivatives. The latter effect has been observed with related catalysts, and rationalized in terms of electronic interactions between the metal center and the aromatic ring of styrene, which directs the insertion reaction toward the alpha-phenylalkyl intermediate.²⁴

Table 1. Results from the hydrosilylation of olefins catalyzed by [DADMB]YMe(THF)₂ (1).

Olefin	Silane	Turnover rate, h-1	Products	Product ratio
	PhSiH3	~100		92% 8%
	PhMeSiH2	4.3		>99%
	PhSiH3	0.66		76% 24%
	PhMeSiH2	0.12		64% 36%
	PhSiH3	~30		>99% (90% ee)

Since the yttrium species involved in the catalysis are chiral, we examined the enantioselectivity of their catalytic action. A resolved version (S-1) of the methyl complex 1 was prepared using a sample of enantiopure (S)-2,2'-diamino-6,6'-dimethylbiphenyl,³³ following the previously reported synthetic procedure¹ (see Chapter 2). An unexpected confirmation of the preserved enantiopurity of the prepared complex S-1 came from its ¹H NMR spectrum. While most chemical shifts for S-1 were identical to those of racemic 1,¹ two separate multiplets due to diastereotopic hydrogen atoms were observed for the alpha-H protons of the coordinated THF in S-1, as compared to a single multiplet in 1. As shown in Scheme 2, coordination of THF to a chiral metal center places the diastereotopic CH₂ protons in two magnetically inequivalent environments. Since all THF coordination sites in an enantiopure system are of the same chirality, exchange of THF between different molecules would preserve the difference in chemical shift between the diastereotopic THF protons. In a racemic mixture, however, an exchange of THF between coordination sites of opposite chirality would lead to an averaged environment for the same protons. A variable temperature NMR study of 1 further confirmed that the single multiplet observed at room temperature is due to a fast intermolecular exchange of THF. This exchange can apparently be slowed down at lower temperature, so that two separate multiplets are observed (Figure 1).

The transition temperature in toluene-d₈ was determined to be 280 K, from which an activation energy deltaG_act = 54 kJ/mol (13 kcal/mol) for THF dissociation can be calculated (rate constant k = 1.8 x 10³ s^-1). The rate of THF dissociation is thus faster than any of the kinetically important steps in the catalytic system (vide infra).

Figure 1. Temperature dependence of the chemical shift of the diastereotopic THF alpha-protons in 1.

Evaluation of the catalyst enantioselectivity was hindered by the fact that most of the prochiral olefins tested did not react, or gave a mixture of regioisomers (vide supra). The substrate chosen for enantioselectivity studies was norbornene, since it gives exclusively the exo-diastereoisomer on reaction with PhSiH₃. The enantiomeric excess of the PhSiH₂-norbornane product was found to be 90.4% (GC analysis with a chiral column), in favor of the 1S-enantiomer, as determined by oxidation of the silane to exo-norborneol following standard literature procedures.^24,26 Highly enantioselective hydrosilylation of norbornene and other olefins using late transition metal (Pd, Pt, Rh) catalysts has been previously reported, with ee's often exceeding 90%.^34-38 The only other study of enantioselective olefin hydrosilylation by early transition metal or lanthanide catalysts, however, utilized the chiral ansa-cyclopentadienyl complexes (R)-Me₂Si(C₅Me₄)[(-)-menthyl-C₅H₄]SmCH-(SiMe₃)₂ and (S)-Me₂Si(C₅Me₅)[(-)-menthyl-C₅H₄]SmCH(SiMe₃)₂, and resulted in observation of enantiomeric excesses of 68% and 65% for the hydrosilylation of PhEtC=CH₂ with PhSiH₃.²⁴

The mechanism of hydrosilylation as catalyzed by early transition metal complexes has been studied for several cyclopentadienyl based systems.^24,25 The active metal species is thought to be a monomeric d⁰ metal hydride. The catalytic cycle is proposed to occur via fast, irreversible insertion of olefin into the metal-hydrogen bond to give an alkyl species, which then reacts with the silane in a slow, rate-determining step (Scheme 1). The overall reaction rate has been found to be zeroth order in olefin, and first order in silane and catalyst precursor.²⁴

To probe the mechanism of hydrosilylation catalysis by [DADMB]Y complexes, we studied the kinetics of this process. The substrates chosen were 1-hexene and PhMeSiH₂, as that reaction was found to proceed at a rate convenient to follow by NMR spectroscopy.

