J. Phys. Chem. 1983, 8 7 , 750-754
750
Ail-Electron Relativistic Calculations on AgH. An Investigation of the Cowan-Griffin Operator in a Molecular Species Richard L. Martin Theoretical Division. T- 12, MsJ569, Los Alamos National Laboratory, Los Alamos, New Mexico 87545 (Received: October 4, 1982)
The Cowan-Griffin relativistic operator, which retains only the mass-velocity and one-electron Darwin terms of the Breit-Pauli Hamiltonian, has been shown to provide a remarkably good approximation to the more sophisticatedDirac-HartreeFock (DHF) approach for many atomic systems. In this paper, the result of applying the Cowan-Griffinoperator in first order to a nonrelativistic self-consistent-field(SCF) wave function expanded in a basis of Gaussian functions is reported for Ag and AgH. Comparisons are made with previously reported numerical wave functions for Ag and with the recent Slater-type-orbital expansion of the DHF wave function for AgH reported by Lee and McLean. The results of the present method are in excellent agreement with the DHF work and suggest that the approach might provide an attractive alternative to the DHF approximation. Further support for the contention of Ziegler, Snijdeh, and Baerends that the relativistic bond length contraction in AgH is caused by the mass-velocity term, as opposed to an atomic orbital contraction, is also presented.
I. Introduction Lee and McLeanl have recently reported the results of DiraeHartree-Fock (DHF) calculations in a Slater-typeorbital (STO) basis for AgH and AuH. These first accurate DHF results represent a significant advance in relativistic capabilities for molecular species. In order to correctly treat the relativistic operators, however, the DHF approach requires a much larger basis than a nonrelativistic approach, and significantly more computational effort. The possibility that a simple first-order perturbation-theory approach utilizing the nonrelativistic wave function and easily computed one-electron relativistic integrals might provide an adequate alternative in certain instances motivated the work described here. The usual starting point for perturbation-theory approaches is the Breit-Pauli Hamiltonian, which is an expansion in powers of ( u / c ) ~of the Dirac-Fock Hamiltonian augmented by the Breit retardation term.2 In the nonrelativistic limit, if terms of order ( u / c )are ~ retained, the Hamiltonian takes the form Pj2 ;K=
Ej
2m
1
e?
ZAe?
- j r, A +- j
