<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=GB2312">
</head>
<body text="#000000" bgcolor="#FFFFFF">
Dear All:<br>
I want to modify the code of ph.x to calculate phonon limited
mobilities of electrons in semiconductors.<br>
I know ph.x can calculate phonon line-width, which contains the
integral of the electron-phonon coefficients g_{q,v}(k,i,j). And the
relaxation time can be calculated from these coefficients too:<br>
1/\tau_{k,i} = (2\pi)/(\hbar) sum_{q,v,i} |g_{q,v}(k,i,j)|^2
( (f_{k+q,j}+n_{q,v})\delta(e_{k,i}-e_{k+q,j}+\hbar \omega_{q,v})
......)<br>
The main difference between computations of phonon line-width and
relaxation time is one integrated in k-space but another in q-space.
Thus I think I need only to add some lines in subroutine elphsum()
to collect and output g_{q,v}(k,i,j) and then integrate them in
q-space.<br>
But I encounter a problem: I'm confused with some variables or
functions in elphsum():<br>
<br>
(1) el_ph_mat(jbnd,ibnd,ik,ii), notes in the code says it means
<\psi(k+q)|dV_{SCF}/du^q_{i a}|\psi(k)><br>
Is u^q_{i a} a normal coordinate or the i'th descartes
coordinate of a'th atom or none of them?<br>
<span style="color: rgb(0, 0, 0); font-family: Arial; font-size:
14px; font-style: normal; font-variant: normal; font-weight:
normal; letter-spacing: normal; line-height: 22px; orphans: auto;
text-align: start; text-indent: 0px; text-transform: none;
white-space: normal; widows: auto; word-spacing: 0px;
-webkit-text-size-adjust: auto; -webkit-text-stroke-width: 0px;
background-color: rgb(255, 255, 255); display: inline !important;
float: none;"></span><br>
(2) What is the function of symm()?<br>
Why must g(ik,:,:) be symmetrized by it ?<br>
Note that
g(ik,ii,jj)=CONJG(el_ph_mat_collect(jbnd,ibnd,ik,ii))*el_ph_mat_collect(jbnd,ibnd,ik,jj).<br>
<br>
(3) I also have tried to use the following formula as
|g_{q,nu}(ik,ibnd,jbnd)|^2:<br>
DO mu=1,3*nat<br>
DO vu=1,3*nat<br>
gq2(ispin,jbnd,ibnd,ik,nu)=gq2(ispin,jbnd,ibnd,ik,nu) +
DBLE(conjg(dyn(mu,nu)) * &<br>
gq(ispin,jbnd,ibnd,ik,mu,vu) *
dyn(vu,nu))<br>
ENDDO ! vu<br>
ENDDO ! mu<br>
gq2(ispin,jbnd,ibnd,ik,nu)=gq2(ispin,jbnd,ibnd,ik,nu)/2/omega<br>
<br>
where gq(ispin,jbnd,ibnd,ik,mu,vu) was gotten from:<br>
a.
g(jbnd,ibnd,ik,ii,jj)=CONJG(el_ph_mat_collect(jbnd,ibnd,ik,ii))*el_ph_mat_collect(jbnd,ibnd,ik,jj)<br>
......<br>
b. map g to the whole brillouin zone,
g(jbnd,ibnd,ik,ii,jj)--> gq(ispin,jbnd,ibnd,ik,mu,vu)<br>
......<br>
c. symmetry gq(ispin,jbnd,ibnd,ik,mu,vu) by symm()<br>
<br>
This process is very similar to the process of calculating
phonon line-width, the main difference is I do it for each k point.
But I also failed to get the right g_{q,v}(k,i,j).<br>
Could you give me some advise? Thank you very much!<br>
<br>
Regards<br>
<br>
Zhida Song<br>
State Key Laboratory of Electronic Thin Films and Integrated
Devices,<br>
University of Electronic Science and Technology of China,<br>
Chengdu 610054, P.R. China<br>
Tel:+86-28-83201475<br>
Fax:+86-28-83204938
</body>
</html>