Indo calculations of spin-hamiltonian constants in transition-metal complexes. Influence of structural distortions and adduct formation on g-tensor of copper (II) bis-acetylacetonate

Conclusions

On the whole, the results of the work reported here demonstrate the adequacy of the wave functions of the INDO method for use in calculating the g-tensor in transition-metal complexes. Along with this, they show that certain elements of the calculational scheme previously proposed for analogous purposes in radicals [1] do require modification. For example, it is necessary in certain cases to account for configuration interaction (CI). In addition to the cases already mentioned in this article in which Eq. (1) is inapplicable because of neglect of CI, let us note that an accounting for CI is necessary in principle in systems with S> 1/2. Also, corrections to the g-factors in transition-metal complexes may be so large in some cases that it becomes necessary to diagonalize the matrix of the Hamiltonian, including spin and nonspin interactions. In the most natural form, the two modifications are combined when they are introduced simultaneously.

We have repeated many of the calculations using the system of parameters of [11], and have also repeated calculations within the framework of the CNDO method. In all cases, in the calculation of the g-tensor, the INDO method in the original parametrization [2] proved to be preferable.