CCL:G: def2sv and def2svpp in gaussian

Dear Michael,


I think the two are the same for the following reasons:

a. In the paper by Florian Weigend and Reinhart Ahlrichs, they did NOT provide a basis set termed "def2-SV".  I did not find one, either, by web search.  This term does not appear in the EMSL basis set library.  I can not find the name in ORCA, MOLPRO nor Q-CHEM manual, neither.

b. I used two two basis sets, and diff the two log files.  I use GFprint option of gaussian to print or the basis sets.  It turns out at least for elements H, C, Cr and Pd, no difference was found.

c. Some notes online said "def2SV a.k.a. def2SV(P)".

Therefore, I conclude that "defSV" is yet another cute alias of def2-SV(P) by Gaussian Ltd. guys.

However, I also found that there is yet another difference between the two outputs aside from date and time thing:


<  3/5=43,11=9,16=1,24=100,25=1,30=1/1,2,3;
>  3/5=43,***7=202,***11=9,16=1,24=100,25=1,30=1/1,2,3;


I am not sure about this part surrounded by ***.  I think other guys can provide more info about Gaussian internal messages.

The tests was done with Gaussian 09, Revision B.01.



On 12/26/2018 07:18, Michael Morgan michaelmorgan937]![ wrote:

Dear all,


Are def2sv and def2svpp basis sets the same in Gaussian?


Form manual, "Def2SVPP corresponds to the "def2-SV(P)" basis set". So I guess def2sv is without polarization but Def2svpp has polarization on heavy atoms but not H. However it seems test run with both basis sets generate the same results. Do I misunderstand something there?