Dear Jeffei, I a normal calculation, you would want the Hartree to *match* between the leads and the scattering region. Since the potential is defined up to a constant, this matching is done *by hand*, by using some appropriate flags (don't know them by heart). So that at a given location (a minima of the potential, you tell the program what the Hartree --let's call it VH(z)-- is).
Now suppose that instead of telling the program the value you obtained for your Hartree potential VH(z), you (with hindsight) feed it VH(z)+DeltaV, where DeltaV is chosen by you. The program will run without complaining, but in your mind you know that DeltaV has shifted the reference potential, hence providing you with a gate voltage, because this shift of bias will occur across your system with the exception of the leads, where the Hartree is fixed to a bulk value plus/minus the chemical potential there. That's all there is to it. In my mind the problem is not is setting the gate voltage, but in converging your calculation once you've done so. It would be nice if you let us know how convergence looks like once you've set some suitable shift. Best regards, -Salvador. ----- Original Message ----- From: "Jeffei Xu" <jeffei...@gmail.com> To: siesta-l@uam.es Sent: Wednesday, November 4, 2009 8:24:17 AM GMT -05:00 US/Canada Eastern Subject: Re: [SIESTA-L] How to apply gate voltage in transiesta Dear Jose, Thank you very much for your reply. For transport calculation, gate voltage is sure an important parameter. In the manual of TranSIESTA-C, it says: ''the electrostatic effect of the gate electrode is simulated by simply shifting the scattering region part of the Hamiltonian with the gate voltage (converted into an electrostatic potential energy).'' I am a interested user, but definitely not skilled, I don't know how to include this effect in the .TSHS file of SR. I will try to read and understand the source code of SIESTA. However, if anybody could do or could give more detailed help, that will be very appreciated. This will also benefit the whole SIESTA community. Jeffei On Tue, Nov 3, 2009 at 12:54 PM, <jose.torres.alo...@uam.es> wrote: > > Dear Jeffei: > > I think the implementation of the Gate Voltage in TranSIESTA-C > consisted of a simple shift of the Energy reference at the > Scattering Region, which simulated the effect of some > electrostatic field locally applied. > > In the TranSIESTA modules within SIESTA that Gate Voltage model > has not been implemented, and I am sorry to say that it is not > in the near-future road map of implementations. > > Nevertheless, it should not be too hard for the interested skilled > user to include such a shift in the SCF loop. I encourage him to > try out and let us know the result ... > > Yours, > > Jose a. > > -- > Jose A. Torres, Ph.D. > Manager of the SIESTA Software > > > Jeffei Xu <jeffei...@gmail.com> escribió: > >> Dear all, >> >> Can we apply gate voltage in transiesta (siesta 3.0 beta)? >> I don't find any description about this in the manual. >> I know it is possible in transiesta-C. >> >> Best regards, >> >> Jeffei >> > > > > -- Salvador Barraza-Lopez Postdoctoral Fellow School of Physics The Georgia Institute of Technology
