Fully Condensed Polyhedral Oligosilsesquioxanes (POSS): From Synthesis to Application

Paul D. Lickiss*; Franck Rataboul    Chemistry Department, Imperial College, London SW7 2AZ, UK
* Corresponding author. email address: p.lickiss@imperial.ac.uk

I INTRODUCTION

The widespread applications of compounds containing Si–O–Si linkages in both inorganic compounds such as silicates and aluminosilicates, and in organometallic compounds such as silicone polymers have been known for many years and have formed the basis of significant industrial effort. In addition to these well-known compounds, the last couple of decades have seen a rapid increase in interest in discrete molecular oligosiloxane species, often formed from the hydrolytic condensation reactions of trifunctional silicon monomers RSiX3 (where X is usually a halide or alkoxide group). The year 2006 saw the 60th anniversary of the first publication by Scott reporting the isolation of a volatile silsesquioxane, probably Si8O12Me8, from the thermal rearrangement of products derived from the cohydrolysis of Me2SiCl2 and MeSiCl3.1 Much earlier work on the hydrolysis of trichlorosilanes had probably led to some polyhedral oligomeric silsesquioxanes type products being formed as components of complicated mixtures but these were generally wrongly identified as silicon analogs of carboxylic acids or their anhydrides.2 Work by Kipping3 did however, recognize the polymeric nature of these compounds and he identified them as “anhydrides” with formulae such as [(C6H11SiO)2O]n. Soon after the report by Scott on a volatile silsesquioxane, Barry and Gilkey4 correctly identified the polyhedral nature of some of the volatile silsesquioxanes derived from hydrolysis of RSiCl3 compounds (R=Et, Pr, Bu) and a more detailed examination of a range of crystalline polyhedral silsesquioxanes was published in 1955.5 Detailed studies on the condensation reactions of c-C6H11SiCl36 and PhSiCl37 showed that condensation reactions of trifunctional organosilicon species can give rise to a wide range of products of general formula (RSiO3/2)n including random polymeric networks, 1, and ladder polymers, 2,8 incompletely condensed polyhedral species, 3,9 and fully condensed polyhedral species, for example, structure 4 as shown in Figure 1.

The random structures, 1, have no long-range order and the ladder polymers, 2, contain no polyhedra as do the oligomeric species, 4, but the incompletely condensed compounds, 3, can readily be seen as precursors to the fully condensed compounds. Significant early work on the organometallic derivatives of species such as 3 (R=cyclohexyl) was carried out by Feher and coworkers9 and led to widespread studies involving the use of the reactive silanol functions as models for such groups at solid silica surfaces.10 Early work on the synthesis of silsesquioxanes was reviewed in 198211 and so this chapter will concentrate on work since then. More recent work has concentrated on the study of compounds, 4, which have been found to have many applications utilizing their highly symmetrical, three-dimensional nature and their potential to act as nano-sized building blocks for the assembly of ordered two- and three-dimensional structures. This review will concentrate on compounds of type 4, for reviews of the chemistry of compounds of type 1, 2, and 3 see Refs. (9, 10c, and 10g). In keeping with the title of this book this chapter also concentrates on organometallic compounds of type 4; the many inorganic silicates and zeolites containing Tn rings will only be mentioned briefly. The nomenclature to describe structures such as 4 is complicated, for example, the full Chemical Abstracts name for 4 (R=H) is pentacyclo[9.5.1.13,9.15,15.17,13]octasiloxane. Fortunately, the nomenclature for describing siloxane polymers is convenient for describing the polyhedra, a “T” unit denoting a silicon atom with three siloxane oxygen atoms attached to it, such that structure 4 (R=H) may be labeled as T8H8. The general stoichiometry of these compounds is (RSiO3/2)n and they are usually thus called silsesquioxanes, although this term has largely been superseded in recent years by the term Polyhedral Oligosilsesquioxanes, abbreviated as POSS by those working in this area and trademarked by Hybrid Plastics.12 These compounds, having a SiO3/2 stoichiometry, can be seen as intermediate between the inorganic SiO2 formula for silica and the SiO repeat unit for the backbone in polysiloxanes or silicones.

