Warning: The calculation did not converge to the requested tolerance!

[tara]

I use quantumwise atk to do the parallel calculation and using slater koster for the calculation.

 I've attached my script in this post.I am getting this warning.I have used k point sampling KA=300 for transmission spectrum and density of states so as to have an effective width of approx =(300*0.7nm)= 210nm

WARNING :calculation did not converge to the requested tolerance!


Can you help me out as to what could be reason for this non convergence???

[Umberto Martinez]

The convergence of your SCF part has nothing to do with the transmission spectrum or dos parameters.
Actually, I suggest you to run these analysis after you manage to converge your SK calculation properly.

You dope left and right electrodes with the same doping type.
According to your tags you should dope the left electrode wit p-type.

If you still have convergence issue you can try to take a look here here

[tara]

Sir,

My calculation for doping concentraton=1e19, converged on increasing fermi poles to 12
contour_parameters = DoubleContourIntegralParameters(
    circle_points=50,
    integral_lower_bound=2.5*Hartree,
    fermi_function_poles=12,

 ,however when i increased my doping value  to 2e21 ,my slater koster calculation did not converge for using same number of fermi poles and for below mention values

    contour_parameters = DoubleContourIntegralParameters(
    circle_points=50,
    integral_lower_bound=2.5*Hartree,
    fermi_function_poles=12,
)

what could be the reason for it?how can i  converge my   calculation for e21 doping concentration?

[tara]

no reply???

[Anders Blom]

There may be many other reasons why it doesn't converge, such as a too short screening region (too short central region, simply). This is particularly important with doping.

[Anders Blom]

Also, you seem to have no hydrogen atoms terminating the dangling surface bonds, so even if it did converge, you would probably get the wrong results.

[tara]

My calculation for doping concentraton=1e19, converged on increasing fermi poles to 12
contour_parameters = DoubleContourIntegralParameters(
    circle_points=50,
    integral_lower_bound=2.5*Hartree,
    fermi_function_poles=12,

however when i tried to sweep the gate bias  from  vg=-0.2v to 0.4v , for drain bias of vd=0.1v , my calculation did not converge to required tolerance!

 +------------------------------------------------------------------------------+
| 199 E = -1604.47 dE =  4.225174e-07 dM =  6.558203e-06 dH =  5.744619e-04    |
+------------------------------------------------------------------------------+
################################################################################
#                                                                              #
# Warning: The calculation did not converge to the requested tolerance!        #
#                                                                              #
################################################################################


what  could i possibly do to make the calculation converge at other biases??

i have attached below the 2 files