From chemistry-request-0at0-server.ccl.net Sun May 21 17:27:03 2000 Received: from ivory.trentu.ca (trentu.ca [192.75.12.103]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id RAA06087 for ; Sun, 21 May 2000 17:27:02 -0400 Received: from trentu.ca ([204.225.13.50]) by trentu.ca (PMDF V5.2-32 #29543) with ESMTP id <01JPO4JUWASY0027WD (+ at +) trentu.ca> for chemistry (+ at +) ccl.net; Sun, 21 May 2000 17:27:01 EDT Date: Sun, 21 May 2000 17:31:07 -0400 From: elewars Subject: SEMIEMPIRICAL FREQUENCY SCALING SUMMARY To: chemistry <-at-> ccl.net Message-id: <3928559B.8CD8C263 (- at -) trentu.ca> MIME-version: 1.0 X-Mailer: Mozilla 4.7 [en] (WinNT; I) Content-type: text/plain; charset=iso-8859-1 Content-transfer-encoding: 8BIT X-Accept-Language: en 2000 May 20 Here is my summary of the answers to my question on semiempirical vibrational freqs scaling factors. I failed to remember that the "canonical reference" of Scott and Radom also gives AM1 and PM3 factors (Table 10: AM1 0.9532; PM3 0.9761; these are briefly discussed then shunted aside for a lengthy treatment of ab initio and DFT freqs, which with their ZPEs are said to be more accurate). Note that for AM1, PM3 and SAM1 a single correction factor for the whole frequency range is not as good as for ab initio or DFT: I think that if you are interested in C=O stretch, for example, it's better to find 20 or so C=O compounds for which the freqs are experimentally known, calculate the stretching freqs with AM1 or PM3, then plot exp vs calc freqs and get the slope of the best-fit line through the origin (exp=k calc). The resulting factor of about 0.85 is a good way from the broad-range ca. 0.95 factor. Thanks very much to all who responded (#1--#8, below). EL ===== The Question 2000 May 1 Hello, I would appreciate it if someone could send me a ref for AM1 and PM3 vibrational frequency scaling factors, i.e the numbers to multiply a calculated IR frequency by to get (approximately) the observed, experimental frequency ; I know these numbers are about 0.9. If the scaling for SAM1 is available I would welcome getting that too. (If anyone is interested, the canonical reference for ab initio and DFT scaling factors--freqs and ZPEs-- is A. P. Scott and L. Radom, J Phys Chem, 1996, 100, 16502-16513). Thanks E. Lewars ==== The Answers #1 Please see: M. B. Coolidge, J. E. Marlin, and J. J. P. Stewart, "Calculations of Molecular Vibrational Frequencies Using Semiempirical Methods", J. Comp. Chem., 12, 948-952 (1991). ( &$at$& &$at$& ) .-------------oOOo----(_)----oOOo-------------------------------------. | James J. P. Stewart | | | Stewart Computational Chemistry | E-mail: jstewart ( ( at ) ) fujitsu.com | | 15210 Paddington Circle | WWW: http://home.att.net/~MrMOPAC | | Colorado Springs CO 80921-2512 | | | USA .ooo0 | Phone: USA +(719) 488-9416 | | ---------- #2 I am somewhat puzzled by your question: According to "the canonical reference" (i.e., Scott & Radom, Table 1) the IR frequency scaling factors for AM1 and PM3 are equal to 0.9532 and 0.9761, respectively. The corresponding rms errors are quoted as 126 and 159 cm-1. Yours, Jens >--< =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= JENS SPANGET-LARSEN Phone: +45 4674 2000 (RUC) Department of Chemistry +45 4674 2710 (direct) Roskilde University (RUC) Fax: +45 4674 3011 P.O.Box 260 E-Mail: JSL-0at0-virgil.ruc.dk DK-4000 Roskilde, Denmark http://www.rub.ruc.dk/dis/chem/psos/ =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= ------------- #3 Dear Professor Lewars, I may be able to help you if you just want a scaling factor. I don't have a reference, but the CCCDBD (http://srdata.nist.gov/cccbdb/) has a few pages for playing around with vibrational frequency scaling factors. Unfortunately it does not have the AM1 and PM3 data in the public version. In a month I hope to have another release which will allow users to choose a set of molecules and calculate a scaling factor for a much broader selection of methods and basis sets, including AM1 and PM3. If there is a certain class of molecules in which you are interested I can query the internal version here and return a scaling factor. For example, if I choose everything in the database with a CH bond and exclude radicals I get 65 species, 1124 vibrations and a factor of 0.9235 ± 0.0076 for AM1. For PM3 I get 68 species, 1171 vibrations and a factor of 0.9451 ± 0.0071. For all the molecules with experimental vibrational frequencies in the CCCBDB I get AM1 224 species, 1834 vibrations, 0.9199 ± 0.0060 PM3 246 species, 1867 vibrations, 0.9435 ± 0.0058 Hope this is useful. Russ Johnson russell.johnson %-% at %-% nist.gov ========= #4 You are in principle right about the values of ca. 0.9, but this is very tentative. For example, you can take a look for sf's for some nitriles in my papers V.A.Basiuk. IR spectra simulation as auxiliary tool for gas chromatography/Fourier transform IR spectroscopy/mass spectrometry identification of unknown compounds. 