The mechanism of the catalyst initiation step (eq 1) in benzene solution was studied by monitoring the disappearance of 1 at different concentrations of PhMeSiH₂, the silane being kept in large excess. This reaction should also represent a good model for the product-forming step in the proposed catalytic mechanism (Scheme 1). A linear decrease of ln[1] with time was observed, which is consistent with a rate law involving first-order dependence on 1 (Figure 2). A plot of the observed rate constant vs. [PhMeSiH₂] was also found to be linear (Figure 3), consistent with the expected first order dependence on silane concentration. The overall rate constant was found to be k_H = 3.8(2) x 10^-4 L/(mol.s) at 298 K. To determine the isotope effect of this sigma-bond metathesis process, several kinetic measurements were performed using deuterated silane (PhMeSiD₂), giving a rate constant kD = 3.4(3) x 10^-4 L/(mol.s). An approximate value of k_H / k_D = 1.1(1) can be estimated, but the significant scatter in the experimental data does not allow meaningful conclusions about the isotope effect of this reaction.

The actual hydrosilylation process was studied in benzene by monitoring the consumption of PhMeSiH₂ at different concentrations of 1-hexene (in most cases kept in at least five-fold excess relative to the silane, so that the decrease in olefin concentration during the reaction is insignificant) and the catalyst precursor 1 (typically about 5% relative to PhMeSiH₂). The linear dependence of ln[PhMeSiH₂] vs. time over the whole range of initial concentrations of 1 and 1-hexene (Figure 4) suggests a first order rate law with respect to the silane, as expected from the mechanism of Scheme 1. The dependence of the overall rate on the olefin concentration was probed by conducting several kinetic runs at different 1-hexene concentrations (ranging from 0.33 to 4.01 mol/L). The observed rate constant was found to be practically independent of the olefin concentration (Figure 5), again consistent with the proposed mechanism which implies zeroth order with respect to olefin.

Figure 2. Pseudo-first order plots for disappearance of 1 at different PhMeSiH₂ concentrations (298 K, benzene-d₆).

Figure 3. Pseudo-first order rate constant for disappearance of 1 as a function of phenylmethylsilane concentration (298 K, benzene-d₆).

Figure 4. Pseudo-first order plots for consumption of PhMeSiH₂ at different concentrations of 1 and 1-hexene (298 K, benzene-d₆).

Figure 5. Observed pseudo-first order rate constant for PhMeSiH₂ consumption as a function of the 1-hexene concentration, at constant catalyst precursor concentration (298 K, benzene-d₆).

Figure 6. Observed pseudo-first order rate constant for PhMeSiH₂ consumption as a function of the catalyst precursor concentration, at constant initial 1-hexene concentration (298 K, benzene-d₆).

The dependence of the observed rate constant on the initial concentration of 1, however, was found to deviate significantly from the expected linear correlation for a first-order rate law with respect to yttrium catalyst (Figure 6). Precipitation of insoluble yttrium hydride dimer at later stages of the reaction, which can irreversibly remove some of the active catalyst from the solution, is one possible cause of the saturation behavior observed at higher catalyst concentrations. Alternatively, an order of 1/2 would lead to a qualitatively similar behavior, and can be explained if the concentration of the active yttrium species in the rate-determining step is controlled by an equilibrium leading to dimer formation. An example of such half-order dependence on catalyst concentration has been reported for the hydrogenation of olefins catalyzed by Cp*-organolanthanide complexes,³⁹ a mechanistically very similar process. In the latter hydrogenation, the half-order dependence was rationalized by invoking a fast equilibrium between the reactive metal alkyl intermediate and an inactive dimeric (alkyl bridged) species. Interestingly, the corresponding catalytic hydrosilylation process using the same catalysts, is reported to exhibit regular first order dependence on catalyst concentration.²⁴ Although in our system no evidence is available to suggest that 1 itself could form a dimer in solution, the dimeric nature of the hydride 2, as well as the existence of many examples of dimeric alkyl and mixed hydrido-alkyl yttrium species in the literature,^39-41 suggests that such equilibria are not to be ignored. The complexity of the system, however, and uncertainty in the structure of the active species, prevent us from formulating a more detailed mechanistic picture at this time.

