From reimo@chem.ut.ee Thu Aug 1 06:17:45 1996 Received: from mega.chem.ut.ee for reimo@chem.ut.ee by www.ccl.net (8.7.5/950822.1) id FAA09625; Thu, 1 Aug 1996 05:44:20 -0400 (EDT) Received: by mega.chem.ut.ee (Smail3.1.28.1 #1) id m0uluIb-00001cC; Thu, 1 Aug 96 12:44 EET DST Date: Thu, 1 Aug 1996 12:44:09 +0300 (EET DST) From: Reimo To: chemistry@www.ccl.net Subject: Unsubscribe Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII unsubscribe thank you ============================================================================ Reimo Rander student of BIOCHEMISTRY, 3rd grade, University of Tartu, ESTONIA Office: Room 127, Inst of Molecular and Cell Biology, Dpment of Biochemistry Vanemuise 46 EE2400 TARTU Phone: (372) 7430-235 Fax: (372) 7430-235 Email: reimo@chem.ut.ee From csilmt12@area.ba.cnr.it Thu Aug 1 07:17:48 1996 Received: from area.ba.cnr.it for csilmt12@area.ba.cnr.it by www.ccl.net (8.7.5/950822.1) id GAA09674; Thu, 1 Aug 1996 06:21:20 -0400 (EDT) Received: by area.ba.cnr.it; (5.65/1.1.8.2/16Jul96-0442PM) id AA21392; Thu, 1 Aug 1996 12:21:01 +0200 From: Massimo Trotta Message-Id: <9608011021.AA21392@area.ba.cnr.it> Subject: dipolar moments To: CHEMISTRY@www.ccl.net Date: Thu, 1 Aug 1996 12:21:01 +0200 (MET DST) Cc: hyperchem@hyper.com X-Mailer: ELM [version 2.4 PL24 PGP3 *ALPHA*] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Dear all, I'm studying the dipolar moments of some organic molecules in the ground and first excited states using semi-empirical calculations (on Hyperchem 4.0) and I have a couple of question regarding those calculations: a) I decided to use ZINDO/1 as semiempirical method (available methods are CNDO, INDO, MINDO3, MNDO, AM1, PM3, ZINDO/1 and ZINDO/s), but the choice was not made on how appropriate is the method for evaluating dipolar moment. Am I doing right? is there any reason for choosing this or that SE method for the dipolar moment calculations? b) In evaluating the dipole moment in the excited state I select the option _Next Lowest_ in the SE option dialog box. Should I work with _Unrestricted Hartree Fox_ to get more accurate number or _Restrictes HF_ gives the same accuracy as well?. c) How can I display, in hyperchem or with some other software (freeware) the dipolar moments as spacial vectors overlapped on the molecules? Any indication is welcome. will summarize. massimo -- !==============================================================! ! Massimo Trotta ! ! Centro Studi Chimico Fisici sull'Interazione Luce Materia ! ! c/o Dept. of Chemistry ! ! V. Orabona, 4 I-70126 Italy ! ! e-mail: csilmt12@area.ba.cnr.it ! ! http://www.ba.cnr.it/~csilmt12 !==============================================================! From tp@elptrs7.rug.ac.be Thu Aug 1 07:17:54 1996 Received: from elptrs7.rug.ac.be for tp@elptrs7.rug.ac.be by www.ccl.net (8.7.5/950822.1) id HAA09723; Thu, 1 Aug 1996 07:00:24 -0400 (EDT) Received: by elptrs7.rug.ac.be (AIX 4.1/UCB 5.64/4.03) id AA16568; Thu, 1 Aug 1996 13:01:43 +0200 Date: Thu, 1 Aug 1996 13:01:43 +0200 (DFT) From: "Park, Tae-Yun" To: Computational Chemistry List Subject: Summary of my earlier questions, answers & thanks(part 2 of 3) Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII ------------------------------------------------------- PART III Subject: Transition state search using MOPAC. ------------------------------------------------------- ####################(questions)######################## Dear all, I need some help from people who uses MOPAC frequently, as I'm a real beginner of using MOPAC. I want to search the transition state(TS) of the following reaction with MOPAC93: H H H H H H | \ / | | | H-C(+) + C=C ---->(TS?)----> H-C-C-C(+) | / \ | | | H H H H H H gas-phase gas-phase methyl ethylene primary propyl carbenium ion carbenium ion After reading the session of transition state location(SADDLE) in MOPAC manual, I have prepared the following input data for MOPAC: *************************************************************** UHF SADDLE DENSITY GRAPH XYZ CHARGE=1 BAR=0.03 Me--->p methylation activated complex C .000000 0 .000000 0 .000000 0 0 0 0 H 1.113983 1 .000000 0 .000000 0 1 0 0 H 1.113983 1 120.001495 1 .000000 0 1 2 0 H 1.113968 1 119.999756 1 -179.453217 1 1 2 3 C 5.210007 1 129.371857 1 -179.367538 1 1 2 3 C 1.336975 1 114.626526 1 -2.741700 1 5 1 2 H 1.099976 1 120.501572 1 -179.999634 1 5 6 1 H 1.100006 1 119.749710 1 -179.453217 1 5 6 7 H 1.099976 1 120.499832 1 179.999634 1 6 5 7 H 1.099976 1 119.749710 1 .547287 1 6 5 7 0 0.000000 0 0.000000 0 0.000000 0 0 0 0 C .000000 0 .000000 0 .000000 0 0 0 0 C 1.506600 1 .000000 0 .000000 0 1 0 0 C 1.418946 1 116.679237 1 .000000 0 2 1 0 H 1.118213 1 111.091591 1 -60.563965 1 1 2 3 H 1.119848 1 109.781792 1 179.985870 1 1 2 3 H 1.118203 1 111.102783 1 60.532600 1 1 2 3 H 1.150792 1 110.664017 1 174.500137 1 2 1 4 H 1.150722 1 110.665337 1 64.389313 1 2 1 4 H 1.109605 1 121.234596 1 179.993790 1 3 2 1 H 1.110235 1 121.565987 1 -.003556 1 3 2 1 0 0.000000 0 0.000000 0 0.000000 0 0 0 0 **************************************************************** The result with this input data was terrible; in the transition state, H2 and H were separated! The data were prepared as follows: 1. I've drawn two reactants using CS Chem 3D Pro and save as MOPAC input. I could not optimize the geometry of these two molecules in MOPAC simulatneously, since there are two molecules exist. 