From chemistry-request@ccl.net Thu May 16 16:42:46 1991
Date: 16 May 91   15:53 LCL
From: PA13808%UTKVM1.BITNET@OHSTVMA.ACS.OHIO-STATE.EDU
To: CHEMISTRY@ccl.net
Subject: BITNET mail follows
Status: R

Topic G86 SPLIT VALENCE EIGENVALUES
   The request on this topic does not make much sense to me.the a 6-31G s
set for carbon has no 3s or 3p component etc it only has an inner
sp component and an outer sp component. To see what basis set you have yo
should insert the request GFPRINT.Also see Poples book or one of the rece
recent reviews by D. FELLER and E.S. DAVIDSON especially the one in
REVIEWS IN COMPUTATIONAL CHEMISTRY ED BY K.B. LIPKOWITZ AND D. B. BOYD
published by VCH. Possibly  Mr Chin is using an unconventional terminolog
y for the principsl quantum number. To get ds into S one has to use a*
and then you will only get a ds for a 3d shell. to put in higher ones
you have to use the MASSAGE OPTION OR put in a new basis from scratch
using GEN option.Remember when using Gaussians one does not specify the p
prin. quantum number and one identifies the shell by the eigenvalue
and by relating it to the atomic shell. Perhaps you are misinterpreting
this relationship. To clarify run some atomic calc. on G 86 with your
basis and compare them with standard tables of atomic energy values.
   John E. Bloor (PA13808 at UTKVM1 on CMS)