The structure of tetrasilylhydrazine, (SiH,),N2, has been determined by the sector microphotometer method of electron diffraction. The data are consistent with planar Si,NN groups, and a dihedral angle of 82-5 f 0.8"; the deviation of this angle from 90" may possibly be explained by torsional effects. The Si-H, N-Si, and N-N bond lengths are 1.487 f 0.01 4, 1 a731 f 0.004. and 1.457 f 0-01 6 a respectively; the Si-N-Si angle is 1 29.5 f 0.7" and the N-Si-H angle is 109.0 f 1.4".A STUDY of the vibrational spectra of tetrasilylhydrazine has suggested that the most likely conformation of the molecule in the vapour phase is of D,, symmetry, with planar co-ordination about nitrogen. In view of this and the interesting structural features found in a number of other di~ilazanes,~-~ we have undertaken an electron diffraction study of tetrasilylhydrazine.
EXPERIMENTALTetrasilylhydrazine was prepared by a modification of the original method :anhydrous hydrazine vapour was blown stepwise by a stream of dry nitrogen into a 5 1. bulb containing an excess of silyl bromide. The product was purified by repeated fractional condensation a t -64" ; the i.r. and n.m.r. spectra revealed no impurities. Photographs were taken on the Balzers KD G2 gas diffraction apparatus at the University of Manchester Institute of Science and Te~hnology,~ and the photographic intensities transferred to punched paper tapes using an automated Joyce-Loebl microdensitometer. The wavelength was determined, from measurements of the accelerating voltage and from the diffraction pattern of powdered thallium(1) chloride, to be 0.05582 Jz 0*00003 %L for the data obtained a t a camera distances of 50 cm. and 0.05569 f 0.00003 A for the data a t 100 and 25 cm. Data from 100, 50, and 25 cm. camera distances gave an overall range of 1.2-29.2 A-1 in the scattering variable s. In the data reduction and least-squares refinement our usual techniques were employed: 358 all computations were carried out using the Cambridge University ' Titan ' computer.Refinement.-The refinement is best discussed with reference to the experimental radial distribution curve, P(Y)/Y, depicted in Figure 1. The three bonded peaks N-N, Si-N, and Si-H form a barely resolved complex: in addition there are three further strong peaks a t 2.70, 3.13, and 3.67A. These can be assigned to Si * N, Si * * (N) -Si and Si 9 . (N-N) -* Si, leading to an Si-N-Si angle of 129" and an Si-N-N angle of 115.5". When the planarity a t nitrogen had been established, it was apparent from the existence of essentially only one Si ---(N-N) -Si distance that the dihedral angle, i.e. the angle between the two Si,N planes, was close to 90'.The molecular model (of D, symmetry) assumed that the silyl groups lie on local axes of three-fold symmetry along