To avoid the potential effects of the insolubility of the yttrium hydride dimers in benzene on the kinetics of the hydrosilylation process, several runs were also performed using THF-d₈ as a solvent. The consumption of PhMeSiH₂ was followed at different yttrium catalyst precursor concentrations, while keeping the 1-hexene concentration constant. A plot of ln[PhMeSiH₂] vs. time, however, showed that an observable decrease in the reaction rate occurs after some period of time (Figure 7), which is likely due to

Figure 7. Pseudo-first order plots for consumption of PhMeSiH₂ at different concentrations of 1 (298 K, THF-d₈).

Figure 8. Observed rate constant for consumption of PhMeSiH₂ (initial rate) as a function of the initial concentration of 1 (298 K, THF-d₈).

catalyst decomposition in the THF solvent. Unlike the approximately half-order rate law observed in benzene, however, the dependence of the observed rate constant on the catalyst precursor concentration in the initial period was found to be linear (Figure 8), suggesting a first order rate law (k = 1.1(1) x 10^-3 L/(mol.s) at 298 K) with respect to catalyst precursor when the reaction is conducted in THF. The observed rate constants in THF are lower than those in benzene, which is consistent with the idea that the catalytically active species is a coordinatively unsaturated complex, and THF coordination can inhibit its reactivity.

Conclusions

We have explored the hydrosilylation activity of some yttrium complexes with bis(silylamido)biphenyl ligands and have performed kinetic and mechanistic investigations of this catalytic process. Although hydrosilylation catalysis by d⁰ transition metal complexes is well known, most of the research on such catalysts has traditionally been done using cyclopentadienyl ligands. The present system is the first to demonstrate the use of non-Cp ligands in the catalytic hydrosilylation of olefins by a d⁰ metal. Mechanistic investigations indicate that this hydrosilylation occurs by the mechanism generally accepted to operate for other d⁰ systems, involving fast olefin insertion into the reactive metal hydride bond, followed by a slow metathesis reaction with a silane molecule. As with the Cp-based systems studied earlier, the diamido catalyst 1 exhibits a high regioselective preference toward terminal addition in case of aliphatic olefins. However, a lower preference for 2,1-addition in case of aromatic olefins was observed, presumably due to the different steric requirements of the bis(silylamido)biphenyl ligand as compared to the bis-Cp systems. The [DADMB]Y catalyst is also reactive enough to allow a secondary silane such as PhMeSiH₂ to be employed in the hydrosilylation, in addition to the more typically employed PhSiH₃. Significantly, the enantioselectivity observed with the enantioresolved [DADMB]Y catalyst in the hydrosilylation of norbornene (90% ee) is impressively high for an unoptimized system, and compares favorably with some of the best late transition metal-based catalysts.

Experimental Section

General. All reactions with air-sensitive compounds were performed under dry nitrogen, using standard Schlenk and glove box techniques. Reagents were obtained from commercial suppliers and used without further purification, unless otherwise noted. Olefin-free pentane, benzene, and toluene were prepared by pretreating with concentrated H₂SO₄, 0.5 N KMnO₄ in 3 M H₂SO₄, NaHCO₃ and finally anhydrous MgSO₄. Solvents (pentane, diethyl ether, benzene, toluene, tetrahydrofuran) were distilled under nitrogen from sodium benzophenone ketyl. Benzene-d₆ and tetrahydrofuran-d₈ were distilled from Na/K alloy. Commercial silanes, 1-hexene and PhCH=CH₂ were dried over molecular sieves and distilled before use. Deuterated phenylmethylsilane was obtained by reduction of PhMeSiCl₂ with LiAlD₄ (98% D). The syntheses of 1 and 2 have been reported in Chapter 2. Enantiopure (S)-2,2'-diamino-6,6'-dimethylbiphenyl³³ (99.9+% ee as determined by polarimetry and HPLC) was provided by prof. A. Togni (ETH-Zurich). NMR spectra were recorded at 300 or 500 MHz (¹H) with Bruker AMX-300 and DRX-500 spectrometers, or at 100 MHz (¹³C{¹H}) with an AMX-400 spectrometer, at ambient temperature and in benzene-d₆, unless otherwise noted. Signal multiplicities are reported as follows: s - singlet, d - doublet, t - triplet, q - quartet, qn - quintet, m - multiplet. GC/MS data was obtained with a HP 6890 GC/MS system, equipped with a JW DB-XLB column.

