From owner-chemistry@ccl.net Sun Apr 8 01:24:01 2007 From: "Chun-Yi Sung juneyi1-x-yahoo.com.tw" To: CCL Subject: CCL:G: Visualizing NBO Message-Id: <-34006-070408012200-28716-sWg8F8a/77EqIElgG4L4Xw,,server.ccl.net> X-Original-From: "Chun-Yi Sung" Date: Sun, 8 Apr 2007 01:21:56 -0400 Sent to CCL by: "Chun-Yi Sung" [juneyi1_._yahoo.com.tw] Hi, Julia I figured how to visualize orbitals, but even when I included "SaveNBOs", I still saw the default MOs using MO editor, and I didn't see any section entitled anything like "NBO coefficients" in the .fchk file. I wanted to visualize NBOs from NBO analysis like these : (Occupancy) Bond orbital/ Coefficients/ Hybrids --------------------------------------------------------------------------- 1. (0.99261) BD ( 1) H 1- C 2 ( 38.34%) 0.6192* H 1 s(100.00%) 1.0000 0.0081 ( 61.66%) 0.7853* C 2 s( 32.48%)p 2.08( 67.52%) -0.0001 0.5686 -0.0386 0.0003 -0.1998 -0.0074 0.7969 0.0149 0.0064 0.0002 2. (0.99704) BD ( 1) C 2- C 3 ( 50.73%) 0.7122* C 2 s( 39.26%)p 1.55( 60.74%) 0.0001 0.6265 -0.0007 -0.0003 -0.5227 -0.0084 -0.5780 -0.0130 -0.0014 0.0000 ( 49.27%) 0.7019* C 3 s( 26.98%)p 2.71( 73.02%) 0.0003 0.5189 0.0229 0.0002 0.6263 -0.0169 0.5809 -0.0122 -0.0010 0.0002 I guess I did not save NBOs successfully in chk file. My route section looks like this : opt ub3lyp/sdd nosymm scf=(qc,maxcycle=800) Pop=(nbo,saveNBOs) Do you see anything weird? It'd be very convenient if I could just use GaussView. I'd appreciate your help very much. thanks! Chunyi Sent to CCL by: ysubboti^ucalgary.ca Hi Chunyi! Actually you can do it easily with any program reading CHK/FCHK files to visualize orbitals. You can find how to do it here: http://educ.gaussian.com/visual/Orbs/index.htm and here with GaussView: http://educ.gaussian.com/visual/Orbs/html/OrbsGaussView.htm. Just be sure that you have put in the line SaveNBOs keyword to save natural bond orbitals in the checkpoint file . Please take a look at gaussian manual to check how to control nbo module and for some other options. Good luck, Julia > > Sent to CCL by: "Chun-Yi Sung" [juneyi1--yahoo.com.tw] > Hi, > > I did some NBO analysis with G03, and I wonder if there > is a way I can visualize those natural bond orbitals with > GaussView or I should go for other visualization software like > NBOview. > > thanks > > chunyi> > > > > From owner-chemistry@ccl.net Sun Apr 8 09:56:01 2007 From: "Dipankar Roy dipankarroy!A!iitb.ac.in" To: CCL Subject: CCL:G: Visualizing NBO Message-Id: <-34007-070408060323-14228-pzYVOMTxrbtVkkD1ukl2+g__server.ccl.net> X-Original-From: "Dipankar Roy" Content-Transfer-Encoding: 8bit Content-Type: text/plain;charset=utf-8 Date: Sun, 8 Apr 2007 14:33:05 +0530 (IST) MIME-Version: 1.0 Sent to CCL by: "Dipankar Roy" [dipankarroy-,-iitb.ac.in] Hello, Please try with POP=(NBOREAD,SAVENBOS) and after your coordinate spacifications add the line $NBO $END Hope this may work -dipankar > > Sent to CCL by: "Chun-Yi Sung" [juneyi1_._yahoo.com.tw] > Hi, Julia > > > I figured how to visualize orbitals, but even when I included > "SaveNBOs", I still saw the default MOs using MO editor, and I didn't see > any section entitled anything like "NBO coefficients" in the .fchk file. > > I wanted to visualize NBOs from NBO analysis like these : > (Occupancy) Bond orbital/ Coefficients/ Hybrids > --------------------------------------------------------------------------- > 1. (0.99261) BD ( 1) H 1- C 2 > ( 38.34%) 0.6192* H 1 s(100.00%) > 1.0000 0.0081 > ( 61.66%) 0.7853* C 2 s( 32.48%)p 2.08( 67.52%) > -0.0001 0.5686 -0.0386 0.0003 > -0.1998 > -0.0074 0.7969 0.0149 0.0064 > 0.0002 > 2. (0.99704) BD ( 1) C 2- C 3 > ( 50.73%) 0.7122* C 2 s( 39.26%)p 1.55( 60.74%) > 0.0001 0.6265 -0.0007 -0.0003 > -0.5227 > -0.0084 -0.5780 -0.0130 -0.0014 > 0.0000 > ( 49.27%) 0.7019* C 3 s( 26.98%)p 2.71( 73.02%) > 0.0003 0.5189 0.0229 0.0002 > 0.6263 > -0.0169 0.5809 -0.0122 -0.0010 > 0.0002 > > I guess I did not save NBOs successfully in chk file. My route section > looks like this : > opt ub3lyp/sdd nosymm scf=(qc,maxcycle=800) > Pop=(nbo,saveNBOs) > Do you see anything weird? > It'd be very convenient if I could just use GaussView. > I'd appreciate your help very much. > > thanks! > > Chunyi > > > > Sent to CCL by: ysubboti^ucalgary.ca > Hi Chunyi! > Actually you can do it easily with any program reading CHK/FCHK files to > visualize orbitals. You can find how to do it here: > http://educ.gaussian.com/visual/Orbs/index.htm > and here with GaussView: > http://educ.gaussian.com/visual/Orbs/html/OrbsGaussView.htm. > Just be sure that you have put in the line SaveNBOs keyword to save > natural bond orbitals in the checkpoint file . Please take a look at > gaussian manual to check how to control nbo module and for some other > options. > Good luck, > Julia > > > > Sent to CCL by: "Chun-Yi Sung" [juneyi1--yahoo.com.tw] > > Hi, > > > > I did some NBO analysis with G03, and I wonder if there > > is a way I can visualize