From tp "-at-" elptrs7.rug.ac.be Thu Aug 1 07:18:09 1996 Received: from elptrs7.rug.ac.be for tp.,at,.elptrs7.rug.ac.be by www.ccl.net (8.7.5/950822.1) id GAA09711; Thu, 1 Aug 1996 06:58:09 -0400 (EDT) Received: by elptrs7.rug.ac.be (AIX 4.1/UCB 5.64/4.03) id AA16544; Thu, 1 Aug 1996 12:59:35 +0200 Date: Thu, 1 Aug 1996 12:59:34 +0200 (DFT) From: "Park, Tae-Yun" To: Computational Chemistry List Subject: Summary of my earlier questions, answers & thanks(part 1 of 3) Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Dear all, During last a few weeks, I posted several questions on my Ph.D. work to CCL, which is related to comput- ataional chemistry field. Hereby, I summerized my questions and answers from many CCLers, which gave me a great help. I have decided to post this summary since there are other CCLers who want to know what I have got from this e-mail communication. Let me express my deep thanks to all the CCL members who are interested in my questions and sending me impressive and helpful answers. Especially, I'd like to express my sincere thanks to Mr. Ernest Chamot who sent me a detailed and valuable message which give me a great help and encoregement. What I'm afraid now is that this summary would be too long to send by e-mail so that it occupies too much disk space. Please execuse me if this message creates any disk space problem on your mail box. Sincerely, Park, TAE-YUN =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= State University of Ghent Laboratorium voor Petrochemische Techniek Krijgslaan 281, Blok S5 9000 Gent, Belgium TEL:+(32)-0(9)-264-4527 FAX:+(32)-0(9)-264-4999 e-mail: tp { *at * } elptrs7.rug.ac.be =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= ------------------------------------------------------- PART I subject: HELP!(urgent):theoretical discrimication & software ------------------------------------------------------- ####################(questions)######################## *** About the structure of surface carbenium ions and symmetry number *** First thing I have to calculate is the symmetry number of surface carbenium ions on the zeolite surface, which is formed by protonation of various olefins. 1. How can we describe the bonding of carbenuim ions to the surface? Is the bonding more ionic or covalent? I know the proton is surely covalently bound, but it is less clear for the case of hydrocarbons(more exactly olefins) adsorbed on zeolite. 2. What is the symmetry number for such surface carbenium ions comparing with corresponding gas phase carbenium ions? In most cases, I believe, the symmetry number of surface ions are identical with those in gas phase, if there is no 2-fold axes. Consider the following examples. C \ C-C+-C (secondary ion in gas phase) / C symmetry number=3*3*3=27 C \ C-C-C (secondary ion on the surface) / | C | --(+)-- symmetry number=3*3*3=27 If there is two-fold axes, however, C \ 2-fold axes.....C+-C-C (tertiary ion in gas phase) / C symmetry number=(3*3*3)*2=54 C \ Not a 2-fold axes(?)...C+-C-C (tertiary on the surface) /| C | --(+)-- symmetry number=3*3*3=27 I think the surface ion loose its 2-fold axes due to the bonding to the catalyst surface. I just want to know whether those approach to calculate the symmetry number of surface carbenium ions is theoretically correct. To check this point, I think, the bonding of carbenium ions to the surface should be defined first, in a certain way of theoretical approach. **** About the structure of activated complex ****** Next problem that I have to solve is concerning the structure of activated complex in transition state for beta-scission. That is, an elementary step representing beta-scission, R1 ---> (Transition State) ----> R2 + O where R1 and R2 are the reactant and product carbenium ion, respectively, and O is the product olefin. What I want to know is the structure of the activated complex in the transition state, as detailed as possible in a certain theoretical way; how many bonds are related, which bonds are to be broken, which bond will be formed, and what does the activated complex look like(is it more like carbenium ion or more like olefin?), etc. **** About heat of formation for gas phase carbenium ions *** My final question is about the calculation of thermodynamic data, more exactly, the heat of formation of gas-phase carbenium ions. Recently, a solution to solve the problem associated with the large number of the parameters has been considered, i.e, by applying the "Polanyi relation". This relation needs only relative differences in heat of reactions in each elementary steps. This means that the heat of formation for the surface carbenium ions are not necessary, but, at least, heat of formation of gas-phase carbenium ions appearing in the reaction network should be given. Unfortunately, only limited amount of data can be found in the literature, so that some analytical method to calculate the heat of formation is necessary. Benson's group contribution method has been known as one of the most approprate way to calculate the thermodydamic properties. The group contribution value of positively-charged carbon atom involved in the carbenium ions is, however, not available at the moment. The only solution I found recently is using a certain quantum chemistry package such as "MOPAC". My question is that is this a suitable package (accurate enough) to calculate the heat of formation data I need? If so, how can I obtain a copy of this package? If not, is there any other software in the field of computational chemistry to calculate these data? *************** End of my questions ************************* #####################(answers)########################## From echamot(-(at)-)xnet.comSat Jul 27 10:23:18 1996 Date: Thu, 27 Jun 96 14:25:24 EDT From: Ernest Chamot To: "Park, Tae-Yun" Subject: Re: CCL:M:HELP!