Disproportionation of chloromethyldisilanes using Lewis-base heterogeneous catalysts - a way to influence the polymer structure

Thomas Lange*, Norbert Schulze, Gerhard Roewer, Robin Richter

Institut für Anorganische Chemie

TU Bergakademie Freiberg

Leipziger Str. 29, D - 09596 Freiberg, Germany

Tel.: (0049 3731) 39-4302

E-mail: Thomas.Lange@chemie.tu-freiberg.de

Keywords: disilane / Lewis base / Lewis acid / heterogeneously catalyzed disproportionation / polysilane

Summary: The heterogeneous disproportionation of 1,1,2,2-tetrachlorodimethyldisilane to chloromethylsilanes and oligo(chloromethylsilane)s is catalyzed by Lewis bases like bis(dimethylamid)phosphoryl compounds and N-heterocycles. The oligosilanes undergo branching and crosslinking reactions controlled by reaction temperature and time schedule forming poly(chloromethylsilane)s that show a 3D polysilyne-type polymer skeleton.

Lewis acids, such as triphenylboron, dissolved in the starting disilane, prevent branching of the polymer backbone during the reaction course.

Introduction

During the last years several synthetic pathways to different polysilane backbones have been extensively studied. One interesting polysilane synthesis has been developed based on the disproportionation of chloromethyldisilanes, that are by-products of the industrial chloromethylsilane production (Müller-Rochow-Synthesis) [1].

The disproportionation of 1,1,2,2-tetrachlorodimethyldisilane (2) leads to trichloromethylsilane (1) and oligo(methylchloro)silanes catalyzed by Lewis bases (Eq. 1).

Eq. 1

Formation of Oligo(chloromethylsilane)s derived from
1,1,2,2-Tetrachlorodimethyldisilane

The heterogeneous catalytic disproportionation offers the access to a poly(chloromethylsilane) free of catalyst, due to a perfect phase separation between the catalyst, the starting chloromethyldisilane and the reaction products [1]. It is thus possible to avoid subsequent uncontrollable cross-linking reactions.

The catalytic active entities, such as bis(dimethylamid)phosphoryl compounds and N-heterocycles were grafted onto the surface of a silica carrier via siloxane bonds, as shown in the following simplified scheme (Figure 1):

8

9

10

Fig. 1 Catalytic entities: bis(dimethylamid)phosphoryl compound (8), dimethylpyrazole groups (9) and benzimidazole groups (10) grafted on the surface of a silica carrier

1,1,2,2-tetrachlorodimethyldisilane is evaporated (bp 155 °C), and then put in contact with the catalyst stored in a fixed-bed reactor. The disilane disproportionates into a mixture of trisilane 3 and tetrasilane 4 and into trichloromethylsilane MeSiCl₃ onto the catalyst surface. The monosilane is distilled off due to its low boiling point (66 °C). The oligomer mixture obtained at 175 °C in the reaction pot contains beside oligomers 3 and 4, higher branched oligomers 5-7 (Figure 2).

3

4

5

6

7

Fig. 2 Oligo(chloromethylsilane)s formed during the first period of the disproportionation reaction- trisilane, tetrasilane, pentasilane, hexasilane and heptasilane (from left to right)

The composition of the oligomer mixture depends on the catalyst. Between 155 °C and 180 °C grafted N-heterocycles (9 and 10) generate oligomer mixtures rich in 3 with no hexa- and heptasilane (6 and 7) in contrast to grafted bis(dimethylamid)phosphoryl groups (8) [2]. We suppose that the basicity of the electron pair donors is not the decisive criterion for the catalytic efficacy. The one-electron donor capability is correlated with the value of the first ionization potential of the Lewis bases. The lower the ionization potential the higher the electron donor capability is expected. Our investigations have shown that Lewis bases with first ionization potentials smaller than 8.5 eV are suitable catalysts for Si-Si bond cleavage in Si₂Cl₄Me₂.

The Si-Si bond cleavage generates donor-stabilized silylene species (:SiClCH₃) that insert into Si-Cl bonds. It is suggested that 3 and 4 are formed in such a way [3]. Due to the functionality of 2, the formed oligo(chloromethylsilane)s show a branched structure. The reactions that lead to higher oligo(chloromethylsilane)s 5-7 in the reaction pot are less understood so far. Their formation is probably caused by condensation involving 3 and/or 4 with formation of MeSiCl₃.

