Menu
▾
▴
Re: [Jmol-users] Displaying multiple NBOs
|
From: Bruce T. <bru...@ne...> - 2019-03-12 18:24:36
|
John My impression when I was working on this was that the answer was ‘no’ with respect to orbital surfaces generated using the MO command. Only one was allowed at a time per frame. One could generate more frames using ‘load append filename.36’ and have different orbitals in each, then superimpose them. For a medium-sized molecule and several overlapping orbitals, I found that waiting times were really too long for the user, which is why I switched from this method, when more than one orbital was to be displayed at once, to using stored isosurfaces. Waiting times for contour plots on a plane through the orbital were even worse than for orbital surface plots. Loading pre-calculated isosurfaces was much faster in both cases. I have set out how I did it, towards the bottom of https://www.staff.ncl.ac.uk/bruce.tattershall/jmolapps/orbitalmethods.html If you do use pre-stored surfaces, notice the warning about also saving the molecular geometry from the same source. The default view direction of a molecule loaded from ‘filename.36’ is different from that for the molecule saved as an .xyz file containing atomic coordinates from Gaussian, and I could never find a sequence of rotates to connect them up so that orbital surface moved with the molecular skeleton as required. Storing an .xyz file from Jmol from the structure loaded from ‘filename.36’ gets round that. When there were lots of orbitals but I wanted only to show one at a time in a window, then I have avoided the work of setting up pre-stored surfaces and the storage and download overheads they entail, as in my orbitals for teaching web pages such as https://www.staff.ncl.ac.uk/bruce.tattershall/teaching/maingrps/pf3mo/pf3.php These use ‘filename.40’ from an NBO calculation to show delocalised MOs rather than NBOs. Students get the quicker orbital surface model by default and the slower contour plot in a plane only if they ask for it. Incidentally, have you thought of using PNBOs from filename.36, rather than orthogonalised NBOs from filename.37 which you mention? I think PNBOs give a more ‘realistic’ picture of e.g. lone pair overlaps. Bruce From: John Keller [mailto:jwk...@al...] Sent: 12 March 2019 17:18 To: jmo...@li... Subject: Re: [Jmol-users] Displaying multiple NBOs Hi, Much as I appreciate Bruce’s feedback on this issue, it does not really address my question. Bruce is creating the .jvxl NBO surface files externally, saving them locally, and loading the surface files with “isosurface” commands. I can do that. My questions is, “Is it possible to display *multiple* NBO’s after loading the Gaussian/NBO plot output files (.37 and .47 by implication)?” If so, this would vastly simplify, and probably speed up, the NBO surface display process. So far, in answer to my own question, this appears to be impossible with the current version of Jmol. John K. Sent from Mail<https://go.microsoft.com/fwlink/?LinkId=550986> for Windows 10 From: Bruce Tattershall<mailto:bru...@ne...> Sent: Monday, March 11, 2019 12:37 PM To: jmo...@li...<mailto:jmo...@li...> Subject: Re: [Jmol-users] Displaying multiple NBOs John Yes, all the surfaces are in the same folder as the structure etc. I used the filename.36 from the NBO analysis in Gaussian to provide the structure in each frame and then a Rasmol-type .spt file to set up ball sizes, orientation etc. Have to go home now, as it is 8.37 pm here! Bruce From: John Keller [mailto:jwk...@al...] Sent: 11 March 2019 20:13 To: jmo...@li... Subject: Re: [Jmol-users] Displaying multiple NBOs Bruce, That’s a nice page! Am I correct that your .jvxl surface files were generated beforehand and saved in the same folder as the .html file? Actually, in this exercise I am trying to avoid that whole process by just loading the NBO plot output file, then using the “nbo n” command to display nbo’s. Part of the problem seems to be that the mo command, and presumably the nbo command, depends on loading and reading a WebMO “.mo” file, or Spartan’s .spartan file. Recent versions of WebMO do not save surfaces in the .mo format as they once did. Currently only .cub files can be saved in WebMO. Also, many folks do not have access to Spartan, and the /examples-11/mo.htm is focused on this file format. John Keller Sent from Mail<https://go.microsoft.com/fwlink/?LinkId=550986> for Windows 10 From: Bruce Tattershall<mailto:bru...@ne...> Sent: Monday, March 11, 2019 11:06 AM To: jmo...@li...<mailto:jmo...@li...> Subject: Re: [Jmol-users] Displaying multiple NBOs Further to this: I generated all eight isosurfaces one at a time, and saved them to disc, so that during this process there was never more than one present at a particular instant. Once you have the isosurfaces, you can load as many as you wish at once, providing they have different file names and you give them all different isosurface names after reloading . Putting them into separate frames is a convenient way to switch them on and off at will. My example gets a bit cluttered if all four are on at once, whether for the orbital surfaces or for the contour plots, but it is very instructive for the user to switch them on or off experimentally, one at a time, until they understand what they are looking at. The web page works well in Chrome as the browser. I think it is OK in Safari as well, if you are Apple based. Bruce From: Bruce Tattershall [mailto:bru...@ne...] Sent: 11 March 2019 17:33 To: jmo...@li... Subject: Re: [Jmol-users] Displaying multiple NBOs John Why not load your different NBOs into separate frames and then superimpose them using ‘frames all’? I have done this successfully in https://www.staff.ncl.ac.uk/bruce.tattershall/structs/p5e2x/showjmol.php?use=HTML5&filestem=p5s2cl&lpnow=true You can see enough of the code for this by View Source Best wishes Bruce Newcastle University, England From: John Keller [mailto:jwk...@al...] Sent: 10 March 2019 22:01 To: jmo...@li... Subject: [Jmol-users] Displaying multiple NBOs Hi All, Displaying multiple NBOs, where one is an electron pair donor and the other an acceptor, is quite useful in the analysis of various non-covalent complexes. I know one can do this in Jmol using .cub files generated by say, WebMO. However, now I am trying to use Jmol to display NBOs directly from Gaussian output. I am using Jmol 14.29.26 for Linux and G09 linked to NBO6. For example, the following input file was used to calculate the NBOs of water, and the resulting h2o.37 file was opened in Jmol. The various NBOs can be displayed using the “NBO n “ command. And an NBO can be modified using cutoff, color, mesh, nofill, resolution, and other modifiers as one would do with the “MO” command. #N B3LYP/6-31G(d,p) SP GFINPUT POP=(FULL,NBO6Read) title molecule specification $NBO file=h2o plot archive $END However, I am not able to create an NBO isosurface, or display multiple NBOs as isosurfaces, or visualize two NBOs by loading h2o.37 twice, as described for MOs on the /examples-11/mo/htm page. Nor can I generate a .jvxl file using a “write” command from a displayed NBO. So, is it possible to display multiple NBOs in Jmol starting with the G09/NBO6 output? (Also, in Jmol I have not been able to display ANY NBO generated by G09 and the built-in NBO3 function, i.e. the above input file with “NBORead” in place of “NBO6Read”.) John Keller University of Alaska Fairbanks P.S. If you don’t have local access to Gaussian, short jobs can be run on the WebMO working demo site https://www.webmo.net/demoserver/cgi-bin/webmo/login.cgi (but only NBO3: not NBO6). P.P.S. To see the Gaussian job referred to above, visit the UAF WebMO site at https://corsair2.cns.uaf.edu/~frank/cgi-bin/webmo/login.cgi . Log on as guest, pwd webmo. It’s job 13202. Sent from Mail<https://go.microsoft.com/fwlink/?LinkId=550986> for Windows 10 |
