<div dir="ltr">Dear quantum-ESPRESSO users,<div><br></div><div>I am running some calculations with espresso 5.1 and the van <span class="" id=":2bx.1" tabindex="-1">der</span> <span class="" id=":2bx.2" tabindex="-1">Waals</span> </div><div>rVV10 functional. </div><div><br></div><div>First, with the program: </div><div><br></div><div>generate_rVV10_kernel_table.x</div><div><br></div><div><br></div><div>I generated the universal file:</div><div><br></div><div>rVV10_kernel_table<br></div><div><br></div><div><br></div><div>Then I tried three systems up to now,<br></div><div>1. two water molecules (insulating)</div><div>2. a surface with some organic molecules on top (insulating). This is the system I am actually interested in.</div><div>3. graphite, with some smearing</div><div><br></div><div>In the first two cases </div><div>the energy seems to reach convergence but then</div><div>suddenly the <span class="" id=":2bx.3" tabindex="-1">scf</span> accuracy increases , like this</div><div><br></div><div><div>estimated <span class="" id=":2bx.4" tabindex="-1">scf</span> accuracy < 0.06260677 Ry</div><div> estimated <span class="" id=":2bx.5" tabindex="-1">scf</span> accuracy < 0.04294545 Ry</div><div> estimated <span class="" id=":2bx.6" tabindex="-1">scf</span> accuracy < 0.02031181 Ry</div><div> estimated <span class="" id=":2bx.7" tabindex="-1">scf</span> accuracy < 0.01888999 Ry</div><div> estimated <span class="" id=":2bx.8" tabindex="-1">scf</span> accuracy < 0.01019674 Ry</div><div> estimated <span class="" id=":2bx.9" tabindex="-1">scf</span> accuracy < 0.00222739 Ry</div><div> estimated <span class="" id=":2bx.10" tabindex="-1">scf</span> accuracy < 0.00317661 Ry</div><div> estimated <span class="" id=":2bx.11" tabindex="-1">scf</span> accuracy < 0.00370097 Ry</div><div> estimated <span class="" id=":2bx.12" tabindex="-1">scf</span> accuracy < 0.00396386 Ry</div><div> estimated <span class="" id=":2bx.13" tabindex="-1">scf</span> accuracy < 0.00018021 Ry</div><div> estimated <span class="" id=":2bx.14" tabindex="-1">scf</span> accuracy < 1.54950129 Ry</div><div> estimated <span class="" id=":2bx.15" tabindex="-1">scf</span> accuracy < 1.54973018 Ry</div><div> estimated <span class="" id=":2bx.16" tabindex="-1">scf</span> accuracy < 1.54679045 Ry</div></div><div><br></div><div>...and so on and does not converge after 100 steps. I tried to reduce the mixing to 0.1, or even smaller, without any effect. (see attachments)</div><div><br></div><div>For the more complex surface+molecules system, the behaviour is similar.</div><div><br></div><div>For the graphite with some smearing, the <span class="" id=":2bx.17" tabindex="-1">eigenvectors</span> of the <span class="" id=":2bx.18" tabindex="-1">Hamiltonian</span> fail to converge in the <span class="" id=":2bx.19" tabindex="-1">diagonalization</span> process. (see attachment)</div><div><br></div><div>I would appreciate some help to ascertain if: </div><div><br></div><div>1. I am doing something wrong somewhere? </div><div>2. there is an issue in the convergence of rVV10 with insulators?</div><div>3. is there a (different) issue with rVV10 and smearing? </div><div><br></div><div>I am aware that water calculations with rVV10 have been successfully performed, so I would appreciate very much your suggestions.</div><div><br></div><div>Thanks</div><div><br></div><div>Best,</div><div>Andrea</div><div><br></div><div>-- <br> Dr Andrea Floris<br> Research Associate<br> King's College London<br> Strand, London WC2R 2LS<br> United Kingdom<br> Phone: +44 (0) 207 848 2064<br> Fax : +44 (0) 207 848 2420<br> Location: Strand Building, 4th floor, Room 4.02 <br> Emails: <span class="" id=":2bx.22" tabindex="-1">andrea</span>.<span class="" id=":2bx.23" tabindex="-1">floris</span>@<span class="" id=":2bx.24" tabindex="-1">kcl</span>.ac.<span class="" id=":2bx.25" tabindex="-1">uk</span>, an.<span class="" id=":2bx.26" tabindex="-1">floris</span>@<a href="http://gmail.com">gmail.com</a><br>
</div></div>