From owner-chemistry@ccl.net Sun Dec 3 11:15:01 2006 From: "Wai-To Chan chan]|[curl.gkcl.yorku.ca" To: CCL Subject: CCL:G: Diradicaloid Transition State Message-Id: <-33160-061203111114-4681-vd0m9J3fYrN9dpxmwxDzAQ- -server.ccl.net> X-Original-From: Wai-To Chan Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii Date: Sun, 3 Dec 2006 11:16:51 -0500 (EST) MIME-Version: 1.0 Sent to CCL by: Wai-To Chan [chan(-)curl.gkcl.yorku.ca] <<<<<<<<<<<<<<<<<<<< I am trying to reproduce some work from the literature. It is the isomerization of vinylcyclopropanes which has been shown to proceed through a diradicaloid transiton state. The authors indicate they find the transition state using HF UB3LYP/6-31G* in Gaussian 94. I have tried everything to reproduce this transiton structure with no avail. I have tried GAMESS, Gaussian03 and Spartan all do not converge on the transition structure when using UHF. In GAMESS I can get RHF to work however the energy is a good bit lower than that of the UHF reported. I have tried UHF and ROHF, as singlets and triplets (and for kicks as a doublet). I have optimized many transition structures in the past (although only RHF) and i understand how important the initial guess is and some of the other intricacies (i.e. start with good hessian etc). The authors state in a reference "Transition structures initially obtained by a closed-shell wavefunction, proved to be UHF unstable. Upon release of the spin restriction constraint, they collapsed to the UB3LYP geometries." I must be missing something here. How do I release the spin restriction constraint? Doesn't release of the spin restiction constraint just mean UHF? Thank you. >>>>>>>>>>>>>>>>>> It is probably a case of ambiguity in usage of English. What "release of spin restriction constraint" means is to relax the restriction of the spatial orbitals of the alpha and beta electrons to be identical. This can be done for a closed-shell system by switching from RHF to UHF with guess=mix and/or stable=opt. Similar procedure for DFT. Whether the transition state you are trying to locate exists on a UB3LYP, UHF or any UHF-based PES I can't tell. According to the authors of the work you are trying to duplicate, the TS was initially obtained by a closed-shell wavefunction. If this TS can be located with higher level method like RHF-MP2 and RHF-QCISD I won't be worried about not being able to locate it with UHF and UB3LYP. It would seem logical to switch to a UHF or UHF-based methods to locate a stationary in the region of a PES where the wavefunction exhibits a RHF---->UHF instability. This is indeed the case for the calculation of the dissociation of a closed-shell X-Y molecule to two radical fragments X and Y. You have to calculate the energies of X and Y with UHF or UHF-MP2 and so forth. Of course you would only use UHF-based methods for pure diradicals. On the other hand UHF-based methods are not necessarily more accurate than RHF-based methods for diradicaloids or partial diradicals. In fact the errors of the high-level UHF-based methods such as UCCSD(T) are significantly higher than RCCSD(T) in the potential curve for the bond dissociation of a closed-shell molecule in the vicinity of the point where the RHF--->UHF instability kicks in. You may however run into difficulties if the PES features other diradical intermediates. In that case the RB3LYP energy diradicaloid TS might not be accurate relative to the UB3LYP energies of the intermediates. I won't be concerned if it is only the energy level of the TS relative to the closed-shell reactant that has to be calculated. Wai-To Chan From owner-chemistry@ccl.net Sun Dec 3 15:08:01 2006 From: "Cory Pye cpye{:}crux.smu.ca" To: CCL Subject: CCL:G: Diradicaloid Transition State Message-Id: <-33161-061203082623-12424-G37svMopWrxRz6tPpuWe3Q|server.ccl.net> X-Original-From: Cory Pye Content-Type: TEXT/PLAIN; charset=US-ASCII Date: Sun, 3 Dec 2006 09:02:46 -0400 (AST) MIME-Version: 1.0 Sent to CCL by: Cory Pye [cpye|*|crux.smu.ca] Hello, If you try to do a calculation using a UHF wavefunction on such a beast, typically the wavefunction reverts back to the RHF solution. In Gaussian, first do a single point calculation at your best geometry using something like RB3LYP/XXX Stable=Opt SCF=(Save,Nosymm) . This checks the stability of the wavefunction, and optimizes it according to any instabilities found. The Save/nosymm will force the optimized, possibly unsymmetrical or UHF wavefunction to be saved on the checkpoint file. Sometimes QC and MaxCyc are needed in the SCF. Then, in the next step do something like --Link1-- %blablabla #N UB3LYP/XXX Opt=(ts,CalcFC) geom=check guess=read scf=(save,nosymm) This starts off with the proper guess from the previous job, evaluates the Hessian and goes on. Check the energies at the beginning of this job and the end of the last job to make sure that they match. -Cory On Sat, 2 Dec 2006, Carl LeBlond carl.leblond[]iup.edu wrote: > The authors state in a reference "Transition structures initially obtained by a closed-shell wavefunction, proved to be UHF unstable. Upon release of the spin restriction constraint, they collapsed to the UB3LYP geometries." I must be missing something here. How do I release the spin restriction constraint? Doesn't release of the spin restiction constraint just mean UHF? > Thank you. > > Carl > ************* ! Dr. Cory C. Pye ***************** ! Associate Professor *** ** ** ** ! Theoretical and Computational Chemistry ** * **** ! Department of Chemistry, Saint Mary's University ** * * ! 923 Robie Street, Halifax, NS B3H 3C3 ** * * ! cpye,crux.stmarys.ca http://apwww.stmarys.ca/~cpye *** * * ** ! Ph: (902)-420-5654 FAX:(902)-496-8104 ***************** ! ************* ! Les Hartree-Focks (Apologies to Montreal Canadien Fans) From owner-chemistry@ccl.net Sun Dec 3 23:57:01 2006 From: "Agalya Govindasamy agalya81::gmail.com" To: CCL Subject: CCL: Oscillator strengh for excitation between different multiplicities Message-Id: <-33162-061203235559-6342-zgh25tqwrZSEExbjTO2RLw() server.ccl.net> X-Original-From: "Agalya Govindasamy" Date: Sun, 3 Dec 2006 23:55:58 -0500 Sent to CCL by: "Agalya Govindasamy" [agalya81]=[gmail.com] Dear CLL users, Can anyone tell me, is there any softwares available to calculate oscillator strength for excitations between two different mulitplicites. For example i want to calculate excitation from ground state septet to excited state quintet multiplicity. Is there any theories and softwares available to do this kind of calculation, which includes spin-orbit coupling. Its not possible by TDDFT? Thanks in advance, Agalya