calculateSpinCurrent — Calculate the spin current for a two-probe system for a given set of bias voltages using the linear-response approximation.
The function calculateSpinCurrent calculates the spin current in a two-probe system for a given set of bias voltages, using the linear-response approximation. The spin current is calculated as the difference between the linear-response current for spin-up and spin-down electrons. For more details about the linear-response approximation, please see calculateLinearResponseCurrent.
List of arguments
An object returned from a previously performed self-consistent calculation for a TwoProbeConfiguration.
Default:
None
Values of voltages where each entry is in units of Volt. The voltage is symmetrically superimposed on the chemical potentials in the electrodes. For example, in the specific case of a positive value of the voltage V, the superimposed voltages are -V/2 and +V/2 on the left and right electrodes, respectively.
Default:
None
The k-point sampling used for integrating the transmission spectrum in the 2D Brillouin zone transverse to the transport direction.
The imaginary infinitesimal added to the energy of the retarded Green's function.
Default: 1.0e-5*eV
The number of energy-points used for the integration.
Default: 100
The function calculateSpinCurrent returns a
NumPy Array of dimension
len(voltages),
holding the spin current corresponding to the input bias voltages.
Each element of the array is a
PhysicalQuantity with units of Ampere.
Please see ATK.Units for the query methods of a PhysicalQuantity.
from ATK.TwoProbe import * from ATK.Consortia import MSM scf = restoreSelfConsistentCalculation('myfile.nc') bz_params = brillouinZoneIntegrationParameters((1,1)) voltages = [-0.10,-0.05, 0.00, 0.05, 0.10, 0.15, 0.20, 0.25] * Units.Volt spin_current = calculateSpinCurrent( scf, voltages, bz_params) for entry in zip(voltages,spin_current): print entry[0], entry[1]
The default k-point sampling (Γ point only) is a crude approximation if the electrodes are three-dimensional (e.g. Au fcc), however, it is completely sufficient if the electrodes are effectively one-dimensional (atomic chain, carbon nanotube, etc).
For high bias voltages the linear-response approximation is no longer valid.