From raf.bruyndonckx@unifr.ch Fri Mar 20 09:19:03 1998 Received: from siufuxsun03.unifr.ch for raf.bruyndonckx@unifr.ch by www.ccl.net (8.8.3/950822.1) id IAA19015; Fri, 20 Mar 1998 08:29:31 -0500 (EST) Received: from ufper6.unifr.ch [134.21.14.66] by siufuxsun03.unifr.ch with smtp (Exim 1.82 #4) id 0yG1rJ-0000mv-00; Fri, 20 Mar 1998 14:29:17 +0100 Received: from unifr.ch by UFPER6.UNIFR.CH via Pony Express SMTP with TCP (v8.1.1-dmr001); Fri, 20 Mar 98 14:29:16 MET Sender: rbruyn@unifr.ch Message-ID: <35126F2B.311E790E@unifr.ch> Date: Fri, 20 Mar 1998 13:29:15 +0000 From: Raf Bruyndonckx Organization: University of Fribourg X-Mailer: Mozilla 4.02 [en] (X11; I; IRIX 5.3 IP22) MIME-Version: 1.0 To: chemistry@www.ccl.net Subject: Exercise for HFS calculation. Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dear CClers, could anyone suggest me a kind of schoolbook example/exercise for the calculation of the isotropic and/or anisotropic hyperfine tensor using Gaussian94? The idea is that this exercise could be given to a public interested in computational chemistry, but still at a beginning phase. The total preparation, calculation and result analysis should fit into an exercise session of max. 2 hours. Myself, I was thinking of an organic radical of, let's say, max. 20 atoms, performing a single point and thus calculating the hyperfine tensor, preferably with the conclusion that they are close to experiment. (Is it as easy as this ?) Since I don't have much experience in this field, could you also give me any remarks and comments on the way this can be calculated (with Gaussian94). I have been screening the CCL archive library and came up with a few messages indicating that this can be done using the following options in the input : #P UHF/6-31g** PROP=efg IOP(6/17=2,6/26=4) Also MELDF package should be able to calcutate this properties. I have access to this package but never worked with it, so ... any comment is welcome. Thanks, Raf -- Raf Bruyndonckx Department of Inorganic and Analytical Chemistry Univ Fribourg (Suisse) Perolles Tel: ++41 26 300 87 49 Fax: ++41 26 300 97 38 mailto:Raf.Bruyndonckx@unifr.ch http://sgich1.unifr.ch/ac/phd/rbruyn/ From gabriele@indigo.chem.polimi.it Fri Mar 20 13:19:05 1998 Received: from indigo.chem.polimi.it for gabriele@indigo.chem.polimi.it by www.ccl.net (8.8.3/950822.1) id MAA20170; Fri, 20 Mar 1998 12:36:51 -0500 (EST) Received: from indigo (localhost [127.0.0.1]) by indigo.chem.polimi.it (950413.SGI.8.6.12/950213.SGI.AUTOCF) via SMTP id SAA03082; Fri, 20 Mar 1998 18:37:03 +0100 Sender: gabriele@indigo.chem.polimi.it Message-ID: <3512A93F.41C6@indigo.chem.polimi.it> Date: Fri, 20 Mar 1998 18:37:03 +0100 From: Gabriele Valerio Organization: Politecnico di Milano X-Mailer: Mozilla 2.0S (X11; I; IRIX64 6.2 IP28) MIME-Version: 1.0 To: mforster CC: chemistry@www.ccl.net Subject: Re:summary intermolecular interactions References: <199803171529.PAA05126@chalsig.nibsc.ac.uk> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dear Mark, and CCL netters, I am sorry for sending a late reply to your question about ab initio calculation of the intermolecular interaction, after your recently posted summary. Gianinetti, Raimondi and Tornaghi have recently adapted the Roothan equations to exclude the basis set superposition error from intermolecular interactions (ref. Int. J. Quantum Chem. vol. 60, 157-166 (1996)). The calculation produces BSSE free Hartree-Fock energy for interacting fragments, without any a posteriori corrections. The computational cost is not appreciably higher than an ordinary HF. The method has benn implemented in latest version of GAMESS-USA by Famulari. In some cases the BSSE error can be larger than the correlation energy. Best wishes Guido Raos and Gabriele Valerio -- Gabriele VALERIO Politecnico di Milano Tel: ++39-(0)2-23993051 Dipartimento di Chimica Fax: ++39-(0)2-23993080 via Mancinelli, 7 20131 Milano (ITALY) mailto:gabriele@indigo.chem.polimi.it From daizadeh@kappa.ucdavis.edu Fri Mar 20 19:19:06 1998 Received: from kappa.ucdavis.edu for daizadeh@kappa.ucdavis.edu by www.ccl.net (8.8.3/950822.1) id TAA22366; Fri, 20 Mar 1998 19:00:08 -0500 (EST) Received: (from daizadeh@localhost) by kappa.ucdavis.edu (950413.SGI.8.6.12/950213.SGI.AUTOCF) id QAA24240; Fri, 20 Mar 1998 16:06:00 -0800 From: "Iraj Daizadeh" Message-Id: <9803201606.ZM24238@kappa.ucdavis.edu> Date: Fri, 20 Mar 1998 16:05:59 -0800 X-Mailer: Z-Mail (3.2.3 08feb96 MediaMail) To: chemistry@www.ccl.net Subject: One-electron Integral (core hamiltonian) Cc: guo@indigo.ucdavis.edu, leee@indigo.ucdavis.edu, antony@indigo.ucdavis.edu, stuchebr@indigo.ucdavis.edu Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Hello. (As you are well aware, I