Firefly and PC GAMESS-related discussion club

 
Learn how to ask questions correctly  
 
 

 

Alex Granovsky
gran@classic.chem.msu.su

this is the known problem. The reason of this problem
is that in the case of Ga atom, Huckel guess assumes that ECP
removes 10 core electrons. This is correct for small core ECPs
while you are using the large core one.

Use guess=hcore instead. In the case of poor convergence,
converge cationic system first, then the original one using
guess=moread. In addition, you may give a try to "combo" utility
(http://classic.chem.msu.su/gran/gamess/combo.html)

Hope this helps.

Kind regards,
Alex Granovsky







On Sat Mar 24 '12 10:57pm, Vladislav Ivanistsev wrote
-----------------------------------------------------
>I'm trying to model a simple system Ga-OH2 using Lanl08 ECP for Ga and 6-31G for H and O. All basis sets are taken from https://bse.pnl.gov/bse/ Similar calculations for Hg-OH2 and Bi-OH2 have finished successfully. However, for Ga-OH2 the output states:

> CONFUSION IN -HUCKEL- WITH ATOM=   1 NUCZ=  31 KIND= 1
> NCORE,NVAL=  10   2 NDVAL=   1   3
> NCECP=   5 NOTECP= F
> LLIM,IULIM=   1   4 NEXTRA= -10 NEXTRA SHOULD BE 0

>What could be the reason for the error?

[ This message was edited on Fri Mar 30 '12 at 5:18pm by the author ]