
2:31:21 PM PDT  Tue, Jul 23rd 2013 

I noticed this a little while ago and always meant to mention it. While sorting out some things this afternoon, I found a piece of paper that prompted me to raise this issue. Here goes:
In the 6311G(2df,2pd) and 6311++G(2d,2p) basis sets for F, the first d exponent is 2.396. I think it should be 3.50.
Reasoning/background:
(1) If you run Gaussian and print the internallystored basis, you will see 3.50 is the larger d exponent.
(2) If you go back to the literature (e.g. J. Chem. Phys. 80, 3265 (1984)), you will see that the recommendation to form the 2d set is to take the d exponent in 6311G(d) and double it and halve it. The exponent in 6311G(d) is 1.75, so the d exponents for 2d are 3.50 and 0.875.
(3) The 6311G(3df) set for F is correct and follows the type of rule followed in the paper mentioned in (2): to go from 1d to 3d, multiply by 4, multiply by 1, and divide by 4. Hence, 1.75 leads to 7.0, 1.75, and 0.4375.
I am not aware of errors with other 6311G(2d) or related sets, but I did not check many.