am slowly but surely dissecting my way through G94 code.) My problem is this: I have written a line or two in my link that prints out the so-called core hamiltonian (1 electron integral) as follows: C Call FileIO(2,-IRwH,NTT,V,0) do 214 i=1,NTT If(Abs(V(i)).lt.Thresh) V(i) = Zero 214 write (IH,2010) V(i) C Here, IRwH is set to 515, and thresh = 1e-6. The results for H2 (Hf/sto-2g) at 8 ang. separation is: -0.528103461738361310000000000000E+00 0.000000000000000000000000000000E+00 -0.512985653829729940000000000000E+00 In G94 standard output file, the core hamiltonian has the values: ****** Core Hamiltonian ****** 1 2 1 -0.520545D+00 2 0.000000D+00 -0.520545D+00 It seems that G94 takes the average of the matrix elements in the one-electron hamiltonian. Is this true and if so why? Thank you in advance. Iraj. -- Iraj Daizadeh Department of Chemistry University of California One Shields Ave. Davis, CA 95616-5295 Phone: 530.754.8695 Fax: 530.752.8995 email: daizadeh@kappa.ucdavis.edu From takanori.kanazawa@pharma.novartis.com Fri Mar 20 21:19:16 1998 Received: from i3se0101.is.chbs.ciba.com for takanori.kanazawa@pharma.novartis.com by www.ccl.net (8.8.3/950822.1) id VAA23043; Fri, 20 Mar 1998 21:08:24 -0500 (EST) From: Received: from mailhub by i3se0101.is.chbs.ciba.com; (5.65v3.2/1.1.8.2/12Mar96-0208PM) id AA10222; Sat, 21 Mar 1998 03:07:46 +0100 Received: from nts1.novartis.com ([192.168.50.131]) by mta3.is.chbs (8.8.5/8.8.5) with SMTP id DAA29302 for ; Sat, 21 Mar 1998 03:08:14 +0100 (MET) Received: by nts1.novartis.com(Lotus SMTP MTA SMTP MTA v1.1.04 (495.1 10-24-1997)) id 412565CE.000BB055 ; Sat, 21 Mar 1998 03:07:40 +0100 X-Lotus-Fromdomain: PH@N1 To: CHEMISTRY@www.ccl.net Message-Id: <492565CE.0009D3C0.00@nts1.novartis.com> Date: Sat, 21 Mar 1998 11:07:33 +0900 Subject: Summary:G94 geom. opt. with constraints Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Dear CCLers, Some time ago, I posted a question concerning geometry optimization of a dimeric molecule, applying dihedral constraints. I've got responses from the following people. Douglas J. Fox Stefan Fau Mark A. Zottola John Waite Krzys Radacki Guido Germano Larry Cuffe Victor I really appreciate their kindness. Currently, I'm testing some of the methods suggested by them and it will take time before I can find a best solution. So here I just summarize their suggestions. Upon finding the best solution, I will post here the method. My original question was >>This time I have a question about geometry optimization of a dimeric molecule. >>Each monomer unit is an almost planar molecule except one methyl group. >>I did geometry optimization of the dimer by fixing the dimer molecule on a plane >>(dihedral angles between the monomer units is fixed to 0 or 180). >>In the course of the geometry optimization, intermolecular geometry begins to deviate >>from planarity and finally the optimization job stops with the following error messages >> --------------------------------------------------------------------------- ------- >>Convergence failuer -- run terminated. >>Error termination via Link1e in /home/Gaussian/g94/l502.exe. >>Job cpu time: .... >>............ >> --------------------------------------------------------------------------- ------- >> >>Before this termination, messages like the following are repeated >>------------------------------------------------------------------ >>Energy Rises -- skip Quadratic search. >>Quartic linear search produced a step of -0.99578. >>....... >>....... >>Iteration 10 RMS(Cart)=************ RMS(Int)=************ >>Iteration 11 RMS(Cart)=************ RMS(Int)=************ >> >>......... >>------------------------------------------------------------------ >> >>I guess the reason for this job failuer is that the dihedral angle between monomer units >>changed abruptly from 0 to 180 or 180 to 0, and the geometry optimization became uncontrollable. >>Is this right? Is there any way to overcome this problem? >>I used the default redundant internal coordinate optimization and the keyword "Opt"(not Fopt) 1. See if the final geometry is still interaction range etc. If the geometry is reasonable, restart with a copy of a checkpoint file %chk=copy_mychk #P HF/chkbas geom=AllCheck SCF=Tight and see what is going on with the SCF Check the energy at various cycles of the geom. opt. 2. Fix a dihedral angle to 0 or 180 with ADDREDUNDANT keyword 3. Optimize intermolecular geometries with Popt and then do full optimization 4. Try the keyword "opt=z-matrix" 5. Take the XYZ coordinates form the best step in the previous job and restrat using these as input with opt=(calcfs) 6. Optimize at a lower level of theory, then do a frequency job and restrat >from this geometry using the