Quantumtransport through CdS-Nanowire

[Ikgut]

Hallo, I have posted some questions in the forum and there good advices, thank you all very much....

But I have a general question. I want to build in modell a device with Schootky(Au) and Ohmic (In/Au) contact. Between them CdS Nanowire( There can beo other materials like Au or Pt for electodes). I have pricipal question: Is that poosible with ATK do buid a modell for that device and how would I do this?

P.S.: Maybe you I can tall me a bit a background for this calculation? I read some literaure from Datta to NEGF and Extend Huecke approach...Is that the way how ATK calculate electric properties like I-V-plots

Thank you very much

Ikgut

[zh]

For the model of Au-CdS naowire-Au, it can be definitely built and simulated by the ATK. The steps of model setup and simulation are basically same with those in the following paper:
Man-Fai Ng et al., Geometry Dependent I−V Characteristics of Silicon Nanowires, Nano Letters, 8, 3662 (2008), http://pubs.acs.org/doi/abs/10.1021/nl801668p

Maybe you I can tall me a bit a background for this calculation? I read some literaure from Datta to NEGF and Extend Huecke approach...Is that the way how ATK calculate electric properties like I-V-plots

Basically, they are same. The more detailed background can be found in "M. Brandbyge, J.-L. Mozos, P. Ordejón, J. Taylor, and K. Stokbro, Phys. Rev. B 65, 165401 (2002)" and "K. Stokbro, D. E. Petersen, S. Smidstrup, A. Blom, M. Ipsen and K. Kaasbjerg, Phys. Rev. B 82, 075420 (2010)". In ATK, the I-V curve can be calculated by the NEGF combined with the DFT or the Extend Huecke approach.

[Ikgut]

Thank you for advice...I try to build a device.....but my I_V curves are only linear....I dont know why?

I can´t load a build of my device...But mybe I can describe what I did....I build a CdS-Nanowire with Custom Builder....and make in Build Repitation....than I labeled all atoms at both end of the wire and make Au Atoms and other Ti-Atoms(I have delete S atoms). This procedre I make for 3 Layers at the end of wire, than I make  a Device Mode. And I have a device like in a paper that you recommend on me. Than I made usually steps to simulate I-V curves like in tutorial...

But I see only linear characteristic? Why I dont see diode characteristic of this device? How I can change parameters than I see I-V-curves with physical reality?

Thank you

IKgut

[Anders Blom]

Sounds like you've gone through the right steps, but without exact details it's a bit hard to say for sure. It's important for you to understand all details of the system. For instance, if you see pure linear I-V you must have a transmission spectrum which is independent of energy, in the bias window under consideration. Is that the case? What is the bias range? And, if this is the case, you can for a 1D system map the transmission to the band structure of the central region and electrodes. Does the central region exhibit a metallic character (when computed with the same parameters as the transport system)? And, finally - it seems you are using Huckel parameters. Whenever you apply a semi-empirical method, one must verify that the parameters of the model provide physically relevant results (by e.g. comparing the band structure of CdS with DFT, etc).

[Ikgut]

...Okay...I think I understand a bit what do you mean because in Landauer formalism transmission is function of energy and bias voltage, they must be the same, is that right? or how I can prove the transmission dependance? My bias range are plus/minus 1V and second calculation plus/minus 2 V. I have very short nanowire, because at the momant I not calculate on cluster, maybe thats is the reason?

I try to realize your advices ....Thynk you

IKgut

[Anders Blom]

You cannot really judge the linear I-V based on 2-3 points :-) Try to ramp up the voltage in steps of about 0.1 V to get a real curve. And do check the transmission spectrum separately as well, plus the band structure.