(S)-[DADMB]YMe(THF)₂ (S-1). To a cold (0 ^oC) solution of (S)-2,2'-diamino-6,6'-dimethylbiphenyl (0.68 g, 3.20 mmol) in THF (50 mL) was added dropwise 4.0 mL (6.4 mmol) of 1.6 M ⁿBuLi. A white precipitate formed initially, but dissolved completely after all of the ⁿBuLi had been added. The solution was allowed to warm to room temperature, and was then stirred for 3 h. A solution of ^tBuMe₂SiCl (1.01 g, 6.72 mmol) in 10 mL of THF was then added dropwise. The mixture was heated at reflux for 1 h, which resulted in the formation of a white precipitate. After cooling to room temperature, a second portion of ⁿBuLi (4.0 mL, 6.4 mmol) was added and the mixture was stirred overnight at room temperature. The THF was removed under vacuum to give an oily white solid. Extraction with hexane (2 x 50 mL) gave a light yellow solution, which was concentrated in vacuo until crystals appeared, and then cooled to -78 ^oC to obtain 1.53 g (83%) yield) of (S)-Li₂[DADMB].(THF)₂ as colorless crystals. ¹H NMR: d 7.06 (d, 2 H, J = 7.7 Hz), 6.96 (t, 2 H, J = 7.7 Hz), 6.48 (d, 2 H, J = 7.1 Hz biphenyl H's), 3.26 (m, 4 H, THF), 3.06 (m, 4 H, THF), 1.99 (s, 6 H, Me), 1.24 (m, 8 H, THF), 1.17 (s, 18 H, ^tBuMe₂Si), 0.50 (s, 6 H, ^tBuMe₂Si), 0.10 (s, 6 H, ^tBuMe₂Si). ¹³C{¹H} NMR: d 157.5, 140.1, 133.2, 128.0, 121.8, 116.6, 68.6 (THF), 28.7 (CMe₃), 25.4 (THF), 21.8, 21.7 (Me; CMe₃), 1.0, -0.5 (^tBuMe₂Si). A portion of this product (1.28 g, 2.15 mmol) mixed with 0.91 g (2.22 mmol) of YCl₃(THF)₃ in 50 mL of THF and the solution was heated at reflux for 3 h. The solvent was removed in vacuo and the resulting white powder was extracted with hexane / THF mixture. The filtrate was concentrated to about 15 mL and more hexane (20 mL) was added to initiate crystallization of the product as a white precipitate. After cooling to -78 ^oC, the solution was filtered and the product was then dried in vacuo to obtain 1.01 g (66% yield) of (S)-[DADMB]YCl(THF)₂ as a white crystalline powder. ¹H NMR: d 7.07 (d, 2 H), 6.95 (t, 2 H), 6.56 d, 2 H, aromatic H), 3.73 (m, 4 H, THF), 3.42 (m, 4 H, THF), 1.88 (s, 6 H, Me), 1.24 (m, 8 H, THF), 1.11 (s, 18 H, ^tBuMe₂Si), 0.52, 0.50 (s, 6 H each, ^tBuMe₂Si). ¹³C{¹H} NMR: d 152.4, 142.4, 130.5, 130.0, 123.5, 120.1 (aromatic C), 71.6 (THF), 28.3 (Me₃C), 25.5 (THF), 22.3 (MeAr), 21.5 (Me₃C), 1.8, -1.5 (Me₂Si). A portion of this product (0.82 g, 1.15 mmol) was dissolved in 50 mL of THF, the solution was cooled in ice bath, and MeLi (0.72 mL, 1.15 mmol) was added. The resulting pale yellow solution was allowed to warm to room temperature and was stirred overnight, the solvents were removed in vacuo, and the oily residue was extracted with 2 x 30 mL of hexane. The (S)-[DADMB]YMe(THF)₂ product was isolated by crystallization at -78 ^oC (0.34 g, 43% yield) as a white crystalline powder, mp 128 - 132 ^oC (cf. 145 - 150 ^oC for racemic 1). ¹H NMR: d 7.07 (m, 2 H), 6.96 (m, 2 H), 6.57 (m, 2 H, aromatic H), 3.66 (m, 4 H, THF), 3.29 (m, 4 H, THF), 1.88 (s, 6 H, Me), 1.24 (m, 8 H, THF), 1.14 (s, 18 H, ^tBuMe₂Si), 0.53, 0.41 (s, 6 H each, ^tBuMe₂Si), -0.42 (d, 3 H, YMe). ¹³C{¹H} NMR: d 153.2, 142.1, 131.1, 129.4, 123.4, 119.5, 70.9 (THF), 28.4 (Me₃C), 25.5 (THF), 22.4 (MeAr), 21.7 (YMe), 21.7 (Me₃C), 2.0, -1.9 (Me₂Si).