command readfc so start the higher level job. I hope this helps people confronting similar problems. With best regards, Takanori *************************************************************** Takanori Kanazawa, Ph.D. E-mail : takanori.kanazawa@pharma.novartis.com Phone : +81-797-74-2597 FAX : +81-797-74-2598 Drug Discovery Group / CAMM Takarazuka Research Institute Novartis Pharma K.K. 10-66 Miyuki-cho, Takarazuka, Hyogo 665 JAPAN *************************************************************** From takanori.kanazawa@pharma.novartis.com Fri Mar 20 22:19:10 1998 Received: from gatekeeper.chbs.ciba.com for takanori.kanazawa@pharma.novartis.com by www.ccl.net (8.8.3/950822.1) id VAA23268; Fri, 20 Mar 1998 21:48:45 -0500 (EST) From: Received: from mailhub by gatekeeper.chbs.ciba.com; (5.65v3.2/1.1.8.2/12Mar96-0208PM) id AA17922; Sat, 21 Mar 1998 03:48:11 +0100 Received: from nts1.novartis.com ([192.168.50.131]) by mta3.is.chbs (8.8.5/8.8.5) with SMTP id DAA05734 for ; Sat, 21 Mar 1998 03:48:39 +0100 (MET) Received: by nts1.novartis.com(Lotus SMTP MTA SMTP MTA v1.1.04 (495.1 10-24-1997)) id 412565CE.000F63DF ; Sat, 21 Mar 1998 03:48:06 +0100 X-Lotus-Fromdomain: PH@N1 To: chemistry@www.ccl.net Message-Id: <492565CE.000F3E8A.00@nts1.novartis.com> Date: Sat, 21 Mar 1998 11:47:54 +0900 Subject: Summary:G94 geom. opt. with constraints Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Dear CCLers, Some time ago, I posted a question concerning geometry optimization of a dimeric molecule, applying dihedral constraints. I've got responses from the following people. Douglas J. Fox Stefan Fau Mark A. Zottola John Waite Krzys Radacki Guido Germano Larry Cuffe Victor I really appreciate their kindness. Currently, I'm testing some of the methods suggested by them and it will take time before I can find a best solution. So here I just summarize their suggestions. Upon finding the best solution, I will post here the method. My original question was >>This time I have a question about geometry optimization of a dimeric molecule. >>Each monomer unit is an almost planar molecule except one methyl group. >>I did geometry optimization of the dimer by fixing the dimer molecule on a plane >>(dihedral angles between the monomer units is fixed to 0 or 180). >>In the course of the geometry optimization, intermolecular geometry begins to deviate >>from planarity and finally the optimization job stops with the following error messages >>------------------------------------------------------------------------- --------- >>Convergence failuer -- run terminated. >>Error termination via Link1e in /home/Gaussian/g94/l502.exe. >>Job cpu time: .... >>............ >>------------------------------------------------------------------------- --------- >> >>Before this termination, messages like the following are repeated >>------------------------------------------------------------------ >>Energy Rises -- skip Quadratic search. >>Quartic linear search produced a step of -0.99578. >>....... >>....... >>Iteration 10 RMS(Cart)=************ RMS(Int)=************ >>Iteration 11 RMS(Cart)=************ RMS(Int)=************ >> >>......... >>------------------------------------------------------------------ >> >>I guess the reason for this job failuer is that the dihedral angle between monomer units >>changed abruptly from 0 to 180 or 180 to 0, and the geometry optimization became uncontrollable. >>Is this right? Is there any way to overcome this problem? >>I used the default redundant internal coordinate optimization and the keyword "Opt"(not Fopt) 1. See if the final geometry is still interaction range etc. If the geometry is reasonable, restart with a copy of a checkpoint file %chk=copy_mychk #P HF/chkbas geom=AllCheck SCF=Tight and see what is going on with the SCF Check the energy at various cycles of the geom. opt. 2. Fix a dihedral angle to 0 or 180 with ADDREDUNDANT keyword 3. Optimize intermolecular geometries with Popt and then do full optimization 4. Try the keyword "opt=z-matrix" 5. Take the XYZ coordinates form the best step in the previous job and restrat using these as input with opt=(calcfs) 6. Optimize at a lower level of theory, then do a frequency job and restrat >from this geometry using the command readfc so start the higher level job. I hope this helps people confronting similar problems. With best regards, Takanori *************************************************************** Takanori Kanazawa, Ph.D. E-mail : takanori.kanazawa@pharma.novartis.com Phone : +81-797-74-2597 FAX : +81-797-74-2598 Drug Discovery Group / CAMM Takarazuka Research Institute Novartis Pharma K.K. 10-66 Miyuki-cho, Takarazuka, Hyogo 665 JAPAN ***************************************************************