[CPMD-list] CHARGES for solids

[Jiri Houska]

Dear all,
when I tried to calculate atomic charges for TiN solid material I  
received the error "ESP CHARGES| NUMBER OF FITTING POINTS 0". I found  
that when this error was reported previously in 2004, prof. Hutter  
replied "You have to use [for solids] one of the other possibilities  
available in CPMD for this (Mulliken, Davidson, Hirshfeld)."

Athough I do not fully understand that (I calculated ESP charges  
succesfully for other solids on the same platform), the Hirshfeld  
charges would be hopefully OK. However, is there any method to  
calculate with cpmd ONLY the Hirsfeld ("INT")charges and not the ESP  
charges? My understandng is that both is called by the same keyword  
CHARGES ...

One more question: please does anybody have an experience which of the  
available charges is best to use for calculation of Coulombic forces  
between atoms?

Thank a lot for any advice.
Jiri Houska
Dpartment of Engineering Physics
Ecole Polytechnique de Montreal

Input file:

&CPMD
   PROPERTIES
   COMPRESS READ WRITE
   RESTART WAVEFUNCTION LATEST
&END

&SYSTEM
   ANGSTROM
   SYMMETRY
    1
   CELL
    8.588 1.0 1.0 0.0 0.0 0.0
   CUTOFF
    60. (same error for various cutoffs)
   SCALE
&END

&DFT
   NEWCODE
   FUNCTIONAL BLYP
&END

&ATOMS

TITANIUM
*Ti_MT_BLYP_NLCC.psp KLEINMAN-BYLANDER NLCC RAGGIO=1.6
LMAX=D LOC=S SKIP=P
   32
.....
&END

&PROP
   CHARGES
&END