energy from GEO_OPT job does not equal ENERGY job?

[Aaron Ledray]

Hello! 

I'm trying to run a DFT calculation for a porphyrin bound to a small molecule and compare energy to that of the unbound porphyrin state and the small molecule alone: 

One thing I am wondering is if it is necessary to perform an energy  job after the geo_opt job, or if I can use the total force_eval energy from the final optimized structure of geo_opt as my energy evaluation? 

I am finding that the energy I am getting at the end of geo_opt jobs are consistently different by about ~0.05 a.u. from energy jobs of the same molecule (the small-molecule bound state), with the geo_opt being lower in energy. Is this expected, is there something I'm not considering?

I've attached my input files. Energy.inp runs from the restart file made by geo_opt.inp.

Thank you kindly,

Aaron

[Jürg Hutter]

Hi

energies at the end of a converged geometry optimization should be the same as the
energy from a restart from the converged geometry using a energy only run.
If this is not the case, there must be an error in your setup or you have to send us
the full input for further investigation.

regards

JH

________________________________________
From: cp...@googlegroups.com <cp...@googlegroups.com> on behalf of Aaron Ledray <aaron...@gmail.com>
Sent: Friday, May 12, 2023 5:09 PM
To: cp2k
Subject: [CP2K:18807] energy from GEO_OPT job does not equal ENERGY job?

--
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/21d053ca-b333-4f4d-a89f-d8494af5dc15n%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/21d053ca-b333-4f4d-a89f-d8494af5dc15n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Aaron Ledray]

Thanks for the response,

I have attached the relevant (I think) files for my geo_opt and energy jobs.

I did a test for the energy on a more simple system that resulted in equal energies between geo_opt and energy, but I'm not sure why that is not the case for the files I have provided.

Best,

Aaron

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/dSHPIfjpYyY/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/ZR0P278MB07597110A4020F75F59641179F789%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM.