2
$\begingroup$

I have obtained DFT results for a bulk TiO2 anatase structure using ICSD 9852 but my formation energy calculations are much higher than the literature values. How can I validate my bulk structure and calculate the correct formation energies for bulk Anatase TiO2?

I have used the following equations to calculate the formation energy: Formation energy = Total Energy of Bulk TiO2 - [(# of O atoms in bulk *Total energy of O atom) + (# of Ti atoms in bulk * Total energy of Ti atom)]

A sample of the input file of the TiO2 bulk is mentioned below.

&control
calculation = 'vc-relax',
prefix = '_anatase',
restart_mode = 'from_scratch',
PSEUDO_DIR = './',
outdir = './',
tstress = .TRUE.
/
&system
ibrav = 0,
nat = 6,
ntyp = 2,
nspin = 2,
ecutwfc = 60.,
ecutrho = 240.,
occupations = 'smearing',
smearing = 'mv',
degauss = 0.01,
starting_magnetization(1) = 0.0
starting_magnetization(2) = 0.5
/
&electrons
conv_thr = 1.0d-8,
/
&ions
ion_dynamics = 'bfgs'
/
&cell
/
ATOMIC_SPECIES
O 15.999 o_pbe_v1.01.uspp.F.UPF
Ti 47.867 ti_pbe_v1.uspp.F.UPF 

ATOMIC_POSITIONS (crystal)
  O       0.79222389202102   0.79222389202102   0.00000000000000  ! // O 
  O       0.54222389202102   0.04222389202102   0.50000000000000  ! // O 
  O       0.95777610797898   0.45777610797898   0.50000000000000  ! // O 
  O       0.20777610797898   0.20777610797898   0.00000000000000  ! // O 
  Ti      0.00000000000000  -0.00000000000000   0.00000000000000  ! // Ti 
  Ti      0.75000000000000   0.25000000000000   0.50000000000000  ! // Ti
 
CELL_PARAMETERS (angstrom)
  -1.93542293504944   1.93542293504944   4.84316761246253
   1.93542293504944  -1.93542293504944   4.84316761246253
   1.93542293504944   1.93542293504944  -4.84316761246253

K_POINTS (automatic)
4 4 4 1 1 1
Improve this question
$\endgroup$
4
  • $\begingroup$ +1 and welcome to MMSE! Could you please edit your post to add how you calculated the formation energy and the energy values of TiO2, isolated Ti atom and O atom? $\endgroup$
    – Abdul Muhaymin -Free Palestine
    Commented Jun 17 at 6:30
  • 1
    $\begingroup$ In particular, you need to specify which computing code you are using, which exchange-correlation functional you are using, and how you have verified the convergence. It is usually easiest to simply upload the input file, so that we can look through the whole thing. $\endgroup$
    – naturallyInconsistent
    Commented Jun 17 at 7:22
  • 1
    $\begingroup$ Your edit has included the input file, but it is missing stuff. For example, if you don't tell us which pseudopotential library you are using, we cannot tell if 60Ry is a suitable wavefunction energy cutoff. You might have used vc-relax for the bulk, but when you are comparing, you should take the vc-relax final results and run an SCF without the vc-relax, just to be extremely sure that things are done properly. Little things like that are helpful. $\endgroup$
    – naturallyInconsistent
    Commented Jun 17 at 8:26
  • $\begingroup$ There might be other issues in your calculations of isolated atoms but there are already two major issues in the input file you provided. First, you are using a very coarse k-grid. For such a small system, you should at least use $12\times 12\times 12$ or for accurate results, use $16\times 16 \times 16$. Second, since you are using ultrasoft pseudopotentials, the ecutrho value should be 8 to 12 times larger than ecutwfc. So, maybe set at least ecutrho=600. Ideally, instead of guessing these values, you should always perform convergence tests to find the appropriate values. $\endgroup$
    – Abdul Muhaymin -Free Palestine
    Commented Jun 17 at 16:12

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .