CompuServe Thread

#pb

8 messages in this thread
#49645From: Larry JonesJul 14, 1993 3:31 PM
hyperchem! I want to: 1) compare the structure I have modeled with the crystallographic data from the literature…..how can I load the crystallographic coordinates and do an overlay? 2) pi bond an olefin to a titanocene metal center that has a growing polymer chain on it….(for rotational studies using the dihedral excel macro) more later!
#49716From: Tom Slee [Hypercube]Jul 15, 1993 8:54 AM
..and more replies.. To load the crystallographics coordinates, you would need to format them according to HIN or some other format. Although the HIN file format can look pretty intimidating, you can get by with very little. Here is an example of a minimal HIN file for ethylene: mol 1 atom 1 – C ** – 0.0 -0.67 0.0 0.0 0 atom 2 – C ** – 0.0 0.67 0.0 0.0 0 endmol You need the – and **'s, but the column numbers donb't matter. The numbers are charges (the first 0.0), coordinates (next three numbers) and a 0 for the number of neighbours. This doesn't connect them, of course. The trick then is to save the structure as a Z-matrix, and read it in again. Coupled with the z-matrix reader is a component that takes a guess at bonding and bond orders. Then do an "Add hydrogens" to add on the hydrogen atoms. Hope this helps. Tom
#49731From: Larry JonesJul 15, 1993 10:44 AM
but once you "add hydrogens" does'nt the structure change to some other optima? and I don't want a central "dummy" atom, I need a pi bonded olefin….next release? R / Cp = \ / M P / \/\/\/\/\/ Cp \ R' where = is a substituted olefin, and is pi bonded to the metal center, and spins around so that the effects of R (or R') on the energy of the system, (and the optimum orientation) can be mapped….will a dummy atom or disconnected Cp system work? How do I measure the distance between the Cp centroid and the metal center? how do I measure the olefin metal center bond distance, and look at the effect of the R (or R') group? more later, ps, thanks!
#49754From: Tom Slee [Hypercube]Jul 15, 1993 2:24 PM
Larry, The structure changes after you "Add Hydrogens and Model Build", but there is also an option to simply add hydrogens, which leaves all other atoms in unchanged positions. As for your pi-bonded olefin: hmmm… let me think till tomorrow on that one.. Tom
#49732From: Larry JonesJul 15, 1993 10:49 AM
also, lets put these crystallographic coordinates into an excel file (with the unit cell parameters, or cartesian coordinates or whatever…..even maybe have one of those nice little windows like the periodic table window for the elements you made in the latest release….) and then have that go into the readable Hyperchem modeling formatted whatever….. more later….
#49755From: Tom Slee [Hypercube]Jul 15, 1993 2:24 PM
I'm not sure if I know what you mean here: but given a spreadsheet with something like C 1.1 1.2 3.2 O 2.1 1.2 3.0 etc., an Excel macro could be written that would send those coordinates into HyperChem, using the script commands. At least, as long as the coordinates are cartesian… Is this the kind of thing you had in mind? Tom
#49838From: Larry JonesJul 16, 1993 7:29 AM
yes, that's exactly what I had in mind, and I hope this kind of thing can be included in a future release…(or even better, written by an interested party and uploaded to the Hyperchem library forum) I don't know how to write those macros……I do think that the utility of hyperchem is greatly increased by having this type of feature avaliable (just like the dihedral macro did)….it is a formula to convert from crystallographic to cartesian coordinates….this type of calculation could be included with the macro? ie "will you enter crystallographic or cartesian coordinates? a: crystallographic, Q: please enter the angular components of the unit cell: a____,b_____,c_____ etc…)….. Yes!
#49717From: Tom Slee [Hypercube]Jul 15, 1993 8:54 AM
To load in a structure from crystallographic data, you need to get it into a format that HyperChem can read. I would recomend a HIN file format. ALthough the HIN file can get pretty complicated, it ca also be pretty simple . Here is an example. mol 1 atom 1 – C ** – 0.0 -0.67 0.0 0.0 0 atom 2 – C ** – 0.0 0.67 0.0 0.0 0 endmol This gives you two carbon atoms. You need the -'s and **'s, but can ignore column numbers: it is just white-space separation that matters. The first 0.0 in each line is the charge. The next three numbers are coordinates, and the final 0 is the number of neighbours. To add the bonding information to this, there is a trick. Save the file as a z-matrix and read it in again. The z-matrix reader has a bond-inferring component: not completely reliable, but it makes a good first guess. Then choose "Add hydrogens" under Build to add the hydrogens and you have ethylene. Hope this helps, TOm