Hydrosilylation Catalysts

One class of platinum compounds used as hydrosilylation catalysts are Pt(0) complexes containing vinyl-siloxane ligands (12, 13). An example of such a catalyst is Karstedt’s catalyst, Pt_x(M^viM^vi)_y, formed by the reaction of divinyltetramethyldisiloxane (M^viM^vi) with chloroplatinic acid, H₂PtCl₆, Equation (iv).

(iv)

Karstedt’s catalyst is a Pt(0) complex which contains both bridging and chelating di-vinyl ligands (14, 15). The designation of Pt_x(M^viM^vi)_y, and M^viD_xM^vi as “solution A” is taken from the work of Lappert’s group and is used to describe the mixture of products containing platinum and vinylsiloxane oligomers (16, 17).

In the late 1960s and early 1970s, Willing of Dow Coming (18) and Karstedt of GE (15) described the reaction of chloroplatinic acid and vinyl-containing siloxane monomers to make siliconesoluble platinum complexes. Little was then known (19) about the structure of the platinum complexes which resulted from the reaction shown in Equation (iv) until the work of Lappert and co-workers (16, 17). The Lappert group reported that the platinum product, 1, of Equation (v):

(v)

contained a bridging M^viM^vi and each platinum had a chelating M^viM^vi group. ¹⁹⁵Pt NMR spectroscopic analysis of “solution A” showed the presence of two resonances which Lappert suggested was due to two isomers of 1. We proposed that the two ¹⁹⁵Pt NMR resonances were caused by the presence of 1 and 2, see Equation (vi) (20).

(vi)

When “solution A” was combined with vinyl-stopped polydimethylsiloxane oligomers, a single ¹⁹⁵Pt resonance was observed. Furthermore, analysis of “solution A” by field desorption mass spectroscopy (FDMS) showed the presence of 2 and vinyl-stopped oligomers, as shown in Equation (iv).

The reduction of Pt(IV) to Pt(0) by a silicon-vinyl group, in Equation (iv), appeared to be a new reaction. When M^viM^vi was the reducing agent, some of the silicon vinyl functionality was converted to a silicon-oxygen group. The net result was the conversion of an M^vi group into a D (dimethylsiloxane) group. The vinyl group was changed primarily into either butadiene or ethylene. Water was the source of the new oxygen bond to silicon. Similarly, when D^vi is the source of reducing agent, as in the reaction of Equation (vii), a D^vi group is converted into a T group.

The overall platinum conversion occurring in Equation (iv), is from Pt(IV) to Pt(0); the platinum in “solution A” being in the zero oxidation state. We attempted to observe the imputed Pt(II) intermediate from this vinyl-silicon-mediated reduction in Equation (iv). Typically, in this reaction, chloroplatinic acid is reacted with M^viM^vi in the presence of some ethanol, which aids in the dissolution of H₂PtCl₆. Sodium bicarbonate was added to remove chloride. Both ethanol and NaHCO₃ may aid the reduction. A ¹⁹⁵Pt NMR spectrum of “solution A” showed the previously mentioned resonances at –6200 ppm.

However, when the reduction was carried out without ethanol or NaHCO₃, a broad ¹⁹⁵Pt resonance was observed at –3470 ppm, which may be due to a Pt(II) intermediate. In order to characterise further the potential Pt(II) intermediates and to improve understanding of the mechanism of the reduction process, H₂PtCl₆ was reacted with several silicon-vinyl-containing species. The product of the reaction of H₂PtCl₆ and dimethyldivinylsilane, (CH₃)₂Si(CH=CH₂)₂, gave two products, see Equation (viii), a Pt(II) complex, 3, and a Pt(0) structure, 4. Complex 3 was isolated and a single crystal X-ray

(vii)

(viii)

structure, see Figure 1, showed it to be a dinuclear complex with two bridging chlorine atoms and a bridging (CH₃)₂Si(CH=CH₂)₂ ligand. Additionally each platinum atom in 3 contained a η¹:η² -(CH₃)₂Si(CH=CH₂)(CH₂CH₃) ligand. Compound 3 had a ¹⁹⁵Pt NMR resonance at –3603 ppm, while the Pt(0) product, 4, had a ¹⁹⁵Pt NMR resonance at –6152 ppm. Although the structure of 4 was not determined directly, addition of PPh₃ to solutions containing 4 produced (M^vi M^vi)Pt(PPh₃) suggesting that 4 was Pt₂(M^viM^vi)_x((CH₃)₂Si(CH=CH₂)₂)_y(20).

Inhibitors

Another important aspect of industrial hydrosilylation is the use of inhibitors (21–23). The crosslinking reaction of Equation (iii) will usually occur with as little as 10 ppm platinum, at ambient temperature in a matter of minutes. The rapidity of this reaction leads to the need for inhibitors to give some control over the reaction. Typical industrial applications require long work times at low temperatures, followed by fast curing times at elevated temperature.

