<div>Hello,</div><div><br /></div><div>I want to do a RESP fit to determine the partial charges for a simple molecule in cp2k. My molecule has 18 atoms. I already looked in past discussions on RESP charges and my questions are not resolved. My procedure was:</div><div><br /></div><div>-Optimize the geometry in cp2k</div><div>-Do a single point calculation with the optimized geometry <br /></div><div>-Fit the RESP charges to obtain the partial charges</div><div><br /></div><div>I do have the partial charges of this molecule from a paper and from a DFT calculation using gamess (with different XC/BASIS) so I can compare the cp2k results to the "ground truth". <br /></div><div>In my opinion the charges obtained via cp2k are quite off. I fixed lots of charges by using the REPEAT method in a periodic box, but my carbons are off!</div><div>I am unsure if this is due to some settings I use, like an inappropriate Basisset/potential when doing my SCF calculation or if my RESP fit is poor, although my RMS seems good.</div><div><br /></div><div>WHAT I TRIED:</div><div>-If I restrict my charges close to the target charges I get more reasonable results, but obviously this approach is nonsense when doing an unknown system (which will be a COF structure in the future). So far I have been using a strength of 0.001. I also tried using constraints to impose equal charges, but still the charges on my molecule deviate from my expected results (this symmetry based assumption I can use in the future for the COF system).</div><div><br /></div><div>-I tried different basis set and XC but it didnt seem to improve much. <br /></div><div><br /></div><div>-I used the optimized geometry from gamess and cp2k as input and both result in unrealistic charges</div><div><br /></div><div>-I used REPEAT METHOD in a periodic box, even though my testsystem was originally in a nonperiodic box and this works much better, here some Carbons are off! In particular, atoms which are not directly bonded to large charge sinks (like S, N) deviate from my expected charges. I use CAMB3LYP as the XC, but I am unsure if I use it correctly. I looked at previous discussions for the use of CAMB3LYP, but I would be happy if anyone could confirm whether my approach is valid!</div><div><br /></div><div>Attached are the script for the geometry optimization and the script for the single point calculation where I determine the RESP charges using REPEAT. I also provide the resulting output file for the charges and the "target" charges from gamess and my Basis/Potential files.</div><div><br /></div><div>I would like to understand if this is a fitting issue, a basisset/potential issue or something in between. There are lots of options to turn on/off in cp2k and maybe I lack a crucial ingredient for RESP calculations. Maybe dispersion? How would I implement that? Any advice is appreciated. <br /></div><div><br /></div><div>Thank You in advance. Sincerely, <br /></div><div>Andreas Doell</div><div><br /></div><div> <br /></div><div><br /></div>
<p></p>
-- <br />
You received this message because you are subscribed to the Google Groups "cp2k" group.<br />
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>.<br />
To view this discussion visit <a href="https://groups.google.com/d/msgid/cp2k/52edb18a-8359-418e-b76b-2f78d2a3bfe5n%40googlegroups.com?utm_medium=email&utm_source=footer">https://groups.google.com/d/msgid/cp2k/52edb18a-8359-418e-b76b-2f78d2a3bfe5n%40googlegroups.com</a>.<br />