QuantumATK Forum
QuantumATK => General Questions and Answers => Topic started by: skg on July 11, 2016, 11:56
-
Dear sir,
I am trying to calculate the mobility for Silicon, following the tutorial given in link:
http://docs.quantumwise.com/tutorials/mobility.html#id11
For silicon electronic band structure calculation, I am applying MGGA (since LDA or GGA underestimates the band gap) , but ATK is not allowing me to calculate the dynamical matrix elements using MGGA. Throwing an error message "Back Engine Exception : MGGACalculator does not implement this method."
Can you please tell me how to proceed on that and calculate the mobility for silicon.
-
The band gap is not relevant for the mobility, what matters most is that you get the correct effective mass in the conduction band (if you are calculating the electron mobility), and LDA does a great job at that, actually.
-
The reason for the error you get is that mGGA is not implemented for force calculations, which you need for the dynamical matrix. Please see the Reference Manual for further details: http://www.quantumwise.com/documents/manuals/latest/ReferenceManual/index.html/chap.atkdft.html
-
Thanks for the quick reply.
In the given tutorial we are calculating the mobility using Boltzmann transport equation, using RTA approximation, which is independent of effective mass. My confusion is whether the relaxation time (scattering time) is independent of LDA or GGA.
-
In short: No, the relaxation time is not independent of the functional. The relaxation time depends on the electron-phonon coupling, as shown in equation 5 in the tutorial, and this will always depend on the functional. However, it is difficult to say a priori whether that dependence is strong or not.