2. The data for the product carbenuim ions was prepared with CS Chem3D Pro again, and the geometry of this molecule was minimized with MOPAC keywards, "SYMMETRY CHARGE=1 AM1 GEO-OK EF PRECISE" and then saved as an optimezed data for the product carbenium ion. 3. Then I combined these two files and modified according to the decription in the MOPAC manual for SADDLE calculation. 4. And I've got the terrible result. What's wrong with my input data for the SADDLE calculation? Is it really possible to get the optimized geometry of transition state for the reaction I mentioned above by using MOPAC? If so, could anyone PLEASE tell me how I can get a reliable structure of the transition state theoretically? ANY suggestion/advice will be GREATLY APPRECIATED! Thank you very much in advence. ####################(Answers)######################## From tajkhors@mbp-sgi7.inet.dkfz-heidelberg.deSat Jul 27 11:07:03 1996 Date: Tue, 9 Jul 1996 21:32:47 -0600 From: Emadeddin Tajkhorshid Reply to: E.tajkhorshid@dkfz-heidelberg.de To: "Park, Tae-Yun" Subject: Re: CCL:M:Transition state search using MOPAC. Hi Tae-Yun As far as I exprienced MOPAC, you should use the same order of atoms in your input file for the geometry of reactants and products. Otherwise you will find a nonsense result. Because MOPAC tries to reconstruct the reaction coordinate toward saddle point via considering the satrting and ending geometry of each atom and is not able to automatically find the right counterpart in the second strusture. On the other hand it is recommended to use XYZ key word even if you use internal coordinate. This can prevent some kind of special errors which may occur in this routine Good luck Emad -- E. Tajkhorshid German Cancer Research Center; DKFZ Tel: +49 6221 42 2339 Dept. Molecular Biophysics (0810) FAX: +49 6221 42 2333 P.O.Box 101949 Email: E.Tajkhorshid@dkfz-heidelberg.de 69009 Heidelberg, FRG ********************************************************************** * It is nice to be important, but it is more important to be nice! * ********************************************************************** From lldmpc::mrgate::a1::moralega@lldmpc.dnet.dupont.comSat Jul 27 11:07:45 1996 Date: Tue, 9 Jul 96 16:18:10 EDT From: "lldmpc::mrgate::a1::moralega"@lldmpc.dnet.dupont.com To: tp@elptrs7.rug.ac.be Subject: RE: CCL:M:Transition state search using MOPAC. From: NAME: GUILLERMO A. MORALES FUNC: Research and Development TEL: 695-1053 To: NAME: tp@elptrs7.rug.ac.be <"tp@elptrs7.rug.ac.be"@ESDS01@MRGATE@LLDMPC> Hi Tae-Yun, I've located transition states for Diels-Alder reactions using CS Chem 3D Pro and MOPAC. I haven't done any work with carbocations, though. Anyways, this is the method I used. Check it out and compare it with yours. 1) Using CS Chem 3D Pro build first the expected product and minimize it. 2) Check that the atom numbers are continuous (1,2,3,etc.). If for some reason one is skipped then renumber ALL the atoms manually (check the manual on how to do this). 3) Save the molecule in MOPAC format. 4) Now with the pointer selec the CH3 unit on the far left (like when you do cut/paste in windows), drag it to the left side of the screen and delete the C-C bond. This is going to be your CH3+ ion. 5) The rest of the molecule is going to be converted automatically into CH3-CH3 Make it manually CH2=CH2 6) I think if you select the CH2=CH2 with the pointer you can minimiza ONLY that molucule EVEN THOUGH the CH3+ is present. I'm not really sure of this. 7) BEFORE you save this window as a MOPAC file MAKE SURE the atom numbers ARE THE SAME. In other words, if in the CH3+ the C is atom 1, then the other H are 2,3,4 MAKE SURE in the expected product the carbon on the left is atom 1 and it's corresponding Hs ae 2,3 and 4. Same fopr the CH2=CH2 unit. This step is CRUCIAL for the saddle calculation since the connectivity in starting materials MUST BE the same as in final products. 8) Now combine both MOPAC files the way you did it before, include the appropriate keywords and run the calculations. 