PtO₂ as a “self-dosing” hydrosilylation catalyst

Abstract

The reaction behavior of PtO₂ in the hydrosilylation of n-octene with heptamethyltrisiloxane was examined. PtO₂ proved to be a highly active and selective hydrosilylation catalyst precursor reaching turn-over frequencies in the order of magnitude of 10⁵ h⁻¹. In contrast to usual homogeneous catalysts PtO₂ dissolves only to a small degree after reaction with silane. The un-reacted solid catalyst could be easily separated from the reaction mixture by simple decantation or filtration and could be used for subsequent runs. The observation of an induction period in every cycle indicates that the active species is formed in situ before the reaction can take place and new material has to dissolve for each run. The active species is formed by reduction of PtO₂ with the silane and is soluble in the reaction mixture. The solubility behavior together with the high activity allows a “self-dosing” of the catalyst—leading to little waste of precious metal in contrast to other “homogeneous” (i.e. better soluble) Pt-based catalysts.

---

1. Introduction

The hydrosilylation of olefins (and acetylenes) has been studied extensively for a long period of time, because the reaction is most important for the preparation of certain organosilicon compounds both in laboratory and in industry.¹ Furthermore, the functionalization and/or activation of silicones by introducing organic, chemically active side groups is of major industrial importance² and silane-functionalized olefins, dienes or polymers have gained substantial industrial interest.^3,4 Among a variety of catalysts, which enable the addition of hydrosilanes to carbon–carbon multiple bonds, hexachloroplatinic acid (H₂[PtCl₆], ‘Speier's catalyst’)⁵ and [Pt₂(sym-tetramethyldivinyldisiloxane)₃] (‘Karstedt's catalyst’)⁶ are still by far the most commonly used catalysts for this reaction. Other catalysts may be nucleophiles, Lewis acids, supported metals, metals reduced in situ or transition metal complexes attached to inorganic or polymeric supports.^7,8

A clear indication of the diverse and continuing interest in the hydrosilylation reaction is the great number of articles and reviews that appeared on this subject over the decades. Extensive reviews on the catalyzed hydrosilylation of unsaturated carbon–carbon multiple bonds were provided by Ojima et al. in 1981,⁹ 1989,¹⁰ and 1998.¹¹ Further comprehensive reviews focusing on different aspects and applications of the hydrosilylation reaction followed subsequently by Voronkov et al.¹² (1992), Brook¹³ (2000), Reichl and Berry¹⁴ (1999), Roy¹⁵ (2007), Marciniec (1992,³ 2002⁴ and 2009⁷) and us⁸ (2010). Many new strategies have been developed to improve reaction conditions and reaction efficiency. Thus, completely new ligand classes have been developed for homogeneous catalysis. Lanthanide compounds were established as efficient hydrosilylation catalysts. Asymmetric hydrosilylation has become an important tool in chiral synthesis, and new materials accessible via hydrosilylation ranging from block copolymers to dendrimers and functionalized silicones have been developed.^7,8

Compared to homogeneous catalysts, heterogeneous catalysts used in hydrosilylation reactions are quite rare. Although they could be easily removed by filtration and reused in several cycles their large-scale application is quite limited. Many systems suffer from significant leaching or lose their activity after only a few runs. However, especially for the very expensive and biological hazardous Pt-catalysts separating them from the reaction mixture and recycling would be highly desirable.

Accordingly, the catalytic activity and recyclability of PtO₂ is examined. This compound was described as a potent hydrosilylation catalyst, especially for the hydrosilylation of aminated alkenes by Mioskowski et al. in 2002¹⁶ and is since then occasionally used for this purpose.^17–19

2. Experimental

2.1. General

All reactions were carried out under an argon-atmosphere using standard Schlenk techniques. Heptamethyltrisiloxane (HMTS, 97%) and n-octene (98%) were used as received from ABCR and Acros Organics. PtO₂ (81–83% Pt, >60 m² g⁻¹) was used as received from Sigma Aldrich. Karstedt's catalyst was used as diluted solution (1 mg Pt mL⁻¹n-hexane) as received from standard commercial sources. H₂PtCl₆ and Pt(PPh₃)₄ were used as received from Sigma Aldrich and were dissolved in isopropanol and dichloromethane, respectively, with a concentration of 1 mg Pt mL⁻¹.