Hydrosilylation of ethylene with PhSiH₃. A sample of 1 (ca. 5 mg) was dissolved in benzene-d₆ in a J. Young NMR tube. The tube was evacuated briefly and refilled with C₂H₄ (5 - 10 psi) several times. An excess of PhSiH₃ (ca. 50 uL) was added. After 20 min at room temperature, the silane was completely consumed and formation of PhH₂SiCH₂CH₃ was observed. The NMR spectrum of PhH₂SiCH₂CH₃ was consistent with the literature data.⁴² A control sample not containing 1 revealed no reaction between PhSiH₃ and C₂H₄ after 8 h.

Hydrosilylation of 1-hexene with PhSiH₃. A sample of 1 (9.6 mg, 0.014 mmol) was dissolved in 0.8 mL of benzene-d₆. To this solution were added 1-hexene (35 uL, 0.28 mmol) and PhSiH₃ (35 uL, 0.28 mmol). Monitoring the reaction by ¹H NMR spectroscopy showed that 80% conversion to products had occurred within the first 10 min after mixing, corresponding to turnover rate of about 100 h^-1. The spectrum of the hydrosilylation product was consistent with that reported for the PhH₂Si(CH₂)₅CH₃ isomer.²⁴ Analysis of the products was also performed by GC/MS, after quenching and diluting the reaction mixture with pentane, which revealed the presence of a small amount (ca. 8%) of the PhSiH₂(CH)(CH₃)(CH₂)₃CH₃ isomer.

Hydrosilylation of 1-hexene with PhMeSiH₂. A sample of 1 (10.0 mg, 0.015 mmol) was dissolved in benzene-d₆ (0.8 mL). To this solution were added 1-hexene (36 uL, 0.29 mmol) and PhMeSiH₂ (40 uL, 0.29 mmol). An initial turnover rate of 4.3 h^-1 was determined by following the disappearance of the starting materials by ¹H NMR spectroscopy for the first 7000 s. The ¹H NMR spectrum and the GS/MS data indicated exclusive formation of a single hydrosilylation product, identified as PhMeHSi-(CH₂)₅CH₃. ¹H NMR: d 7.50 (m, 2 H, Ph), 7.20 (m, 3 H, Ph), 4.58 (m, 1 H, SiH), 1.10 - 1.30 (m, 8 H, CH₂), 0.87 (t, 3 H, CH₃), 0.70 - 0.80 (m, 2 H, CH₂), 0.26 (d, ³J_HH = 3.8 Hz, SiMe).

Hydrosilylation of PhCH=CH₂ with PhSiH₃. A sample of 1 (9.6 mg, 0.014 mmol) was dissolved in 0.8 mL of benzene-d₆. To this solution were added PhCH=CH₂ (34 uL, 0.30 mmol) and PhSiH₃ (35 uL, 0.28 mmol). An initial turnover rate of 0.66 h^-1 was determined by following the disappearance of the starting materials by ¹H NMR spectroscopy for the first 12,000 s. After 3 days, ca. 80% conversion was observed. Complete consumption of the starting materials was observed after 1 week. The product ratio PhCH(CH₃)SiH₂Ph to PhCH₂CH₂SiH₂Ph as determined by ¹H NMR integration was 3.1:1. The NMR spectrum was consistent with the literature data.²⁴ The identity of the products was also confirmed by GC/MS (m/z = 212).

