Ceilidh ... Re: HESSIAN NaN,

Firefly and PC GAMESS-related discussion club

Re: HESSIAN NaN,

First time you got NaN is the first movement of the second atom (copper). In this step, SCF failed to converge again, which means that your "molecule" (which is in fact a fragment of a crystal lattice and, therefore, can hardly exist on its own) is unstable both geometrically and electronically. Since one of the necessary steps in hessian calculation gave NaN, so, the final answer is NaN, too. Even if you force the SCF convergence on each step of the hessian calculation, you'll most probably find several imaginary frequencies, because this structure is unlikely to correspond to any real (i.e., stable) molecule.

Remember that small clusters, especially not properly capped or embedded, cannot simulate bulk solid. Try capping peripheral copper atoms with something resembling their neighbors in the crystal lattice.

On Wed Jan 13 '10 3:17pm, luxiaohua wrote
-----------------------------------------
>When I caculate hessian about some cluster. it shows NaN. output below:
>what is wrong?

Wed Jan 13 '10 4:22pm

This message read 908 times