From moshe_o "-at-" VNET.IBM.COM Tue Nov 28 10:08:42 1995 Received: from VNET.IBM.COM for moshe_o <-at-> VNET.IBM.COM by www.ccl.net (8.6.10/950822.1) id JAA23211; Tue, 28 Nov 1995 09:55:50 -0500 Message-Id: <199511281455.JAA23211 &$at$& www.ccl.net> Received: from HAIFASC3 by VNET.IBM.COM (IBM VM SMTP V2R3) with BSMTP id 4117; Tue, 28 Nov 95 09:55:42 EST Date: Tue, 28 Nov 95 16:36:37 IST From: "Moshe Olshansky" To: chemistry (+ at +) www.ccl.net Subject: combining different basis sets - an addition Dear netters! After posting the summary of the responses I got an additional note which may be of interest to some of you. Here it is: - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Date: Mon, 27 Nov 1995 16:23:15 -0330 (NST) From: Uli Salzner Subject: basis set mixing Dear Moshe, I just read read the answers you received according basis set mixing and I think one important point was not addressed. It is definetly important that a basis set has to be balanced but that does not mean that you have to use the same basis set for each atom. In fact, to use the "same" basis set for different atoms may be the perfect way to get an unbalanced basis set. For example: consider the calculation of an ionic compound such as LiF. If you use 6-31G* on both atoms, you have a very good basis set for Li. Since Li looses about 90% of its 2s electron (maybe even more) when bound to F, it has almost no electrons in the 2s and 2p. For such a cation a valence double zeta basis set is very good. On the other hand fluorine is an anion with a diffuse charge distribution. For an anion the 6-31G* basis set quite bad. What you get is spurious charge transfer from F to Li. Thus, one uses different basis sets for different atoms all the time even if they have the same name. What you have to figure out is how to combine different basis sets in the right way. For the above example it would be a improvement to use 6-31+G* on F. This gives a better balanced basis set than 6-31G* on both atoms. Similar arguments hold for polarization functions. You have to decide which atoms need them and which don't. For instance for third row systems you definetely need d-functions. But there is no problem to combine a sulfer basis set with d-functions with hydrogen basis set that contains only s-functions because hydrogen d-orbitals are so high in energy that the effect would be negligible. Bets wishes, Uli Salzner uli $#at#$ smaug.physics.mun.ca - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - P.S. and now I have an additional question: I am a mathematician, not a chemist, so let me look at the basis sets purely mathematically. If one has a complete (and hence necessarily infinite) basis set, he/she gets a limit of Hartree-Fock model. Otherwise (with limited basis set) one gets some approximation to this limit and the more complete the basis set is the better is the approximation. Now assume one uses a certain "standard" basis set and gets some result (from Hartree-Fock model). And now we add ANY additional function to this basis set. This does not make the basis set less complete and so it should lead to at least as good (or even better) an approximation as the original basis did (it is also possible to get the original solution by taking that additional function with zero coefficient for every electron). Is there anything wrong with this statement? Moshe Olshansky IBM Israel Science & Technology