those natural bond orbitals with > > GaussView or I should go for other visualization software like > > NBOview. > > > > thanks > > > > chunyi> > > > *********************************************** Dipankar Roy Graduate Student of Prof. R. B. Sunoj Computational Chemistry Laboratory Dept. of Chemistry Indian Institute of Technology, Bombay India - 400076 Phone: +91-22-2576-4130(lab) URL: http://www.geocities.com/dipankar_roy79/dipankar.html *********************************************** A SCIENTIST IS SOMEONE WHOSE CURIOSITY SURVIVES EDUCATION'S ASSAULTS ON IT. - SIR HERMANN BONDI From owner-chemistry@ccl.net Sun Apr 8 16:00:01 2007 From: "dedey**alumni.bilkent.edu.tr" To: CCL Subject: CCL: significance of very small vibrational frequencies Message-Id: <-34008-070408155812-10414-/ZbNOpqCyYI5P2UCF4uNiA.:.server.ccl.net> X-Original-From: Date: Sun, 08 Apr 2007 22:57:58 EEST Sent to CCL by: [dedey---alumni.bilkent.edu.tr] Hi all, My question is related with the significance of normal modes with very minor positive or negative frequencies. I do not doubt about considering a large molecule which has many -sort of- soft modes, due to rotations of groups like methyls,phenyls and so on, and exhibits frequencies of approximately -1.0 to -10.0 (cm-1) or smaller, as being a minimum. But avoiding such frequencies in optimizations may fail and what you end up becomes a TS mathematically. I mean a rotational TS. Tightening the convergence criteria, changing the optimization algorithms and playing with the step sizes may not help. Do you agree? I wonder if there is a published work -I could not manage to find- about the significance of very small negative frequencies,especially about semi empirical methods. If this is not present do you know about any published work which presents such structures as being a minimum and comments on the small negative frequencies as being insignificant or unimportant. Or what are your opinions about disregarding a vibration of about -1.0 to -10.0 cm-1 and declaring such a structure as a minimum. And finally, what would be the true strategy to get the optimized structure of a large molecule with many soft modes as I mentioned above, especially if we know an experimental geometry and can not locate it with various levels of theory. I would appreciate any comments. Thanks a lot... -- Yavuz Dede Comp. Chem. Lab. Dep. of Chem. METU 06531 Ankara TURKEY +90 312 210 3187 From owner-chemistry@ccl.net Sun Apr 8 19:34:01 2007 From: "Arthit Vongachariya varthit:gmail.com" To: CCL Subject: CCL:G: How can run scan relax with lowmethod and singlepoint with highmethod Message-Id: <-34009-070408022601-19859-SoR0Y54P1oRXGVeU6GOPBQ..server.ccl.net> X-Original-From: "Arthit Vongachariya" Date: Sun, 8 Apr 2007 02:25:57 -0400 Sent to CCL by: "Arthit Vongachariya" [varthit**gmail.com] I used gaussian03 c0.2 for run the job I want to run scan relax for 91 Step with low method(mp2/cc-pvdz) and i want to run singlepoint after relax on the every step with high method ( mp4(SDTQ)/cc-pvdz) How can i do ? i test many command for try to run the last,my input is opt(z-matrix,Guided(mp2/cc-pvdz)) mp4 But it was not to success after that I try ot used chkpoint file to try it But it was not to success too. Could And bodyhelp me ? Please From owner-chemistry@ccl.net Sun Apr 8 21:50:11 2007 From: "Kin-Yiu Wong kywong**chem.umn.edu" To: CCL Subject: CCL:G: Gaussian: Freq with fixing some atoms Message-Id: <-34010-070408201409-1345-wfj7sD7Fi0lDGmsq44M44g(!)server.ccl.net> X-Original-From: "Kin-Yiu Wong" Date: Sun, 8 Apr 2007 20:14:05 -0400 Sent to CCL by: "Kin-Yiu Wong" [kywong{=}chem.umn.edu] Hi there, I have a quick question about the frequency analysis (Freq) in Gaussian. If I fix some atoms by putting "-1" in between the elements and the x-coordinate in the geometry specification, from the output of the frequency analysis. I notice that Gaussian would diagonalize the Hessian matrix by removing those elements associated with the atoms fixed by "-1" . However, it looks like to me that Gaussian would still first calculate the whole 3N x 3N Cartesian Hessian matrix before removing those elements. As a result, the computational time for the Freq cannot be reduced by fixing some atoms by "-1". And it seems like to me it is sort of waste of computational time. So my question is: Is it possible to ask Gaussian not to calculate the entire 3N x 3N Hessian elements, instead Only calculate those Hessian matrix elements which are not associated with the atoms fixed by "-1" ? If Gaussian can do that, then certainly it wll save me a lot of computational time from expensively calculating the whole 3N x 3N matrix and then throw most of the elements away. Thank for any help in advance! Kiniu ---- Kin-Yiu WONG (Kiniu) Chemical Physics PhD Student Department of Chemistry The University of Minnesota at Twin Cities USA Primary Email: kywong!^!chem.umn.edu Permanent Email: kiniu!^!alumni.cuhk.net