<div dir="ltr"><div class="gmail_default" style="font-family:monospace,monospace">Dear Yuriii:</div><div class="gmail_default" style="font-family:monospace,monospace">Treating U as a response property may not be sufficient. Response properties are not bound neither from above nor from below, and a particular (valid!) value of U may still correspond to unphysical values of energy due to the approximate nature of DFT+U. The situation is somewhat similar to using self-consistent vs HF densities in DFT to get the total energies (see K.Burke works on that).<br></div><div class="gmail_default" style="font-family:monospace,monospace"><br></div><div class="gmail_default" style="font-family:monospace,monospace">But in general, the study of this interesting question seems to be a real scientific challenge.</div><div class="gmail_default" style="font-family:monospace,monospace"><br></div><div class="gmail_default" style="font-family:monospace,monospace">--</div><div class="gmail_default" style="font-family:monospace,monospace">Ilya Ryabinkin,</div><div class="gmail_default" style="font-family:monospace,monospace">Senior Research Fellow,</div><div class="gmail_default" style="font-family:monospace,monospace">OTI Lumionics Inc</div><div class="gmail_default" style="font-family:monospace,monospace">Toronto, Canada.<br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Thu, Oct 29, 2020 at 1:37 PM Timrov Iurii <<a href="mailto:iurii.timrov@epfl.ch">iurii.timrov@epfl.ch</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div>
<div dir="ltr">
<div id="gmail-m_664054172435930906x_divtagdefaultwrapper" dir="ltr" style="font-size:12pt;color:rgb(0,0,0);font-family:Calibri,Helvetica,sans-serif">
<p>Dear Giuseppe,</p>
<p><br>
</p>
<p>Thank you for your comment!</p>
<p><br>
</p>
<p><font size="2"><span style="font-size:10pt">> In principle you are right, but if you try to compute, e.g., formation
<br>
> heats or formation energy of defects by using different U values (I <br>
> did it long time ago in the case of TiO2 related studies for metallic <br>
> Ti, for TiO2, for Ti2O3) you fall sometimes very far from experimental <br>
> references, farther then using non corrected GGA.</span></font></p>
<p><font size="2"><span style="font-size:10pt"><br>
</span></font></p>
<p><font size="2"><span style="font-size:10pt"><span style="font-size:12pt;font-family:Arial,Helvetica,sans-serif">Ok, I see. So</span><span style="font-size:12pt;font-family:Arial,Helvetica,sans-serif"> you used semiempirical U and things got worse, right?
It would be interesting to see what happens if ab initio U is used for each structure.</span><br>
</span></font></p>
<p><font size="2"><span style="font-size:10pt"><br>
</span></font></p>
<p><font size="2"><span style="font-size:10pt">> I've used semiempirically averaged values in a lot of studies (including the
<br>
> correction of O- and N-related bands in oxides and nitrides) and <br>
> things are a bit semiempirical, obviously, but still quite improved <br>
> with respect to GGA. As long as you motivate what you are doing, I do <br>
> not see particular harm in semiempirical U corrections!</span></font><br>
</p>
<p><br>
</p>
<p>Ok. Of course, one is free to choose whether to use a semiempirical or ab initio U. But I see the following issues when using semiempirical U:</p>
<p>- choice of the Hubbard manifold. U depends a lot on the Hubbard manifold (atomic, ortho-atomic, Wannier functions, etc.). Different codes use different Hubbard manifolds. So if one takes U from some paper where it was determined semiempirically (e.g. using
VASP) and use it in QE (with a different manifold) the results can be totally different</p>
<p>- choice of the pseudopotential (see Appendix in J. Chem. Phys. 129, 134314 (2008)). From this reference you can see that the pseudopotentials generated in different oxidation states require different U. It is easy to compute U for a given pseudo, but how
to determine it semiempirically (try to reproduce some experimental quantity? and what to do if there is no experimental data for the material under investigation?)</p>
<p>- choice the exchange-correlation functional: PW91, PBE, PBEsol, ... - each time the U correction is not exactly the same. For each functional U must be tuned by hand if done semiempirically.