9) Even if the final geometry is not the real transition state (thing I doubt), then take the generated final geometry and refine it. This should do the trick. If you have any more questions just let me know, OK? Good luck. Guillermo. ------------------------------------------------------------------------------- Guillermo A. Morales, Ph.D. Postdoctoral Fellow The DuPont-Merck Pharmaceuticals Co. ------------------------------------------------------------------------------- Note: The ideas/comments herein expressed are mine only. From oldziej@IRO.UMontreal.CASat Jul 27 11:11:41 1996 Date: Tue, 9 Jul 1996 16:45:36 -0400 (EDT) From: Stanislaw Oldziej To: tp@elptrs7.rug.ac.be Subject: Transition state: Hi, 1. You should remember that order of atoms in substrate compex and in products should be identical. If you see in substrate atom 2 it is hydrogen in products it is carbon ect. 2. Why you using UHF ? With best regards Stanislaw Oldziej From laidig@pg.comSat Jul 27 11:12:46 1996 Date: Tue, 9 Jul 1996 17:30:25 -0400 From: Bill Laidig To: "Park, Tae-Yun" Subject: Re: CCL:M:Transition state search using MOPAC. Tae-Yun, I would do this calculation as follows since I do not know what the TS will look like in advance. I would optimize the geometry of the product ion first. Next, I would do a series of constrained minimizations in which the CH3-C2H4+ r(C-C) was systematically stretched (i.e. the first calculation has r(C-C) at the equilibrium distance (re), next fixed at re+0.1, next fixed at re+0.2... After a maximum along the stretching coordinate was located, I would next use this geometry unconstrained in a TS calculation. Bill -- ****************************************************************************** * "Like jewels in a crown, the precious stones glittered in the queen's * * round metal hat." - Jack Handey * * * * Bill Laidig * * The Procter & Gamble Co. tel 513-627-2857 fax - 1233 * * Miami Valley Laboratories laidig@pg.com (preferred) * * P.O. Box 538707 wd_laidig@pg.com * * Cincinnati, OH 45253-8707 laidig@qtp.ufl.edu * ****************************************************************************** From Steve.Bowlus@sandoz.comSat Jul 27 11:13:31 1996 Date: Wed, 10 Jul 1996 01:21:48 +0200 From: Steve.Bowlus@sandoz.com To: tp@elptrs7.rug.ac.be Subject: M: SADDLE First, the SADDLE command does not give an optimized geometry of the TS -- only an approximate geometry. The "best" TS approximation must be refined using the TS keyword, or NLLSQ, or some combination (perhaps in several steps), followed by a FORCE calculation to assure you are at a TS (one -ve root to the Hessian), and ideally followed by reaction coordinate calculations (IRC=1 and IRC=-1) to generate the product and starting material. Only through all this work can you assure you have a TS for the reaction you are interested in. This applies to all TS, incidentally, whether you generate them by the SADDLE command or by guessing. Since your results are so spectacularly strange, it seems most likely that your starting geometry for the separate pieces is indigestible. I would do some minimization in Mopac on the starting side before trying the saddle calculation. I expect there are several "reasonable" geometries for this, and you might have to try several before you get it right. Given a "reasonable" starting geometry, you might try running the calculation in AMPAC, if you have access to the program. I believe the algorithm for locating TS is different from that used in Mopac, but the two programs should give otherwise very similar results (with minor modification, the AMPAC output can be transferred to Mopac as well). I think the keyword is CHAIN. sb From jig@qorg.unizar.esSat Jul 27 11:14:17 1996 Date: Wed, 10 Jul 96 9:22:32 METDST From: Jose Ignacio Garcia To: tp@elptrs7.rug.ac.be Subject: SADDLE calculation My guess for your problem is that the "terrible" result you get is mainly due to the fact that you are trying to calculate an ion-molecule reaction IN THE GAS PHASE. As ions are VERY unstable in the gas phase, its reactivity can be also very different that that observed in solution, where the ion is comfortably solvated. Thus, in general, in an isolated system calculation as yours, an ion always will tend to be "solvated" by the neutral molecule, through a polarisability mechanism, and will approach to its hydrogen atoms. Often, this approach ends with a hydrogen abstraction reaction, as you observed. I have not experience with cation reactions, but this is exactly what happens with the paradigmatic SN2 reaction between chloride ion and methyl chloride. If one wants to correctly describe the SN2 attack, one must be careful in designing the chloride approach trajectory. Maybe the inclusion of solvation effects through a continuum model would help to fix this problem. I hope this helps you. Best regards. Jose I. Garcia -- -------------------------------------------------------------------------------- Dr. Jose Ignacio Garcia-Laureiro Phone : 34-(9)76-762077 Departamento de Quimica Organica 761210 Instituto de Ciencia de Materiales de Aragon Fax : 34-(9)76-761159 C.S.I.C.