Question: Can NPT simulations be used to get IFCs in TDEP

[rashidrafeek]

This is a question on the TDEP method itself, and might not be applicable to the tdep code. But I thought this might be a good place to ask this question. Please let me know if it not.

I was wondering whether NPT simulations, where the volumes are fluctuating can be used to fit the IFCs. I noticed that the TDEP formalism is developed for constant volume simulations. And in the publications generally several NVT simulations at different volumes are done to obtain the pressure and thereby the Gibbs Free energy. For example, it is mentioned in this publication that,

Though it is possible to carry out simulations in the NPT ensemble, they are currently not well suited for use with the TDEP method because the underlying formalism assumes a constant volume when fitting results of the MD simulation to obtain the force-constant matrix

Is this the only way to compute the Gibbs Free energy? Is there any alternative to this method if I only have NPT simulations? Any information regarding this would be useful. Thanks.

[mjv500]

Hi Rashid, no - in a word. Olle convinced me there are some pretty fundamental issues which impose using NVT simulations for TDEP, and it's best to do several different volume NVTs rather than mess with the NPT. The indexing of the atoms and the reference positions are not well defined, so the displacements to match the forces and stress are not uniquely defined either. What we do is NPT to get equilibrated volumes for the T we want, then stop and run an NVT from a well thermalized starting point. This is all easier with an MLIP of course, with explicit DFT you are better off doing canonical configurations, then adjusting the full Free energy in the V-T space to find the equilibrium volume etc. You can get Gibbs from F or H doing the appropriate thermodynamic transforms. Best Matthieu

…

On Thu, Jun 27, 2024 at 1:10 PM Rashid Rafeek ***@***.***> wrote: This is a question on the TDEP method itself, and might not be applicable to the tdep code. But I thought this might be a good place to ask this question. Please let me know if it not. I was wondering whether NPT simulations, where the volumes are fluctuating can be used to fit the IFCs. I noticed that the TDEP formalism is developed for constant volume simulations. And in the publications generally several NVT simulations at different volumes are done to obtain the pressure and thereby the Gibbs Free energy. For example, it is mentioned in this publication <https://pubs.acs.org/doi/full/10.1021/acs.jpca.6b08633> that, Though it is possible to carry out simulations in the NPT ensemble, they are currently not well suited for use with the TDEP method because the underlying formalism assumes a constant volume when fitting results of the MD simulation to obtain the force-constant matrix Is this the only way to compute the Gibbs Free energy? Is there any alternative to this method if I only have NPT simulations? Any information regarding this would be useful. Thanks. — Reply to this email directly, view it on GitHub <#73>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABWOT35QPQYB4DCJMDP3Y2TZJPXKVAVCNFSM6AAAAABJ7WIEXCVHI2DSMVQWIX3LMV43ASLTON2WKOZSGM3TOOBWGA4DCNQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Professor Matthieu J Verstraete Fellow, American Physical Society Chair, Steering Committee, European Theoretical Spectroscopy Facility www.etsf.eu Alumnus Fellow, Young Academy of Europe yacadeuro.org Institute for Theoretical Physics (ITP) Department of Physics Buys Ballot Gebouw/Building, Princetonplein 5, office University of Utrecht, 3584 CC Utrecht ITP Secretariat: +31 30 253 5928 E-mail: ***@***.*** Group web page (Liège): http://www.nanomat.ulg.ac.be/ Nanomat lab, Q-Mat center, Université de Liège Département de Physique, Bat. B5a, 4/50 Allée du 6 août, 19 B-4000 Sart Tilman, Liège Belgium Phone : +32 4 366 90 17 European Theoretical Spectroscopy Facility (ETSF) Mail : ***@***.*** ***@***.*** ***@***.***

[rashidrafeek]

Hi Matthieu,

That makes sense. As we have built MLIPs for the system, I think it would be possible to do NVT simulations using the equilibrium volume. I was wondering whether I can use some of the AIMD NPT simulations we already have for the calculations. Seems not. Thank you very much for letting me know.

Regards,
Rashid.