Charge schemes

[icamps]

Hello,

I would like to know which are the charge scheme used by xtb.

If running with GFN1-xTB, we got the Mulliken and CM5 charges but when running with GFN2-xTB, I just got the charges, without any other information about the scheme used.

Regards,

Camps

[haneug]

These charges should be Mulliken charges which are briefly mentioned in the GFN2-xTB paper.

[haneug]

You can also find this in the scc routine:

xtb/src/scc_core.f90

Lines 484 to 489 in 1922dd3

	! new q
	call mpopsh(n,ndim,nshell,ao2sh,S,P,qsh)
	qsh = zsh - qsh
	! qat from qsh
	call qsh2qat(ash,qsh,q)

and

xtb/src/scc_core.f90

Lines 1304 to 1326 in 1922dd3

	!> Mulliken pop shell wise
	subroutine mpopsh(n,nao,nshell,ao2sh,S,P,qsh)
	integer nao,n,nshell,ao2sh(nao)
	real(wp) S (nao,nao)
	real(wp) P (nao,nao)
	real(wp) qsh(nshell),ps
	integer i,j,ii,jj,ij
	qsh=0
	do i=1,nao
	ii =ao2sh(i)
	do j=1,i-1
	jj =ao2sh(j)
	ps=p(j,i)*s(j,i)
	qsh(ii)=qsh(ii)+ps
	qsh(jj)=qsh(jj)+ps
	enddo
	ps=p(i,i)*s(i,i)
	qsh(ii)=qsh(ii)+ps
	enddo
	end subroutine mpopsh

[icamps]

The Mulliken charges are used even when I add mulliken=false in my input file?

[haneug]

The Mulliken charges are a core component of GFN2-xTB and can not be turned off to my knowledge.