Hydrosilylation of PhCH=CH₂ with PhMeSiH₂. A sample of 1 (9.3 mg, 0.013 mmol) was dissolved in 0.8 mL of benzene-d₆. To this solution were added PhCH=CH₂ (31 uL, 0.27 mmol) and PhMeSiH₂ (37 uL, 0.27 mmol). An initial turnover rate of ca. 0.12 h^-1 was determined by following the disappearance of the starting materials by ¹H NMR spectroscopy over the first 80,000 s. After 2 days, 50% conversion was observed. Starting materials were still present even after 10 days. The product ratio PhCH(CH₃)SiHMePh to PhCH₂CH₂SiHMePh as determined by ¹H NMR integration was 1.8:1. The identity of the products was also confirmed by GC/MS (m/z = 226), which also showed the presence of both diastereoisomers of PhCH(CH₃)SiHMePh, in 1.2 : 1 ratio. ¹H NMR (some peaks overlap): d 7.45 (m, 2 H, Ph), 7.32 (m, 2 H, Ph), 6.95-7.20 (m, 6 H, Ph), 4.50-4.55 (m, 2 H, SiH), 2.60 (t, 2 H, PhCH₂CH₂SiHMePh), 2.28-2.35 (m, 1 H, PhCH(CH₃)SiHMePh), 1.30 (m, 3 H, PhCH(CH₃)SiHMePh), 1.05-1.10 (m, 2 H, PhCH₂CH₂SiHMePh), 0.19 (d, SiMe(H), overlaps with PhMeSiH₂), 0.12 (d, SiHMe, 3 H).

Preparative scale hydrosilylation of norbornene with PhSiH₃. To a solution of 1 (0.248 g, 0.36 mmol) and norbornene (1.13 g, 12.0 mmol) in 30 mL of C₆H₆ was added 1.50 mL of PhSiH₃ (12.0 mmol) and the mixture was stirred at room temperature for 48 h. The cloudy solution was diluted with 50 mL of Et₂O and the resulting mixture was poured into 75 mL of saturated aqueous NH₄Cl. The organic layer was separated, the aqueous layer was extracted with Et₂O (3 x 50 mL), and the combined Et₂O extracts were dried over MgSO₄ and then concentrated under vacuum. The resulting oil was redissolved in 30 mL of hexane and filtered through a short silica column. Removal of the volatiles with a rotovap produced 2.80 g (quantitative yield) of exo-phenylsilylnorbornane as a colorless oil (pure by ¹H NMR spectroscopy, containing some residual solvents). Only the exo-isomer was produced, as identified by NOESY ¹H NMR spectroscopy and GC/MS (m/z = 202). The NMR spectrum was in agreement with the published literature data.²⁵ Using the same procedure, enantioselective hydrosilylation of norbornene (0.77 g, 8.2 mmol) with PhSiH₃ (1.0 mL, 8.2 mmol) in presence of (S)-[DADMB]YMe(THF)₂ (0.17 g, 0.25 mmol) produced 0.94 g of exo-phenylsilylnorbornane.

Oxidation of exo-phenylsilylnorbornane to exo-norborneol. The racemic product from the hydrosilylation of norbornene (2.80 g, ca. 12.0 mmol) was dissolved in CHCl₃ (140 mL), the solution was cooled in ice bath, and 4.1 mL of HBF₄.Et₂O was added. The reaction mixture was stirred for 3 h, the volatiles were removed in vacuo and the residue redissolved in 1:1 CH₃OH / THF mixture (200 mL). To this solution were added KF (3.75 g), KHCO₃ (6.7 g) and 33 mL of 30% H₂O₂. The mixture was stirred for 1 h at room temperature and then refluxed overnight, which resulted in the formation of copius white precipitate. After reducing the volume of the solution in vacuo, it was poured into 400 mL concentrated aqueous NaCl and the mixture was extracted with Et₂O (3 x 100 mL). Removal of volatiles produced pale yellow oil, which was purified by flash chromatography on silica using 3:1 pentane / Et₂O mixture to give 0.75 g of racemic exo-norborneol as a white powder. Under the same reaction conditions, oxidation of the enantioenriched exo-phenylsilylnorbornane product (0.94 g) afforded 0.20 g of exo-norborneol. The enantiomeric excess of the produced exo-norborneol was measured to be 90.4 %, as determined by GC analysis with a chiral column (Supelco beta-DEX 120, at 80^oC, He eluent at 1.4 mL/min flow)³⁴ at the laboratory of prof. A. Togni at ETH-Zurich.

