<html xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40"> <head> <meta http-equiv="Content-Type" content="text/html; charset=Windows-1252"> <meta name="Generator" content="Microsoft Word 15 (filtered medium)"> <style></style> </head> <body lang="en-CH" link="blue" vlink="purple" style="word-wrap:break-word"> <div class="WordSection1"> Hi Daniel<o:p></o:p> <o:p> </o:p> Did you try to decrease EPS_DEFAULT and to increase the PW cutoff? These input parameters mainly determine the accuracy of the atomic forces rather than the convergence thresholds for the force calculation like MAX_FORCE or MAX_DR which can also show convergence for poor forces.<o:p></o:p> <o:p> </o:p> HTH<o:p></o:p> <o:p> </o:p> Matthias<o:p></o:p> <o:p> </o:p> <div id="mail-editor-reference-message-container"> <div> <div> <div style="border:none;border-top:solid #B5C4DF 1.0pt;padding:3.0pt 0cm 0cm 0cm"> From: 'daniel Storm' via cp2k <cp2k@googlegroups.com> Date: Tuesday, 8 October 2024 at 14:50 To: cp2k <cp2k@googlegroups.com> Subject: [CP2K:20756] Inconsistent geometry optimisations<o:p></o:p> </div> Hi everyone,<o:p></o:p> Just to provide some context, in our group we use CP2k to model molecular reactivity in the solid state, so slightly different than “standard” solid state chemistry. We run optimisations for minima and transition states (TS), whose character we confirm by running phonon calculations.<o:p></o:p> Now, to the issue. I am working with a crystal structure of an iridium organometallic salt, [Ir(PONOP)(H)Me][BArF4], however, I am encountering issues with achieving consistent optimisations. Despite starting from the same experimental structure, I am getting very different SCF energies (up to 3 kcal/mol) and I have observed minor differences in the resulting geometries. I think that those geometries are close enough to not show that big of an SCF difference, but my main worry is the fact that the starting point is always the same.<o:p></o:p> The convergence criteria that I am currently using is MAX_FORCE 1.0 x 10-4. I have attempted to improve this by tightening the criteria (MAX_FORCE, MAX_DR, RMS_DR and RMS_FORCE) to 1.0 x 10-6, but this didn’t fix the issue. All the optimisations are done with the PBE-D3 functional and DZVP-MOLOPT-SR-GTH basis set.<o:p></o:p> So, to sum up, I run optimisations with the same starting geometry that converge to different structures, and I am not sure why or which criterion to use to select one structure over the other, etc. <o:p></o:p> I have worked with other organometallic salts in the past, and this is the first time I find this issue. I am using CP2k version 2023.2 in the HPC cluster Archer-2.<o:p></o:p> Has anyone seen this before? Any advice on how to proceed?<o:p></o:p> Thanks in advance,<o:p></o:p> Daniel.<o:p></o:p> -- 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 <a href="mailto:cp2k+unsubscribe@googlegroups.com">cp2k+unsubscribe@googlegroups.com</a>. To view this discussion on the web visit <a href="https://groups.google.com/d/msgid/cp2k/d50cb0ef-42e9-465d-81c7-2fdacdab6d41n%40googlegroups.com?utm_medium=email&utm_source=footer"> https://groups.google.com/d/msgid/cp2k/d50cb0ef-42e9-465d-81c7-2fdacdab6d41n%40googlegroups.com</a>.<o:p></o:p> </div> </div> </div> </div> </body> </html> -- 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 <a href="mailto:cp2k+unsubscribe@googlegroups.com">cp2k+unsubscribe@googlegroups.com</a>. To view this discussion on the web visit <a href="https://groups.google.com/d/msgid/cp2k/ZRAP278MB0827F79512B3329F4B818BCEF47E2%40ZRAP278MB0827.CHEP278.PROD.OUTLOOK.COM?utm_medium=email&utm_source=footer">https://groups.google.com/d/msgid/cp2k/ZRAP278MB0827F79512B3329F4B818BCEF47E2%40ZRAP278MB0827.CHEP278.PROD.OUTLOOK.COM</a>.