tert-Butyldimethylsilanol hemihydrate

The crystal structure of the title compound, C6H16OSi·0.5H2O, reveals an asymmetric unit containing two molecules of the silanol and a single water molecule. There is evidence of hydrogen bonding between the three molecules in the asymmetric unit. The H atoms of the silanol OH groups and the water H atoms are each disordered equally over two positions.


Comment
The stucture of the title compound tert-butyldimethylsilanol hemihydrate is shown below (Fig. 1, 2); dimensions are available in the archived CIF. This compound has previously been characterized by gas-phase electron diffraction of both the free silanol and its hemihydrate (Lickiss et al., 1995). It has also been structurally characterized within lanthanoid complexes (McGeary et al., 1991) and in recent structures of complexes with several main group and transition metals (for examples see Mansfeld, Mehring & Schürmann, 2005;Mansfeld, Schürmann & Mehring, 2005, Veith et al., 2006. However direct crystallographic characterization of the silanol has hitherto remained elusive. tert-Butyldimethylsilanol hemihydrate was isolated in crystalline form during the synthesis of biomimetic ligands for iron-mediated hydrocarbon oxidation (Krall et al., 2005, Barry & Rutledge, 2008. The title compound was obtained as the side product of reactions to prepare a compound incorporating the tert-butyldimethylsilyl ether as a protecting group (2-(bromomethyl)-6-((tert-butyldimethylsilyloxy)methyl)pyridine).
The assymetric unit contains two molecules of the silanol and one water molecule, linked by hydrogen bonding. The hydrogen atoms of the water molecule and the silanol O-H groups are disordered between two alternative occupancies ( Fig. 2).

Experimental
The title compound crystallized serendipitously as colourless needles from a sample of the silyl ether 2-(bromomethyl)-6-((tert-butyldimethylsilyloxy)methyl)pyridine (isolated as an oil from a pentane:ether solvent mixture), upon standing at room temperature for several weeks.
Large crystals of the silanol were obtained (1.00 × 1.00 × 0.80 mm) and data was collected from a crystal at the upper size limit of the beam used. Moreover, the crystals of tert-butyldimethylsilanol hemihydrate that have been isolated exhibit remarkably low density (0.996 g cm -3 ), a property that may be linked to the high volitility of this compound.

Refinement
A full sphere of the reciprocal space was scanned by φ-ω scans. Hydrogen atoms of the silanol molecules were added at calculated positions and refined using a riding model. C-H distances were assumed to be 0.98 Å, O-H distances to be 0.84 Å. The water protons were located in the difference Fourier map. The distance of these protons to the oxygen atom was restrained to be 0.84 Å using the DFIX command. In the same way the H-O-H angles were restrained to be 114°, the value to which a preliminary refinement of one component converged. U iso (H) = 1.5 U eq (carrier) for all H atoms.
The site occupation factor of the disordered hydrogen atoms was fixed to 0.5. Attempts to refine the occupation factors were unsuccessful. However, electron densities in the difference Fourier map suggest a fairly even distribution between the two disordered parts.
Discrepancies between the expected and reported values of the maximum and minimum transmission (T max /T min ) are thought to have arisen from the large size of the crystal relative to the beam, and because the crystal mount has given rise to some absorption during data collection. However this is not thought to impact significantly on the dataset given that there is an almost fourfold redundancy with the collection of a full sphere, and the capacity of SADABS to handle data collected from large crystals (Görbitz, 1999, Sheldrick, 2000. Fig. 1. The molecular structure of the title compound with atom labels and thermal ellipsoids drawn on the 50% probability level. Carbon atoms are shown in grey, the silicon in black and the oxygen in blue.   Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating Rfactors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.