(urgent):theoretical discrimication & software Hi TAE-YUN, I'm sure you realize you are taking on quite a task for your research: >I'm a ph.D. student of State University of Ghent, Belgium. >I'm working with zeolite HZSM-5. The goal of my work is >to develop a detailed kinetic model for hydrocarbon >transformation process over HZSM-5 catalyst. A tremendous amount of research has gone into working out the kinetics for the strictly thermal processes of cracking hydrocarbons to olefins. The catalytic process is, of course, more complicated. And zeolites in particular have been subject to a great deal of study, both theoretically and experimentally, just to understand the nature of their catalysis, not to mention working out the detailed kinetics that result as a consequence. I have done some work in both areas, so I offer the following observations in addition to some specific responses to your query. 1) The mechanism that leads to the formation of olefins (among other products) from hydrocarbons is very complicated. The beta scission reaction to eliminate an olefin from a free radical is the dominant process in the thermal reaction of light hydrocarbons. This cascade takes place in competition with an array of rearangement reactions and cyclizations (which lead to aromatics and ultimately to coke). I am attaching a .gif file of part of the mechanism showing this cascade and a few of the rearrangement reactions. (I had a more complete mechanism, showing the steps leading to various cyclics and aromatics and coke precursors, but I can't lay my hand on it at the moment.) 2) The same mechanism applies to the carbenium ion process (initiated by "acid" catalysis), except the relative rates will be different. The rearrangement and cyclization reactions are known to be much faster, relative to the beta-scission. (The major reason a cracking catalyst for hydrocarbons to light olefins has never been commercialized is due to the fact that zeolites and other acid catalysts all coke virtually instantaneously at temperatures at which olefins are favored thermodynamically over paraffins.) 3) One way zeolite catalysts exhibit selectivity is by imposing geometrical constraints upon the "activated complex" of the reactant and the catalytic site, as you propose studying. More commonly, observed selectivities are due to geometric constraints on either the ingress of reactants or the egress of products: all possible products can form within the pores, but only the ones that can fit through the pores to get out are observed. Other products stay inside and continue to reversibly interconvert back to reactants and forward to the mix of potential products. Hence, the energetics of the "activated complex" as affected by the geometric constraints about the acidic site inside a zeolite pore will only be relavent for reactions with both reactants and products small enough to pass completely unhindered into and out of the zeolite pores. So far as your specific questions: >I have some urgent questions for my ph.D. work, which have >been discussed with many people who were in the field of >chemical engineering/organic chemistry/electrical chemistry >in my university or in the internet. > >So far, I could not get reliable solutions, which have stuck >me for relatively long time. During searching for the solutions, >I realized that the best people who can give me an advice are >in the field of computational chemistry. > >Please take some time to read the following questions >and give me some informaion/advice if it doesn't bother >you too much. > > >*** About the structure of surface carbenium ions and symmetry number *** > >First thing I have to calculate is the symmetry number >of surface carbenium ions on the zeolite surface, which >is formed by protonation of various olefins. > >1. How can we describe the bonding of carbenuim > ions to the surface? Is the bonding more > ionic or covalent? I know the proton is surely > covalently bound, but it is less clear for the > case of hydrocarbons(more exactly olefins) adsorbed > on zeolite. If you really mean to model the "surface" of the zeolite, you will have to consider a combination of sites, all contributing to an overall reaction. There are several sites within a zeolite that can be acidic, depending on where the trivalent atom (Al, B, etc.) substitutes