Find information on common issues.
Ask questions and find answers from other users.
Suggest a new site feature or improvement.
Check on status of your tickets.
By Gloria Wahyu Budiman1, Yunfei Gao1, Xufeng Wang1, Siyu Koswatta2, Mark Lundstrom1
1. Purdue University 2. IBM
Simulate 2-D electrons transport in CNTFET
You must login before you can run this tool.
Version 0.9.6 - published on 15 Aug 2014
doi:10.4231/D39C6S20D cite this
View All Supporting Documents
Write a review
11 Feb 2010
3.0 out of 5 stars
The main benefits of this tool is that it is the only NEGF CNT simulator that I have found on nanoHUB and that it runs reasonably fast for a NEGF simulator. While a simulation on a long channel CNT device – such as 100nm – took about 35min to simulate, this is very tractable in comparison to other NEGF codes that I have simulated. The ability to take the code and export it easily to MATLAB is handy for additional analysis, and the fact that the data points are given in the plots is also useful for back-of-the-envelope estimates using the tool.
The tool allows a great deal of control over many important parameters, such as the tight-binding energy and the work function of the gate, but it is imperative to have control over the gate, source, and drain biases for performing further analyses. Also, knowing which scattering mechanisms (e.g. phonons) are present in the device in the simulator’s menu would be helpful. There seem to be some issues with output presentation – the simulator seems to freeze after accessing DOS end of simulation results, and it would be useful not to state 100% finished when there is plotting of results to follow, as it seemed to take about 2-3 more minutes just to plot results.
With regards to the channel length, I believe that there is a lower limit on the length which is higher than the tooltips denote, as the Vd vs. Ids, Vg, and capacitance curves break down for small channel lengths (e.g. 2nm). Moreover, despite the description of the device stated that it was helpful for simulating band-to-band tunneling, the effects of this behavior are not especially apparent to me.
In the future, if the tool had a way to mapping CNT (n,m) to a diameter and band gap using a tool like CNTbands, that would prove to be very useful. Furthermore, estimations of running time – for problem tractability – and mobility for cylindrical wrap-gated geometry using charge control model – for CNT device characterization – would be useful. Based on the complexity and computational expense of modeling of different geometries using NEGF, simulating that might be out of scope for the tool. A back-gated CNTFET transport simulation using NEGF could prove to be very enlightening, however.
When comparing this tool against the literature – especially the references that it cites – show some inconsistencies in the tool port. The references include Fermi levels in band diagrams from the Green’s functions – these are not apparent in the ending plots. Also, the Ids vs. Vgs plot does not have the same dip at negative gate biases per the plots in Appl. Phys. Lett. 87, 253107 (2005), which should be due to increased phonon scattering. Since I could not determine which scattering mechanisms were present, I could not intuit what was the cause of the inconsistency.
Overall, this tool was useful in finding out more about the band structure of the CNTFET in a wrap-gated geometry as well as some of the transport parameters. Further modifications would make this a great tool for nanoHUB, as I believe this tool’s functionality will improve as the physics of CNTs, graphene, and graphene nanoribbons are more well-understood.
Joshua Wood (UIUC)
Please login to vote.