From chemistry-request@ccl.net Fri Oct 17 00:50:30 2003 Received: from nankai.edu.cn (sun.nankai.edu.cn [202.113.16.21]) by server.ccl.net (8.12.8/8.12.8) with ESMTP id h9H4nodH014105 for ; Fri, 17 Oct 2003 00:49:54 -0400 Received: (eyou send program); Fri, 17 Oct 2003 12:50:31 +0800 Received: from unknown (HELO 192.168.0.101) (unknown@128.195.87.184) by 202.113.16.21 with SMTP; Fri, 17 Oct 2003 12:50:31 +0800 Date: Thu, 16 Oct 2003 21:49:27 -0700 From: John Lee X-Mailer: The Bat! (v1.62r) Reply-To: John Lee X-Priority: 3 (Normal) Message-ID: <1455746254.20031016214927.-at-.joyie.com> To: gaussian.com!csd.-at-.gaussian.com (Cust. Service Doug) CC: chemistry.-at-.ccl.net Subject: Re[4]: CCL:GDIIS and Z-MATRIX In-Reply-To: References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Spam-Status: No, hits=-1.8 required=7.0 tests=EMAIL_ATTRIBUTION,FROM_ENDS_IN_NUMS,IN_REP_TO, QUOTED_EMAIL_TEXT,REFERENCES,REPLY_WITH_QUOTES version=2.55 X-Spam-Checker-Version: SpamAssassin 2.55 (1.174.2.19-2003-05-19-exp) Hi Cust., It does help me a lot. I used Int=ultrafine and opt=calcall, with tight option. although the RMS Displacement still didn't reach the threshold, the RMS force reached. I need to perform a freq calculation to make sure it is a minimum. Now all the distances between neighbor oxygens are the same (4 water cluster). The distances beteen O-H and O...H are also the same. Thank you very much. Best regards John Lee Wednesday, October 15, 2003, 9:13:24 AM, you wrote: CSD> My guess is that it was hard because the Berny algorithm needed CSD> to have the integration quality improved to go along with the tighter CSD> integration criteria and the small forces involved in these clusters. CSD> In such a case I am dubious that GDIIS is really giving you a "better" CSD> structure, but I can believe it is willing to claim the criteria CSD> are met. It deals better with inaccuracies in the gradient and so CSD> I have suggested its use with solvation models and as you can see from CSD> the manual with semi-empirical models. CSD> To use the Z-matrix I would recommend you try a couple of CSD> approaches. First use Int=Ultrafine and see if OPT=Tight does not CSD> behave better. Second, I would suggest doing this as OPT=(CalcFC,Tight) CSD> if you are starting close to an optimized structure. Or possibly CSD> OPT=(CalcAll,Tight) because the force constants here are small and CSD> it is vital to have them correct for the optimization to work. >> >> Hi Cust., >> >> I the simulation of 4-8 water cluster, I want to keep all the >> distances between neighbor oxygen the same. However it was hard to >> converge with B3LYP 6-311++G** and tight if I didn't use GDIIS. If I use GDIIS >> the distances will not be the same. >> >> Thanks for your reply. >> >> John >> >> Tuesday, October 14, 2003, 5:27:52 PM, you wrote: >> >> >> CSD> John, >> >> CSD> Why do you need to use GDIIS? >> >> CSD> Note that this would break symmetry only if you are using the Z-matrix >> CSD> to define symmetry which is not really in the molecule? >> >> >> >> >> >> Hi CCLers, >> >> >> >> Z-matrix optimization only work with Berny algorithm. >> >> What if I have to use GDIIS and still want to control the zmatrix >> >> symmetry. Is there a way to do it? >> >> >> >> >> >> >> >> -- >> >> Best regards, >> >> John >> >> >> >> >> >> >> >> -= This is automatically added to each message by the mailing script =- >> >> To send e-mail to subscribers of CCL put the string CCL: on your Subject: line >> >> and send your message to: CHEMISTRY.-at-.ccl.net >> >> >> >> Send your subscription/unsubscription requests to: CHEMISTRY-REQUEST.-at-.ccl.net >> >> HOME Page: http://www.ccl.net | Jobs Page: http://www.ccl.net/jobs >> >> >> >> If your mail is bouncing from CCL.NET domain send it to the maintainer: >> >> Jan Labanowski, jkl.-at-.ccl.net (read about it on CCL Home Page) >> >> -+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> -- >> Best regards, >> John Lee >>