In order to avoid contaminations, all experiments except the recycling experiment were carried out in fresh glassware with a new magnetic stir bar. NMR experiments were performed in CDCl₃ (Deutero) with a Bruker Avance 400, in situ IR studies were performed on a Mettler Toledo ReactIR™.

2.2. Comparison of homogeneous and heterogeneous catalysts

4.00 g HMTS (18.0 mmol) and 2.22 g n-octene (19.8 mmol, 1.1 eq.) were mixed at room temperature, the corresponding amount of catalyst was added and the reaction mixture was moved into an 80 °C preheated oil bath. For the homogeneous catalysis 10 ppm (0.06 mg Pt) of platinum (with respect to the total weight of the reaction mixture) were used. In the case of PtO₂ 100 ppm (0.622 mg) of platinum (0.8 mg PtO₂, 3.5 μmol) were used. Samples for ¹H-NMR spectroscopy were taken every 15 min. Yields were calculated from the ratio of the Si–H signal of the silane at 4.7 ppm and the Si–CH₂ signal of the product at 0.5 ppm.

¹H-NMR (400 MHz, CDCl₃): δ = 0.02 (s, 3H, Si–CH₃), 0.11 (s, 18H, OSi(CH₃)₃), 0.47 (m, 2H, Si–CH₂), 0.91 (t, 3H, ³J = 6.7 Hz, CH₃), 1.29 (m, 12H, CH₂) ppm.

2.3. Determination of TOFs

4.00 g HMTS (18.0 mmol) and 2.22 g n-octene (19.8 mmol, 1.1 eq.) were mixed at room temperature, the corresponding amount of catalyst was added and the reaction mixture was moved to a 60 °C preheated oil bath. 1000 ppm correspond to 7.8 mg (35 μmol) PtO₂, 100 ppm to 0.8 mg PtO₂ (3.5 μmol) and 10 ppm to 0.08 mg PtO₂ (here, approx. 0.1 mg were used). Samples were taken and analyzed as described above.

2.4. Recycling and leaching experiments

3.00 g HMTS (13.5 mmol) and 1.51 g n-octene (13.5 mmol, 1.0 eq.) were mixed at room temperature, 1.4 mg PtO₂ (300 ppm Pt, 6.0 μmol) were added, the reaction mixture was moved to a 85 °C preheated oil bath and the in situ IR measurement was started. After complete conversion, the clear and colorless product solution was carefully removed with a syringe and fresh substrates (3.00 g HMTS and 1.51 g n-octene) were added to the solid catalyst.

For the leaching experiments the supernatants of the first 4 cycles were filtered through a 0.45 μm syringe filter, fresh substrates (2.00 g HMTS and 1.01 g n-octene) were added and the IR measurement was started. After complete conversion (50 min) the supernatant of cycle 1 was again mixed with fresh substrates (2.00 g HMTS and 1.01 g n-octene). This procedure was repeated once again.

2.5. Pretreatment of HMTS and n-octene with PtO₂

3.00 g HMTS or 1.51 g n-octene, respectively, were stirred with 1.4 mg PtO₂ for 2 h at 85 °C. The liquid phase was removed with a syringe and filtered over a 0.45 μm syringe filter. 1.51 g n-octene or 3.00 g HMTS, respectively, were added and the reaction mixture was moved to an 85 °C preheated oil bath and the in situ IR measurement was started.

3. Results and discussion

As a test reaction to explore the potential of PtO₂ as a catalyst for the hydrosilylation of olefins, we choose the hydrosilylation of n-octene with 1,1,1,3,5,5,5-heptamethyltrisiloxane (HMTS) according to Scheme 1.

