On the use of a variable-step method for the computation of diatomic eigenvalues near dissociation: The Lennard-Jones potential (Q583858)

The authors discuss the application of the variable-step method introduced by the first two authors [J. Comput. Phys. 77, 501-512 (1988; Zbl 0654.65064)] for the computation of diatomic eigenproblems. Their variable-step method, which can be seen as a numerical realization of the Picard iteration is aimed to treat with the one-dimensional Schrödinger equation \(y''(r)=(U(r)-E)y(r)\), where U(r) is a given potential function. Through numerical experiments they argue that the method considerably reduces the CPU time, especially in the case for the Lennard-Jones potential near dissociation.