TURBOMOLE Users Forum

TURBOMOLE Modules => Ridft, Rdgrad, Dscf, Grad => Topic started by: cneiss on March 02, 2022, 09:16:42 AM

Title: DFT-D4 energy for charged systems
Post by: cneiss on March 02, 2022, 09:16:42 AM

Dear all,

it seems that turbomole calculates the DFT-D4 correction always assuming a neutral molecule by default.
As the D4 correction depends on the charge of the molecule, the correction is wrong for charged systems.

At the moment one can fix this by stating the charge explicitly, e.g. for a negatively charged molecule:

$disp4 chrg -1

It would be nice however, if the correct charge would be adopted automatically by the dftd4 code.

Best wishes,
Christian

Title: Re: DFT-D4 energy for charged systems
Post by: uwe on March 02, 2022, 09:25:05 AM

Hello,

this is fixed in the 7.5.1 version and also in 7.6. Which version do you use?

Regards,
Uwe

Title: Re: DFT-D4 energy for charged systems
Post by: cneiss on March 02, 2022, 10:37:09 AM

Hi Uwe,

that is strange, I am using a locally modified version of TM 7.6 (based on master commit 90139e895b1c43bf131ebaba8b16c34a9f1cf569; I did not make any modifications in the ridft / dftd4 code).

Attached, you can find my input; I am running the ridft SMP version.
Maybe, I did some input error?

Best,
Christian

Title: Re: DFT-D4 energy for charged systems
Post by: uwe on March 02, 2022, 10:45:32 AM

Hi,

if you use a locally modified developers version, it might be better to discuss that within the developer community and not on the user forum?

Cheers