Second negative frequency

From: "Ivan" <ivanhome (- at -) stonline.sk>
Subject: Second negative frequency
Date: Mon, 12 Feb 2001 00:42:42 +0100

Dear CCL'ers
 I am working with programme GAMESS and an present time I am facing some
 problems with TS.I am simulating H-transfer from NH3 bonded to the rest of
 molecule to another "free" NH3 forming "free"NH4.
 In TS there are two imaginary frequencies ( 1. - value about 975, that
 correspond to H-transfer and 2.- value about 36 )  and  can't get rid of
 this second little negative frequency. Firstly I tried to use keywords
 PURIFY in $STATPT and $FORCE groups to purify hessian matrix but with no
 results. Then I tried several approachs to find better TS, but all resulted
 to the same TS. And IRC calculations in both directions are correct, because
 according to IRC's this TS connects appropriate minimas on both side of the
 hill.Finally I wanted to to perform IRC folowing the second negative
 frequency, but GAMESS did not accept the specified frequency ( in $IRC
 group ).
 I would very appreciate any help,advice or trick.
 Stanislav Ivan
 student