Two inhibitors, dimethyl fumarate and dimethyl maleate, are commonly added to platinum-catalysed crosslinkable silicone formulations to permit long, work life by the end user at ambient temperature with a rapid cure at elevated temperature. The reaction of “solution A” with four equivalents of dimethyl fumarate, relative to platinum, resulted in formation of a platinum-fumarate complex, 5, see Equation (ix). Complex 5 was characterised on the basis of its ¹³C NMR spectrum and by comparison of its spectrum with those of the maleate analog, 6, and the previously well characterised compound PPh₃Pt(M^viM^vi), 7. Compound 7 was prepared by the reaction of “solution A” with one equivalent of PPh₃, in the presence of 10 equivalents of M^viM^vi. In effect the fumarate (or maleate) replaces the bridging, but not the chelating M^viM^vi ligandin Karstedt’s catalyst.

(ix)

A comparison of the ¹³C NMR spectra for 5, 6 and 7 in the olefin-platinum region is shown in Figure 2. All the ¹³C olefin resonances had ¹⁹⁵Pt satellites. The spectrum of 5 showed four overlapping triplets from 68 to 72 ppm. These four resonances derived from the four different olefinic carbons present in the chelating M^viM^viligand.

By contrast, the ¹³CNMR spectrum for 6 had two resonances for the olefinic carbons of the M^viM^vi ligand because the two double bonds are chemically equivalent. The downfield triplet resonance at 69.2 ppm (J_Pt−c = 56 Hz) was assigned to the methylene carbon of the M^viM^vi ligand of 6, while the methyne carbon was the triplet upfield at 71.94 ppm (J_Pt−c = 42 Hz). Both 5 and 6 showed peaks in the ¹³C NMR due to the olefinic carbons of fumarate and maleate, respectively, with large coupling constants: J_Pt−c = 89 and 104 Hz, respectively. Compound 7, where maleate or fumarate was replaced with PPh₃ had two resonances for those of the olefinic carbons of M^viM^vi, but shifted upfield relative to those in 6 due to the presence of the PPh₃ ligand.

Platinum Colloid Formation

The mechanism of the hydrosilylation reaction at the molecular level has been studied for many years (1–8), and several reviews have been devoted to the reaction (24–28). The Chalk-Harrod mechanism, developed during the 1960s, is the most often cited mechanism for hydrosilylation and is based on fundamental steps of organometallic chemistry (8); these include oxidative addition of a Si-H group to a metal-olefin complex, insertion steps and reductive elimination. However, a number of phenomena are not explained by the Chalk-Harrod mechanism and among these is the presence of an induction period (the time between the start of a chemical reaction and its observable occurrence), formation of coloured bodies and co-catalysis by oxygen.

In the 1980s Lewis and co-workers proposed a mechanism based on the intermediacy of platinum colloids (13, 14, 25, 26), which had been observed via transmission electron microscopic (TEM) analysis of reaction solutions following platinum-catalysed hydrosilylation (27).

Reaction solutions from a platinum-catalysed hydrosilylation reaction were originally analysed by first evaporating the reaction solutions and then recording the TEM images (12–14, 22–25). TEM analysis showed, in some cases, that colloidal platinum had formed after the hydrosilylation reaction. More recently, Pt EXAFS of reaction solutions from hydrosilylation reactions has shown that the type of platinum species formed in solution at the end of a reaction depends on:

• the ratio of Si-H to vinyl used (21), and

• the nature of the olefin (hydrocarbon or silicon-vinyl).

The two types of platinum end products observed are illustrated in Equation (x) where 8 was the species formed at high vinyl concentrations and 9 was the platinum product formed when the concentration of Si-H was greater than, or equal to, that of vinyl. Compound 8 was the designation used for platinum species containing only Pt-C bonds (at a typical Pt-C:olefin distance of around 2.17 Å and with a co-ordination number, number in parentheses, of six). Compound 9 was the species that contained both Pt-Pt and Pt-Si single bonds as determined by EXAFS.

(x)

(xi)

(xii)

(xiii)

(xiv)

Note that D₄ is cyclooctamethyltetrasiloxane

Platinum species 9 could be converted to 8 in the presence of excess olefin (if the olefin was a silicon-vinyl compound, that is, if more MD^viM was added). Compound 8 was formed when silicon-olefin was the olefin source, that is MD^viM.

However, if a Si-H compound was reacted with an olefinic hydrocarbon, such as hexene, which does not contain silicon-vinyl, then 9 was formed regardless of the stoichiometry of olefin and Si-H. Similarly, with neo-hexene (3,3-dimethyl-l-hexene) platinum reaction product 9 is formed from the platinum-catalysed hydrosilylation reaction between neo-hexene and a number of different Si-H compounds, even if the concentration of Si-H:neo-hexene is 1:2.

