How to decide the size of the central region at the interface of the electrode

[kalewu]

Hi all,

I want to construct the two-probe device using Device from Bulk in the DEVICE TOOLS. It is easy to select the electrode lengths. However, I don't know how to selece the size of the central region at the interface of the respective electrode.

Can you give me some advices?

Thanks in advance.

Kale

[Anders Blom]

First read http://quantumwise.com/publications/tutorials/mini-tutorials/99-how-to-set-up-the-electrodes-properly-in-a-two-probe-system. You should really set the size of the central region before using the Device from bulk tool, but I hope things are clearer after you read that page. What you probably refer to are the +- buttons at the bottom; those are only used to make the central region larger/smaller in case you need to, in order to be able to increase the electrodes.

[kalewu]

First read http://quantumwise.com/publications/tutorials/mini-tutorials/99-how-to-set-up-the-electrodes-properly-in-a-two-probe-system. You should really set the size of the central region before using the Device from bulk tool, but I hope things are clearer after you read that page. What you probably refer to are the +- buttons at the bottom; those are only used to make the central region larger/smaller in case you need to, in order to be able to increase the electrodes.

Thanks for your reply.

I have read the page you gave. In my calculations, I just set the left(right) electrode (one primitive unit cell containing 64 atoms) length to be 6.97 A. I checked the electrode length using the script downloaded from that page and found that the truncation distance is 8.2088 Ang. It seems that 6.97 A is a litter smaller. However, if I use two primitive unit cells (the length is 6.97*2=  13.94 A) as the electrode, the system is too large to calculate.

I just want to know whether the results is not right if I just use the electrode length of 6.97 Ang.

Thanks in advance.

[Anders Blom]

If depends a bit on which calculator you use, but if it's DFT it's probably not too bad. It's an approximation, but as long as all direct matrix elements are accounted for, it's probably not worse than a few other approximations made in the calculation.

[kalewu]

If depends a bit on which calculator you use, but if it's DFT it's probably not too bad. It's an approximation, but as long as all direct matrix elements are accounted for, it's probably not worse than a few other approximations made in the calculation.

Thanks. I just used the DFT to calculate the T(E).