754makes it clear that the mass-velocity term is responsible for both the relativistic decrease in re and the increase in De, a fact first discovered by Ziegler, Snijders, and Baer-ends14 in their relativistic X a calculations utilizing the operator described in the Introduction, and confirmed by Snijders and Pykko15 via single-center expansion DHF calculations. As the molecule forms the bond, the kinetic energy initially decreases, but near equilibrium it is significantly larger than in the atoms, and the mass-velocity term is therefore stabilizing. The one-electron Darwin term is of opposite sign but smaller in magnitude, and so the relativistic decrease of the kinetic energy dominates the correction. The present work thus provides yet more confirmation for the view espoused by Ziegler et al. that the bond contraction is due to the p4 term, and not (in this series) dominated by an orbital contraction effect.(15) J. G. Snijders and P. Pyykko, Chem. Phys. Lett., 75, 5 (1980).
75, 1 (1980).The results presented here suggest that the RPTl approach can provide a reasonably accurate, yet economical, alternative to DHF calculations. The fact that good agreement was obtained with the DHF results without the inclusion of the spin-orbit operator is probably related to the dominance of the H( ls)-Ag(5s) bonding interaction; those bonding situations where p or d shells of the heavy atom play a role will no doubt require the introduction of the spin-orbit operator as a perturbation. In order to refine the present approach for -the first two transition series, or to extend the method to the third row, one must find a way around the p4 divergence. Contracting the inner shells does not prevent the collapse; it just makes it somewhat less dramatic. At present, it appears that the simple expedient utilized by Cowan and Griffin in their atomic work, i.e., forcing the wavefunction to behave properly near the nuclei, may be the simplest and most easily implemented approach.Registry No. Silver, 7440-22-4; silver hydride (AgH), 13967-01-6.Structures of vanadium oxide supported on Ti02 have been investigated by using the rectangular pulse technique coupled with various physicochemical measurements including X-ray diffraction and IR, ESR, and UV-visible spectroscopy. In situ I R spectra of adsorbed ammonia have also been measured to investigate the acid properties of the catalysts. It has been found that modification of the TiOz support-anatase, rutile, or a mixture of anatase with rutile-does not affect significantly the structure of the supported vanadium oxide catalyst. The V=O species or the (010) face of Vz05 has been selectively exposed on the surface irrespective of the kind of Ti02 support. When the Vz05 content increases to 5 mol %, the surface of Ti02 is covered by 1-3 layers of Vz05 lamellae; however, the TiOz surface is not completely covered by V205 but it is exposed to the catalyst surface. There are both Bronsted and Lewis acid sites on the Ti02 surface, while only the Bronsted acid site is formed on the unsupported V205...