for the tetravalent atom (Si), forcing the adjacent oxygen to be capped with an acidic hydrogen: O O O O O O | H+ | | | | | - Si - O - Al(-) - O - Si - vs. - Si - O - Si - O - Si - | | | | | | O O O O O O These are referred to as "T sites" based on the labeling of the definitive XRD data: T1, T2, T3, ... T8 for the MFI class of zeolites (ZSM-5) for instance. A great deal of computational studies have gone into trying to determine which sites are the catalytically active ones, but although several claim to have determined the preferred sites or the active sites, none are without serious questions. The only calculations (in my opinion) that aren't seriously flawed by either unrealistic assumptions about the geometry or by electronically unrealistic choice of model compounds (to make the calculations tractable) are those by Tony Hess's group, and the most recent work of van Santen's group: Teunissen, Jansen, and R. A. van Santen, J. Phys. Chem., 99, 1873-9 (1995) (and references therein to A. Hess). "Surface" sites will be constructed from each of these T-sites with one or more connections missing. Here again, there is a continuing argument as to what you end up with. Some have assumed everything ends up with hydroxyl caps. Others are convinced that there are extra bridging oxygens, etc. H H+ H / / / O O O O O | | | / \ / \ - Si - O - Al(-) - O - Si - vs. -Si- O -Si- O -Si- O -Si- | | | | | | | O O O O O O O There is also continuing discussion of whether the "acidic" site is a Bronsted acid (H+) or a Lewis acid (RO3Al), usually depending on whether the reaction conditions are such that elimination of water would be expected. Basically, although there is plenty of opinion, there is no consensus on either where "the" active acid site is in a zeolite, or what that site looks like. You will need to pick what you believe is most reasonable and either prove it, or at least be consistent with it when you associate an olefin with the site, or create an ion pair by protonation of the olefin next to the site. >2. What is the symmetry number for such surface carbenium > ions comparing with corresponding gas phase carbenium > ions? In most cases, I believe, the symmetry number > of surface ions are identical with those in gas phase, > if there is no 2-fold axes. As there is not a consensus on the structure of the acidic site, the structure of the activated complexes which incorporate the site is not known. Much modeling has assumed there is symmetry (at least a plane of symmetry, and sometimes 2 planes of symmetry), but this is definitely not correct. In a zeolite the different T sites each occupy unique assymetric centers. The only way there could be any element of symmetry in an absolute sense would be for every unit cell of the zeolite to also have a similarly bonded activated complex at the same time. This is not a reasonable expectation. Local symmetry may be more important than absolute symmetry, however. For this, you would have to consider each of the T sites in the zeolite you are interested in, and decide how locally you want to define the system. Then there may be a few sites with a local plane of symmetry, but for the most part these won't be symmetric either. Calculations I had done on the T2 site in MFI zeolites showed that you will have to include at least 2 layers of silicon-oxygen bonds away from the acidic site to even directionally calculate the relative acidities of aluminosilicates and borosilicates. E. Chamot, Modeling Acid Sites in MFI Zeolites with Realistic Geometric Constraints," Preprints Division of Petroleum Chemistry, Inc. Vol 37, No. 2, ACS San Francisco Meeting, March, 1992. I assume the symmetry "number" you are asking about is defined by the engineering software or method you are planning on using to predict energies (CHETAH? Benson?) Chemically, the symmetry group would be C1: no elements of symmetry, unless by coincidence one of the sites nearly has 2-fold symmetry when considered locally. (Orienting the carbenium ion next to the active site will not increase the symmetry, obviously, only decrease the symmetry.) > >**** About the structure of activated complex ****** > >Next problem that I have to solve is concerning the structure >of activated complex in transition state for beta-scission. > >That is, an elementary step representing beta-scission, > > R1 ---> (Transition State) ----> R2 + O > >where R1 and R2 are the reactant and product carbenium ion, >respectively, and O is the product olefin. > >What I want to know is the structure of the activated complex >in the transition state, as detailed as possible in a certain >theoretical way; how many bonds are related, which bonds are to >be broken, which bond will be formed, and what does the activated >complex look like(is it more like carbenium ion or more like >olefin?), etc. > >Is there any reference/software to describe the structure >of activated complex for this reaction in a theoretical way? This is something that would be quite amenable to computation, once one determined the appropriate structure of