Do Gaussian and Dalton define functional and basis set the same?

[wanghong]

Professors: do Dalton and Gaussian have the same definition of the same basis set? In other words, do Gaussian and Dalton have to use the same functional and basis set for the same computing task?

[kennethruud]

If your question is whether you will get the same result with seemingly the same basis set and DFT functional, the answer is not necessarily. The default definition of the B3LYP functional differ in the two programs (you can get the same functional as used in Gaussian by the B3LYPg functional (see page 312 in the Dalton 2020 manual).

As regards basis set, they should in principle be the same, but make sure both programs use either spherical or Cartesian Gaussian. Both programs would by default use spherical Gaussians, but there are a number of exceptions to this rule in the Gaussian program for the Pople-style basis ses, as described in detail here: https://gaussian.com/basissets/. This may also lead to differences in the results obtained with the two programs.

And finally, the two programs use different thresholds for convergence.


Kenneth

[wanghong]

Professor Kenneth, thank you very much for your kind reply. Your answer has solved my confusion.
Wish you all the best.