-Universidad de Zaragoza e-mail: jig@qorg.unizar.es E-50009 ZARAGOZA (SPAIN) jig@msf.unizar.es jig@posta.unizar.es -------------------------------------------------------------------------------- "And all this science I don't understand it's just my job five days a week..." ELTON JOHN - Rocket man -------------------------------------------------------------------------------- From borkent@caos.kun.nlSat Jul 27 11:15:07 1996 Date: Wed, 10 Jul 1996 09:40:09 +0200 (MDT) From: Hens Borkent To: tp@elptrs7.rug.ac.be Subject: mopac ts calc. Dear Tae-Yun Park, Your question is a rather frequent one these days, so at our site we have developed a tutorial on this subject: http://www.caos.kun.nl/~borkent/compcourse/comp.html You can also have a look at the Re-view pages for more suggestions: http://http1.brunel.ac.uk:8080/depts/chem/ch241s/re_view/ In your case it boils down to drawing the product, then increase the reaction bond length gradually, in which you should observe a maximum, optimize this conformation using the keyword TS, check frequencies with FORCE. I'll give it a try as well. Sincerely, -- ***** J.H. (Hens) Borkent, CAOS/CAMM Center, *CAOS * P.O. Box 9010, 6500 GL Nijmegen, The Netherlands * / * Tel 0031 24 36 52137 Fax 0031 24 36 52977 * CAMM* e-mail: borkent@caos.kun.nl ***** http://www.caos.kun.nl/staff/borkent.html From wallenborn@phys.chem.ethz.chSat Jul 27 11:15:38 1996 Date: Wed, 10 Jul 1996 11:14:33 +0200 From: Ernst Wallenborn To: tp@elptrs7.rug.ac.be Subject: Re: Transition state search using MOPAC. Hi tere, i guess your MOPAC input is corrupt. See: UHF SADDLE DENSITY GRAPH XYZ CHARGE=1 BAR=0.03 Me--->p methylation activated complex C .000000 0 .000000 0 .000000 0 0 0 0 ^^^^^ a zero here means "do not optimize this coordinate". This is what you want for 6 coords (since you have 3N-6 degrees of freedom), but not more. Here's already three of them. H 1.113983 1 .000000 0 .000000 0 1 0 0 H 1.113983 1 120.001495 1 .000000 0 1 2 0 six zeroes so far, including this one. Now the spatial position of your complex is fixed. Any more zeroes will distort the result. H 1.113968 1 119.999756 1 -179.453217 1 1 2 3 C 5.210007 1 129.371857 1 -179.367538 1 1 2 3 C 1.336975 1 114.626526 1 -2.741700 1 5 1 2 H 1.099976 1 120.501572 1 -179.999634 1 5 6 1 H 1.100006 1 119.749710 1 -179.453217 1 5 6 7 H 1.099976 1 120.499832 1 179.999634 1 6 5 7 H 1.099976 1 119.749710 1 .547287 1 6 5 7 0 0.000000 0 0.000000 0 0.000000 0 0 0 0 First of all, what's Oxygen doing here in the first place. You gave it a bond distance of 0.00 not to be optimized, to atom 0. This cannot lead to any sensical results. C .000000 0 .000000 0 .000000 0 0 0 0 This is even worse. Now you have two atoms, on a non-defined position in space, which do not belong there anyway, and which cannot be moved! C 1.506600 1 .000000 0 .000000 0 1 0 0 C 1.418946 1 116.679237 1 .000000 0 2 1 0 H 1.118213 1 111.091591 1 -60.563965 1 1 2 3 H 1.119848 1 109.781792 1 179.985870 1 1 2 3 H 1.118203 1 111.102783 1 60.532600 1 1 2 3 H 1.150792 1 110.664017 1 174.500137 1 2 1 4 H 1.150722 1 110.665337 1 64.389313 1 2 1 4 H 1.109605 1 121.234596 1 179.993790 1 3 2 1 H 1.110235 1 121.565987 1 -.003556 1 3 2 1 0 0.000000 0 0.000000 0 0.000000 0 0 0 0 I guess you mistake was using Chem3D as z-matrix generator. Try to save as xyz file and read this into mopac (you'll have to manually edit the input file, naturally). Or settle for a different z-matrix editor altogether (i use molden, cause it's unix, free and gives good control over the actual variables) Hope this helps. -- -ernst wallenborn. i'm not a bug. i'm an undocumented feature. From jig@qorg.unizar.esSat Jul 27 11:16:34 1996 Date: Wed, 10 Jul 96 14:14:42 METDST From: Jose Ignacio Garcia To: tp@elptrs7.rug.ac.be Subject: SADDLE calculation Hello! It's me again. Have you tried to use an alternative method to SADDLE? For instance, a reaction coordinate can be estimated by starting from the product and breaking the C-C bond. As the C-C distance increases, the total energy will change, and a maximum will probably appear. The structure corresponding to this maximum can be then used as an starting point for a TS or NLLSQ calculation. By the way, I am still concerned with solvation. It is sure that in the approach of the cation to the neutral molecule, an energy minimum corresponding to an ion-molecule pair will be observed. Best regards. Jose I. Garcia -- -------------------------------------------------------------------------------- Dr. Jose Ignacio Garcia-Laureiro Phone : 34-(9)76-762077 Departamento de Quimica Organica 761210 Instituto de Ciencia de Materiales de Aragon Fax : 34-(9)76-761159 C.S.I.C.-Universidad de Zaragoza e-mail: jig@qorg.unizar.es E-50009 ZARAGOZA (SPAIN) jig@msf.unizar.es jig@posta.unizar.es -------------------------------------------------------------------------------- "And all this science I don't understand it's just my job five days a week..." ELTON JOHN - Rocket man -------------------------------------------------------------------------------- From borkent@caos.kun.nlSat Jul 27 11:18:10 1996 Date: Wed, 10 Jul 1996 14:38:10 +0200 (MDT) From: Hens Borkent To: tp@elptrs7.rug.ac.be Subject: M:Searching transition state Dear T-Y Park, As promised I had a look at your molecule and found that there is a complication in the form of a cyclic intermediate, which is more stable than the product. Furthermore it is possible that you find no TS, because in vacuum the methyl cation is high in energy and the energy is falling if it approaches the ethylene, continuously. So in fact you have to explore the pot.energy surface by varying the C-C