All versions of transiesta allow one to populate the electrons according to a specific temperature smearing related through the Fermi function.

In 4.1 and newer versions one can also set different temperatures in each of the electrode baths.

See e.g. Sec. 3.7 in 10.1016/j.cpc.2016.09.022.

Also, let me stress that 4.0.1 has been released. You are highly encouraged to at least use that version (it is the latest stable release).

Thank you @Nick Papior

I will look into it.

@Nick Rübner Papior

Dear Nick,

As far as I know, the temperature smearing in DFT code can not represent the temperature effect on the electronic structures. It just speeds up the convergence of electronic structure SCF. On the other hand, the temperature in TBTrans broadens the bias window. In this sense, it partially takes account of the temperature effect on electron transport, since effects from temperature induced changes in electronic structures are maybe more significant than this.

With best regards,

/Guang-Ping Zhang

My understanding is that temperature parameters in transiesta only accounts for its effect on the electronic structure, as what Guang Ping has said.

If you want to simulate STM experiments, you will have to take into account the electronic transport of various molecular geometries that may arise due to temperature effects.

If you can find out the time window over which the STM measurement was made, might be possible to use kinetic monte carlo (or other techniques) to simulate nuclei movements over the time period. Taking snapshots over the time window, and simulating the electronic transport of each molecular configuration might give you a way to simulate the ensemble IV character of the STM junction.

Respected all,

I have a further question:

To study the IV characteristics of any system (say, benzene), what should be set the electronic temperature during the TranSIESTA / TbTrans run ?

I tried for 10K and 300K (default), both giving different transmission.

Regards.