The format of this review is based on the number of T units in the polyhedral structure and looks at methods for POSS synthesis, structure, reactivity, and their relatively recent applications in technology. The recent surge in interest in POSS chemistry has been enabled by new syntheses to replace the original, lengthy, and sometimes erratic preparations. The majority of work in the area of POSS chemistry has been carried out on T8 derivatives (at the end of 2006, a Chemical Abstracts substructure search revealed 23 T4 derivatives in 26 publications, 87 T6 derivatives in 87 publications, 1826 T8 derivatives in 1186 publications, 50 T10 derivatives in 74 publications, and 23 T12 derivatives in 33 publications) and these compounds will be described in groups according to the nature of the pendent group at the silicon vertices. There are now a very large number of studies on POSS molecules and some discretion has been required in what to include in the review. Many details of the syntheses and physical data are tabulated for convenience and discussion has been limited to the most important compounds. Reviews of the uses of POSS molecules in polymers, copolymers, and nanocomposites have been published13 and so have computational simulations of POSS organic/inorganic hybrid materials.14

II T2 COMPOUNDS

The smallest possible fully condensed silsesquioxanes are the hypothetical T2R2 species (Figure 2). Such compounds have not been isolated and they would be expected to be extremely reactive toward oligomerization and addition reactions as are the monocyclic compounds containing a single Si2O2 ring.15 The T2H2 structure has, however, been the subject of ab initio calculations that show it to have an unusually short ⋯Si distance of 2.060 Å and an Si–O–Si angle of 74.7°.16

It is unlikely that T2R2 compounds can be prepared by the usual hydrolytic routes to siloxanes, but it might be possible to prepare one by dehydration of a very sterically hindered silanetriol, RSi(OH)3 which can readily be isolated,17 using a powerful dehydrating agent. This route has, however, been found previously to give T6R6 species for R=tBu or tHexyl,18 but even larger substituents may promote the formation of smaller rings.

III T4 COMPOUNDS

A Synthesis

The smallest synthetically accessible POSS compounds are the T4 species based on just four Si atoms in an approximately tetrahedral arrangement. There are few of these compounds that have been well described, the problem being that the obvious synthetic routes to them, for example hydrolysis of RSiCl3, much more readily give rise to compounds containing Si4O4 rings, that is, T8 species rather than the more strained Si3O3 rings required in T4 compounds. Although the subject of several computational studies (see below) T4H4 has not been isolated, the hydrolysis of HSiCl3 being well known to provide polymeric materials and low yields of T8H8 (see Section V.B). Attempts to prepare T4R4 with R=Me, Et, iPr, or tBu by hydrolysis of RSiCl3 (Figure 3) depend very much on the size of the substituent R, the larger groups giving good yields of polyhedral product.

This formation of smaller polyhedra when larger substituents are present is not surprising when the stability of intermediate silanols formed in these reactions toward polycondensation is taken into account. The methylsilanols condense rapidly and simple methylsilanols are difficult to isolate,17 whereas tBuSi(OH)3,20 and [tBuSi(OH)2]2O21 can both be readily isolated. This bulk effect is clearly consistent with the well-known ability of bulky group to stabilize small rings and low coordination numbers in other areas of organometallic chemistry. Attempts to prepare T4 derivatives by using intramolecular hydrolysis/condensation reactions of a suitable precursor, such as MeSi(OSiMeX2)3 (X=Cl or OEt), already containing the four silicon atoms, have also been unsuccessful leading only to T8 POSS derivatives being isolated.22 Adamantanoid structures of Si with larger chalcogens and small substituents have however, been known for many years, the reaction between alkylsilanes or alkyltrichlorosilanes and H2S or H2Se giving, for example, Si4S6Me4 or...