QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: atk_user on February 19, 2014, 05:16

Title: I-V Calculation Density Matrix
Post by: atk_user on February 19, 2014, 05:16

Dear all,

I try to calculate I-V characteristics of two terminal devices. I set the voltage range from 0 to 1 V (11 points).
The calculation results shows 0, 0.1, 0.2, 0.3, ... 1.0 V. The log represent the calculation performed step by step from 0 V to 1V.

I have a question about I-V calculation procedure.
After calculation of specific voltage point (0 V or something), the next step voltage (0.1V or something) continuously calculate by considering the density matrix of previous step voltage (0V)?  ???

or

Each voltage steps are independently calculated? ???

Title: Re: I-V Calculation Density Matrix
Post by: Anders Blom on February 19, 2014, 11:58

Both :) Each one is separately calculated, but for 0.1 V it will use the solution from 0.0 V as starting guess to make it converge faster. This is also why it's important not to do the scan from -2 V to + 2 V in one go, since the -2 V calculation will come first, and it will be very hard to converge. Instead it should always be split into two steps, starting at 0 V going up and down, separately.

Title: Re: I-V Calculation Density Matrix
Post by: Dipankar Saha on February 19, 2014, 15:13

1) Is there any effect of temp. change on the G_el calculation?? /
2) How the transmission spctrm. data are being found in ATK?

Title: Re: I-V Calculation Density Matrix
Post by: Anders Blom on February 19, 2014, 16:15

1.There is a temperature effect on the Fermi distribution of the electrons in the electrodes, but there is no thermal coupling.
2. See the standard refs http://quantumwise.com/support/faq/87

Title: Re: I-V Calculation Density Matrix
Post by: Dipankar Saha on February 19, 2014, 18:24

Dr. A. Blom,

Thank you sir..... / Let me go through these... :)

Regards__
Dipankar Saha

Title: Re: I-V Calculation Density Matrix
Post by: atk_user on February 20, 2014, 03:27

Dear Dr. Anders Blom

Thank you for your help.  :)

I have more questions about the number of KPOINTS for I-V calculations.

I've checked the current values are different as increase the number of KPOINTS. It's not finished but from the specific KPOINTS the current is incased with increase the KPOINTS.
(very small bias from 0V to 0.15V; 7 points / the current is very small about 60 nA)


Is there any guide lines to determine the number of KPOINTS?

Title: Re: I-V Calculation Density Matrix
Post by: Anders Blom on February 20, 2014, 12:40

The more k-points the more accurate, so the only rule is to increase until the results stop changing.

Title: Re: I-V Calculation Density Matrix
Post by: Dipankar Saha on February 22, 2014, 10:47

"atk-user",

for any further details you can see these previous discussions.......

http://quantumwise.com/forum/index.php?topic=2628.msg12335#msg12335

or,

http://quantumwise.com/forum/index.php?topic=2523.msg11776#msg11776

Regards__
Dipankar Saha