Scheme 1 Hydrosilylation of n-octene with heptamethyltrisiloxane.

HMTS can be seen as a model for poly(dimethyl-co-hydromethyl)siloxanes, which are important intermediates in the functionalization of silicones.²

At first, we compared the catalytic activity of PtO₂ with that of the well established homogeneous systems, namely Karstedt's catalyst, H₂PtCl₆ and Pt(PPh₃)₄. All catalysts were examined under standard hydrosilylation conditions, i.e. silane and a slight excess of n-octene (1.1 eq.) were stirred at room temperature under argon, the catalyst was added and the reaction mixture was moved to a 80 °C preheated oil bath. For the homogeneous catalysts 10 ppm of platinum (with respect to the total weight of the reaction mixture (0.002 mol%)) were used. In the case of PtO₂ 100 ppm of platinum were applied because smaller amounts are problematic to be weighed correctly. With the homogeneous catalysts, the reaction is strongly exothermic and the product solution turns yellow due to the formation of colloidal platinum.²⁰ With PtO₂ no increase in temperature can be observed, the product solution stays colorless. In all cases only the desired terminal addition product 1 (Scheme 1) is formed. Except for traces of 2-octene—as a result of isomerization—no byproducts are formed.

To follow the reaction progress by ¹H-NMR spectroscopy samples were taken every 15 min. Yields were calculated based on the ratio of the Si–H signal of the silane at 4.7 ppm and the Si–CH₂ signal of the product at 0.5 ppm. The corresponding conversion plots at 80 °C are shown in Fig. 1.

Fig. 1 Comparison of homogeneous hydrosilylation catalysts and PtO₂ at 80 °C.

With all three classical homogeneous catalysts, the reaction is complete within 15 min. With PtO₂ an induction period of approx. 30 min is observed, followed by a fast and complete reaction. 100% conversion is reached after 60 min.

To study the kinetics in more detail, we varied the amount of PtO₂. Fig. 2 shows the corresponding conversion–time plots at 60 °C for 1000 ppm Pt (0.2 mol%), 100 ppm Pt (0.02 mol%) and approx. 10 ppm Pt (0.002 mol%). Samples were taken every 15 min and the reaction progress was followed by ¹H-NMR spectroscopy as described above.

Fig. 2 Variation of the amount of PtO₂ at 60 °C.

Obviously the reaction is completed much faster when increasing the concentration of the catalyst. For a catalyst loading of 1000 ppm the reaction is complete within 30 min. However, even with very low catalyst loading (10 ppm, 0.002 mol% Pt) the reaction proceeds within 90 min. Again, in all cases an induction period is observed, ranging from 15 min (1000 ppm) to 45 min (10 ppm). After that, the reaction proceeds smoothly, indicating that a catalytic active species has to be formed in situ before the reaction can take place. However, even when using only 10 ppm PtO₂, the material does not dissolve completely. It appears that still only a small portion is used for the reaction.

Unfortunately, it is very difficult to determine the amount of the dissolved platinum species by classical quantitative analysis because of a strong interference with the silicone matrix.

Anyway, PtO₂ has to be regarded as a catalyst precursor, which is reduced in situ to transfer the Pt in a lower oxidation state (II or 0) which is required for oxidative addition of the silane.²¹ A certain amount of the active species has to be generated before the reaction can occur, for a higher amount of the catalyst precursor this takes a shorter period of time than for lower catalyst loading—probably with larger PtO₂ amounts available, the particles more easily to dissolve or react with HMTS are more abundant. The particle sizes and shapes in the commercially available PtO₂ vary greatly and a more detailed picture concerning which particles dissolve preferably could not be obtained.

The slope of the conversion–time curves at the times of maximal turnover, i.e. after the induction period, is nearly the same in all cases. It is to assume that always the same, small amount of Pt is dissolving. From this experiment also the turn-over-frequencies (TOFs) of PtO₂ were determined (Table 1). The highest obtained number indicates a “lower estimate” of the real activity of the active species.

