From owner-hyperchem@hyper.com Mon Jul 1 09:40:25 1996 Date: Mon, 1 Jul 1996 10:02:16 GMT X-Sender: haticeo@chem.leeds.ac.uk To: hyperchem@hyper.com From: haticeo@chemistry.leeds.ac.uk Subject: the computer type I would like to ask the computer(PC Pentium process) properties for running HyperChem Relase 4. Thanks a lot. Hatice Ozdemir _______________________________________________________________________________ From owner-hyperchem@hyper.com Mon Jul 1 09:53:21 1996 Date: Wed, 20 Mar 1996 08:33:08 +0000 (GMT) From: Peter Gedeck To: hyperchem@hyper.com Subject: A couple of questions Hello, I have a couple of questions: HyperChem: 1.) How is the environment included in mixed QM/MM systems if CI calculations are carried out? 2.) I marked an atom of an aromatic residue in a protein and selected "expand until sp3". This way the aromatic system of the residue was selected. I then tried a semi-empirical SCF calculation followed by CI. But this failed (problems with capped atoms). What did I do wrong? HyperNMR: 1.) Are the atom types used for TNDO calculations the same as the atom types of the AMBER force field? 2.) Is there a possiblity to reduce the complexity of the simulated NMR spectrum in order to allow the calculation of NMR spectra for a larger number of NMR atoms? Thank you very much in advance. Yours sincerely, Peter Gedeck Dr. Peter Gedeck Centrum f. Computer Chemie - Institut f. Organische Chemie I Naegelsbachstrasse 25 - D-91052 Erlangen Deutschland / Germany Tel: ++9131 - 85 6581 Fax: ++9131 - 85 6565 E-Mail: gedeck@organik.uni-erlangen.de WWW: http://organik.uni-erlangen.de/clark/gedeck _______________________________________________________________________________ From owner-hyperchem@hyper.com Tue Jul 2 15:04:25 1996 Date: Tue, 2 Jul 1996 15:12:17 +1000 (EST) From: Chris Kladis To: hyperchem@hyper.com Subject: MM+ Missing Parameters I am using HyperChem's MM+ force field to do some inorganic molecular mechanic calculations. I'm missing several parameters, and I was wandering if there is anyone who can provide me with these parameters, so that I don't have to use HyperChem's default parameters. Any literature reference would be much appreciated. TORSIONIAL TYPE NH-C4-CO-O1 NH-C4-CO-O2 C4-C4-S2-LP H-C4-S2-LP LP-S2-C4-H N3-C4-CO-O2 O2-CO-C4-NH C4-N2-C3-NC N2-C3-NC-HN HV-N2-C3-NC BOND LENGTH TYPE S2-LP BOND ANGLE TYPE C4-S2-LP LP-S2-LP LP-S2-C4 N2-C3-NC Regards, Mr. Chris Kladis Dept. of Environmental Management Victoria University of Technology Melbourne, Australia _______________________________________________________________________________ From owner-hyperchem@hyper.com Tue Jul 2 16:17:39 1996 Date: Tue, 2 Jul 1996 13:29:04 -0400 (EDT) From: Williams Brian W To: hyperchem@hyper.com Subject: Contour Plots, HOMO/LUMO Charges or Charge Densities I new to computational chemistry, and am attempting to use Hyperchem to better understand charge distributions in molecules which may show solvent sensitivity in their UV or fluorescence spectra. Although Hyperchem lets you see contour plots after completing calculations, the manual is a bit sketchy on how to determine the meaning and numerical magnitude corresponding to a particular contour. I am assuming these correspond to charge densities; are these in atomic units, and what values are taken as defaults for the lowest and highest values? Can these values be changed? I am also particularly interested in finding the differences in charges or charge densities between HOMO and LUMO's- is/are there an easy way to do this? Can a log file give the individual atom charges or charge densities for both HOMO's and LUMO's, so I could then do the calculation manually? Or is their a graphical way to subtract one contour plot from another? I apologize to more sophisicated users if these are common questions-thanks for indulging a tyro. Brian Williams Chemistry Bucknell University _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 3 16:04:59 1996 From: "Thomas Futterer" To: hyperchem@hyper.com Date: Wed, 3 Jul 1996 17:39:12 +100 Subject: ionic radii My question is: Is it possible to render the alkali ions as spheres with the relative ION RADII? I assume, that the default radii are relative ATOM RADII. Thank you for reading, Thomas Futterer _________________________________________________ Thomas Futterer AK Prof.Dr.A.Merz Universitaetsstrasse 31 93053 Regensburg Germany Tel: 0/941/9434633 Fax: 0/941/9434505 __________________________________________________ _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 3 17:15:33 1996 Date: Wed, 3 Jul 96 12:36:56 -0400 From: polowin (Joel Polowin) To: Peter Gedeck , hyperchem@hyper.com Subject: Re: A couple of questions > Date: Wed, 20 Mar 1996 08:33:08 +0000 (GMT) [I think that the date is incorrect!] > From: Peter Gedeck > 1.) How is the environment included in mixed QM/MM systems if > CI calculations are carried out? > > 2.) I marked an atom of an aromatic residue in a protein and > selected "expand until sp3". This way the aromatic system > of the residue was selected. I then tried a semi-empirical > SCF calculation followed by CI. But this failed (problems > with capped atoms). > What did I do wrong? The inclusion of CI does not affect either of the above. The environment (i.e., non-selected atoms) is included in QM calculations in HyperChem by the charges on those atoms contributing to a static electric field. Atoms without charges do not contribute. Non-selected atoms which are bonded to selected atoms are treated as "capping atoms", with different parameters intended to give a proper boundary to the QM region. But depending on the atom types of the capping atoms, and the semi-empirical method that you choose, there may not be parameters in the standard data set for those capping atoms. You may need to look at the parameter files with a text editor -- the parameters for the capping atoms are given at the ends of the files, with negative atomic numbers indicating the capping atom lines. The files and formats are described in Appendix C of the main _Reference_ manual. > HyperNMR: > 1.) Are the atom types used for TNDO calculations the same as > the atom types of the AMBER force field? For the most part, yes. There are a few extra atom types in the HyperNMR parameter set, for a "C1" carbon and for elements in the first two periodic rows which are not included in AMBER. > 2.) Is there a possiblity to reduce the complexity of the > simulated NMR spectrum in order to allow the calculation > of NMR spectra for a larger number of NMR atoms? For a given structure, it is possible to select which atoms are to be included in the semi-empirical calculation (as with the mixed-mode calculations in HyperChem). Of those atoms, you can select which should be included in the NMR calculation for determining shifts and coupling constants. You can then reduce the number of atoms in the NMR set, keeping the atoms that you want to focus on (if necessary by cutting and pasting the tables of shifts and coupling constants) for the spectral calculation. The problem is that the interactions of all of the spin states must be considered before their respective contributions to the final spectrum can be determined (as I understand things), so it is not possible to reduce the computational problem by setting some kind of cutoff value for the spectrum. We are working on an updated version of HyperNMR that will allow calculations of spectra for larger systems, but we are not yet sure when it will be ready. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 3 20:00:16 1996 Date: Wed, 3 Jul 96 17:51:00 -0400 From: polowin (Joel Polowin) To: "Thomas Futterer" , hyperchem@hyper.com Subject: Re: ionic radii > From: "Thomas Futterer" > Date: Wed, 3 Jul 1996 17:39:12 +100 > > Is it possible to render the alkali ions as spheres with the relative > ION RADII? > I assume, that the default radii are relative ATOM RADII. In HyperChem, atoms are drawn with their default radii regardless of the applied charges. The default drawing radius of an atom is 0.75 times its van der Waals radius except when that value was not available, in which case the covalent radius is used. In order to change the drawing radius, it is necessary to use a package such as the ChemPlus Molecule Presentations module which allows finer control over the rendering. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. _______________________________________________________________________________ From owner-hyperchem@hyper.com Thu Jul 4 16:02:00 1996 Date: Thu, 4 Jul 96 12:41:04 -0400 From: polowin (Joel Polowin) To: Williams Brian W Subject: Re: Contour Plots, HOMO/LUMO Charges or Charge Densities Cc: hyperchem@hyper.com > Date: Tue, 2 Jul 1996 13:29:04 -0400 (EDT) > From: Williams Brian W > > I new to computational chemistry, and am attempting to use Hyperchem to > better understand charge distributions in molecules which may show > solvent sensitivity in their UV or fluorescence spectra. Although > Hyperchem lets you see contour plots after completing calculations, the > manual is a bit sketchy on how to determine the meaning and numerical > magnitude corresponding to a particular contour. I am assuming these > correspond to charge densities; are these in atomic units, and what > values are taken as defaults for the lowest and highest values? Can these > values be changed? Well, the contours correspond to charge densities if that's what you're plotting (as opposed to electrostatic potential or spin density). The units are atomic units. The default values for the starting contour and increments depend on the range of values in the plot; you can find out what the values are by displaying the plot, then opening the Grid Control dialogue box and switching the "Starting value" and "Increment" from "Default" to "Other" -- the current values will be displayed in the boxes. You can change these values if you wish. > I am also particularly interested in finding the differences in > charges or charge densities between HOMO and LUMO's- is/are there an easy > way to do this? Can a log file give the individual atom charges or > charge densities for both HOMO's and LUMO's, so I could then do the > calculation manually? Or is their a graphical way to subtract one contour > plot from another? I apologize to more sophisicated users if these are common > questions-thanks for indulging a tyro. There is no easy way to do the subtraction. The log file can give the individual atom charges (see that section of Appendix A) but not charge density data. HyperChem for SGI is able to save 3D grid data, including charge density data, in a HIN file, but the current versions of HyperChem for Windows do not have that function. Better output of this kind of information is in our development wish list. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. _______________________________________________________________________________ From owner-hyperchem@hyper.com Sat Jul 6 13:20:43 1996 Date: Sat, 6 Jul 1996 16:09:09 +0000 (GMT) From: Peter Gedeck To: hyperchem@hyper.com Subject: Some more questions on HyperNMR I have some more questions on HyperNMR 1.) In NMR_SHFT.ABP and NMR_CUPL.ABP entries for bromine exists. Still it is not possible to include Br in HyperNMR calculations. Why? 2.) I tried to calculate the molecule CHFBrCl with HyperNMR. As pointed out in question (1) Br can't be included in the SCF calculations. Therefore I included only C, H, F and Cl as quantum atoms and H as NMR atom. Starting the calculation gives the Error Message: ERROR: Atomic number 35 is too high when running a semi-empirical NMR calculation, the allowed maximum atomic number is 18. Why do I get this error message? I did not include Br (35) in the calculation. 3.) If one gets a molecule from HyperChem into HyperNMR, the molecular force field and thereby the atom types are changed to Amber. As was pointed out in the answer to my last question, the atom types of HyperNMR are an extension of the Amber types. Does HyperNMR reacalculate the atom types after the geometry is obtained or does it rely on the assignments of HyperChem? Is it better to always calculate atom types in HyperNMR? 4.) In a log file the shielding constants of all (quantum ?) atoms are given, not only the selected NMR atoms. Why? It seems that HyperNMR calculates all possible information but behaves as if only a selection is calculated. Assuming that one wants to first calculate the H-NMR spectra for one part of the molecule and than for another part of the molecule. So I would specify the hydrogens of the first part as NMR atoms and calculate shielding, coupling and the spectrum. Then I specify th hydrogens of the second part as NMR atoms and again have to calculate shielding and coupling before the spectrum is calculated. And this although this shielding and coupling information is already vailable (see Log file). Is this observation correct? 5.) Is a shielding of 0 ppm the shielding of a single atom? 6.) It is necessary to select one or more C atoms as labelled atoms in order to be able to calculate a C13-NMR spectrum. Is this true or am I doing something wrong? If not this leads to the inclusion of all C-C couplings in the C13-NMR spectrum. But, experimentally C-C couplings are usually not observed. An experimental C-NMR spectrum is more like a mixture of NMR spectra of molecules where only one carbon is a C-13 and not all. Regards, Peter Gedeck gedeck@organik.uni-erlangen.de _______________________________________________________________________________ From owner-hyperchem@hyper.com Mon Jul 8 19:23:59 1996 Date: Mon, 8 Jul 96 16:30:18 -0400 From: polowin (Joel Polowin) To: Peter Gedeck Subject: Re: Some more questions on HyperNMR Cc: hyperchem@hyper.com > Date: Sat, 6 Jul 1996 16:09:09 +0000 (GMT) > From: Peter Gedeck > > 1.) In NMR_SHFT.ABP and NMR_CUPL.ABP entries for bromine exists. Still it is > not possible to include Br in HyperNMR calculations. Why? It appears that the NMR parameter files were created by editing the parameter files used for CINDO by HyperChem. The lines for bromine in both of the HyperNMR files are identical (apart from formatting) with the bromine line in CINDO.ABP . All of the values in those lines in the HyperNMR files should have been set to 0, but it really does not matter much since HyperNMR will not do calculations with Br anyways. > 2.) I tried to calculate the molecule CHFBrCl with HyperNMR. As pointed out > in question (1) Br can't be included in the SCF calculations. Therefore I > included only C, H, F and Cl as quantum atoms and H as NMR atom. Starting the > calculation gives the Error Message: > ERROR: Atomic number 35 is too high when running a semi-empirical > NMR calculation, the allowed maximum atomic number is 18. > Why do I get this error message? I did not include Br (35) in the > calculation. A non-quantum atom which is bonded to a quantum atom is designated as a "capping" atom; it is included in the quantum calculation, and is given special parameters because of its special status at the quantum-classical boundary. Please see the section "Quantum Atoms" on page 8 of Chapter 3, and the grey box on page 72, in the HyperNMR reference manual; see also information about capping atoms in the HyperChem manuals. > 3.) If one gets a molecule from HyperChem into HyperNMR, the molecular force > field and thereby the atom types are changed to Amber. As was pointed out in > the answer to my last question, the atom types of HyperNMR are an extension > of the Amber types. Does HyperNMR reacalculate the atom types after the > geometry is obtained or does it rely on the assignments of HyperChem? Is it > better to always calculate atom types in HyperNMR? When a structure is transferred from HyperChem to HyperNMR, the HyperChem molecular mechanics force field type is changed to AMBER. If the initial force field setting was some other type -- even if HyperChem was actually set up for a semi-empirical or _ab initio_ calculation -- this will recalculate the atom types for AMBER. If the initial force field setting was for AMBER, the types are not recalculated. Then the structure, with whatever atom types are attached, is transferred to HyperNMR. If all of the atoms in HyperChem have types with AMBER, then it is not necessary to calculate types in HyperNMR; the AMBER types in HyperChem are a proper subset of those in HyperNMR, and an atom whose type has been determined by the AMBER rules in HyperChem will not have that type changed by HyperNMR. If the structure in HyperChem was set up with the AMBER setting but did not have its types calculated, then when that structure is transferred to HyperNMR its atom types may be incomplete and incorrect. In general, then, it does not hurt and may help to recalculate the atom types in HyperNMR. The only times that I can think of when recalculating the atom types would be bad are if you have changed the typing rules in HyperNMR and/or in HyperChem, or if you have manually set atom types in HyperChem, and you want to use the old types instead of whatever types would be set up by the automatic process in HyperNMR. > 4.) In a log file the shielding constants of all (quantum ?) atoms are given, > not only the selected NMR atoms. Why? It seems that HyperNMR calculates all > possible information but behaves as if only a selection is calculated. > > Assuming that one wants to first calculate the H-NMR spectra for one part > of the molecule and than for another part of the molecule. So I would specify > the hydrogens of the first part as NMR atoms and calculate shielding, > coupling and the spectrum. Then I specify th hydrogens of the second part as > NMR atoms and again have to calculate shielding and coupling before the > spectrum is calculated. And this although this shielding and coupling > information is already vailable (see Log file). Is this observation correct? The log file contains shielding constants for all of the quantum atoms, since that is a trivial calculation from the quantum calculation. And it contains coupling constants for each NMR atom with each quantum atom, since once the calculation is set up for an NMR atom, its coupling with any of the quantum atoms can be derived easily. But the log file does not contain coupling constants for all quantum atoms with all other quantum atoms. > 5.) Is a shielding of 0 ppm the shielding of a single atom? The default reference points are with respect to TNDO calculations on methane. Assuming that you use the default shielding constants and set up the calculation properly, CH4 will give shift values of 0 for both proton and 13C. If you prefer some other reference point (for example, TMS), you just have to do the calculation on that structure and apply the result as a "correction factor" to all of your other results. > 6.) It is necessary to select one or more C atoms as labelled atoms in order > to be able to calculate a C13-NMR spectrum. Is this true or am I doing > something wrong? This is correct. By default, all atoms are assumed to be of the most common natural isotope: 1H, 12C, etc. > If not this leads to the inclusion of all C-C couplings in the > C13-NMR spectrum. But, experimentally C-C couplings are usually not > observed. An experimental C-NMR spectrum is more like a mixture of NMR > spectra of molecules where only one carbon is a C-13 and not all. There is no simple way for the program to know if the user wants a spectrum in which atoms have natural abundances (i.e., two adjacent 13C atoms are extremely rare) or one in which the structure is artificially enriched. I have added to our development wishlist that it would be nice to find some way to distinguish the two cases. For now, the simple answer is to do the calculation of shifts and coupling constants, then edit the table of coupling constants to set all of the C--C values to 0. To simulate a broadband decoupled spectrum, you would uncheck the "Fermi spin-spin coupling constants" box in the "Shielding & Coupling Options" dialogue box. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 10 14:31:28 1996 From: Harald Lanig Subject: Script problem To: hyperchem@hyper.com Date: Wed, 10 Jul 1996 12:40:33 +0200 (MET) Dear HyperChem users, a short question about script programming: Is it possible to select an atom via a dialog box? Can I tell HyperChem values of variables within a script? My idea is writing a script that asks for an atom number and a radius to select all residues within a 3D sphere. To do this, there should be the possibility to use a dialog box for filling a variable. I know the commands to activate select by sphere, but I found no possibility to supply the radius... Example: select all residues of p450cam within 15 Angstroms around the Fe atom! Thanks for any information -Harry --------------------------------------------------------------------------- Dr. Harald Lanig Computer Chemie Centrum der Dipl.-Chem. Universitaet Erlangen/Nuernberg Institut fuer Organische Chemie I phone: +49(0)9131-856581 Naegelsbachstr. 25 fax: +49(0)9131-856565 D-91052 Erlangen email: lanig@organik.uni-erlangen.de Germany http: //www.organik.uni-erlangen.de/clark/lanig --------------------------------------------------------------------------- _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 10 18:49:29 1996 Date: Wed, 10 Jul 96 16:09:02 -0400 From: polowin (Joel Polowin) To: Harald Lanig Subject: Re: Script problem Cc: hyperchem@hyper.com > From: Harald Lanig > Date: Wed, 10 Jul 1996 12:40:33 +0200 (MET) > > Is it possible to select an atom via a dialog box? Yes: the Select/Select... dialogue box allows you to select an atom by molecule and atom number. Or, if a name has been applied to that atom, it can be selected by that name. > Can I tell HyperChem values of variables within a script? HyperChem 4 does not have anything like this. HyperChem 4.5 has a few functions for working with variables in scripts, but for the most part they do not work properly. To work with variables, at the moment, you need to use controlling software with a bit more power in its programming abilities -- Visual Basic, Excel, etc. Even Microsoft Word can be used in this way. > My idea is writing a script that asks for an atom number > and a radius to select all residues within a 3D sphere. > To do this, there should be the possibility to use > a dialog box for filling a variable. > I know the commands to activate select by sphere, but I > found no possibility to supply the radius... There is not yet such a script command; the suggestion is on the development wish list. The existing commands only switch back and forth between selection by sphere and selection by rectangle, with the mouse. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 10 22:08:31 1996 Date: Wed, 10 Jul 1996 14:28:55 -0700 X-Sender: lavelle@ewald.mbi.ucla.edu To: hyperchem@hyper.com From: lavelle@ewald.mbi.ucla.edu (Laurence Lavelle) Subject: Atom Limit for HC. What is the atom limit for HyperChem 4.0 and how can one change it. Thanks Laurence Lavelle """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""" Laurence Lavelle University of California Los Angeles Molecular Biology Institute Los Angeles, CA90095 Email:LAVELLE@MBI.UCLA.EDU Phone:(310)206-8270 Fax:(310)825-0982 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""" _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 10 23:57:50 1996 Date: Wed, 10 Jul 96 22:30:38 -0400 From: polowin (Joel Polowin) To: lavelle@ewald.mbi.ucla.edu (Laurence Lavelle), hyperchem@hyper.com Subject: Re: Atom Limit for HC. > Date: Wed, 10 Jul 1996 14:28:55 -0700 > From: lavelle@ewald.mbi.ucla.edu (Laurence Lavelle) > > What is the atom limit for HyperChem 4.0 and how can one change it. What do you mean by "atom limit"? If you're referring to the maximum number of atoms that the program can handle, that depends on the structure(s). The fundamental limitation is due to a data structure limited to 32,000 items: atoms, residues, molecules, structural restraints, etc. This is built into the program and can't be changed. It leads to a maximum of about 16,000 atoms for structures built from residues (since these include residue information and a lot of structural restraints) and perhaps 25,0000 to 28,000 atoms for other kinds of structures, depending on whether you have a lot of little molecules or a few large ones. The SGI version of HyperChem 4.5 does not have these limitations. Obviously, structures as large as these will cause other problems in geometry optimisations because of the memory requirements for the calculations. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. _______________________________________________________________________________ From owner-hyperchem@hyper.com Thu Jul 11 16:07:10 1996 Date: Thu, 11 Jul 1996 18:26:55 +0001 From: Konstantin Pokhodnia Subject: 3D contour plot X-Sender: pokhodnia@cathy.ijs.si To: hyperchem@hyper.com Dear HyperChem users, Is it possible to plot with ChemPlus a 3D contour of spin or charge density distribution? I'll appreciate any help you can offer. Thank you K.Pokhodnia _______________________________________________________________________________ From owner-hyperchem@hyper.com Thu Jul 11 20:10:25 1996 Date: Thu, 11 Jul 96 16:59:21 -0400 From: polowin (Joel Polowin) To: Konstantin Pokhodnia , hyperchem@hyper.com Subject: Re: 3D contour plot > Date: Thu, 11 Jul 1996 18:26:55 +0001 > From: Konstantin Pokhodnia > > Is it possible to plot with ChemPlus a 3D contour of spin or charge density > distribution? Not easily. That information cannot be sent directly from HyperChem to the Molecule Presentations module for display. If you can get the data in the necessary 3D grid form, then you could reformat that data in a way that the module could read and display, but there is no simple way to get the necessary data from HyperChem. (I can send information about the 3D grid data format that Molecule Presentations can read; or see our WWW home page under Product Support, Frequently Asked Question #16.) HyperChem Lite has improved graphics, and can display charge density and spin density as the 3D surfaces. These functions will be available in the next release of HyperChem. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. _______________________________________________________________________________ From owner-hyperchem@hyper.com Fri Jul 12 19:00:48 1996 Date: Fri, 12 Jul 1996 13:55:32 -0700 X-Sender: lavelle@ewald.mbi.ucla.edu To: hyperchem@hyper.com From: lavelle@ewald.mbi.ucla.edu (Laurence Lavelle) Subject: HyperChem 5.0 Hi Joel Thanks for the excellent support. What follows are a couple of comments and questions about PC's and Windows versions of HyperChem. HyperChem is a good product, with its strength being ease of use as a computational tool. However I find its graphical interface limiting, e.g., limited ease of manipulating one molecule to interact with a second molecule. Also with respect to Ver. 5 for Windows, the general trend is that PC versions are usually limited with respect to other versions (e.g., SG). I hope with the rapid increase in both computational and graphical ability of PC's that future Windows versions of HyperChem will be as sophisticated as those for e.g., SG. This trend has not gone unnoticed, for example, Evans & Sutherland now make a sophisticated plug in graphics card for PC's called, "Freedom Graphics" that contains the Evans & Sutherland 3D graphics engine and up to 24MB of RAM. My own PC, which will be 'out of date' by the time HyperChem Ver. 6 is released, and is as good as or faster than many SG machines, is a Dual Pentium Pro 200MHZ, 1MB 2nd cache and 4MB video RAM, with enough potential RAM (up to 192 MB) and storage space (up to 4 HD) to satisfy most needs. (Integrated PC/SCSI-2 controller, CD ROM SCSI interface, magnetic tape 4.0 GB DAT SCSI interface). Of course many of these are on my upgrade wish list. And speaking of upgrade wish lists when is HyperChem Ver. 5 to be released and what are the changes? Will Ver. 5 allow for long filenames and what MS operating systems is it designed for (and what will it work with) e.g., the expected release of Windows NT Workstation 4.0 (9/96) and Windows NT Workstation CAIRO in 97 which both support SMP. Thanks for your support and I look forward to Ver. 5. Sincerely Laurence Lavelle """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""" Laurence Lavelle University of California Los Angeles Molecular Biology Institute Los Angeles, CA90095 Email:LAVELLE@MBI.UCLA.EDU Phone:(310)206-8270 Fax:(310)825-0982 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""" _______________________________________________________________________________ From owner-hyperchem@hyper.com Mon Jul 15 21:51:11 1996 Date: Mon, 15 Jul 96 17:53:35 -0400 From: polowin (Joel Polowin) To: lavelle@ewald.mbi.ucla.edu (Laurence Lavelle), hyperchem@hyper.com Subject: Re: HyperChem 5.0 > Date: Fri, 12 Jul 1996 13:55:32 -0700 > From: lavelle@ewald.mbi.ucla.edu (Laurence Lavelle) > > HyperChem is a good product, with its strength being ease of use as a > computational tool. However I find its graphical interface limiting, e.g., > limited ease of manipulating one molecule to interact with a second molecule. Well, we always welcome suggestions on how to do this kind of thing better. We can't promise to implement them (obviously) but new ideas are welcome. > Also with respect to Ver. 5 for Windows, the general trend is that PC > versions are usually limited with respect to other versions (e.g., SG). I > hope with the rapid increase in both computational and graphical ability of > PC's that future Windows versions of HyperChem will be as sophisticated as > those for e.g., SG. > > This trend has not gone unnoticed, for example, Evans & Sutherland now make > a sophisticated plug in graphics card for PC's called, "Freedom Graphics" > that contains the Evans & Sutherland 3D graphics engine and up to 24MB of RAM. We're always working on this sort of thing. But it's a simple fact that the built-in operations in something like an SGI machine make it much easier to do some kinds of graphical operations than on a PC. The UNIX versions of HyperChem are moving to OpenGL, which leads to much simpler portability. It's hard to say what will happen in the future with the Windows versions. But one thing that we don't want to do is to require that our users purchase extra hardware to run our software; we want to keep everything as generic as possible. > And speaking of upgrade wish lists when is HyperChem Ver. 5 to be released > and what are the changes? Will Ver. 5 allow for long filenames and what MS > operating systems is it designed for (and what will it work with) e.g., the > expected release of Windows NT Workstation 4.0 (9/96) and Windows NT > Workstation CAIRO in 97 which both support SMP. I do not have much information yet which is sufficiently stable/reliable to release to the public. HyperChem 5 will be available later this year. Based on our historical pattern of a new release every 12 to 14 months, it will probably be out in early autumn. It will be fully 32-bit (in comparison with the current Windows versions, which have 32-bit calculations but a 16-bit graphical front end). This includes native support for Windows 95 and Windows NT. It appears that among the side effects of this will be the elimination of the structure-size limitations in the current Windows versions, use of long filenames, etc. HyperChem 5 will feature considerably improved graphics. This will include the new graphics options in HyperChem Lite, such as ball-and-cylinders rendering, and 3D isosurfaces for orbitals and the related calculated results such as electrostatic potential. There will be other new graphics features as well but the details have not been settled. Similarly, there should be some new computational functions, but I don't have details yet. We are working to make it easier to get HyperChem to work with other software, but it is not yet clear how much of that will be due to modifications in the software and how much to our providing information and support for interfacing. There will be formal announcements in the usual places -- this mailing list, our WWW page, etc. -- when we get closer to the release. We always walk that fine line between wanting to keep people interested, and not wanting to release information prematurely. We don't want to tell people that some feature will be included before we have shown that it can be made to work reliably, and we don't want to make a big promotion about a release date and then not have the software ready until eight months later. