<p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt"><span lang="EN-US" style>Dear all,</span></p>
<p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt">I am trying to
define the value of the electron affinities of oxygen vacancy (i.e., the energy
gain when the electron from the bottom of the conduction band is trapped at the
defect) in bulk crystal of transition metal oxides, as follows:</p>
<div><br></div><div>E_tot(perfect,
q=-1) + E_tot(defect, q= 0) - E_tot(perfect, 0) - E_tot(defect, q=-1)</div>
<p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt"> </p>
<p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt"><span lang="EN-US" style>Whether this approach is valid for
direct using in ESPRESSO? I know the total energy of charged systems have no
physical meaning in ESPRESSO because of the interaction with the balancing
background of charge. </span>However in such approach there is difference: E(perfect,
q=-1)-E(defect, q=-1), and I suppose that errors are cancelled. <span lang="EN-US" style>Is it true supposition? I am exploiting
B3LYP functional and have no problem with band gap value. </span></p>
<p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt">For
preliminary results, I obtained electron affinities for oxygen vacancy
in HfO2 and Ta2O5
polymorphs are close to zero. Moreover, the results are sensitive to supercell
size.</p>
<p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt"> </p>
<p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt"><span lang="EN-US" style>My second question is follows. Is it
possible to define the value of the electron affinities using the above
equation adapted with Janak’s theorem? The simplified according to the mean
value theorem for integrals Janak’s theorem stating that:</span></p>
<div><span lang="EN-US" style><br></span></div><div><span lang="EN-US" style>E_tot(q=-1) - E_tot(q= 0) ~= [e(h+1,N) +
e(h+1,N+1)]/2,</span></div>
<div><span lang="EN-US" style><br></span></div><div><span lang="EN-US" style>where e(h+1,N) is the Kohn-Sham
eigenvalue of the lowest unoccupied state for the neutral system</span></div>
<p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt"><span lang="EN-US" style>e(h+1,N+1) is the Kohn-Sham
eigenvalue of the highest occupied state for the -1 charged system.</span></p>
<p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt"><span lang="EN-US" style> </span></p>
<p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt">
</p><p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt"><span lang="EN-US" style>Best Regards, Timofey Perevalov,</span></p>
<p class="MsoNormal" style="margin-bottom:0cm;margin-bottom:.0001pt"><span lang="EN-US" style>Rzhanov Institute of Semiconductor
Physics SB RAS</span><br></p><p></p>