[Pw_forum] Cannot reproduce standard CO/Au(111) adsorption energy calculations

[Janesko, Benjamin]

Dear QE experts, 

 I am just starting to use pwscf for surface chemistry research. I've begun by attempting to reproduce the CO/Au(111) adsorption energy in de Gironcoli et al, PRB 77, 085414 (2008). My problem is that my PBE adsorption energy is an order of magnitude lower than any I've seen in the literature! And I can't figure out what I'm doing wrong. 

I've prepared a 5-layer slab of Au(111) at the experimental bulk lattice parameter (2.884 Angstrom Au-Au spacing): 

--------------------------------------- 
&control
    calculation='relax'
    prefix='au5-3x3',
    pseudo_dir = '/home/bjanesko2/espresso/pseudo//',
    outdir='/tmp/espresso/'
 /
 &system
    ibrav=4,
    celldm(1)=9.439624066,
    celldm(3)=8.0,
    nat=15,
    ntyp=1,
    ecutwfc = 40.0,
    ecutrho = 250.0,
    occupations='smearing', smearing='marzari-vanderbilt', degauss=0.01
 /
 &electrons
    conv_thr=0.0000001
 /
 &ions
   ion_dynamics='bfgs',
   path_thr=0.03
 /
ATOMIC_SPECIES
 Au  196.966  Au.pbe-nd-van.UPF
ATOMIC_POSITIONS alat
Au      0.000000        0.000000        4.000000 1 1 1
Au      0.500000        0.288675        4.000000 1 1 1
Au      1.000000        0.577350        4.000000 1 1 1
Au      0.333333        0.000000        3.528595 1 1 1
Au      0.833333        0.288675        3.528595 1 1 1
Au      1.333333        0.577350        3.528595 1 1 1
Au      0.666667        0.000000        3.057191 0 0 0
Au      1.166667        0.288675        3.057191 0 0 0
Au      1.666667        0.577350        3.057191 0 0 0
Au      1.000000        0.000000        2.585786 0 0 0
Au      1.500000        0.288675        2.585786 0 0 0
Au      2.000000        0.577350        2.585786 0 0 0
Au      1.333333        0.000000        2.114382 0 0 0
Au      1.833333        0.288675        2.114382 0 0 0
Au      2.333333        0.577350        2.114382 0 0 0
K_POINTS automatic
8 8 1 0 0 0
----------------------------------------------------


and CO adsorbed to the fcc site, in a sqrt(3) x sqrt(3) overlayer (1/3 ML coverage)

-------------------------------------------------------
&control
    calculation='relax'
    prefix='coau5-3x3-hcp',
    pseudo_dir = '/home/bjanesko2/espresso/pseudo//',
    outdir='/tmp/espresso/'
 /
 &system
    ibrav=4,
    celldm(1)=9.439624066,
    celldm(3)=8.0,
    nat=17,
    ntyp=3,
    ecutwfc = 40.0,
    ecutrho = 250.0,
    occupations='smearing', smearing='marzari-vanderbilt', degauss=0.01
 /
 &electrons
    conv_thr=0.0000001
 /
 &ions
   ion_dynamics='bfgs',
   path_thr=0.03
 /
ATOMIC_SPECIES
 Au  196.966  Au.pbe-nd-van.UPF
 C   12.0107  C-pbe-van_ak.UPF
 O   15.9994  O-pbe-van_ak.UPF
ATOMIC_POSITIONS alat
Au      0.000000        0.000000        4.000000 1 1 1
Au      0.500000        0.288675        4.000000 1 1 1
Au      1.000000        0.577350        4.000000 1 1 1
Au      0.333333        0.000000        3.528595 1 1 1
Au      0.833333        0.288675        3.528595 1 1 1
Au      1.333333        0.577350        3.528595 1 1 1
Au      0.666667        0.000000        3.057191 0 0 0
Au      1.166667        0.288675        3.057191 0 0 0
Au      1.666667        0.577350        3.057191 0 0 0
Au      1.000000        0.000000        2.585786 0 0 0
Au      1.500000        0.288675        2.585786 0 0 0
Au      2.000000        0.577350        2.585786 0 0 0
Au      1.333333        0.000000        2.114382 0 0 0
Au      1.833333        0.288675        2.114382 0 0 0
Au      2.333333        0.577350        2.114382 0 0 0
C        0.333333333   0.000000000   5.309331423 1 1 1
O        0.333333333   0.000000000   5.544873192 1 1 1
K_POINTS automatic
8 8 1 0 0 0
---------------------------------------------------

The pseudopotentials are all from the Quantum Espresso website, and are used without modification. I get total energies at optimized geometries of -43.3739979804 Ry for an isolated CO molecule, -1530.4827281024 Ry for the Au slab, and -1573.8569359340 Ry for CO on the slab. Everything looks reasonable in XCrysDen. Am I doing something obviously wrong? 

 Thanks, 

  Ben J. 


===================
Benjamin G. Janesko
Assistant Professor
Department of Chemistry
Texas Christian University
Office: SWR 428 
http://personal.tcu.edu/bjanesko/
b.janesko at tcu.edu
Phone: 1-817-257-6202
Fax: 1-817-257-5851 (attn:janesko)