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 17 07:23:40 1996 Date: Wed, 17 Jul 1996 10:53:10 +0200 (MET DST) From: Birgit Grossmann To: hyperchem@hyper.com Subject: Question conc. spin multipl. for transmet Hello :-) We`ve a problem. We want to calculate a manganese d5-highspin porphyrin complex. The max. spin multiplicity in Hyperchem 3 and 4 is only 4. Are there any extension programms? Or otherwise, does Hyperchem 4.5 have a wider range for spin mult.? Thanks for help _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 17 15:52:17 1996 Date: Wed, 17 Jul 96 12:17:11 -0400 From: polowin (Joel Polowin) To: hyperchem@hyper.com Subject: Re: Question conc. spin multipl. for transmet > Date: Wed, 17 Jul 1996 10:53:10 +0200 (MET DST) > From: Birgit Grossmann > > We`ve a problem. We want to calculate a manganese d5-highspin > porphyrin complex. The max. spin multiplicity in Hyperchem 3 and 4 is > only 4. Are there any extension programms? Or otherwise, does Hyperchem > 4.5 have a wider range for spin mult.? HyperChem releases 4 and 4.5 allow spin multiplicities of not more than 4. I asked our semi-empirical specialist about this, and he said that the errors due to the half-electron approximations could be serious for larger multiplicities. I suggested to him that as long as the user was warned, we should allow the user to use higher values if s/he wants to, since some people do want to calculate high-spin systems. We have also received some literature references about avoiding the errors due to the half-electron approximation. They go beyond my limited understanding of these calculations; I do not know when or if they will be implemented as improvements to HyperChem. With HyperChem Lite, the limit was increased to 6, and I think that the limit in future releases of HyperChem will be at least as high as that. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself. _______________________________________________________________________________ From owner-hyperchem@hyper.com Tue Jul 23 02:28:42 1996 Date: Mon, 22 Jul 96 22:28 MET DST To: hyperchem@hyper.com, listserv@iubvm.ucs.indiana.edu, mailbase@mailbase.ac.uk, listserv@umdd.umd.edu, charmm-bbs-sysop@emperor.harvard.edu, MOL-DIVERSITY@LISTSERV.ARIZONA.EDU Subject: NOW AVAILABLE !!! X-Sender: 0924191216-0001@t-online.de (Dr. Heinz Hofmann) From: Science_direct@t-online.de (Heinz Hofmann) Four weeks after the announcement of the availability of the first in a series of tutorials in computational chemistry, edited by Tim Clark, Science Learning Ltd. has made available an electronic version of "Tutorials in Computational Chemistry : Semiempirical MO Calculations". Priced at US $39, the electronic version is equivalent to the physical book in content and make up. You simply download the electronic book in Adobe Portable Document Format (PDF) and read it with the Adobe Free Reader, Version 2.1, a free program that is available over the Internet (http://www.adobe.com/acrobat/readstep.html). If you like what you see, you pay $39 to obtain a password that unlocks the entire document. For more details, table of contents and an example chapter, please visit the Electronic Bookstore at Science Learning's home page at http://www.sciencelearn.com and download the electronic book file (book1.exe). ** Note : This message is being sent to multiple lists ** ** Our apologies if you receive it more than once ** _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 24 03:16:12 1996 Date: Tue, 23 Jul 1996 22:55:06 -0600 (MDT) From: Paul Scott X-Sender: scottp@gpu3.srv.ualberta.ca To: hyperchem@hyper.com Subject: Hydration energy calculations I want to do some approximate calculations on changes in hydration energy on oligomerization of small protein subunits and seem to have run into a bug (or memory limitation?) in the hydration energy module of ChemPlus 1.5 (ruuning with HyperChem 4.0 on a Pentium 133 with 16 MB RAM). Once the size of the protein reaches around 2100 atoms the calculation crashes with the enigmatic and unhelpful message "Error in DoCalculation". Surprisingly calculations of other additive Qsar properties such as LogP, surface area (grid) and volume keep going but surface area (approximate) crashes at the same time as the hydration energy calculation. Is there a fix or a work-around for this problem? Any help would be much appreciated. Paul G. Scott Department of Biochemistry University of Alberta Edmonton Alberta T6G 2H7 CANADA Tel: 403-492-5755 Fax: 403-492-6361 e-mail: scottp@gpu.srv.ualberta.ca. _______________________________________________________________________________ From owner-hyperchem@hyper.com Mon Jul 29 09:20:47 1996 Date: Mon, 29 Jul 96 11:22:13 BST From: szr@tcxs1.dl.ac.uk (S. Rogers) To: hyperchem@hyper.com Subject: Proton Affinities Hi, I've come accross what seems to be a problem with semi-empirical QM. I have been calculating proton affinities for a range of organic compounds as the enthalpy change for the reaction: R + H+ ---> RH+ Most compounds seem to give sensible answers but for acetonitrile (CH3CN) this reaction is predicted to be highly endothermic using AM1 or PM3. Ab initio treatment gives it as exothermic which is what I would expect since the nitrogen is generally thought to be quite basic (eg forms strong hydrogen bonds). Has anyone got any useful comments on this anomoly? The semi-empirical charges also look suspect. Steve Rogers _______________________________________________________________________________ From owner-hyperchem@hyper.com Mon Jul 29 10:53:58 1996 From: "Jens Spanget-Larsen" Organization: Roskilde Universitetscenter To: hyperchem@hyper.com, szr@tcxs1.dl.ac.uk (S. Rogers) Date: Mon, 29 Jul 1996 13:59:57 +0100 Subject: Re: Proton Affinities > Date: Mon, 29 Jul 96 11:22:13 BST > From: szr@tcxs1.dl.ac.uk (S. Rogers) > To: hyperchem@hyper.com > Subject: Proton Affinities > > Hi, > > I've come accross what seems to be a problem with semi-empirical QM. I have > been calculating proton affinities for a range of organic compounds as the > enthalpy change for the reaction: > > R + H+ ---> RH+ > > Most compounds seem to give sensible answers but for acetonitrile (CH3CN) this > reaction is predicted to be highly endothermic using AM1 or PM3. Ab initio > treatment gives it as exothermic which is what I would expect since the nitrogen > is generally thought to be quite basic (eg forms strong hydrogen bonds). > > Has anyone got any useful comments on this anomoly? The semi-empirical charges > also look suspect. > > Steve Rogers > Hello Steve! I have no precise answer to offer, but I can tell you that structural elements with CN bonding seems to be problematic with procedures such as AM1 and PM3; the energetics of CN mono-, double- and triple-bonding is not reproduced correctly. Now and then, papers mention this discrepancy, but I have no reference at hand. Since you are calculating Proton Affinities, the following paper may be of interest to you: J. Spanget-Larsen: J.Phys.Org.Chem. 8, 496-505 (1995) Yours, Jens >--< =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- JENS SPANGET-LARSEN Department