From aps501(+ at +)anugpo.anu.edu.au  Wed Jun  7 20:17:55 1995
Received: from anugpo.anu.edu.au  for aps501 -8 at 8- anugpo.anu.edu.au
	by www.ccl.net (8.6.10/930601.1506) id UAA29740; Wed, 7 Jun 1995 20:06:05 -0400
Received: from huxley.anu.edu.au (aps501%!at!%huxley.anu.edu.au [150.203.2.12]) by anugpo.anu.edu.au (8.6.12/8.6.12) with ESMTP id JAA12718; Thu, 8 Jun 1995 09:58:21 +1000
From: Anthony P Scott <Anthony.Scott $#at#$ anu.edu.au>
Received: (aps501-!at!-localhost) by huxley.anu.edu.au (8.6.12/8.6.12) id KAA07312; Thu, 8 Jun 1995 10:04:40 +1000
Date: Thu, 8 Jun 1995 10:04:40 +1000
Message-Id: <199506080004.KAA07312 -x- at -x- huxley.anu.edu.au>
To: CHEMISTRY-!at!-ccl.net, sanja-!at!-indigo.irb.hr
Subject: Re:  CCL:scaling MP2 freq.


The original poster asked about scaling factors to be applied to mp2/6-31g**
calculated frequencies. He asked if our mp2/6-31g* scaling factor is 
applicable and if the scalinf factor varies with basis set.

Our worl in this area, which is nearing completion, indicates that the mp2 
level of theory is not particularly sensitive to basis set in predicting
frequencies. While the 6-31g** basis set was not studied, we have data for the
6-311g** basis set. The appropriate scaling factor for the mp2 level of theory
with this basis set is 0.9496 (with about the same overall RMS error as wefound for the mp2/6-31g* level of theory).

We have actually studied a very extensive range of methods (including DFT) with
various basis sets. If readers would like further information then please
contact me.

Hope this is of some assistance,

Kind Regards,

Anthony P. Scott
Research Officer
Computational Chemistry Group
Research School of Chemistry
Australian National University
Canberra, ACT, Australia