How to calculate PVR (Ip/Iv) in NDR curve and effective mass?

[Sukhbir]

Hello,

I have done some DFT calculation of device and bulk and I have following question regarding it.
1. How to find peak valley ratio (PVR) Ip/Iv of negative differential region, If Iv lies in negative value( see attachment IVplot).?
2. How to find effective mass and what else details are provided in it (see attachment (Effective mass)) ?     

[Jess Wellendorff]

1) Perhaps you could export the I-V data from the widget and measure the peak-to-valley ratio manually.
2) The Effective Mass widget gives you of course the effective mass, but also shows you information about how the mass is derived from a fit to the bandstructure in the vicinity of the k-point in question.

[Anders Blom]

Of course, this plot is from an older version of ATK - in 2014 and 2015 you should use the LabFloor IV curve plugin instead. We don't really provide support for older versions.

But unfortunately this plot is very misleading, due to the spline interpolation, which sometimes is a very bad idea.
It would be best to extract the current for each bias and just plot the I-V relation manually. From looking at the transmission spectra, I suspect the curve will look very different compared that that odd oscillation plot. You should be aware of the fact that the current cannot possibly switch sign unless also the bias switches sign. That is, for positive bias, the current can only be zero or positive, never negative. The fact that it is, in the plot, is due to the interpolation. Perhaps with a large step size the oscillations go away, but I doubt it - there are simply too few points for the spline interpolation to work properly. The transmission plot is correct though.

[Sukhbir]

Hello Sir,
 Sorry for late reply. But I am confused of split interpolation. What is it and how to correct it ?

Thanking You,
Sukhbir Singh. 

[Anders Blom]

Typo, sorry. Should be "spline".
To fix it, just plot the curve using the raw data (the points) without spline or add more points on the curve - the spline interpolation will be accurate if you have enough points, but with very few points it can go wrong..

[Sukhbir]

Thanks for your kind reply it really works. I m very happy.