Kinetic measurements. Reactions were monitored by ¹H NMR spectroscopy, with a Bruker AMX300 spectrometer, using 5 mm Wilmad NMR tubes, equipped with J. Young Teflon screw caps. Liquid reagents were measured using a 100 uL Hamilton gas-tight syringe. The total volume of the reaction solution was determined by measuring its height in the precalibrated NMR tube. The samples were frozen in liquid N₂ immediately after preparation, and defrosted just before being placed in the preshimmed probe, which was preheated at 25 ^oC. Single scan spectra were acquired automatically at preset time intervals. The peaks were integrated relative to ferrocene as an internal standard. Rate constants were obtained by non-weighted linear least-squares fits of the integrated first-order rate law in logarithmic form, lnC = lnC₀ - k_obst.

Kinetic study of the reaction of 1 with PhMeSiH₂. Samples of 1 (12.3-15.5 mg, 0.0179-0.0226 mmol) and Cp₂Fe (1.0-3.0 mg) were weighed into an NMR tube and dissolved in benzene-d₆ (0.9 mL). To the solution was added a known amount of PhMeSiH₂. The disappearance of the YMe signal as integrated with respect to the ferrocene standard was monitored. Five kinetic runs were performed, using different amounts of silane. The overall rate constant was determined by plotting the observed pseudo-first order rate constant for the consumption of 1 vs. the concentration of PhMeSiH₂.

Kinetic study of the hydrosilylation of 1-hexene with PhMeSiH₂ in benzene-d₆. Samples of 1 (9.6-32.7 mg, 0.014-0.048 mmol) and Cp₂Fe (0.5-3.0 mg) were weighed into an NMR tube and dissolved in benzene-d₆ (ca 0.6 mL). Alternatively, for accurate measuring of the amount of the catalyst precursor at low concentrations, a standard solution of 1 was prepared by dissolving 30.6 mg (0.0445 mmol) of 1 in 3.0 mL benzene-d₆, and aliquots of this solution (50-800 ul) were used to prepare the NMR sample. To the solution were added known volumes of 1-hexene (36.4-500 uL) and PhMeSiH₂ (20-44 uL). The consumption of PhMeSiH₂ was monitored by integrating the SiH₂ signal against ferrocene. The observed pseudo-first order rate constants for silane disappearance were determined at different olefin or yttrium methyl initial concentrations by plotting ln[PhMeSiH₂] vs. time.

Kinetic study of the hydrosilylation of 1-hexene with PhMeSiH₂ in THF-d₈. Samples of 1 (8.2-18.6 mg, 0.012-0.042 mmol) and Cp₂Fe (0.5-2.0 mg) were weighed into an NMR tube and dissolved in THF-d₈ (ca 0.5 mL). To this solution were added 1-hexene (150 uL, 1.20 mmol) and PhMeSiH₂ (40 uL, 0.29 mmol). The consumption of PhMeSiH₂ was monitored by integrating the CH₃ signal against ferrocene.

References:

(1) Gountchev, T. I.; Tilley, T. D. Organometallics 1999, 18, 2896.

(2) Aoyagi, K.; Gantzel, P. K.; Kalai, K.; Tilley, T. D. Organometallics 1996, 15, 923.

(3) Scollard, J. D.; McConville, D. H. J. Am. Chem. Soc. 1996, 118, 10008.

(4) Scollard, J. D.; McConville, D. H.; Rettig, S. J. Organometallics 1997, 16, 1810.

(5) Horton, A. D.; de With, J. Organometallics 1997, 16, 5424.

(6) Baumann, R.; Davis, W. M.; Schrock, R. R. J. Am. Chem. Soc. 1997, 119, 3830.

(7) Warren, T. H.; Schrock, R. R.; Davis, W. M. Organometallics 1998, 17, 308.

(8) Tsuie, B.; Swenson, D. C.; Jordan, R. F. Organometallics 1997, 16, 1392.

(9) Tinkler, S.; Deeth, R. J.; Duncalf, D. J.; McCamley, A. J. Chem. Soc. Chem. Commun. 1996, 2623.

(10) Gibson, V. C.; Kimberley, B. S.; White, A. J. P.; Williams, D. J.; Howard, P. J. Chem. Soc. Chem. Commun. 1998, 313.

