Is my thermal-induced-current script correct?

[wot19920302]

Dear Quantumwise Staffs:           I wanna to calculate  thermal-induced-current, so I make a new script by modifying the script from this topic: http://quantumwise.com/forum/index.php?topic=3411.msg20255#msg20255.  But I have a questions:            what are units of  elementary charge e, plank constant h and energes  in the script, respectively?  I find that I have to add "eV" in this sytanx to calculate current successfully :            

Code

   k1 =1.0/planck_constant*Units.e*1*[b]eV[/b]*sum(f*T1*dE)

. But  If so. it means that there is no unit for energes.             Attachments I   is the script for calculating spin seebeck from the topic             Attachments II  is what I modified using  Attachments one             Could any one help me to inspect if there is mistakes in my script ?             Thanks          

[Jess Wellendorff]

You are making quite a few changes when going from

Code

k1 = -1.0/planck_constant*(boltzmann_constant*temperature)*sum(E*T1*df)*dE

to

Code

k1 =1.0/planck_constant*Units.e*sum(f*T1*dE)

. For example,

Code

boltzmann_constant*temperature

has units of eV, but you remove it, so you may easily be missing the unit charge.

[Anders Blom]

It's very important to note that as a strict and general rule, ATK always uses real units. So the Planck constant has the unit of action (energy*time),  energies are energies (unless removed by "inUnitsOf", which some scripts do for simplicity), the elementary charge has the unit of charge, and so on.

[wot19920302]

thank you, Sir, but I am confused           acccording to the formula in fig1, I write this code:                    

Code

  k1 =1.0/planck_constant*Units.e*sum(f*T1*dE)

planck_constant has the unit of eV*s, Units.e  is elementary charge so in unit of Coulomb(C), f is Fr-Fl so is dimensionless, T1 is electron spectrum so is also  dimensionless, dE is the infinitesimal energy so have the unit of eV.   If I multiply these dimensions  I will get Coulomb/s, which is just the current. Why should I have to add "eV" ?          what's more, E in attachment II is :                

Code

 E = numpy.array(energies+energy_zero-average_fermi_level-energy_shift)

   instead of E in attachment I :              

Code

 E = numpy.array((energies+energy_zero-average_fermi_level-energy_shift)/(temperature * boltzmann_constant))

    I don't know whether I am wrong     If possible, could you please provide a standard script for calculating thermal-induced current?    best          

[wot19920302]

can anyone help me?

[Anders Blom]

A general script is provided in http://quantumwise.com/forum/index.php?topic=3411 that you already referenced. I don't see any reason to modify it. The formula you write gives the current, but for the thermoelectric coefficients you need the conductance.

I suggest you just use the Thermoelectric Coefficient plugin if you are unsure about Python and the formalism.

[wot19920302]

thank you sir, I want to calculate current instead of cnductance. I just want to consult you whether my script (attachment II) is correct or wrong.? could you please answser me?
yours

[Anders Blom]

current = transmission_spectrum.current()
Done

[wot19920302]

thank you,sir! I understand!

well....I make a simple thing complex

best!

[Anders Blom]