2. PM3, AM1, MNDO and MINDO3 simulations for simple nitriles/ Spectrochim. Acta A, 1999, 55, N14, 2771-2782. V.A. Basiuk, R. Navarro-Gonzalez, Y. Benilan and F. Raulin. PM3, AM1, MNDO and MINDO3 semi-empirical IR spectra simulations for some nitriles of interest for Titan's chemistry/ Spectrochim. Acta A, 2000, 56, N6, 1157-1165. But from my experience with other classes of compounds, I can say that sf's can vary significantly depending on the chemical structure. Good luck! *********************************************** Dr. Vladimir Basiuk Instituto de Ciencias Nucleares Universidad Nacional Autonoma de Mexico Circuito Exterior C.U. A. Postal 70-543 04510 Mexico, D.F., MEXICO Tel: (52) 56 16 27 97 (52) 56 22 46 74 Fax: (52) 56 16 22 33 email: basiuk -x- at -x- nuclecu.unam.mx ============ #5 Hi E. Lewars, You asked: >I would appreciate it if someone could send me a ref for AM1 and PM3 >vibrational frequency scaling factors, i.e the numbers to multiply a >calculated IR frequency by to get (approximately) the observed, >experimental frequency ; I know these numbers are about 0.9. >If the scaling for SAM1 is available I would welcome getting that too. The reference you mentioned for ab initio and DFT scaling factors: >(If anyone is interested, the canonical reference for ab initio and DFT >scaling factors--freqs and ZPEs-- is >A. P. Scott and L. Radom, J Phys Chem, 1996, 100, 16502-16513). Is also the best source I know of for general PM3 scaling factors. (If you need some Metal-Main group vibration information, try: K. K. Stavrev, ZCPE Bulletin (1992).) For AM1 and SAM1, though, Andy Holder has done the most thorough main group studies. Originally for AM1 vibrational frequencies: E. F. Healy & A. Holder, J. Mol. Struct. (Theochem), 281, 141-56 (1993). And he has updated this information for SAM1, combining both sets of comparisons in : A. Holder & R. Dennington II, J. Mol. Struct. (Theochem)401, 207-18 (1997). SAM1 is more accurate, (average signed error of only 0.7% vs 4.9% for AM1), except he identified systematic scaling factors for some different vibrational modes: specifically a scaling factor of 0.86 is required for non-H stretching frequencies. EC --- Ernest Chamot Chamot Laboratories, Inc. 530 E. Hillside Rd. Naperville, Illinois 60540 (630)637-1559 echamot (+ at +) chamotlabs.com http://www.chamotlabs.com/cl ============ #6 Professor Lewars, The correction you mention for stretching frequencies is about correct, but things are not as simple as they are for ab initio frequencies. the only way to correct frequencies coherently is by type. The same scaling factor will not apply to all types of vibrations. The paper you mention is good for AM1. the following reference is for a similar study on SAM1. As above, I don't know of a PM3 paper. A.J. Holder* and R.D. Dennington II. "An Evaluation of SAM1 Calculated Vibrational Frequencies", THEOCHEM,, 401/3 (1997) 207-18. 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 | holdera()at()umkc.edu K | (816) 235-2293, FAX (816) 235-6543 -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- #7 Hello, for AM1 and PM3 scaling factors (frequencies and force constants) see for example: Seeger, D. M., C. Korzeniewski, et al. (1991). “Evaluation of Vibrational Force Fields Derived by Using Semiempirical and Ab Initio Methods.” J. Phys. Chem. 95: 6871-6879. Grunenberg, J. and R. Herges (1997). “Calculation of Molecular Vibrations: Selective Scaling Factors for Semiempirical Force Constants.” J. Comp. Chem.18: 2050-2059. Greetings, Jörg -- Dr. Jörg Grunenberg, Org. Chemie, Hagenring 30, 38106 Braunschweig email: Joerg.Grunenberg %-% at %-% tu-bs.de phone: +49 531 391 5252 URL: http://www.tu-bs.de/institute/org-chem/grunenberg/grunenberg.html ========= #8 Hi! Could you send me a summary of the answers you received by now. I am also extremely interested in calculated (both Ab Initio and semiempirical) vibrational frequencies. Best regards, Vasile CHIS -------------------------------------------------------------------- Dr. Vasile CHIS "Babes-Bolyai" University Department of Physics 1, Kogalniceanu RO-3400 Cluj-Napoca ROMANIA Tel: +40 64 194315 / 5153 Fax: +40 64 191906 E-mail: vchis -8 at 8- phys.ubbcluj.ro ------------------------------------------------------------------- ============ ===========