I found an example of a Gaussian input file, where they use split basis set with ECP for heavy atom, see below.
...........B3LYP/GENECP........
coordinate
Ru Au 0
SDD
****
C H O 0
6-31G*
****
blank line
Ru Au 0
SDD
blank line
I interpret that line below, please help me to check if my interpretation is correct:
- They use SDD for outer electrons of Ru and Au; and 6-31Gd for all electron of C H O.
- The inner core of Ru and Au is replaced by ECP, with basis set SDD?
Is that the correct interpretation? And if the input is just like that, how to understand whether the ECP used is small-core or large-core SDD basis set?
Thank you!
Your interpretation is right.
I don't exactly know how to specified small or large core in SDD basis set.
I have heard that SDD costs much time than other ECP due to a number of basis functions.
You can check in output file how ECP treats valence electrons of metal.
The valence shell size determines the level of ECP. (check output file using pseudo keyword)
Oh, thank you. Is that something like, for Ru lets say, Atomic No. 75; Valence electrons: 15; then that counts electrons on 5x and 6x as valance electron, hence it is small-core ECP?
Oleg,
Thank you for the information. On our computer, i can not find MWBxx basis set. Only "SDD" is listed. Should it be okay if i just specify Ru 0 SDD, and then Gaussian read it as 60e ECP and 15 valence electron?
I read that it is possible to incorporate new basis set, but maybe at this stage, i try to use what we already have first, slowly i will try to catch up.
Thank you!
Very much, thank you, Oleg! Really useful information. I will need spend time and go through it.
Thank you!
Dr Oleg,
In case the basis set is specified as
"
Ru 0
SDD
****
blank line
Ru 0
SDD
blank line
"
Then how should i interpret it.
The 28 core electrons was treated with MWB28 . How about the leftover valance electrons? Which basis set is actually Gaussian would read?
- Badges
- Science topic