[CP2K:6637] Re: Thermal properties of metals with CP2K

Alex nedo... at gmail.com
Thu Jun 4 00:21:41 UTC 2015

Hi Ying,

The short answer is no and I decided it was not a very well thought out
idea. The longer answer is that you _may_ be able to get good data,
depending on what satisfies you, how much you know about combining methods,
what data you need, etc, etc. The naive way for, say, heat capacity, or the
expansion coefficient, is to do a set of simulations at various
temperatures and plot the average total energy and lattice spacing
(respectively), and then calculate the local slopes. The lack of k-sampling
aside, neither of these will give you what you want. Well, the heat
capacity will be the "correct" classical 3R value, while the expansion
coefficient is likely to be nonsensical.

A way to get heat capacity is to split it into the phonon and the electron
portions and then use the appropriate partition functions from QM. For
phonons, for instance, this means the ability to obtain the ion core
velocity autocorrelation to get PDOS, then apply Bose-Einstein statistics.
For electrons, it can in principle be done similarly. Problem is, I think
CP2K only outputs the total energy of the system, so please inquire with
the developers and please let me know what they say.

For any thermodynamic properties, the biggest issues are the ensemble size
and the time scale limitation. With atom counts in the 100s and simulated
times in tens of picoseconds, it will be a bit hard to justify the use of
what I described above. So, using DFT directly is probably a bad idea.
Indirect methods exist (e.g. obtain force constants from DFT, calculate
phonon spectra analytically, etc), but that's a whole separate discussion.


On Wed, Jun 3, 2015 at 5:39 PM, Ying Zhang <yingzh... at gmail.com> wrote:

> Hey Alex,
> I have the same question and I am brand new user. Do you figure out the
> way to calculate thermodynamic with CP2K now?
> Thanks,
> Ying
> On Thursday, September 11, 2014 at 2:54:41 PM UTC-6, Alex wrote:
>> Hi all,
>> Is there a way with CP2K to calculate things like thermal expansion
>> coefficient of a metal, or the electronic part of heat capacity and/or
>> conductance? From the theoretical point of view, at least the way I see it,
>> this would require calculating the electronic band structure in combination
>> with existing theory (e.g. http://arxiv.org/abs/cond-mat/0109020 ). It
>> is my understanding that CP2K only calculates the electrons' energy at the
>> gamma point of the Brillouin zone. Are there any other ways to approach
>> this problem with CP2K? I am very new to all of this, so feel free to let
>> me know if I am completely off base here.
>> Thank you,
>> Alex
>>  --
> 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/SOQvHM4XLWA/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> cp2k+uns... at googlegroups.com.
> To post to this group, send email to cp... at googlegroups.com.
> Visit this group at http://groups.google.com/group/cp2k.
> For more options, visit https://groups.google.com/d/optout.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20150603/e6eb8986/attachment.htm>

More information about the CP2K-user mailing list