Unrestricted calculations

[peremiro]

QCxMS does not read spin when performs the initial xTB calculation in EI.

initializing GFN1-xTB ...
initialization successful!
energy = -58.25234
charge = 0
mult = 2
etemp = 298.1

I think the spin is determined by the subroutine getspin in utility.f90. Probably it could be modified to read the .UHF uf present.

  !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
  ! spin value of molecule
  ! returns -1 if no electrons are found (e.g. H+)
  !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
  subroutine getspin(nat,ic,chrg,isp)

  integer nat,i,j,chrg,isp,ic(nat)

  j=0

  do i = 1, nat
     j = j + ic(i)
  enddo

  j = j - abs(chrg)
  isp = 1 + mod(j,2)

  if ( j < 1 ) isp = -1

 end subroutine getspin

[JayTheDog]

That's because it's QCxMS and not xTB. The charge is read through the input file, which should not be overwritten by an exterior file. Having all options in one place makes more sense than implementing something purely xTB specific, as e.g. ORCA wouldn't read the .UHF file anyways.

[peremiro]

The charge is read correctly, but the spin is calculated always as singlet or doublet by QCxMS, even if a higher multiplicity would be the correct one. I cannot find a way to specify multiplicities above 2.