the catalytic site to use when building the activated complex. A Quantum Mechanical method would have no problem determining where bonds actually existed, what the atom-atom distances and angles were, or how the electronics would reorganize to distribute partial charges and bonds. The software is certainly there, but again, I am not aware of any consensus on the correct system to model, however. >**** About heat of formation for gas phase carbenium ions *** > >My final question is about the calculation of thermodynamic data, >more exactly, the heat of formation of gas-phase carbenium ions. . . . >The only solution I found recently is using a certain quantum >chemistry package such as "MOPAC". My question is that is this >a suitable package (accurate enough) to calculate the heat of >formation data I need? Yes, this would be an entirely appropriate use of computational chemistry: calculating a thermodynamic value or series of thermodynamic values that are difficult to come by experimentally. MOPAC will do this quite easily, and should be able to handle molecules and ions large enough for your purposes (up to 70-100 atoms with reasonable compute power.) Moreover, although one normally thinks of zeolites as inorganic, the zeolite structure turns out to be very covalent in nature and MOPAC (which contains parameters for Si and Al) works very well for modeling the system when compared to high level ab initio calculations (see Hess's work). As you point out, you need to consider the accuracy of the calculation. Ab initio methods can always be made more accurate, but you rapidly become limited by how large a molecule you can handle: you may be limited to C6 and less (especially if you are to consider all isomers.) Heats of formation have been calculated with MOPAC using the AM1 parameterization and in the case of a series of cations the calculations reproduced the experimentally available data to within 4.7 kcal on the average: Dewar, Zoebisch, Healy, and Stewart, J. Amer. Chem. Soc., 13, 3901 (1985). Not too exciting, but if you can use RELATIVE energies (especially between "isodesmic" reactions) you can probably do a lot better. > If so, how can I obtain a copy of this >package? If not, is there any other software in the field of >computational chemistry to calculate these data? MOPAC is available in several forms. I believe some version of it is still available from the QCPE for a nominal charge. It is sold as part of the CAChe worksystem (which includes a very easy to use interface, and standard procedures already built for using MOPAC), and by MSI/Biosym. The parameters themselves (AM1 and PM3) are part of both Spartan and HyperChem. The URL's I have for these organizations are: CAChe http://www.oxmol.com/ MSI/Biosym http://www.msi.com/ Spartan http://wavefun.com/ HyperChem http://www.hyper.com/ Good luck. And as a final piece of advice, make sure you carve out a piece of this problem that is doable. EC --- Ernest Chamot Consultant in Computational Chemistry Applications Chamot Laboratories, Inc. 530 E. Hillside Rd. Naperville, Illinois 60540 (708) 637-1559 (Voice & Fax) echamot.,at,.xnet.com http://www.xnet.com/~chamotlb Enclosure: RxnScheme.GIF [14,877 bytes] (I cut off this scheme due to its large size. If someone needs this image, please contact me or Mr. Ernest Chamot) From branch -x- at -x- acetsw.amat.comSat Jul 27 10:26:32 1996 Date: Mon, 24 Jun 1996 08:47:48 -0700 (PDT) From: "Michael A. Branch" To: "Park, Tae-Yun" Subject: Re: CCL:M:HELP!(urgent):theoretical discrimication & software contact Mike Schmidt (mike - at - si.fi.ameslab.gov) and ask him if you can get a copy of GAMESS. It can do the transition state searchs, IRC, etc. good luck. Let me know if I can be of assistance. Mike --------------------------------------------------------- Michael A. Branch "I turn big problems into Process Engineer, HDP-CVD little problems." Applied Materials, Inc. (408) 563-0689 Santa Clara, CA 95051 mbranch-: at :-hammerhead.eecs.berkeley.edu --------------------------------------------------------- From guojx(-(at)-)infoc3.icas.ac.cnSat Jul 27 10:41:06 1996 Date: Tue, 25 Jun 1996 08:16:02 +0900 From: Guo Jian-xin To: "Park, Tae-Yun" Subject: Re: CCL:M:HELP!(urgent):theoretical discrimication & software Dear Park, You can got the MOPAC software from the fellowing address at anonymous FTP FTP.www.ccl.net I am wondering what it mean " The Symmetry Number" in your mail, It seems be connected with your questions. Guo, Jian-xin From JDA03546 (- at -) niftyserve.or.jpSat Jul 27 10:42:37 1996 Date: Tue, 25 Jun 1996 13:24:00 +0900 From: "JDA03546 "at@at" niftyse" To: "Park, Tae-Yun" Subject: RE:CCL:M:HELP!