From owner-chemistry@ccl.net Sun Jun 16 11:12:01 2013 From: "Errol Lewars elewars__trentu.ca" To: CCL Subject: CCL:G: TS optimization help! Message-Id: <-48830-130616111012-2871-FwP/soB0O1X1jLrZz765Yg ~ server.ccl.net> X-Original-From: Errol Lewars Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=ISO-8859-1; format=flowed Date: Sun, 16 Jun 2013 11:10:21 -0400 MIME-Version: 1.0 Sent to CCL by: Errol Lewars [elewars:-:trentu.ca] 2013 June 16 "In order to obtain the TS structure, I first tried QST2 using optimized reactant and product structure. Didn't work out". QST2 is said to work better if instead of using as input the *optimized* reactant and product you distort each toward the other, i.e toward the transition state. You might also try taking one of the two minima and doing a scan, altering a geometric parameter in the TS direction; then submitting the highest point on the IRC-like curve (assuming you get such a curve) to a TS optimization/freq. E. Lewars == Joseph Takayama sjtakayama(~)gmail.com wrote: > Sent to CCL by: "Joseph Takayama" [sjtakayama*gmail.com] > Hi Everyone, > > I'm trying to reproduce the TS structure optimization which is already > reported in the paper. > > First, I prepared the structures of the reactant and the product from the > scratch and optimized them using the same basis functions as reported. > Obtained optimized structures seems similar to those reported and energy > difference between reactant and product also matched to the value in the > paper. > > In order to obtain the TS structure, I first tried QST2 using optimized > reactant and product structure. Didn't work out. > > Some key bond lengths of the TS structure are also reported in the paper. > So next, I freezed 2 key bond lengths using modredundant option and optimized > the rest. > Obtained structure seems quite similar to the TS structure in the paper. > >> From that structure, I tried Opt=(ts, calcfc, noeigentest), with or without > "tight" option and Opt=(QST3, calcfc). > However, every time I tried, optimization goes to the structure far away from > the TS structure that I want. > > Is there any tips to lead the optimization to the TS structure that I want? > Or is there any options that prevents structure to go too far away from the > initial structure? > I'm using Gaussian 09. > > Any comment would be appreciated! > > Joseph Takayama > sjtakayama_-_gmail.com> >