The reactions in Equations (xi) to (xiv), (where (xi) and (xii) describe the platinum catalysed hydrosilylation of triethylsilane with neo-hexene), all give dark-coloured solutions. TEM analyses of the reaction products from these solutions after evaporation all show that platinum crystallites were present.

A representative TEM, from Equation (xiii), is shown in Figure 3; here individual crystallites can be seen, displaying diffraction fringes with spacing which matches that for the (111) plane in crystalline platinum. The diffraction pattern also could be indexed to that of platinum. Finally, energy dispersive X-ray spectroscopy of the spots confirmed that they were composed of platinum.

In the reactions shown in Equations (xi) and (xii), in situ studies showed that platinum of type 9 (from Equation (x)) formed in the reactions with neo-hexene. On evaporation, the type 9 platinum composition is converted to platinum crystallites. The effect of dilution with D₄ or of running the reaction without solvent was examined together with the effect on platinum crystallite size. Reactions between dimethyl-ethoxysilane and either Karstedt’s catalyst (Equation (xiii)) or PtCl₄ (Equation (xiv) (12)) were carried out under conditions where previous reports had shown that colloids formed. The Table shows the results of a statistical analysis of the particle sizes of the platinum crystallites formed from the reactions in Equations (xi) to (xiv).

The particle size of the crystallites may have been affected by the platinum concentration; it can be seen in the Table that the larger particles from Equation (xiv) are produced from a higher platinum concentration. The particles formed from Equations (xi) and (xiii) were similar in size and larger than those from Equation (xii). Equation (xii) did not use D₄ diluent.

When hydrosilylation reactions were run with silicon-vinyl substrates, for example Equation (x) where platinum of type 8 is formed, TEM analyses of the resulting evaporated solutions were different from those obtained from Equations (xi) to (xiv). Note that all of these reactions were carried out in neat solutions and it is not known if the absence of diluent contributed to the observed TEM. TEM of evaporated solutions from Equations (xv) to (xviii) showed the complete absence of the platinum crystallites seen in Figure 3.

(xv)

(xvi)

(xvii)

(xviii)

A darkfield TEM micrograph, prototypical for the evaporated solution from Equations (xv) to (xviii) is shown in Figure 4. Energy dispersive X-ray spectroscopy confirmed the presence of platinum. However, the diffraction pattern did not match that of metallic platinum or any platinum compound in the powder diffraction file. Electron diffraction analysis indicated the presence of a crystalline material which had formed in very thin crystalline sheets that were slightly misaligned with respect to one another. The moiré diffraction fringes due to this misalignment are clearly visible in the micrograph.

Evaporation of the product solution from the reaction between PtCl₄ and a siloxane-like SiH source, M₃T^H, tris(trimethylsiloxy)silane (Me₃SiO)₃SiH, gave a TEM similar to Figure 3, while evaporation of the product solution from the reaction between PtCl₄, M₃T^H and M₃T^vi gave a TEM similar to Figure 4. Further studies are underway to determine the structure of platinum species present during hydrosilylation under the various conditions discussed here. The species in Figure 4 may be explained by these analyses (29).

Summary and Conclusions

Hydrosilylation is widely used industrially for the preparation of monomers with silicon-carbon bonds and for producing crosslinked polymers.

Highly active, silicone-soluble platinum catalysts can be prepared by the reaction of chloroplatinic acid with vinyl-silicon containing compounds. The vinyl-silicon compounds, such as divinyltetramethyldisiloxane (M^viM^vi) reduce platinum(IV), in chloroplatinic acid, to platinum(0) with the concurrent conversion of the silicon-vinyl group to a silicon-oxygen group. A typical platinum catalyst is Karstedt’s catalyst, Pt_x(M^viM^vi)_y.

Inhibitors are widely used in industry to control the platinum-mediated addition reaction of curable systems. The addition of inhibitors prevents hydrosilylation (crosslinking) at low temperature while permitting rapid reaction at elevated temperature. Commonly used inhibitors include those with electron deficient double bonds, such as maleates and fumarates. Dimethyl maleate reacts with Karstedt’s catalyst to make a complex containing a chelating M^viM^vi group and a maleate ligand.

Previous studies reported that formation of platinum colloids was a key step in hydrosilylation. It is now clear that colloid formation occurs as an end stage of the reaction. EXAFS analysis has shown that molecular compounds are present during hydrosilylation. Colloidal platinum is observed by TEM after evaporation of solutions from several reactions which involve platinum, a Si-H compound and either poorly co-ordinating olefins or no olefin. However, in some cases where silicon-vinyl-containing species were present, the reaction product between platinum and a Si-H-containing compound did not give colloidal platinum species; and TEM analysis showed that the crystalline material present was not metallic platinum crystallites.

Further work is in progress to identify the structures of platinum intermediates formed during hydrosilylation and industry continues to search for more highly active catalysts and more effective inhibitors.