(urgent):theoretical discr Hi Park, I'm studying zeolite by means of coputational techniques. Regarding to your post, there are millions of literatures about the computational simulations on zeolites, both by MD and MO. So why don't you search the literature first? Anyway, for the mechanism of the hydrocarbon cracking over zeolite, you may start from the report by R. A. Santen, et. al. appeared on Cat. Lett., 27, p345, 1994, and ibid. 28, p211, 1994. Good luck. --- Teraish Kazuo TEL:0492-66-8375 FAX:0492-66-8359 E-MAIL:JDA03546 - at - niftyserve.or.jp ------------------------------------------------------- PART II Subject: Heat of formation calculation using MOPAC. ------------------------------------------------------- ####################(questions)######################## Hi! I have a question on the parameters in MOPAC. I'm tring to calculate the heat of formation for more or less 100 different gas-phase carbenium ions. Which parameter would be suitable for this calculation? PM3? AM1? I tried to compare the results obtained from MOPAC with a few experimental data available in the literature, and found that AM1 is much closer to those experimental data than the case of PM3 when the molecule contains the cabon atoms up to 5. According to the manual of MOAPC, however, PM3 is more improved parameter than AM1! My question is that the AM1 would also produce better result if the molecule becomes larger (upto C8)? Any suggestion would be greatly appreciated. #####################(answers)########################## From Y0H8797 ( ( at ) ) ACS.TAMU.EDUSat Jul 27 10:55:40 1996 Date: Fri, 5 Jul 1996 12:06:02 -0500 (CDT) From: YONG HUANG To: TP' at \`elptrs7.rug.ac.be Subject: From CCL archive ---------Yong (Texas A&M) From: SMTP%"pitsel $#at#$ chemul.uni.lodz.pl" 2-NOV-1995 06:52:30.77 To: Y0H8797 CC: Subj: CCL:G:summary AM1 vs PM3 Return-Path: Received: from www.ccl.net by VMS2.TAMU.EDU with SMTP; Thu, 2 Nov 1995 6:52:28 -0600 (CST) Received: for chemistry-request %! at !% www.ccl.net by www.ccl.net (8.6.10/950822.1) id DAA27298; Thu, 2 Nov 1995 03:34:28 -0500 Received: from bedrock.ccl.net for owner-chemistry -AatT- ccl.net by www.ccl.net (8.6.10/950822.1) id DAA27135; Thu, 2 Nov 1995 03:20:05 -0500 Received: from chemul.uni.lodz.pl for pitsel -x- at -x- chemul.uni.lodz.pl by bedrock.ccl.net (8.7.1/950822.1) id DAA09331; Thu, 2 Nov 1995 03:19:26 -0500 (EST) Received: by chemul.uni.lodz.pl (5.65/25-eef) id AA18481; Thu, 2 Nov 95 08:19:12 +0100 Message-Id: <9511020719.AA18481 <-at-> chemul.uni.lodz.pl> From: pitsel "-at-" chemul.uni.lodz.pl (Piotr Seliger) X-Mailer: SCO System V Mail (version 3.2) To: chemistry { *at * } ccl.net, jlye { *at * } tx.ncsu.edu, m.t.cronin { *at * } livjm.ac.uk, moralega%a1' at \`lldmpc Subject: CCL:G:summary AM1 vs PM3 Date: Thu, 2 Nov 95 8:19:11 MEZ Sender: Computational Chemistry List Errors-To: ccl()at()www.ccl.net Precedence: bulk This is the summary of the responses I got to my request about AM1 vs PM3 references. Thanks for help. ---------------- ***************************************************************************** Dear Dr. Seliger, Following is a note that I sent to the next over a year ago. Perhaps it will be informative. Our SAM1 papers also extensively documented PMs vs AM1 as well as SAM1. The references for these are: 1. Dewar, M. J. S.; Jie, C.; Yu, G. Tetrahedron 1993, 23, 5003. 2. Holder, A. J.; Dennington, R. D.; Jie, C. Tetrahedron 1994, 50, 627. 3. Holder, A. J.; Evleth, E. M. in Modeling the Hydrogen Bond; Smith, D. A.; American Chemical Society, Washington, DC, 1994; pp 113. Please let me know if I can be of further assistance Regards, Andy Holder -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- UUUU UUU MMM MMKK KKKK CCCC | ANDREW J. HOLDER UU U MM MMK K CC CC | Assoc. Prof. of Comp./Org. Chemistry UU U MMM M MK KK CCC | Dept. of Chemistry UU U M MM MK KK CC CC | University of Missouri-Kansas City UUUUU MMM M MMKK KK CCCC | Kansas City, MO 64110 KK | aholder $#at#$ cctr.umkc.edu K | (816) 235-2293, FAX (816) 235-5502 -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Netters, A few weeks ago, Jeffrey Nauss asked about a comparison between the AM1 and PM3 semiempirical methods. Both of these semiempirical methods are included in most programs that support semiempirical calculations (AMPAC, MOPAC, etc.). Please note that the following discussion is MY OPINION and a compendium of MY EXPERIENCES. I hope you find it somewhat useful. The lead references to each method follows: AM1: Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. J. Am. Chem. Soc. 1985, 107, 3902. PM3: Stewart, J. J. P. J. Comput. Chem. 1989, 10, 209. AM1 stands for "Austin Model 1" and PM3 stands for "Parameterization Method 3". Both methods implement the same basic NDDO theory pioneered by Michael Dewar while at the University of Texas at Austin. The differ- erence is in how the parameters that the semiempirical methods utilize to replace portions of the full ab initio implementation of Hartree-Fock theory. Perhaps the most important difference between AM1 and PM3 is the involvement of the researcher in the parameterization process. PM3 was developed using a largely undirected mathematical optimization process with greatly reduced guidance from chemical knowledge or intuition, an addition that the Dewar methods consider essential. The human researcher knows for which molecules it is necessary to obtain the best fit. For instance, it is useless to obtain parameters