why the charge of 2(H ) and 3(H ) in AIM are inequality？

For highly symmetric system like benzene, the inequality problem will be more evident. The essential reason is that the basin module is based on cubic grid, however, the grid distribution commonly does not conform to molecular symmetry.

by I_was_a_baby

There is a obvious finding that the correction of ADCH charge for Hirshfeld charge is in excess of that of CM5 charge. So, if the regular calculation can be done, the ADCH charge can be recommended.

On the other hand, the AIM charge has been overrated the atomic charge. As the same as the cost contained time and computing resource is high, the AIM method for atomic charge is not suggested.

I totally agree with you.

As an extremely powerful program for realizing electronic wavefunction analysis, Multiwfn also can be applied to calculate the atomic charges of each atoms in molecule. Currently, the CM5 method, charge Model 5 by D. G. Truhlar, has been supported by Multiwfn. It can be seen that, from the definition, the CM5 method is a correction for Hirshfeld charge, just like the ADCH charge by Lu Tian.

It makes me pay more attention on the difference between them. So, by taking a water molecule as examples, there are some atomic charges have been obtained as list below. The level of geometry optimization is B3LYP/6-311G*.

Hirshfeld
1(O ) is   -0.328822
2(H ) is    0.164430
3(H ) is    0.164430
ADCH
1(O ) is   -0.812445
2(H ) is    0.406242
3(H ) is    0.406242
CM5
1(O ) is   -0.661055
2(H ) is    0.330546
3(H ) is    0.330546
AIM
1(O ) is   -1.027696
2(H ) is    0.513844
3(H ) is    0.513852

There is a obvious finding that the correction of ADCH charge for Hirshfeld charge is in excess of that of CM5 charge. So, if the regular calculation can be done, the ADCH charge can be recommended.

On the other hand, the AIM charge has been overrated the atomic charge. As the same as the cost contained time and computing resource is high, the AIM method for atomic charge is not suggested.