QuantumATK Forum
QuantumATK => General Questions and Answers => Topic started by: ramkrishna on March 10, 2011, 13:13
-
I want to make rectangular Silicon nanowires (SiNW) having 10 nm length and 2X3 nm^2 cross-sectional area along <100> direction. Can anyone help me to make such a system??
-
use the nanowire builder in the custom builder and set the (100) surface energy to 0.1
Remember to passivate the wire with the sp3 passivater which can be found under scripts
-
Dear Prof Stokbro,
Please let me know how is the surface energy in nanowire window related to the length , cross sectional area ? In the nanowire Wulff constructor window, how the radius slot is affecting the length and breadth ? how can I specify the dimension of the rectangular nanowire?
thanking you,
Ramkrishna
-
Dear Sir,
Please let me know the answers to the above questions.
Thanking you,
Ramkrishna
-
If you increase the radius, the wire cross-section becomes larger. The surface energies, or rather their relation to each other, determine the shape.
-
Dear Sir,
Thank you for your valuable answers but how can I specify the dimension of the rectangular nanowire in this procedure?
Regards
Ramkrishna
-
I guess you are referring to, how you can make it rectangular rather than quadratic. Well, you can't, the tool in it's current design will only generate regular structures. One alternative is to build a 1x1 nm^2 shape and "glue" them together, 2x3. If you use a metallic material or Si or Ge, you could use the EMT or Brenner potentials in ATK-Classical to optimize the so-obtained geometries. Or build a 3x3 nm^2 structure and remove 1 nm in one dimension.
-
Dear Sir,
Can you please let me know the procedure of making a 1X1 nm^2 Si wire ? Eagerly waiting for your answer.
Regards
Ramkrishna
-
radius 5 Å, 100 surface energy 0.1
-
Dear Sir,
To get 2X2 nm^2 should we have to double the radius only or change has to be made in surface energy also?
Regards
Ramkrishna
-
As mentioned already above, the surface energy controls the shape, not the size. The exact number for the radius if you want 2 nm may not be exactly double, but close.
-
Dear Sir,
Thank you for your valuable reply .
Regards
Ramkrishna
-
Dear Sir,
I have made a hydrogen passivated rectangular Silicon nanowire along (100) with surface energy 0.1 and radius 1.5nm. Attached please find the file.
After running using Extended Huckel method, I find the direct band gap very close to zero. [Direct Band gap 0.0023 eV at [0.0000, 0.0000,0.4114] ]. However we know from theory that the direct band gap should be of the order of 3-4 eV .
Please let me know if there is any mistake in my model.
Regards
Ramkrishna
-
Why do you have (4x4) k-points in X/Y? A more proper k-point sampling would be 1x1x100. Also, why limit the number of iterations to 10? Are you sure it even converged? Do you have a reference for the "theory" value of the band gap? (I formulate my answer as a few questions, because you really need to think about these things, and understand the basics.)
-
Dear Sir,
I have done as per your previous suggestion but I still get very low band gap (almost zero). Please see the following file and FIG 2 of this given link of a journal paper http://prb.aps.org/abstract/PRB/v76/i15/e155435.
Regards
Ramkrishna
-
There is a list of errors in your script.
- First of all your structure is wrong. You have passivated the system in the C-axis. These should be removed.
- The parameters you have choosen, seems a bit random. The only thing you should set for this system, is the basis set and the number of k-points in Z to something bigger than 1 or so.
- The parameters for your analysis is also a poor choice if you want to see a nice bandstructure with a good bandgap
I have allowed myself to correct all the errors in the script. Try running it and see if your results is not much closer to the article.
-
Dear Sir,
I have tried to run the script silicon_wire.py in atk version 11.2 but it gives the following error
Traceback (most recent call last):
File "/tmp/1669493090142845.py", line 415, in <module>
vacuum_level=-7.67047*eV,
File "./zipdir/NL/Calculators/HuckelCalculator/Parameters/HuckelBasisParameters.py", line 70, in __init__
File "./zipdir/NL/Calculators/HuckelCalculator/Parameters/HuckelBasisParameters.py", line 255, in checkOnsiteHartreeShift
NL.ComputerScienceUtilities.Exceptions.NLValueError: The parameter, onsite_hartree_shift, must be an energy - e.g. 2.0*Units.Ry
NanoLanguageScript execution failure
Regards
Ramkrishna
-
Sorry.
I was using a different version. I have updated the script above.
-
Dear Sir,
The file, which you have given, is still giving very very low band gap ,
Direct Band gap 0.0082 eV at [0.0000, 0.0000,0.2855] .
Regards
Ramkrishna
-
Dear Sir,
I am waiting for your valuable answer. Please help me to overcome this problem.
Regards
Ramkrishna
-
It may be easier to just cleave the structure, then make a rectangular surface lattice (for 100 and 110 it's already rectangular), then make a supercell in the XY plane to fit your required size. It may be necessary to cut out the wire from a slightly larger supercell, in that case use the select tool to select the atoms you want keep, the hit Ctrl+I to invert the selection and finally delete the extra atoms now selected.
-
what do you mean with "different dimensions"?
there is a cluster radius parameter in the plugin.
if you mean different dimension of the sides of the square section you will have to use the selection tools and directly delete the selected atoms.