of Chemistry Phone: +45 46757711 Roskilde University (RUC) Fax: +45 46757721 POB 260, DK-4000 Roskilde, Denmark E-Mail: JSL@virgil.ruc.dk =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- _______________________________________________________________________________ From owner-hyperchem@hyper.com Mon Jul 29 20:34:06 1996 Date: Mon, 29 Jul 96 17:33:44 -0400 From: polowin (Joel Polowin) To: szr@tcxs1.dl.ac.uk (S. Rogers) Subject: Re: Proton Affinities Cc: hyperchem@hyper.com > Date: Mon, 29 Jul 96 11:22:13 BST > From: szr@tcxs1.dl.ac.uk (S. Rogers) > > I've come accross what seems to be a problem with semi-empirical QM. I have > been calculating proton affinities for a range of organic compounds as the > enthalpy change for the reaction: > > R + H+ ---> RH+ > > Most compounds seem to give sensible answers but for acetonitrile (CH3CN) this > reaction is predicted to be highly endothermic using AM1 or PM3. Ab initio > treatment gives it as exothermic which is what I would expect since the nitrogen > is generally thought to be quite basic (eg forms strong hydrogen bonds). I don't seem to duplicate your results. I optimised CH3CNH(+) with AM1 and got a total energy of -11758.37 kcal/mol and a heat of formation of 196.0357 kcal/mol. Optimised acetonitrile gave a total energy of -11620.29 and a heat of formation of -11620.29; H+ had a total energy of 0 and a heat of formation of 333.6545 kcal/mol. Either by total energy or by heat of formation, CH3CN + H+ --> CH3CNH+ seems to be exothermic by about 138.08 kcal/mol. I get just about exactly the same result if I start with the protonated species, increase the N--H distance to about 20 A, and do a restrained optimization. Would you please double-check your method -- make sure that you are including all of the atoms in the calculation, for example, and that you have correctly specified the charge and multiplicity -- and let me know if you still get the same result? Joel polowin@hyper.com _______________________________________________________________________________ From owner-hyperchem@hyper.com Mon Jul 29 21:24:03 1996 Date: Mon, 29 Jul 1996 17:39:39 -0500 (CDT) From: Phillip Stafford To: Jens Spanget-Larsen Cc: hyperchem@hyper.com Subject: Endothermic? The CH3CN (Acetonitrile) reaction is endothermic, at least in my experience in the lab. Most reactions I've run across using acetonitrile get very cold immediately after the reaction begins. Phillip Stafford Department of Biological Sciences University of Arkansas http://comp.uark.edu/~pstaffor _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 31 11:04:08 1996 From: kreidler@IRIS.rz.uni-duesseldorf.de (Kay Kreidler) Subject: Problems with semiempirical calculations with partially fixed geometry To: hyperchem@hyper.com Date: Wed, 31 Jul 1996 12:49:52 +0200 (MDT) Dear Hyperchem users! I have a problem with a semiempirical calculation (AM1 or PM3). To get the rotational barrier of for example butane, I fixed the main dihedral angle and optimized the rest of the molecule. When trying to optimze such a structure with a fixed angle, following error message appeared: Error message received from node=0 Capping atom 5 with atomic number 1: the slater exponent of s orbital is less than or equal to zero, please check your parameter file pm3_1.abp. Doing a single point calculation caused the same problem. What could be the reason for this message? The Z-matrix of the molecule is the following: Name: C:\TMP\BUTAN.ZMT MOPAC file created on 31/7 11:43:56 1996 by HYPERCHEM C 00000.0000 0 00000.0000 0 00000.0000 0 0 0 0 C 00001.5400 1 00000.0000 0 00000.0000 0 1 0 0 C 00001.5400 1 00109.4709 1 00000.0000 0 2 1 0 C 00001.5400 1 00109.4709 1 00179.9999 1 3 2 1 H 00001.0899 1 00109.4709 1 00179.9999 1 1 2 3 H 00001.0899 1 00109.4709 1 00299.9994 1 1 2 3 H 00001.0899 1 00109.4714 1 00059.9997 1 1 2 3 H 00001.0900 1 00109.4710 1 00299.9994 1 2 1 5 H 00001.0900 1 00109.4714 1 00059.9997 1 2 1 5 H 00001.0899 1 00109.4709 1 00179.9999 1 3 2 8 H 00001.0899 1 00109.4714 1 00180.0000 1 3 2 9 H 00001.0899 1 00109.4709 1 00060.0005 1 4 3 10 H 00001.0900 1 00109.4709 1 00059.9997 1 4 3 11 H 00001.0899 1 00109.4714 1 00179.9999 1 4 3 11 0 Thanks a lot. Kay. _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 31 14:12:54 1996 Date: Wed, 31 Jul 96 11:05:25 -0400 From: polowin (Joel Polowin) To: kreidler@iris.rz.uni-duesseldorf.de Subject: Re: Problems with semiempirical calculations with partially fixed geometry Cc: hyperchem@hyper.com > Date: Wed, 31 Jul 1996 12:49:52 +0200 (MDT) > From: kreidler@IRIS.rz.uni-duesseldorf.de (Kay Kreidler) > > I have a problem with a semiempirical calculation (AM1 or PM3). > To get the rotational barrier of for example butane, I fixed > the main dihedral angle and optimized the rest of the molecule. > > When trying to optimze such a structure with a fixed angle, > following error message appeared: > > Error message received from node=0 Capping atom 5 with atomic number 1: > the slater exponent of s orbital is less than or equal to zero, please > check your parameter file pm3_1.abp. The problem is that you have fixed the angle by selecting only some atoms in the workspace before running the calculation. When you run a semi-empirical calculation with some atoms selected, only the selected atoms are fully included in the calculation. Unselected atoms which are bonded to selected atoms are included as "capping atoms", with special parameters so that they give a good boundary between the quantum and classical parts of the system. The remaining unselected atoms contribute to the calculation only by their charges modifying the potential field. In your case, there are no PM3 parameters for using H -- atomic number 1 -- as a capping atom. If you wish to do the calculation in this way, you would need to add parameters to the file; see Appendix C of the _Reference_ manual, and comments about capping atoms in _Computational Chemistry_. But the more appropriate way to do the kind of calculation you describe is to fix the torsional angle by adding a restraint to it. You would need to select the angle, use Select/Name Selection to give a name to that set of atoms, and then use Setup/Restraints to add an extra potential function to keep that set of atoms at the desired angle during optimisation. If the energy of the system is of interest to you, then after the optimisation you should remove the restraint and do a single-point calculation on the system to give the energy without the contribution of the extra function. While you are running these calculations, no atoms should be selected. Joel polowin@hyper.com _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 31 17:26:14 1996 Date: Wed, 31 Jul 1996 20:04:26 +0200 From: allin@helios.bph.ruhr-uni-bochum.de (Christoph Allin) To: hyperchem@hyper.com Subject: Convergence Problem Dear Joel, we try to minimize the structure of H2PO4 complexed with one Mg. Our intention is to calculate the normal modes of this system. Therefore we are interested in the well relaxed structure (very small gradient). We used the following input data for the calculations (extract of the log file): ******************************************************************************* Geometry optimization, AbInitio, molecule = /usr/people/roman/hyperch/po4h2mg/poplar4.hin. AbInitio PolakRibiere optimizer Convergence limit = 0.0000100 Iteration limit = 50 Accelerate convergence = YES Optimization algorithm = Polak-Ribiere Criterion of RMS gradient = 0.0010 kcal/(A mol) Maximum cycles = 120 The initial guess of the MO coefficients is from a projected INDO calculation. Shell Types: S, S=P, 6D. RHF Calculation: Singlet state calculation Number of electrons = 60 Number of Doubly-Occupied Levels = 30 Charge on the System = 1 Total Orbitals (Basis Functions) = 108 Primitive Gaussians = 230 Starting HyperGauss calculation for IRIX 5.x versions 4.5 (Serial No. 600-10000000) with 108 basis functions and 230 primitive Gaussians 2-electron Integral buffers will be 32000 words (double precision) long. Two electron integrals will use a cutoff of 1.00000e-10 Raffenetti 1 integral format is used. ****************************************************************************** After a few days of calculation we have got the problem, that the gradient do not reach a value smaller than 0.012 kcal/(A mol). We have read in the reference manual, that by means of numerical round-off errors, there exists a limit for the value of the RMS gradient of 0.001. But our result is still above this limit. Our question is: What is the reason for this discrepancy. And how it is possible to reach a value of 0.001 or better in our case. The following are the latest steps in our calculations: **************************************************************************** E=-527637.1875 Grad=0.013 Conv=NO(0 cycles 10 points) [Iter=1 Diff=0.00002] E=-527637.1875 Grad=0.013 Conv=NO(0 cycles 10 points) [Iter=2 Diff=0.00000] E=-527637.1875 Grad=0.013 Conv=NO(0 cycles 11 points) [Iter=1 Diff=0.00001] E=-527637.1875 Grad=0.013 Conv=NO(0 cycles 11 points) [Iter=2 Diff=0.00000] E=-527637.1875 Grad=0.013 Conv=NO(0 cycles 12 points) [Iter=1 Diff=0.00001] E=-527637.1875 Grad=0.013 Conv=NO(0 cycles 12 points) [Iter=2 Diff=0.00000] E=-527637.1875 Grad=0.013 Conv=NO(1 cycles 13 points) [Iter=1 Diff=0.00000] E=-527637.1875 Grad=0.012 Conv=NO(1 cycles 14 points) [Iter=1 Diff=0.00000] E=-527637.1875 Grad=0.012 Conv=NO(1 cycles 15 points) [Iter=1 Diff=0.00000] E=-527637.1875 Grad=0.012 Conv=NO(1 cycles 16 points) [Iter=1 Diff=0.00000] E=-527637.1875 Grad=0.012 Conv=NO(1 cycles 17 points) [Iter=1 Diff=0.00000] E=-527637.1875 Grad=0.012 Conv=NO(1 cycles 18 points) [Iter=1 Diff=0.00000] E=-527637.1875 Grad=0.012 Conv=NO(1 cycles 19 points) [Iter=1 Diff=0.00000] E=-527637.1875 Grad=0.012 Conv=NO(1 cycles 20 points) [Iter=1 Diff=0.00000] E=-527637.1875 Grad=0.012 Conv=NO(1 cycles 21 points) [Iter=1 Diff=0.00000] E=-527637.1875 Grad=0.012 Conv=NO(1 cycles 22 points) [Iter=1 Diff=0.00000] E=-527637.1875 Grad=0.013 Conv=NO(1 cycles 23 points) [Iter=1 Diff=0.00000] ***************************************************************************** With best wishes Yours Sincerely Christoph _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 31 19:28:18 1996 Date: Wed, 31 Jul 1996 23:18:37 +0200 (MET DST) From: Birgit Grossmann To: Kay Kreidler Cc: hyperchem@hyper.com Subject: Re: Problems with semiempirical calculations with partially fixed geometry Hi Kay :-) > > I have a problem with a semiempirical calculation (AM1 or PM3). > To get the rotational barrier of for example butane, I fixed > the main dihedral angle and optimized the rest of the molecule. > > When trying to optimze such a structure with a fixed angle, > following error message appeared: > > Error message received from node=0 Capping atom 5 with atomic number 1: > the slater exponent of s orbital is less than or equal to zero, please > check your parameter file pm3_1.abp. Sounds for me like you had only the atoms for the fixed angle activated. Hyperchem can`t calculate only substructures, so it gives that error message. Activate or deactivate the whole molecule. Birgit :-) _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 31 20:22:13 1996 Date: Wed, 31 Jul 96 18:21:43 EDT From: tailkink@chem.QueensU.CA (Evan Vaughan) To: hyperchem@hyper.com Subject: Graphics cards Hi! I was wondering what kind of graphics cards you recommend for using with Hyperchem? Thanks! ================================================================= Evan Vaughan tailkink@silicon.chem.queensu.ca "Practice random acts of kindness and senseless beauty" _______________________________________________________________________________ From owner-hyperchem@hyper.com Wed Jul 31 23:13:40 1996 Date: Wed, 31 Jul 96 21:17:32 -0400 From: polowin (Joel Polowin) To: Birgit Grossmann , Kay Kreidler Subject: Re: Problems with semiempirical calculations with partially fixed geometry Cc: hyperchem@hyper.com > Date: Wed, 31 Jul 1996 23:18:37 +0200 (MET DST) > From: Birgit Grossmann > > Sounds for me like you had only the atoms for the fixed angle activated. > Hyperchem can`t calculate only substructures, so it gives that error > message. Activate or deactivate the whole molecule. HyperChem *can* run calculations on substructures, using any of the molecular mechanics or semi-empirical methods. But with the semi-empirical calculations, it is necessary to have parameters for the "capping atoms". We supply parameters for only a few atoms in the standard data sets that come with HyperChem. In many cases, it would be necessary for a user to add his or her own parameters. For the _ab initio_ calculations, it is necessary to include complete molecules in the calculation; we have not yet got a good quantum/classical boundary solution. Joel ------------ Joel Polowin, Ph.D. Manager, Scientific Support Email to: polowin@hyper.com WWW: http://www.hyper.com/ Hypercube Inc, 419 Phillip St, Waterloo, Ont, Canada N2L 3X2 (519)725-4040 Info requests to: info@hyper.com Support questions to: support@hyper.com Email group: Send "subscribe hyperchem" (or "unsubscribe hyperchem") to hyperchem-request@hyper.com; please do not send such administrative messages to the group itself.