<div dir="ltr">Hi all, I want to use DFT within cp2k to get an estimate of the energy of the charge transfer state of 2 molecules. Hence, I was looking for a method that forces molecule A to a charge of +1 and molecule B to a charge of -1. So far, I have been playing around with BECKE_RESTRAINT and DDAPC_RESTRAINT within the DFT->QS section and also within the PROPERTIES->ET_COUPLING section (not simultaneously though). I have a couple of questions regarding my results so far: 1) What do I specify with the target value? The total charge of a fragment? (and is it defined the same for BECKE and DDAPC?) 2) I get completely different values in the output for "Final target value of constraint A:" for Becke and DDAPC. For Becke I was assuming that the value should be (number of valence electrons +/- 1) for fragment A/B, i.e. in my case 83 and 85. (I have obtained these values, but this also results in a far-too-high coupling value, so there is probably something wrong.) With ddapc on the other hand, I get two completely different, and mutually hardly varying values for the "Final target value of constraint A" and of constraint B, both between 0 and 1. So my question is what these final target values tell me, what do they represent, and what/why is the difference between the final target values of the two methods? 3) Am I right if I say that I should tweak the strength-variable till I obtain a correct "final target value"? 4) Can one beforehand tell if one of the sections (QS vs ET_coupling) is more appropriate for this case than the other section? (I don't need the coupling info, but with QS's Becke-restraint, I can only apply one restraint; I guess that this can be countered by setting the overall charge to 0 though) 5) There is also Mulliken and S2_restraint in the QS-section. If anyone knows if one of the four restraint-methods is (most) appropriate/inappropriate or you have good/bad experiences with one of them, please let me know. If you have the answer or experience to one or more of the questions, please let me know, would really help me a lot! Best, W.R </div>