(11) Male, N. A. H.; Thornton-Pett, M.; Bochmann, M. J. Chem. Soc., Dalton Trans. 1997, 2487.

(12) Horton, A. D.; de With, J. J. Chem. Soc. Chem. Commun. 1996, 1375.

(13) Cloke, F. G. N.; Geldbach, T. J.; Hitchcock, P. B.; Love, J. B. J. Organomet. Chem. 1996, 506, 343.

(14) Sinnema, P.-J.; Liekelema, K.; Staal, O. K. B.; Hessen, B.; Teuben, J. H. J. Mol. Catal. A: Chemical 1998, 128, 143.

(15) Scollard, J. D.; McConville, D. H.; Payne, N. C.; Vittal, J. J. Macromolecules 1996, 29, 5241.

(16) Scollard, J. D.; McConville, D. H.; Vittal, J. J. Organometallics 1995, 14, 5478.

(17) Guérin, F.; McConville, D. H.; Payne, N. C. Organometallics 1996, 15, 5085.

(18) Tilley, T. D. Acc. Chem. Res. 1993, 26, 22.

(19) Woo, H.-G.; Tilley, T. D. J. Am. Chem. Soc. 1989, 111, 8043.

(20) Woo, H.-G.; Walzer, J. F.; Tilley, T. D. J. Am. Chem. Soc. 1992, 114, 7047.

(21) Gauvin, F.; Harrod, J. F.; Woo, H. G. Adv. Organomet. Chem. 1998, 42, 363.

(22) Corey, J. ; Larson, G., Ed.; JAI Press, Inc.: Greenwich, Conn., 1991; Vol. 1, pp 327.

(23) Gountchev, T. I.; Tilley, T. D., submitted for publication.

(24) Fu, P.-F.; Brard, L.; Li, Y.; Marks, T. J. J. Am. Chem. Soc. 1995, 117, 7157.

(25) Molander, G. A.; Julius, M. J. Org. Chem. 1992, 57, 6347.

(26) Molander, G. A.; Dowdy, E. D.; Noll, B. C. Organometallics 1998, 17, 3754.

(27) Koo, K.; Fu, P.-F.; Marks, T. J. Macromolecules 1999, 32, 981.

(28) Marciniec, B.; Gulinski, J. J. Organomet. Chem. 1993, 446, 15.

(29) Schumann, H.; Keitsch, M. R.; Demtschuk, J.; Molander, G. A. J. Organomet. Chem. 1999, 582, 70.

(30) Molander, G. A.; Nichols, P. J. J. Am. Chem. Soc. 1995, 117, 4415.

(31) Kesti, M. R.; Waymouth, R. M. Organometallics 1992, 11, 1095.

(32) Carter, M. B.; Schiott, B.; Gutierrez, A.; Buchwald, S. L. J. Am. Chem. Soc. 1994, 116, 11667.

(33) Kanoh, S.; Goka, S.; Murose, N.; Kubo, H.; Kondo, M.; Sugino, T.; Motoi, M.; Suda, H. Polymer Journal 1987, 19, 1047.

(34) Pioda, G.; Togni, A. Tetrahedron: Asymmetry 1998, 9, 3903.

(35) Marciniec, B.; Gulinski, J. J. Organomet. Chem. 1993, 446, 15.

(36) Uozumi, Y.; Hayashi, T. J. Am. Chem. Soc. 1991, 113, 9887.

(37) Bergens, S. H.; Noheda, P.; Whelan, J.; Bosnich, B. J. Am. Chem. Soc. 1992, 114, 2121.

(38) Noyori, R. Asymmetric Catalysis in Organic Synthesis; Wiley: New York, 1994.

(39) Giardello, M. A.; Conticello, V. P.; Brard, L.; Gagné, M. R.; Marks, T. J. J. Am. Chem. Soc. 1994, 116, 10241.

(40) Stern, D.; Sabat, M.; Marks, T. J. J. Am. Chem. Soc. 1990, 112, 9558.

(41) Schaverien, C. J. Organometallics 1994, 13, 69.

(42) Bissinger, P.; Paul, M.; Jürgen, R.; Schmidbaur, H. Chem. Ber. 1993, 126, 2579.

[PREVIOUS][TableOfContents][NEXT]