| 10:08:14 AM PDT - Thu, Aug 13th 2015 |
|
Hello,
I realized a possible inconsistency in at least some basis sets of the cc-pVXZ series. I checked mainly for hydrogen and cc-pVTZ but this seems to be a general effect.
Depending on the format (e.g., Molcas or Gaussian94), the contraction length of the first S function is different. For example, the cc-pVTZ basis set for hydrogen in Gaussian94 format includes three primary basis functions, while in Molcas format there are five (the two additional ones being the other separate S functions). As a result, this AO seems to be not normalized when the shorter contraction is used. I have found a similar behaviour for other atoms and the cc-pVDZ basis sets. Is this difference intended?
Moreover, I compared to a basis set file that, according to the header, was retrieved from the EMSL earlier this year. It contains the longer contractions for the S functions even in Gaussian94 format. Could there have been some sort of change?
Thanks for your help.
Regards,
Caroline
|