Table 1 Turn-over-frequencies for different amounts of PtO₂

Amount of PtO2/ppm	TOF/h^−1
1000	1200
100	12 000
10	95 000

---

For the calculation, the steepest slope of every curve, corresponding to approx. 60% conversion in 15 min has been used. With a TOF of (likely ≫) 95 000 h⁻¹ PtO₂ is a highly active and efficient hydrosilylation catalyst. As stated above, depending on the amount of PtO₂ that is actually converted to the active species, the corresponding TOF is most likely considerably higher, for it is assumed in these calculations that all PtO₂ dissolve to obtain at least a lower limit for the TOF.

After these more fundamental studies, we examined the usability of the remaining PtO₂ with in situ IR spectroscopy at 85 °C. After each cycle the clear and colorless product solution was carefully removed with a syringe and fresh substrates were added to the catalyst. It cannot be avoided, however, to also remove small amounts of the fine catalyst powder. Accordingly 300 ppm of Pt have been used to minimize the effects of the Pt-loss. The reaction behavior of the first seven cycles is shown in Fig. 3. The reaction progress was determined by following the decrease of the Si–H absorption band at 2100 cm⁻¹.

Fig. 3 First seven cycles of the recycling of PtO₂ at 85 °C.

The remaining catalyst can be used for at least six recycling steps without significant loss of activity. In all cases the reaction is complete within 25 min and an induction period of approx. 8 min can be observed. Because the induction period is present in every catalytic cycle, we assume that the catalytically active species has to be formed anew in every cycle to perform the hydrosilylation reaction. The easiest way to explain such a reaction behavior is that only a small part of the PtO₂ is dissolved to form the active species and is removed with the product solution after the reaction. In the next cycle the active species has to be formed again to maintain the catalytic performance.

To further examine this assumption, we also analyzed the reaction behavior of the supernatant solution. For this purpose, the supernatants of the first four recycling cycles were filtered through a 0.45 μm syringe filter to remove all traces of the heterogeneous catalyst and mixed with fresh substrates. The reaction progress was followed by in situ IR spectroscopy as shown in Fig. 4.

Fig. 4 Catalytic behavior of the supernatant of the first four recycling cycles.

The supernatants show high catalytic activity without a significant initiation period. In all cases the reaction proceeds with almost 100% conversion in 40 min. The nearly similar reaction velocity in every cycle suggests that always an equal amount of active species is generated. The doubling of the reaction time—compared to the original run—can be explained by a dilution effect.

After full conversion the reaction can be restarted by addition of fresh substrates to the product solution. This is shown in Fig. 5 for the supernatant of cycle 1.

Fig. 5 Addition of fresh substrates to the supernatant of cycle 1.

Based on these experiments, it seems to be evident that the active species is formed in situ and is soluble in the reaction mixture. It is highly catalytically active and can be “reused” by addition of fresh substrates to the supernatant.

When HMTS is pre-treated with PtO₂ for 2 h at 85 °C dissolution of some of the PtO₂ also occurs. The HMTS phase shows high catalytic activity when removed from the solid catalyst–precursor and mixed with n-octene. Furthermore, no induction period can be observed. When n-octene is stirred with PtO₂ under the same conditions, after removal of PtO₂ and addition of HMTS, nearly no reaction takes place. Therefore it can be concluded that the active species is formed from PtO₂ in the presence of the silane, probably by reduction of Pt(IV) to Pt(II) or Pt(0). With n-octene instead, this species is not formed. Fig. 6 shows the course of catalysis with the PtO₂-pretreated HMTS and n-octene phase compared to normal catalysis with PtO₂.

Fig. 6 Effect of pretreatment.

