Density Corrected DFT

[timdu...@gmail.com]

Hello, 

I am wondering if any knows or has implemented density corrected DFT in CP2K? 

As described here: https://chemrxiv.org/engage/chemrxiv/article-details/62468f9358f7af2f0eae6081

ie compute SCAN energies with HF density?

Thanks very much 

Tim Duignan

[Jürg Hutter]

Hi Tim

two very new developments (Harris functionals and 2nd derivatives for meta-functionals)
make this possible. You can calculate forces and stress tensor for methods like
SCAN@HF now in CP2K.
Here is an input with the important sections:

&XC
&XC_FUNCTIONAL NONE
&END XC_FUNCTIONAL
&HF
&END
&END XC
&ENERGY_CORRECTION
ENERGY_FUNCTIONAL HARRIS
HARRIS_BASIS ORBITAL
&RESPONSE_SOLVER
METHOD MO_SOLVER
PRECONDITIONER FULL_SINGLE_INVERSE
&END
&XC
&XC_FUNCTIONAL
&MGGA_X_R2SCAN
&END
&MGGA_C_R2SCAN
&END
&END XC_FUNCTIONAL
&END XC
&END ENERGY_CORRECTION

best regards

Juerg

________________________________________
From: cp...@googlegroups.com <cp...@googlegroups.com> on behalf of timdu...@gmail.com <timdu...@gmail.com>
Sent: Tuesday, April 5, 2022 6:23 AM
To: cp2k
Subject: [CP2K:16804] Density Corrected DFT

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com<mailto:cp2k+uns...@googlegroups.com>.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/3079692b-a599-4d55-994a-e064eca42ff6n%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/3079692b-a599-4d55-994a-e064eca42ff6n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[timdu...@gmail.com]

Hi Juerg, 

Brilliant thanks so much for that! I will try it 

Best

Tim

[Frederick Stein]

Dear Tim,

I have implemented the additional terms for meta functionals and have to correct Jürg partially:

Forces do work with meta functionals of all kind, the stress tensors do not work (you will receive an abort). I do not know why they don't yet.

Best,

Frederick

[Tim Duignan]

Dear Frederick, 

Great thanks for letting me know

Best

Tim

You received this message because you are subscribed to a topic in the Google Groups "cp2k" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cp2k/0vMLuTmTKUE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/1dd23d46-b595-47b2-ad9d-804c4bdb3873n%40googlegroups.com.

[Jürg Hutter]

Dear Tim

I just realized that I was wrong. We have currently
implemented a Harris functional scheme that allows
you to do a type of SCAN@HF, but this is not
what people have done in the original paper.

regards

Juerg

________________________________________
From: cp...@googlegroups.com <cp...@googlegroups.com> on behalf of Tim Duignan <t...@duignan.net>
Sent: Thursday, April 7, 2022 1:33 AM
To: cp...@googlegroups.com
Subject: Re: [CP2K:16818] Density Corrected DFT

Dear Frederick,

Great thanks for letting me know

Best
Tim

On Wed, Apr 6, 2022 at 7:09 PM 'Frederick Stein' via cp2k <cp...@googlegroups.com<mailto:cp...@googlegroups.com>> wrote:
Dear Tim,

I have implemented the additional terms for meta functionals and have to correct Jürg partially:
Forces do work with meta functionals of all kind, the stress tensors do not work (you will receive an abort). I do not know why they don't yet.

Best,
Frederick

To unsubscribe from this group and all its topics, send an email to cp2k+uns...@googlegroups.com<mailto:cp2k+uns...@googlegroups.com>.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/1dd23d46-b595-47b2-ad9d-804c4bdb3873n%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/1dd23d46-b595-47b2-ad9d-804c4bdb3873n%40googlegroups.com?utm_medium=email&utm_source=footer>.

--

You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com<mailto:cp2k+uns...@googlegroups.com>.

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAA9w_TzwLkpRqM_LHwz6gXxoPiBcdtA9Tq%2BjKmwjgQ%3DBPk%3DEMg%40mail.gmail.com<https://groups.google.com/d/msgid/cp2k/CAA9w_TzwLkpRqM_LHwz6gXxoPiBcdtA9Tq%2BjKmwjgQ%3DBPk%3DEMg%40mail.gmail.com?utm_medium=email&utm_source=footer>.

[Tim Duignan]

Ok no problem thanks for letting me know Juerg

Best

Tim

To unsubscribe from this group and all its topics, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZR0P278MB07591C59038E87488BDABF349FD49%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM.