Dear Colleagues, I trying to estimate (at least on a qualitative level), where the electrons are located in the environment of 3- and 4-fold oxygen coordinated boron atoms. But i found a strange result for the Mulliken charges, since the electro-negativity of boron (approx 2.0) is much lower than the electro-negativity of oxygen (approx. 3.5) i would expect naivly, that the charges are more located near the oxygen but i get following result* for the mulliken charges O approx. 5.9 and B approx. 3.2, i.e. the boron is negative charged and the oxygen is slightly positive charged. Hence I have some questions (i) How good is the (possible) correlation between Difference of the (macroscopic) electro-negativity and the Mullikencharges? (ii) Are the mulliken charges the right meassure for information about local charges (in the strict classical sense, i.e. a charge which acts as a effective source of coulomb interactions). (iii) If not- There is a big number of different definitions of effective charges beside the Mullikencharges. Whats the best choice of effective charges in the view of a optimal compromise between computational effort (i.e. lowest possible expense of programming time to extract the informations from siesta output) and physical meaning (I want to use this to estimate an effective coulomb interaction)?
I thank you in advance Michael * Investigating by small clusters (20-30 aotm) using PBE and a PAO.Energyshift of 0.001 -- Michael Schuch FB. Physik / AG. Statistische Physik Barbarastrasse 7 49069 Osnabrück