for carbon and hydrogen that describe the properties of cubane correctly if the results for benzene are significantly different from experiment. An attentive and knowledgeable chemist can also guide the search into areas of the parameter hypersurface that make sense as far as the absolute magnitude of the parameters themselves are concerned. As with many chemical properties, the parameter values should vary periodically. While this should not unduly constrain the final values, parameters should follow well-defined general trends for proper interaction with other elements. In terms of the actual NDDO model, the actual parameters allowed to vary in the two methods are quite different. In AM1, a large number of values we used from spectroscopy for some of the one-center terms and the other parameters derived with these values fixed. (This is possible only for the lighter elements in the Main Group.) PM3 allowed ALL of these values to float, resulting in substantially more parameters. AM1 also had a quite different concept as to the application of the Guassian functions introduced with AM1 to adjust the core-electron/core- electron repulsion function. Workers in the Dewar group and subsequently in my group see Gaussian functions as PATCHES to the theory, not integral parts. All models fail at some point and the Gaussians were introduced to help with some of the systematic errors in MNDO. Traditionally, these patches were applied to adjust for difficult molecular systems AFTER semiempirical parameters were stabilized. PM3 includes these Gaussian functions (two for each element) FROM THE BEGINNING. Our experience indicates that in such a situtaion, the chemistry os the element will very likely be very strongly effected by the presence of these functions and the importance of the "real", "chemical" parameters will be reduced and swallowed up bu the Gaussians. In short, Gaussians should only be used where absolutely needed, and then viewed with askance. The essence of the difference between the two philosophies is evident: the theoretical basis for the method is either accepted or denied. Significant approximations are made to gain the speed advantage that semiempirical methods enjoy over their ab initio quantum mechanical brethren. But both the ab initio and semiempirical models are based on the Hartree-Fock set of ideas. These ideas possess theoretical rigor as regards solution of the Schrodinger Equation. If one simply views the semiempirical parameters as adjustables within a curve-fit scheme rather than as components of a theoretical model, little faith or importance resides in the meaning of their final values. Simply put, the method of parameterization described above and used so successfully with AM1 and MNDO (and now SAM1) expresses confidence in the theory. With a firmer footing in chemical reality, AM1 parameters are more likely to yield useful results for situations not specifically included in the molecular basis set for parameterization (MBSP). Some Practical Considerations ----------------------------- The differences in errors between the two methods as published are minimal, but that does not relate the real story of how the methods perform differently. Some key points: - PM3 is clearly better for NO2 compounds as a larger number of these were included in the MBSP. - PM3 is usually a little better for geometries, as these were also heavily weighted. - The molecular orbital picture with PM3 is usually different from that expected or that predicted by other methods. This is a direct consequence of the lack of attention paid to the absolute values of Uss and Upp. It can be seen in the lack of performance in ionization potentials. - PM3 charges are usually unreliable, again a result of the rather strange values that some of the parameters take on, even when other experimental data such as heats of formation and geometries are acceptable. This makes PM3 essentially useless for the derivation of molecular m echanics force fields. Perhaps the best known example of this is the case of formamide. The partial charges for the atoms in the molecules are listed below. The lack of any appreciable charge on N has led to a reversal of the actual bond dipole between C and N in this molecule! Atom AM1 PM3 HF/6-31G* --------------------------------------------- O -0.3706 -0.3692 -0.5541 C 0.2575 0.2141 0.5079 N -0.4483 -0.0311 -0.8835 O // H-C \ NH2 - Several papers have been published describing the performance of AM1 vs. PM3: Dewar, M. J. S.; Healy, E. F.; Yuan, Y.-C.; Holder, A. J. J. Comput. Chem. 1990, 11, 541. Smith, D.A. J. Fluor. Chem. 1990, 50, 427 Smith, D.A.; Ulmer, C.W.; Gilbert, M.J. J. Comput. Chem. 1992, 13, 640. - Most reserachers in my experience have stopped using PM3 and have returned to AM1. An Example of Parameterization Values for Aluminum -------------------------------------------------- Parameter AM1 MNDO PM3 Uss, eV -24.353585 -23.807097 -24.845404 Upp, eV -18.363645 -17.519878 -22.264159 zetas, au 1.516593 1.70288 } 1.444161 zetap, au 1.306347 1.073269 betas, eV -3.866822 -0.594301 } -2.670284 betap, eV -2.317146 -0.956550 alpha 1.976586 1.868839 1.521073 Gaussians: Intensity #1, eV 0.090000 - -0.473090 Width #1 12.392443 - 1.915825 Position #1 2.050394 - 1.451728 Intensity #2, eV - - -0.154051 Width #2 - - 6.005086 Position #2 - - 2.51997 The point on the potential surface located by PM3 is significantly different than that located by AM1. This is immediately apparent from the large discrepancy between the Upp values. These are the important one- electron energy values and they have strong influence on the parameter hypersurface. Also, the difference between Uss and Upp for both MNDO and AM1 is about 6 eV. This has been reduced to 2.5 eV in PM3. The real difficulty, however, is in the Beta values. These parameters are the two-center/one- electron resonance terms and are responsible for bonding interactions between atoms. The PM3 values are almost zero, resulting in the conclusion that there is very little bonding between atoms involving aluminum! (Note that the AM1 values for betas and betap spread out around the single MNDO value for beta. This suggests that the MNDO values were reasonable and AM1 adds greater flexibility.) PM3 regains the bonding interactions lost in the low beta values with two strongly attractive Gaussians spanning the bonding region. ******************************************************************************* We did a short note on rotational barriers in branched alkenes: L. A. Burke et al. J. Physical Organic Chem vol 5,614-616(1992). We were surprised at the time that no one else seemed to have published such results. Luke Anthony Burke tel:609-225-6158 Department of Chemistry fax:609-225-6506 Rutgers University e-mail: Camden, NJ 08102 burke &$at$& camden.rutgers.edu USA ****************************************************************************** The comparison of AM1 against PM3 has been quite recently discussed on this list. You may try to search trough archives. My three pens to that discuss may be that AM1 absolutely incorrectly describes interaction in small water clusters; while it was known that it gives not correct hydrogen bond geometry for dimers - I found that it also fails for geometry of larger systems: tetramers, octamers etc. The geometry is absolutely different from what we expect for such clusters (as known from ab-initio and MD studies). In the same time PM3 reproduces these geometries acceptably good - difference in oxygen position between PM3 and HF/6-31G* is ~ 0.1 A for octamers. This is a reason that I am now using PM3 in my Molecular Dynamics studies that use semi-empirical energy surface to derive forces (kind of "quantum" dynamics). Mirek --------------------------------------------------------------- dr Miroslaw Sopek MAKO-LAB Computer Graphics Laboratory ul. Piotrkowska 102a 90-026 LODZ, POLAND tel. (48)(42)332946,322346 fax. (48)(42)332937 e-mail: mako #*at*# pdi.lodz.pl, sopekmir #*at*# mitr.p.lodz.pl --------------------------------------------------------------- ******************************************************************************* 1995 Oct 28 Recently the NET was asked for refs to (1) AM1 compared to PM3, and (2) Sigma-aromaticity. Here are some refs: AM1 cf. PM3 1) Extensive comparison: J Computer-Aided Molecular Design, 4 (1990) Issue 1 (Special issue) ; discusses PM3, AM1 and MNDO 2) W. Thiel, Tetrahedron, 44 (1988) 7393 3) J. J. P. Stewart, J Comp Chem 11 (1990) 543 10 (1989) 209 10 (1989) 221 12 (1991) 320 4) Dewar et al J Comp Chem 11 (1990) 541 5) Smith et al J Comp Chem 13 (1992) 640 6) In a letter to the Net (1995), Andy Holder (SemiChem) said: PM3 is better than AM1 for NO2 compounds and usually a little better for geom's. It is not as good for MO's and is unreliable for charges. Sigma-aromaticity 1) M. J. S. Dewar "Chemical Implications of Sigma Conjugation" J Am Chem Soc 106 (1984) 669 2) Inagaki et al JACS 116 (1994) 5954 3) Ichikawa et al J Phys Chem 99 (1995) 2307 4) Hiberty et al JACS 117 (1995) 7760 =========== Errol Lewars Chem Dept Trent U, Peterborough Ontario Canada ===== *****************************************************************************8 Piotr- We have a paper in press with Spectrochimica Acta comparing AM1 and PM3 for the prediction of carboxylate stretches. Briefly, PM3 is much closer in absolute terms, but AM1 represents differences between compounds more reliably. This is a very limited specific application, of course, and probably only useful to spectroscopists. I would be interested in hearing what others have to say about more general comparisons. Regards, Steve Cabaniss ****************************************************************************** If you mean comparison of conformational energies you might want to have a look at our paper in J.Comp:Chem. 12, 200 (1991). Kind Regards * Klaus Gundertofte * Head, Department of Computational Chemistry * H.Lundbeck A/S * Ottiliavej 9 * DK-2500 Valby - Denmark Fax +45 3630 1385 * Phone +45 3644 2425-3206 * E-mail kgu - at - lundbeck.dk *************************************************************************** Hi