The covalent bond in silicon
A detailed electron distribution study of the covalent bond in silicon is discussed. The study is based on the strategy employed earlier to define the covalent bond in diamond, and estimates of the bonded-atom scattering power, including results for 222, that have been obtained in three different X -ray experiments at room temperature are considered. Two of these involve the conventional approach via intensity measurement, that of Gottlicher & Wolfel (G. W. ) on powders and that of DeMarco & Weiss (D. M. W. ) on perfect single crystals, while the third by Hattori, Kuriyama, Katagaw a & Kato (H. K. K. K. ) involves the measurement of spacing in X -ray Pendellösung fringes observed in wedge-shaped perfect crystal specimens and their interpretation by the dynamical theory of Kato. It is shown that only the results obtained by H. K. K. K. are capable of detailed analysis in term s of the criteria now available from the earlier study of diamond, from which we conclude that the fringe-spacing method has yielded data for the low-angle reflexions which are superior to the data obtained m re conventionally by G. W. and D. M. W. The covalent bond in silicon is described comprehensively by two non-spherical components as in diamond. The major (antisymmetric) component involves the radial function F 3 ( r ) = 1⋅11 r 2 exp (─ 0⋅88 r 2 ), and the minor (cubo-centrosymmetric) the function G 4 ( r ) = ─ 0⋅32 1 r 2 exp (─0⋅88 r 2 ). The origin of these components lies in an electron redistribution of the spherical unbonded atom involving 0⋅127 electron per bond, and this leads to a peak of + 0⋅25 e/Å 3 at the mid-point of the covalent bond. The consequences of this study are considered in relation to future attempts to examine aspects of electron distribution in atoms of different weight. The suitability of diamond and silicon as standards for scaling intensity data to the strict absolute basis so essential to such studies is also noted.