distance and the C-C-C bond angle. There is probably a TS between the bridged intermediate (CH3 symmetrical above the ethylene) and the propyl cation. This is in fact a corner-protonated cyclopropane. This system must have been studied in the literature, using ab initio methods as well. (Von Schleyer, JACS, early 80's). Best regards, -- ***** J.H. (Hens) Borkent, CAOS/CAMM Center, *CAOS * P.O. Box 9010, 6500 GL Nijmegen, The Netherlands * / * Tel 0031 24 36 52137 Fax 0031 24 36 52977 * CAMM* e-mail: borkent@caos.kun.nl ***** http://www.caos.kun.nl/staff/borkent.html From borkent@caos.kun.nlSat Jul 27 11:18:48 1996 Date: Wed, 10 Jul 1996 15:05:18 +0200 (MDT) From: Hens Borkent To: tp@elptrs7.rug.ac.be Subject: c3h7+ reference As a follow-up on my previous mail: the reference is JACS 103, 5649 (1981) -- ***** J.H. (Hens) Borkent, CAOS/CAMM Center, *CAOS * P.O. Box 9010, 6500 GL Nijmegen, The Netherlands * / * Tel 0031 24 36 52137 Fax 0031 24 36 52977 * CAMM* e-mail: borkent@caos.kun.nl ***** http://www.caos.kun.nl/staff/borkent.html Please notice change in the Tel/Fax numbers: 80 -> 243 From Steve.Bowlus@sandoz.comSat Jul 27 11:20:04 1996 Date: Wed, 10 Jul 1996 16:51:24 +0200 From: Steve.Bowlus@sandoz.com To: tp@elptrs7.rug.ac.be Subject: Mopac ts I would try moving the starting structures closer together, so that the reactants "look" more like a ts. The c-c bond length in the ts is (at a guess) on the order of 1.8 - 2.0 A -- try 2.2 - 2.4 A. You may also have to simply guess the ts structure, and then refine using TS, NLLSQ, etc. it will be _crucial_ in this case that you do the IRC calculations to assure you have a ts for the desired reaction. sb From cramer@maroon.tc.umn.eduSat Jul 27 11:20:14 1996 Date: Wed, 10 Jul 1996 09:48:28 -0500 (CDT) From: Christopher J Cramer To: tp@elptrs7.rug.ac.be Subject: Re: CCL:M:Searching transition state with MOPAC; part2 What makes you think there IS a transition state structure? In the gas phase, I would expect this reaction to have a good chance of being barrierless. Rather than attempt to locate a TS structure a priori, you might simply do constrained optimizations pulling off the methyl group by 0.1 angstrom increments (for instance). If you generate a reaction path without a barrier, you will know that it is fruitless to look for a TS. On the other hand, if you do find a barrier, you will know where to start your TS optimization. CJC -- Christopher J. Cramer University of Minnesota Department of Chemistry 207 Pleasant St. SE Minneapolis, MN 55455-0431 -------------------------- Phone: (612) 624-0859 || FAX: (612) 626-7541 cramer@maroon.tc.umn.edu http://pollux.chem.umn.edu/~cramer ------------------------------------------------------- PART IV Subject: Searching transition state with MOPAC; part2 ------------------------------------------------------- ####################(questions)######################## Dear all, I've received a lot of helpful messages from many CCLers, concerning my earlier questions on how to get the structure of transition state of the reaction, H H H H H H | \ / | | | H-C(+) + C=C ---->(TS?)----> H-C-C-C(+) | / \ | | | H H H H H H gas-phase gas-phase methyl ethylene primary propyl carbenium ion carbenium ion Thanks a lot! I greatly appreciate all the messages concerning this problem. Most of the messages pointed out that I made a mistake to prepare the input file for SADDLE calculation; all the atom should be in the same order between reactants and product z-matrix field. I made second try with a corrected input file as follows; ************************************************************** SADDLE DENSITY GRAPH XYZ CHARGE=1 PRECISE Me--->p methylation activated complex C 0.000000 0 0.000000 0 0.000000 0 0 0 0 C 8.471619 1 0.000000 0 0.000000 0 1 0 0 C 1.325912 1 162.328125 1 0.000000 0 2 1 0 H 1.108597 1 65.592545 1 -150.347214 1 1 2 3 H 1.113968 1 120.001495 1 -151.401566 1 1 4 2 H 1.108612 1 119.996246 1 -179.367538 1 1 4 5 H 1.098282 1 122.722214 1 57.159393 1 2 3 1 H 1.098328 1 122.715210 1 179.999634 1 2 3 7 H 1.098267 1 122.715210 1 -0.316483 1 3 2 7 H 1.098206 1 122.715210 1 -179.684021 1 3 2 7 0 0.000000 0 0.000000 0 0.000000 0 0 0 0 C .000000 0 .000000 0 .000000 0 0 0 0 C 1.506600 1 .000000 0 .000000 0 1 0 0 C 1.418947 1 116.688263 1 .000000 0 2 1 0 H 1.118077 1 111.116249 1 -59.952415 1 1 2 3 H 1.119762 1 109.783371 1 -179.449097 1 1 2 3 H 1.118210 1 111.069687 1 61.123405 1 1 2 3 H 1.150598 1 110.672989 1 174.990601 1 2 1 4 H 1.150955 1 110.650314 1 64.874481 1 2 1 4 H 1.109621 1 121.247902 1 -179.729874 1 3 2 1 H 1.110257 1 121.556961 1 .300391 1 3 2 1 0 .000000 0 .000000 0 .000000 0 0 0 0 ************************************************************** This time MOPAC produced a structure of transition state which is almost the same as that of reactants, that is, there was almost no change between reactants and transition state. I've tried to refine the transition state with the key words, "TS", "SIGMA", and "NLLSQ" separately, which gives no change either. Why is the result like this? Could