4. Conclusions

PtO₂ is a highly active and regioselective hydrosilylation catalyst precursor. In the reaction of HMTS with n-octene TOFs of at least 95 000 h⁻¹ were obtained. After complete conversion the remaining, unused PtO₂ can be removed from the reaction mixture by simple decantation or filtration and can be utilized for many (≫7) runs. In every cycle an induction period is observed, indicating that the active species is formed from unused PtO₂in situ before the reaction takes place. The active species is soluble in the reaction mixture and is removed with the product after each cycle. Since only small amounts of catalyst (≪10 ppm) are needed the active species must indeed be highly catalytically active and can be reused by addition of fresh substrates to the supernatant.

When HMTS is treated with PtO₂ this activation also occurs. The HMTS phase shows high catalytic activity when separated from the solid catalyst and mixed with n-octene. In this case no induction period is observed. When n-octene is stirred with PtO₂ alone under the same conditions nearly no reaction takes place. Therefore, it can be concluded that the active species is formed from PtO₂ in the presence of the silane, probably by reduction of Pt(IV) to Pt(II) or Pt(0). The lower oxidation state is required for oxidative addition of the silane to the metal in the catalytic cycle of hydrosilylation.

Due to the low amount of PtO₂ reacting even with large excesses of HMTS the unravelling of the true nature of the active species might be difficult to achieve, at least with the currently available spectroscopic means. While this perspective could be sobering for the scientist, the notion of a largely “self-dosing” catalyst is probably attractive for (industrial) applicants.

Acknowledgements

Financial support of this work by the TUM graduate school is gratefully acknowledged.

Notes and references

1. W. Noll, Chemistry and Technology of Silicones, Academic Press, New York, 1968.
2. L. Abdellah, B. Boutevin and B. Youssef, Prog. Org. Coat., 1994, 23, 201–236.
3. Comprehensive Handbook on Hydrosilylation, ed. B. Marciniec, Pergamon, Oxford, 1992.
4. B. Marciniec, Silicon Chem., 2002, 1, 155–175.
5. J. L. Speier, J. A. Webster and G. H. Barnes, J. Am. Chem. Soc., 1957, 79, 974–979.
6. B. D. Karstedt, US Pat., 3 775 452, Assigned to General Electric Co., 1973.
7. B. Marciniec, Adv. Silicon Sci., 2009, 1, 1–51.
8. S. Putzien, O. Nuyken and F. E. Kühn, Prog. Polym. Sci., 2010, 35, 687–713.
9. I. Ojima and T. Kogure, Rev. Silicon, Germanium, Tin Lead Compd., 1981, 5, 7–66.
10. I. Ojima, in The Chemistry of Organic Silicon Compounds, ed. Z. Li and J. Zhu, John Wiley & Sons, New York, 1989, pp. 1479–1526.
11. I. Ojima, Z. Li and J. Zhu, in The Chemistry of Organic Silicon Compounds, ed. S. Patai and Z. Rappoport, John Wiley & Sons, New York, 1998, pp. 1687–1792.
12. V. B. Pukhnarevich, E. Lukevits, L. T. Kopylova and M. G. Voronkov, Prospects for hydrosilylation, Lituana, 1992, 381–397.
13. M. A. Brook, Silicon in Organic, Organometallic and Polymer chemistry, Wiley, New York, 2000.
14. J. A. Reichl and D. H. Berry, Adv. Inorgmet. Chem., 1998, 43, 197–265.
15. A. Roy, Adv. Organomet. Chem., 2007, 55, 1–59.
16. N. Sabourault, G. Mignani, A. Wagner and C. Mioskowski, Org. Lett., 2002, 4, 2117–2119.
17. P. Reyes, C. Campos and J. L. G. Fierro, J. Chil. Chem. Soc., 2007, 3, 1249–1253.
18. A. Hamze, O. Provot, J.-D. Brion and M. Alami, Synthesis, 2007, 2025–2036.
19. J. Zhou, R. F. Guan and S. Y. Feng, Eur. Polym. J., 2004, 40, 165–170.
20. L. N. Lewis and N. Lewis, J. Am. Chem. Soc., 1986, 108, 7228–7231.
21. A. J. Chalk and J. F. Harrod, J. Am. Chem. Soc., 1965, 87, 16–21.

---