EPW Forum

Dear EPW developers/users:
I was trying to calculate the intrinsic resistivity of a metal (as an extension of fcc Pb in tutorial) and some puzzles appeared in these days.

In my understanding, the total (isotropic) resistivity is calculated by averaging the resistivity from all q points defined by nqf1*nqf2*nqf3. Thus,
1, The very fine q mesh (nqf1*nqf2*nqf3) gives us all positive values in all q points, why we don not need to consider the q points with negative values like (-0.5,0.5,0)?
2, If I want to calculate the resistivity along x axis (Suppose the lattice is orthogonal), I revised the a2f.90 in src to select specific q points which are parallel with x axis to obtain specific a2f(w). Then the resistivity along x is calculated by ziman formula. Is this treatment right?

Thank you in advance!
Jin Zhang

Dear Jin Zhang,

1) The grid can be defined between [0,1]. This is equivalent to [-0.5,0.5] since its a periodic system.

2) For resistivity along special axis, you need a resistivity tensor (a 3x3 matrix). the diagonal component will be xx,yy and zz.

Best,

Samuel

Dear Samuel,
Many thanks for your reply.

I have no idea how to proceed on this part.
Should I just use a revised the Ziman formula or something more on the codes?
Can you introduce some related materials (publications) to me (jzhang1991@iphy.ac.cn)?

Thank you for your help.
Best,
Jin

Hello,

The Ziman formula (as presented in the EPW paper) is indeed isotropic.

I'm working in a transport version of the code with anisotropy. I will let you know when the code is ready.

Best,

Samuel

Hi,

Any news on the transport with anisotropy ? Do you have an idea when it will be released (is it in terms of weeks, months, years ?). Meanwhile, what can we do to compute the resistivity along a special axis ?

Thanks !

Guillaume

Hello,

It will be released before the Mars 2018 ICTP school.

Best,
Samuel