</p>
<p>So for a given computational setup (Hubbard manifold, pseudos, functional, etc.) I would simply compute U from first principles (luckily we have now theories and codes to do that, which was not the case in the past). And when one wants to do a high-throughput
study for thousands of materials, fitting U for each system to the experimental data would kill the whole idea of high-throughput. Just my thoughts.
<br>
</p>
<p><br>
</p>
<p>Thanks!<br>
</p>
<p><br>
</p>
<p>Greetings,</p>
<p>Iurii<br>
</p>
<p><br>
</p>
<div id="gmail-m_664054172435930906x_Signature">
<div id="gmail-m_664054172435930906x_divtagdefaultwrapper" dir="ltr" style="font-size:12pt;color:rgb(0,0,0);font-family:Calibri,Helvetica,sans-serif,"EmojiFont","Apple Color Emoji","Segoe UI Emoji",NotoColorEmoji,"Segoe UI Symbol","Android Emoji",EmojiSymbols">
<div name="x_divtagdefaultwrapper">
<div name="x_divtagdefaultwrapper">
<font size="3" face="'Times New Roman', Times, serif" color="808080">--<br>
Dr. Iurii TIMROV<br>
Postdoctoral Researcher<br>
<font color="808080"><font face="'Times New Roman', Times, serif"></font></font></font></div>
<font color="808080"></font>
<div name="x_divtagdefaultwrapper">
<font size="3" face="'Times New Roman', Times, serif" color="808080">STI - IMX <font color="808080">
<font face="'Times New Roman', Times, serif">- THEOS</font></font></font><font size="3" face="'Times New Roman', Times, serif" color="808080"> and NCCR - MARVEL<br>
</font></div>
<div name="x_divtagdefaultwrapper">
<font size="3" face="'Times New Roman', Times, serif" color="808080"><font size="3" face="'Times New Roman', Times, serif" color="808080">Swiss Federal Institute of Technology Lausanne (EPFL<font color="808080"><font face="'Times New Roman', Times, serif">)</font></font></font><br>
</font></div>
<font color="808080"></font>
<div name="x_divtagdefaultwrapper">
<font size="3" face="'Times New Roman', Times, serif" color="808080">CH-1015 Lausanne, Switzerland<br>
+41 21 69 34 881</font></div>
<div name="x_divtagdefaultwrapper">
<a href="http://people.epfl.ch/265334" id="gmail-m_664054172435930906LPNoLP" target="_blank">http://people.epfl.ch/265334</a><br>
</div>
</div>
</div>
</div>
</div>
<hr style="display:inline-block;width:98%">
<div id="gmail-m_664054172435930906x_divRplyFwdMsg" dir="ltr"><font style="font-size:11pt" face="Calibri, sans-serif" color="#000000"><b>From:</b> users <<a href="mailto:users-bounces@lists.quantum-espresso.org" target="_blank">users-bounces@lists.quantum-espresso.org</a>> on behalf of Giuseppe Mattioli <<a href="mailto:giuseppe.mattioli@ism.cnr.it" target="_blank">giuseppe.mattioli@ism.cnr.it</a>><br>
<b>Sent:</b> Thursday, October 29, 2020 5:58:39 PM<br>
<b>To:</b> Quantum ESPRESSO users Forum<br>
<b>Subject:</b> Re: [QE-users] Thermodynamics with DFT+U</font>
<div> </div>
</div>
</div>
<font size="2"><span style="font-size:10pt">
<div><br>
Dear Iurii<br>
In principle you are right, but if you try to compute, e.g., formation <br>
heats or formation energy of defects by using different U values (I <br>
did it long time ago in the case of TiO2 related studies for metallic <br>
Ti, for TiO2, for Ti2O3) you fall sometimes very far from experimental <br>
references, farther then using non corrected GGA. I've used <br>
semiempirically averaged values in a lot of studies (including the <br>
correction of O- and N-related bands in oxides and nitrides) and <br>
things are a bit semiempirical, obviously, but still quite improved <br>
with respect to GGA. As long as you motivate what you are doing, I do <br>
not see particular harm in semiempirical U corrections!<br>
Best<br>
Giuseppe<br>
<br>
Quoting Timrov Iurii <<a href="mailto:iurii.timrov@epfl.ch" target="_blank">iurii.timrov@epfl.ch</a>>:<br>