anyone PLEASE tell me how can I go further to obtain a reliable structure of the transition state?? Thank you VERY MUCH in advance. ####################(Answers)######################## From olsonl@darwin.pprd.abbott.comSat Jul 27 11:24:00 1996 Date: Wed, 10 Jul 1996 10:56:02 -0500 From: "Leif P. Olson" To: tp@elptrs7.rug.ac.be Subject: mopac ts Hi Tae-Yun, I would suggest taking a different approach altogether, breaking up the problem into several parts: 1) get an approximate transition state for the reaction. Forget about SADDLE and so forth; just do a trajectory calculation where you start with your 1-n-propyl cation and stretch the bond in small increments (say, 0.1 angstroms) from the starting point to about 4 or 5 angstroms. Remember, by microscopic reversibility, the ts for the reverse reaction is precisely the same as the forward reaction. Also, by starting with one molecule rather than two, you can be pretty sure that the program will not get confused about how to assign the electrons in the orbitals, at least in the early stages of the reaction. 2) take the highest point in the trajectory and see if it seems like some reasonable structure. If so, change the coordinate system from internal coordinates (used in the trajectory calculation) to xyz coordinates. Searches using the TS keyword, in my experience, almost always benefit from being done in Cartesian coordinates (despite what the manual says regarding the use of this in only difficult cases). The eigenvector-following algorithm works pretty well (TS keyword). 3) Once you have a refined TS structure, you may want to refine it a bit further; i.e., if you are quite close to a symmetric structure, you may want to go back to internal coordinates and make 179.8 degree dihedrals to 180.000, for example. This must be done judiciously, but can in some cases overcome rounding errors that creep in. Now try another TS calculation (or NLLSQ or SIGMA) using tighter optimization criteria. 4) Do a FORCE calculation to see if you are at a true transition state 5) Do an IRC run to make sure that the transition state you have actually goes to reactants and products (actually this requires two IRC runs). If you used the PRECISE keyword in step 2, you can generally skip step 3 entirely. But the general pattern has proven successful for me in the past. The two most important points seem to be (i) having a good guess at an approximate transition state; and (ii) using the combination of XYZ and TS. As a separate issue: why not do good-quality ab initio calculations instead? You only have three heavy atoms. Leif Olson olsonl@darwin.pprd.abbott.com From newhoir@duc.auburn.eduSat Jul 27 11:24:06 1996 Date: Wed, 10 Jul 1996 11:47:12 -0500 From: Irene Newhouse To: tp@elptrs7.rug.ac.be Subject: Re: CCL:M:Searching transition state with MOPAC; part2 Try your latest structure in a frequency computation & see if there is indeed 1 & only 1 'negative' frequency [it's really imaginary, but is reported as negative by the program]. I've found transition states to be very nonintuitive. Not surprising, since so little experimental work can be done! The 'obvious' choices for TSs, I've found, are often plain wrong!!! If this is not indeed a very reactant-like TS, & you still haven't located it, look at the literature & see if you can find a similar computation somewhere. Use that as a model, no matter how strange it looks, & see where it gets you. For example, see Foresman & Frisch's book on using Gaussian for the Formal- dehyde decomp/formation rxn from CO & H2. I'd not have guessed that TS struc- ture in a month of Sundays, yet that's what it is! Good luck! Irene Newhouse From oldziej@IRO.UMontreal.CASat Jul 27 11:24:11 1996 Date: Wed, 10 Jul 1996 12:49:16 -0400 (EDT) From: Stanislaw Oldziej To: tp@elptrs7.rug.ac.be Subject: Transition state Hi 1. In my opinion you substrates and product are to "far" one to each anothera. Try to optimize substrate compex and products. You should expect that substrates are formed pi-complex stabilized by interaction between empty p-orbital of the carbocation and pi-orbitals of ethylene. I thing that distans carbon-carbon should be in range 3-2.2 A but not 8 (see your input). If you have optimized complexes than run SADDLE. 2. If you not find TS using procedure from point 1. Try PATH option with changing distance between two atoms your input schould look like: CHARGE=1 PRECISE Me--->p methylation activated complex C .000000 0 .000000 0 .000000 0 0 0 0 C 1.500000 -1 .000000 0 .000000 0 1 0 0 C 1.418947 1 116.688263 1 .000000 0 2 1 0 H 1.118077 1 111.116249 1 -59.952415 1 1 2 3 H 1.119762 1 109.783371 1 -179.449097 1 1 2 3 H 1.118210 1 111.069687 1 61.123405 1 1 2 3 H 1.150598 1 110.672989 1 174.990601 1 2 1 4 H 1.150955 1 110.650314 1 64.874481 1 2 1 4 H 1.109621 1 121.247902 1 -179.729874 1 3 2 1 H 1.110257 1 121.556961 1 .300391 1 3 2 1 0 .000000 0 .000000 0 .000000 0 0 0 0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 You can start from product to substrates (like in example) or in reversed direction. After this you should select structure with the highest energy and try to optimized it by NLLSQ and/or TS. 3. You should done calculation using another methods AM1, PM3 Stanislaw Oldziej P.S. Sorry for my still bad English From gford@post.cis.smu.eduSat Jul 27 11:24:16 1996 Date: Wed, 10 Jul 96 12:29 CDT From: "George P. Ford" To: tp@elptrs7.rug.ac.be Subject: Re: CCL:M:Searching transition state with MOPAC; part2 Dear Pack: I have a couple of comments on your study that might be useful. One refers specifically to the transition state problem, the other relates to the philosophy of your study. The use of saddle is best reserved for complex situations were several bonds are being made and broken at the same time. Reactions like yours, where the transition vector can be anticipated to involve essentially the formation of a single bond, are almost always best studied by following a simple reaction coordinate for the reverse reaction. (This might be a bit less clear-cut if you were not using a Z-matrix-based geometry definition). This also has the advantage of immediately identifying those reactions where there is no transition state per se, i.e. the energy simply falls monotonically from reactants to products. A quick look at your system at the AM1 level shows it to be a reaction of this kind. There is no t.s. - the energy rises to a plateau corresponding to the dissociation energy. A saddle calculation should end up at some arbitrary point on that plateau which it apparently did. While this exercise has hopefully illustrated something about locating transition states it probably isnUt the way to study the chemistry of carbenium ions. The trouble is that the potential surfaces given by AM1, and MNDO (I assume also PM3) are completely wrong. The relative ordering, and even the nature (minimum vs t.s.) of most of the stationary points is incorrect. For example, corner protonated cyclopropane which is probably an intermediate on the true reaction path you are seeking is incorreclty predicted by these methods to be a saddle point. Conversely, the primary propyl cation, predicted to be a minimum at the semiempirical level, is actually a saddle point. Hope this helps, George P. Ford ============================================================================ George P. Ford | email: gford@smu.edu Department of Chemistry | web: http://www.smu.edu/~gford/ Southern Methodist University | telephone: (214)768-2479 Dallas, Texas 75275 | fax: (214)768-4089 ============================================================================ From kcousins@wiley.csusb.eduSat Jul 27 11:24:24 1996 Date: Wed, 10 Jul 1996 13:11:17 -0700 (PDT) From: Kimberley Cousins To: tp@elptrs7.rug.ac.be Subject: TS location Hi. I've had limited luck with Saddle, and usually use the "reaction coordinate" approach. Your first Z matrix is set up to do just that. Instead of "1" for C1-C2 bond length opt, use "-1" and list a series of decreasing bond lenghths following a blank line following the Z matrices (like 7.1, 6.1, 5.1 . . .) and look for the highest energy structure generated that doesn't otherwise fly apart. This should (might) be your approx TS structure that you can optimize further using NLLSQ or the like. Kimberley Cousins Department of Chemistry California State University, San Bernardino 5500 University Parkway San Bernardino, CA 92407 (909)880-5391 kcousins@wiley.csusb.edu From Annik.VanKeer@chem.kuleuven.ac.beSat Jul 27 11:24:34 1996 Date: Fri, 12 Jul 1996 17:12:21 +0100 (NFT) From: Van Keer Annik To: "Park, Tae-Yun" Subject: transition state Hello Park, I donnot know if I can help you with your Mopac version because I am a Gaussian user. Firstly I do not understand why the bond between C1 and C2 is so large: 8.47 A? If you take 3 A or 3.5 A as distance between these bonds than you also have two reactants without any interaction. I guess that any change in geometry than will earlier lead to a transition state for which I guess that the initially formed CC bond is about 1.8 A. Secondly, I do not know which optimization procedure you are using, but as I see two inputs one for the reactants and another one for the products I guess your program is looking for a TS via a LST path. In this case you have to identify two identically z-matrices, with of course different parameters, which you didn't. Thirdly, it is possible that the addition of CH3(+) to C2H4 is a barrier free proces. I know that the addition of triplet CH2 to C2H4 is barrierfree. I hope that you can use my suggestions. Good luck, Annik ****************************************************************** Annik Van Keer Katholieke Universiteit Leuven Laboratorium voor Quantumchemie Celestijnenlaan 200F 3001 Heverlee, Belgium Tel:00-32-16-327384 Fax:00-32-16-327992 E-mail:annik.vankeer@chem.kuleuven.ac.be From luo96@cyberramp.net Thu Aug 1 14:14:50 1996 Received: from mailhost.cyberramp.net for luo96@cyberramp.net by www.ccl.net (8.7.5/950822.1) id SAA28290; Thu, 18 Jul 1996 18:45:57 -0400 (EDT) Received: from ramp1-22.cyberramp.net (ramp1-22.cyberramp.net [206.42.4.71]) by mailhost.cyberramp.net (8.7.1/8.7.1/$Revision: 1.2 $) with SMTP id RAA09751 for ; Thu, 18 Jul 1996 17:45:49 -0500 (CDT) Date: Thu, 18 Jul 1996 17:45:49 -0500 (CDT) Message-Id: <199607182245.RAA09751@mailhost.cyberramp.net> X-Sender: luo96@cyberramp.net X-Mailer: Windows Eudora Version 1.4.4 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: chemistry@www.ccl.net From: luo96@cyberramp.net (Yu-Ran Luo) Subject: Which Company Would Develop A Novel Project? >Which Company Would Develop A Novel Project? > > >Hi, Netters, > > Are you interested to predict BOND DISSOCIATION ENERGIES (BDEs) >in large organic species ? Are you interested to develop a novel software? > > You know, the BDEs are considered to be fundamental to chemistry and >biochemistry. However, about thousands of the experimental BDEs values in >organic species are available today although several hundreds of research >groups measured for sixty years. > > We thus need theory to help enlarge our knowledge. Here I would share >a good news with you. > > I found a reliable and simple method for predicting BDEs during the >past nine years. It is most powerful to large organic, organometallic and >biologically active species. Now the values of trillions of BDEs can be >estimated easily by my approach and a calculator (not computer so far!). >The large-size species include biologically active species (eg., Vitamins >A, C, D, E, K), drugs (nicotine, hormones), toxicant, explosives/energetic >materials (eg., TNT, RDX, NTO), fuels/coals and so on. Once an organic >molecular structure is drawn, I can quickly point out where the weakest >bond is, how large this BDE is, what the stability and reactivity of this >molecule are, and which position in this molecule will react first >(active site or reaction center). These predictions will save a lot of >money and time for scientists and engineers, and very useful for the >computer-aided molecular design. > > You may give me a test. Which molecules you are interested? Please show >me and draw the structures. I would give you the answers ASAP. > > I hope my research will be a new and powerful tool for chemistry, >chemical engineering, energy science, molecular biology, medicinal chemistry, >protein/enzyme chemistry. > > Why is the human visual system sensitive only to light of wavelengths >from about 760 to 380 nm? Why is "super-vitamin E" possible? Which >structure is a super-vitamin E? I can give quantitative answers now. >The main reason is that I can describe the species energetics and its >change/transfer. > > My method is different from MO theory, FDT, MM and the group additivity. >(GA). But my results can compare with the FDT calculations based new >Gaussian package and supercomputer. > > Let us make a test. > Let us compete: you operate a super-computer and I use a calculator. > Let us see: WHICH calculations are first, reliable and super. > > My method is a key to open the unknown door of molecular energetics. >Some scientists thus advised me to make an electronic version of my method >and link it with a good software available so that you only play a mouse >and keyboard when predicting the BDEs. This is a good suggestion. > > You know, there are so many chemical/biochemical software companies. >Unfortunately, they have not known my method. > > Which company would consider this project and make this link? Please >tell me. My proposal is > > My Method For BDEs + A Best Structure Software ---> A Novel Program > > Thank you for your attention. > > Yu-Ran Luo, Ph. D. > (3100 Verbena Drive, Plano, TX 75075, Phone and Fax 214-509-2075, > E-Mail: luo96@cyberramp.net) From ccl@www.ccl.net Thu Aug 1 21:17:30 1996 Received: for ccl@www.ccl.net by www.ccl.net (8.7.5/950822.1) id RAA01685; Thu, 1 Aug 1996 17:30:55 -0400 (EDT) Date: Thu, 1 Aug 1996 17:30:55 -0400 (EDT) From: Computational Chemistry Message-Id: <199608012130.RAA01685@www.ccl.net> To: chemistry@www.ccl.net Cc: jkl@www.ccl.net Subject: Backlog of messages CCLer`s: This message is just a warning to you that you will shortly recieve a surge of messages from CCL (about 45). This is due to the script "flagging" numerous messages the past few days and me unfortunately not keeping up with them. Jan has the list set up so that he or I have to go through all the flagged message, make judgement calls on them, and then resend them if appropiate. Although I am not a computational chemist and do not understand a lot of what is said in the messages, I think I have somewhat of an idea what is and is not appropiate and have tried to make the right calls on a lot of these messages. I am sorry if any of them are ones that Jan normally would have not let out on the list. I also apologize for this "mass mailing" of the messages all at once. I hope this will not cause any big problems with anyone and I hope that the circumstances will not arise again that led to this problem. Once again, I apologize and if you have any comments or questions feel free to address them to me. Alan Chalker alanc@ccl.net