[Pw_forum] vc-relax gives good result of lattice constant, but not good in cohesive energy

[Dongsheng Zhang]

Hi,

I am calculating the optimum lattice constant nd cohesive energy of Zr hcp
crystal. The total energy of a sinle Zr atom I got from input file Zr.in is
-5.53133015 Ry, the total energy of Zr hcp crystal I got from input file
Zr.hcp.in is -12.13562915 Ry, so the cohesive energy is -0.536 Ry,
converted to -7.30 eV. From a reference paper, the cohesive energy should
be -6.2eV. Attached below are Zr.in and Zr.hcp.in. Could anyone tell me if
I had made any mistake in those input files? Thanks.

Zr.in:

&control
    prefix='Zr',
    pseudo_dir = '/home/zhang/Software/espresso_PP/',
    outdir='/share/scratch/tmpZr/eCut120SM0.01/',
 /
 &system
    ibrav=  1, celldm(1) =50, nat=  1, ntyp= 1,
    occupations='smearing', smearing='marzari-vanderbilt', degauss=0.01,
    ecutwfc =120,

 /
 &electrons
 /
ATOMIC_SPECIES
 Zr  91.224 Zr.pbe-mt_fhi.UPF
ATOMIC_POSITIONS
  Zr  0.000000000000 0.000000000000 0.000000000000
K_POINTS gamma


Zr.hcp.in

&control
    prefix='Zr',
    pseudo_dir = '/home/zhang/Software/espresso_PP/',
    outdir='/share/scratch/tmpZr/eCut120K18hcpSM0.01/',
    calculation='vc-relax'
 /
 &system
    ibrav=  4, celldm(1) =6, celldm(3) =1.633, nat=  2, ntyp= 1,
    occupations='smearing', smearing='marzari-vanderbilt', degauss=0.01,
    ecutwfc =120,
 /
 &electrons
 /
 &ions
 /
 &cell
 /
ATOMIC_SPECIES
 Zr  91.224 Zr.pbe-mt_fhi.UPF
ATOMIC_POSITIONS crystal
  Zr  0.000000000000 0.000000000000 0.000000000000
  Zr  0.666666666666 0.333333333333 0.50000000000
K_POINTS automatic
18 18 36 1 1 1

PS The optimum lattice constant got from Zr.hcp.in is very close to the
data in literature.


I have tried different value of the smearing parameter ranged from 0.001 to
0.05, none of their cohesive energy is good.

If the smearing value is 0.1, simulation by zr.in (single atom) will stop
at iteration # 10 without any error message. Why does it happen?

Thank you for your help,

Dongsheng
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20130604/b934d436/attachment.html>