<br>
> Dear Malte,<br>
><br>
><br>
> This is not an easy question. Let me express my opinion.<br>
><br>
><br>
>> I read many times in papers that I should not compare total energies of<br>
>> DFT+U calculations with different U values.<br>
><br>
><br>
> Can you give some references please?<br>
><br>
><br>
> Well the problem is that in literature often people still use <br>
> empirical values of U. In this case, I agree that the comparison of <br>
> energies will depend strongly on U which one chooses based on some <br>
> arguments. But if one computes U from first principles, then U is a <br>
> response property of each system (and the response is different in <br>
> each system, hence different U) then it makes sense to me to compare <br>
> energies (but I am not aware of any theoretical proof that it is <br>
> allowed to do so).<br>
><br>
><br>
> For example, in LiCoO2 the computed value of U for Co-3d is 6.91 eV, <br>
> while U for Co-3d in CoO is 4.55 eV (both using 'atomic' orbitals <br>
> and some pseudos on top of PBEsol - see PRB 101, 064305 (2020)). So <br>
> in both cases we consider Co-3d states, but the value of U is very <br>
> different. This is so because Co-3d states require different <br>
> corrections when they are in different chemical environments (these <br>
> Co-3d states hybridize differently with ligands in different <br>
> compounds). If you use an average value of 5.73 eV for both systems, <br>
> then this value is not a response property of any of these systems <br>
> (I do not know what it is in this case).<br>
><br>
><br>
> One can also see DFT+U as DFT with a Hubbard functional: the U <br>
> correction is different for different systems (why should it be the <br>
> same?). It's like hybrids: people tune the value of alpha in PBE0 <br>
> for each system (alpha is related to the inverse of the dielectric <br>
> constant), and alpha is different for different systems (because the <br>
> dielectric constant is different). So here is the same, U is <br>
> different for different systems: different systems need different <br>
> amount of the corrections to restore the piece-wise linearity of the <br>
> energy (PRB 71, 035105 (2005)) and alleviate self-interactions.<br>
><br>
><br>
> The comparison of energies with different ab initio U values was <br>
> done in these papers: PRB 99, 094102 (2019); PRR 2, 023313 (2020); <br>
> PRM 3, 033801 (2019).<br>
><br>
><br>
>> What should I do in case of complex hull calculations of a phase diagram?<br>
><br>
><br>
> I would compute U for each phase and compare energies.<br>
><br>
><br>
>> With QE it is now possible to calculate the U values for each <br>
>> composition by the hp.x code. Can I<br>
> compare the resulting total energies with each other safely?<br>
><br>
><br>
> Well there is no general consensus on this issue, in part because <br>
> still many people use empirical U. I think that if U is computed ab <br>
> initio (i.e. it is really a response property of each material) then <br>
> it makes sense to compare energies with different U values (U <br>
> computed for each system). But people are still investigating this <br>
> issue.<br>
><br>
><br>
> I will be happy to hear other opinions, and to hear <br>
> corrections/remarks if something what I said is misleading.<br>
><br>
><br>
> Greetings,<br>
><br>
> Iurii<br>
><br>
><br>
> --<br>
> Dr. Iurii TIMROV<br>
> Postdoctoral Researcher<br>
> STI - IMX - THEOS and NCCR - MARVEL<br>
> Swiss Federal Institute of Technology Lausanne (EPFL)<br>
> CH-1015 Lausanne, Switzerland<br>
> +41 21 69 34 881<br>
> <a href="http://people.epfl.ch/265334" target="_blank">http://people.epfl.ch/265334</a><br>