Piotr, The performance of this methods in relation to which property ? If you are interested in heats of formation both are OK with about the same results. For minimum energy conformations PM3 has lots of problems. I performed many calculations with PM3, AM1 and ab initio and PM3 is qualitatively wrong in most cases. For electronic properties I didn't tried PM3. Best regards, Edgardo Garcia Cristol Chem & Biochem University of Colorado BOULDER CO USA **************************************************************************** Note: %0 Journal Article %A Gano, J.E. %A Jacob, E.J. %A Roesner, R. %D 1991 %T Evaluation of PM3, AM1, and MNDO for Calculation of Higher Energy Ionization Potentials %B J. Computat. Chem. %V 12 %P 127-134 James E. Gano, Director Instrumentation Center in Arts and Sciences University of Toledo Toledo, Ohio 43606 Instrumentation Center : http://www.icenter.utoledo.edu Department of Chemistry: http://www.chem.utoledo.edu 419-530-7847 419-530-4033 (FAX) ****************************** THE END ********************************* ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Piotr Seliger PPP I TTT SSS EEE L Department of General P P I T S E L and Inorganic Chemistry, PPP I T SSS EE L University of Lodz, P I T S E L Narutowicza 68, P I T SSS EEE LLL 90-136 Lodz, POLAND "The right to knowledge is like E-mail: pitsel-: at :-chemul.uni.lodz.pl the right to life" (G.B.Shaw) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ From ZUILHOF ( ( at ) ) rulgca.LeidenUniv.nlSat Jul 27 10:59:16 1996 Date: Sat, 06 Jul 1996 00:55:02 +0100 (MET) From: Han Zuilhof To: tp -AatT- elptrs7.rug.ac.be Subject: carbenium ions Dear Park, Several years ago I've computed the stability of a variety of alkyl and vinyl cations using AM1, PM3, and several ab initio methods. A preliminary report on this was published: Tetrahedron Lett. 1994, 35, 265. In my experinece AM1 is somewhat better than PM3. It makes no sense to go to ab initio methods when one does not include correlation energy corrections. MP2/6-31G*//6-31G* did best (even better than the MP3 equivalent) of the methods we looked at (AM1, PM3, HF/6-31G*, MP2/6-31G*//6-31G* and MP3/6-31G*//6-31G*). A full report on this issue should be published soon. A copy of the chapter of my Ph. D. thesis -which deals in part with this issue- can be obtained from: dr. Gerrit Lodder; Dept. of Chemistry, Gorlaeus Laboratoria; Rijksuniversiteit Leiden; Postbus 9500; 2300 RA Leiden; Nederland. I hope this will help you. If I can assist you in any other way, please let me know. With best regards, Han Zuilhof ****************************************************************************** ** Dr. Han Zuilhof ** e-mail: ZUILHOF { *at * } chem.chem.rochester.edu ** ** Department of Chemistry ** (optional: ZUILHOF (- at -) rulgca.leidenuniv.nl) ** ** University of Rochester ** ** ** Rochester, NY, 14627 ** Fax: (716) 473-6889 ** ** USA ** Voice: (716) 275-2219 ** ****************************************************************************** ** ** ** "Excite a photochemist!" ** ** ** ****************************************************************************** From echamot;at;xnet.comSat Jul 27 11:01:35 1996 Date: Sun, 7 Jul 96 20:18:00 EDT From: Ernest Chamot To: "Park, Tae-Yun" Subject: Re: CCL:M:Heat of formation calculation using MOPAC. Hi TAE-YUN, I haven't seen a head-to-head comparison of AM1 to PM3 specifically for carbenium ion heats of formation, though there have been several published comparisons of the two methods. I am not surprised, however, that you are finding: >AM1 is much closer to those experimental >data than the case of PM3 when the molecule >contains the cabon atoms up to 5. Even though: >According to the manual of MOAPC, however, >PM3 is more improved parameter than AM1! The main difference, relevant to your use, is that the AM1 parameterization was developed to be theoretically consistent, whereas the PM3 parameterization was developed to most closely reproduce known experimental data. That means that more molecules were used to develop the PM3 parameters than the AM1 parameters (something like 657 vs. 100), but most experimental data is on stable, ground state molecules, so it isn't necessarily as good for species not readily observed: transition states, intermediates, etc. I would expect carbenium ions to be among those difficult to observe species that are not well represented in the parameterizations. An old reference I have shows AM1 coming within 4.7 kcal rms error for a series of cations. Is this comparable to the numbers you are getting? If so, then yes, I would expect that: >the AM1 would also produce >better result if the molecule becomes larger >(upto C8)? EC --- Ernest Chamot Consultant in Computational Chemistry Applications Chamot Laboratories, Inc. 530 E. Hillside Rd. Naperville, Illinois 60540 (708) 637-1559 (Voice & Fax) echamot;at;xnet.com http://www.xnet.com/~chamotlb (Concluding Remarks for part II) I concluded that AM1 is slightly better than PM3 for the calculation of heat of formation for various carbenium ions.