onlyphbte problems

[zhengjian4210]

Dear Nakib,
Thank you for the great work on elphbolt.

I have successfully run the si example. And the result is not much different from the given output example. But I am plagued by some problems.

I modified the qmesh=24 , kmesh=24 , and added onlyphbte = .t.. The calculated phonon thermal conductivity is 135 W/(m K), which is almost the same as that calculated by shengbte (ngird=24) .

But without adding onlyphbte = .t., the phonon thermal conductivity is 109 W/(m K). I tested from 6 to 30, and I found that differences began to gradually appear starting from qmesh=6. I am very confused about this. The files for elphbolt are as follows:
input1.txt
input2.txt
output1.log
output2.log

Could you please provide some insights and help for me? Any advice or suggestions would be greatly appreciated.

Thank you for your time and help.

Best regards,
Jian Zheng

[nakib]

Dear Jian Zheng,

Thank you for your message.

Looking at your input and output, I think the code is doing what it is supposed to. In the input1.txt, you said onlyphbte = .t.. Note that, by default, phonon-electron interaction is turned off (phe = .f.). In input2.txt, you did not specify anything related to the type of BTE. By default, elphbolt will solve the coupled electron-phonon BTEs. Since for the coupled electron-phonon BTEs, the phonon-electron interaction will have to be on, the phonon thermal conductivity will be reduced compared to your first case. (I can see in output2.log that the system is n-doped, with n = 2.38e19 cm^-3.)

The README.org file lists the default values of (almost) all input parameters. Furthermore, the output file will echo the parameters. You will find it useful to go read the entire output file to understand what the code is doing.

Let me know if you need any further assistance.

Best,
Nakib

[zhengjian4210]

Dear Nakib,

Thank you for your timely and informative response.

You are right. I am using elphbolt in accordance with the README.org. But for the seond case, the code not only outputs coupled electron-phonon transport but also outputs dragless phonon/electron transport. For the dragless part, the other phonon-electron interaction still exists. This is the reason why thermal conductivity changes. Can I understand it this way?

Best,
Jian Zheng

[zhengjian4210]

Dear Nakib,

Sorry, I found that I may not have explained clearly at the beginning. Let me add a little more, my confusion about the differences in thermal conductivity mainly concerns the dragless part of the two output files.

Best,
Jian Zheng

[nakib]

Dear Jian Zheng,

The difference is not due to drag, rather it is because the first calculation does not include the effect of ph-e scattering, while the second calculation does. You can try a calculation where you have onlyphbte = .t. and phe = .t.. This will carry out a dragless phonon BTE calculation including the effects of both the ph-ph and ph-e interactions.

Best,
Nakib

[zhengjian4210]

Dear Nakib,

I think I understand now. And I will do as you said. Thanks a lot !

Best,
Jian Zheng