> ________________________________<br>
> From: users <<a href="mailto:users-bounces@lists.quantum-espresso.org" target="_blank">users-bounces@lists.quantum-espresso.org</a>> on behalf of <br>
> Giuseppe Mattioli <<a href="mailto:giuseppe.mattioli@ism.cnr.it" target="_blank">giuseppe.mattioli@ism.cnr.it</a>><br>
> Sent: Thursday, October 29, 2020 3:52:48 PM<br>
> To: Quantum ESPRESSO users Forum<br>
> Subject: Re: [QE-users] Thermodynamics with DFT+U<br>
><br>
><br>
> Dear Malte<br>
> I suppose that, as a trivial rule of thumb, if U values are not so<br>
> different then you can use some average value for all systems.<br>
> HTH<br>
> Giuseppe<br>
><br>
> Quoting Malte Sachs <<a href="mailto:malte.sachs@chemie.uni-marburg.de" target="_blank">malte.sachs@chemie.uni-marburg.de</a>>:<br>
><br>
>> Dear all,<br>
>><br>
>> I read many times in papers that I should not compare total energies<br>
>> of DFT+U calculations with different U values. What should I do in<br>
>> case of complex hull calculations of a phase diagram? With QE it is<br>
>> now possible to calculate the U values for each composition by the<br>
>> hp.x code. Can I compare the resulting total energies with each<br>
>> other safely?<br>
>><br>
>> Best regards,<br>
>> Malte<br>
>><br>
>> --<br>
>> Malte Sachs<br>
>> Anorganische Chemie, Fluorchemie<br>
>> Philipps-Universität Marburg<br>
>> Hans-Meerwein-Straße 4<br>
>> 35032 Marburg (Paketpost: 35043 Marburg)<br>
>> Tel.: +49 (0)6421 28 - 25 68 0<br>
>> <a href="http://www.uni-marburg.de/fb15/ag-kraus/" target="_blank">http://www.uni-marburg.de/fb15/ag-kraus/</a><br>
><br>
><br>
><br>
> GIUSEPPE MATTIOLI<br>
> CNR - ISTITUTO DI STRUTTURA DELLA MATERIA<br>
> Via Salaria Km 29,300 - C.P. 10<br>
> I-00015 - Monterotondo Scalo (RM)<br>
> Mob (*preferred*) +39 373 7305625<br>
> Tel + 39 06 90672342 - Fax +39 06 90672316<br>
> E-mail: <<a href="mailto:giuseppe.mattioli@ism.cnr.it" target="_blank">giuseppe.mattioli@ism.cnr.it</a>><br>
><br>
> _______________________________________________<br>
> Quantum ESPRESSO is supported by MaX <br>
> (<a>www.max-centre.eu<http://www.max-centre.eu</a>>)<br>
> users mailing list <a href="mailto:users@lists.quantum-espresso.org" target="_blank">users@lists.quantum-espresso.org</a><br>
> <a href="https://lists.quantum-espresso.org/mailman/listinfo/users" target="_blank">https://lists.quantum-espresso.org/mailman/listinfo/users</a><br>
<br>
<br>
<br>
GIUSEPPE MATTIOLI<br>
CNR - ISTITUTO DI STRUTTURA DELLA MATERIA<br>
Via Salaria Km 29,300 - C.P. 10<br>
I-00015 - Monterotondo Scalo (RM)<br>
Mob (*preferred*) +39 373 7305625<br>
Tel + 39 06 90672342 - Fax +39 06 90672316<br>
E-mail: <<a href="mailto:giuseppe.mattioli@ism.cnr.it" target="_blank">giuseppe.mattioli@ism.cnr.it</a>><br>
<br>
_______________________________________________<br>
Quantum ESPRESSO is supported by MaX (<a href="http://www.max-centre.eu" target="_blank">www.max-centre.eu</a>)<br>
users mailing list <a href="mailto:users@lists.quantum-espresso.org" target="_blank">users@lists.quantum-espresso.org</a><br>
<a href="https://lists.quantum-espresso.org/mailman/listinfo/users" target="_blank">https://lists.quantum-espresso.org/mailman/listinfo/users</a></div>
</span></font>
</div>
_______________________________________________<br>
Quantum ESPRESSO is supported by MaX (<a href="http://www.max-centre.eu" rel="noreferrer" target="_blank">www.max-centre.eu</a>)<br>
users mailing list <a href="mailto:users@lists.quantum-espresso.org" target="_blank">users@lists.quantum-espresso.org</a><br>
<a href="https://lists.quantum-espresso.org/mailman/listinfo/users" rel="noreferrer" target="_blank">https://lists.quantum-espresso.org/mailman/listinfo/users</a></blockquote></div>