The Formation of Poly(chloromethylsilane)s

If the pot temperature is slightly increased up to 220 °C, the oligosilanes undergo cross-linking reactions into highly branched poly(chloromethylsilane)s.

Using ¹³C as well as ²⁹Si NMR spectral editing techniques, an average composition of MeSiCl_0.73 is obtained which is in rather good agreement with the mass balance analysis (MeSiCl_0.62).

The polymer skeleton is constituted with MeCl₂Si-, MeClSi< and MeSi(Si)₃ groups, as shown in ¹³C and ²⁹Si CP-MAS NMR spectra (Figure 3) [3, 4].

Fig. 3 Influence of the Lewis base (8) on the polymer structure: ²⁹Si CP-MAS-NMR spectrum (left above), ¹³C CP-MAS-NMR spectrum (left below)

Influence of the Lewis acid (BPh₃) on the polymer structure: ²⁹Si CP-MAS-NMR spectrum (right above), ¹³C CP-MAS-NMR spectrum (right below)

(SSB: spinning side bands; CP: cross polarization)

A suggestion of the polysilane structure based on NMR and mass balance data is depicted in Figure 4.

Fig. 4 Structure of a poly(chloromethylsilane); the symbols n, u and l (see also Figure 3 left-hand side) are assigned the Si atoms with different substituents, (CD: cross-linking degree)

Gel permeation chromatography indicates a multi-modal polydispersity with a broad molecular weight distribution. Currently it is not possible to specify the definite values of the average molecular weights due to a lack of comparable standards.

The addition of a weak Lewis acid like triphenylborane (BPh₃) to the oligo(chloromethylsilane)s has a considerable effect on the polymer building procedure, above all on the average molecular weight and on the structural groups of the resulting polymer. These polymers are characterized by higher molecular weights and a small polydispersity, which can be described almost mono-modal. Compared to the polysilane represented in Figure 4 the characteristic end groups MeCl₂Si- and branched points MeSi(Si)₃ are missing or only found in very low percentages (see Figure 3 right-hand side). On the contrary, the linear units MeClSi< are the dominant sites. The presence of different carbosilane units especially (-CH₂)_xSiR_4-x (R = Me or Cl with x > 2) seems typical for triphenylborane modified polymers.

The oligomer formation (3-7) remains unchanged, when BPh₃ is already added to the starting disilane. However the composition of the volatile compounds becomes different. It does not only consist of trichloromethylsilane but also of dichlorodimethylsilane, benzene, trichloroborane, trimethylborane, and other diverse substituted silanes containing different combinations of chloro-, hydrido- and methyl groups.

The reaction mechanism, including the possible incorporation of the triphenylborane, has not yet been understood in detail very well. This complex reaction may compete with the branching process. We think that the donor-acceptor interaction between BPh₃ and the MeCl₂Si- groups of the oligomers induces first a carbosilane formation, which is then followed by a conversion into MeClSi< groups. Therefore the modified polymers has got more linear structural units resulting in lower branching.

Conclusion

The heterogeneously catalyzed disproportionation of 1,1,2,2-tetrachlorodimethyldisilane leads via oligo(chloromethylsilane)s to highly branched poly(chloromethylsilane)s. The disilane derived oligomer formation can be controlled by the nature of the Lewis base catalyst.

Modified polymers synthesized by the addition of the Lewis acid BPh₃ permit the access to lower branched polymer skeleton.

Acknowledgment

The authors are grateful to the „Deutsche Forschungsgemeinschaft“ and the „Fonds der Chemischen Industrie“ for financial support. We gratefully acknowledge the NATO grant for traveling between Freiberg and Paris. We particularly thank Mrs. Dr. Florence Babonneau (CNRS, Université Pierre et Marie Curie, Paris) for CP/IRCP MAS-NMR measurements.

References:

[1] U. Herzog, R. Richter, E. Brendler, G. Roewer, J. Organomet. Chem. 1996, 507, 221.

[2] R. Richter, G. Roewer, U. Böhme, K. Busch, F. Babonneau, H.-P. Martin, E. Müller, Applied Organomet. Chem. 1997, 11, 71.

[3] F. Babonneau, J. Maquet, C. Bonhomme, R. Richter, G. Roewer, D. Bahloul, Chem. Mater. 1996, 8, 1415.

[4] F. Babonneau, R. Richter, C. Bonhomme, J. Maquet, G. Roewer, J. Chim. Phys. 1995, 92, 1745-1748.