QuantumATK => General Questions and Answers => Topic started by: Dipankar Saha on August 26, 2013, 20:55
Title: Phonon Dispersion
Post by: Dipankar Saha on August 26, 2013, 20:55
Can anybody help me to find out the phonon dispersion, for any 2D material (e.g., graphene), using vnl 13.8.b1 ???
Title: Re: Phonon Dispersion
Post by: kstokbro on August 26, 2013, 23:39
Add a calculator (use ATK classical for speed) Add a phonon bandstructure object perform the calculation done
Title: Re: Phonon Dispersion
Post by: Anders Blom on August 27, 2013, 10:28
When you have a DeviceConfiguration in the Scripter, only the Analysis options that are relevant for devices are shown. You can only compute the phonon band structure for a BulkConfiguration, just like you need a bulk structure to calculate the electron band structure.
Title: Re: Phonon Dispersion
Post by: Anders Blom on August 27, 2013, 12:45
It's not so much the point what the device is made of etc. It's the same question as for the electronic band structure, as I wrote above. If you want to calculate a quantity which is a property of a bulk system (like band structure or phonon dispersion) you must use a periodic system, and a device is not periodic (in Z). So if you want to know the phonon dispersion of a 4-AGNR system, then make a 4-AGNR unit cell as a BulkConfiguration and compute it. A device structure doesn't have the property phonon dispersion, just as it doesn't have a band structure.
Title: Re: Phonon Dispersion
Post by: Dipankar Saha on September 5, 2013, 06:02
Hi all ... :) / Can anybody suggest the way out (regarding the problem as described below)...??? ____________________ Structure: Armchair GNR Width: 4 C direction repetition: 10 Passivated with: H+ Now, while trying to find out the Phonon Band Structure, with ATK-extended Huckel(K points 1,16,16; Grid Mesh cutoff 10 Hartree), the problem is... It's is not getting converged in the end. Now if I choose ATK Classical(E0,E1-- 0 eV, 4eV)....then I'm getting the result, but that is not satisfactory at all (Simply a few straight lines between the G, Z points)
Title: Re: Phonon Dispersion
Post by: Nordland on September 5, 2013, 10:09
Never use Huckel for anything with forces except if you have your own pair-potential at hand.
If I follow the recipe given above, this is my results:
Title: Re: Phonon Dispersion
Post by: Dipankar Saha on September 5, 2013, 12:34
Thank u Norland... for all those details.../ Bt, again, there will be the question, that how can I use the calculator (as mentioned in your code), called "TersoffCalculator" ??? _______________________
Error showing in Log:
# ------------------------------------------------------------ # +------------------------------------------------------------------------------+ | | | Atomistix ToolKit 13.8.b1 [Build a1c2415] | | | +------------------------------------------------------------------------------+ Traceback (most recent call last): File "c:\users\user\appdata\local\temp\0223720281437294.py", line 368, in <module> calculator = TersoffCalculator(parameters=Tersoff_CH_2005()) NameError: name 'Tersoff_CH_2005' is not defined +------------------------------------------------------------------------------+ | NanoLanguageScript execution finished | +------------------------------------------------------------------------------+
Title: Re: Phonon Dispersion
Post by: Nordland on September 5, 2013, 15:29
Sorry my bad.
I will upload a new version of the script.
Title: Re: Phonon Dispersion
Post by: Nordland on September 5, 2013, 15:31
Alternative you can also change the calculator to the Brenner calculator
Code
calculator = BrennerCalculator()
Title: Re: Phonon Dispersion
Post by: Dipankar Saha on October 19, 2013, 12:55
Sir while running the Phonon Transmission Spectrum....for a simple two electrode Graphene Device.... I am getting the folllowing error..... May I know, where is it going wrong?? _____________________________
Traceback (most recent call last): File "/tmp/9951790614159924.py", line 396, in <module> self_energy_calculator=KrylovSelfEnergy(), File "./zipdir/NL/Analysis/PhononTransmissionSpectrum.py", line 108, in __init__ File "./zipdir/NL/ComputerScienceUtilities/Timer.py", line 45, in __call__ File "./zipdir/NL/Analysis/PhononTransmissionSpectrum.py", line 107, in <lambda> File "./zipdir/NL/Analysis/PhononTransmissionSpectrum.py", line 333, in calculatePhononTransmissionSpectrum File "./zipdir/NL/Analysis/PhononTransmissionSpectrum.py", line 359, in calculateDevicePhononTransmissionSpectrum File "./zipdir/NL/Calculators/DeviceCalculatorInterface.py", line 145, in _dynamicalMatrix File "./zipdir/NL/Calculators/DeviceCalculatorInterface.py", line 398, in downgrade File "./zipdir/NL/Calculators/SlaterKosterCalculator/Builders/DeviceSlaterKosterBuilder.py", line 351, in downgrade NL.ComputerScienceUtilities.Exceptions.NLError: Internal Error - Trying to use non EquivalentBulk density.
Title: Re: Phonon Dispersion
Post by: Anders Blom on October 20, 2013, 02:05
It's an odd error message, but my suspicion is that your Slater-Koster method doesn't have a pair potential and hence cannot be used to compute phonon properties (or perform geometry optimization, etc).
Title: Re: Phonon Dispersion
Post by: Dipankar Saha on October 23, 2013, 11:33
That may be...let me check..../ Thanks... :) / In this regard, let me ask you something..... Considering the calculation time, complexity and obviously the accuracy....can you please suggest that, which calculator will be a better option to choose while calculating the Phonon Transmission Spectrum of any simple device structure (say, for example 8-AGNR with 12 repetitions in c direction)??
Title: Re: Phonon Dispersion
Post by: Anders Blom on October 23, 2013, 18:30
Classical potentials is more or less the only sensible option unless you have a very powerful cluster to parallelize on. The results should be ok - but of course one needs to be aware of the limitations of an empirical method.
Title: Re: Phonon Dispersion
Post by: Dipankar Saha on October 23, 2013, 18:50