1
General Questions and Answers / Re: Moment Tensor Potential Training Attributes
« on: February 28, 2023, 10:20 »
Are you using the SP1 version of 2022.03? If not, it should work with that.
Show Posts
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Pages: [1]
2
General Questions and Answers / Re: Moment Tensor Potential Training Attributes
« on: February 27, 2023, 12:50 »
Just checked, and the command 'moment_tensor_potential_training._nlplotScatter(fit_index=0)' works fine in the current release. So, I am not sure why you are getting the error message. Can you give/share additional information?
3
General Questions and Answers / Re: PackMole file for amorhpous HfO2 structure
« on: February 15, 2023, 18:18 »
Hi Jiwon,
The best solution is to use packmol builder plugin to generate the initial structure.
For a script solution, check out this manual page https://docs.quantumatk.com/manual/Types/runPackmol/runPackmol.html for more examples.
For active learning, one can use packmol structure itself as initial geometry. Active learning annealing will result in an amorphous structure in the end.
Of course, the amorphous structure generated from a regular MD can also be used as the initial geometry for active learning.
I hope this helps.
With regards,
Suresh
The best solution is to use packmol builder plugin to generate the initial structure.
For a script solution, check out this manual page https://docs.quantumatk.com/manual/Types/runPackmol/runPackmol.html for more examples.
For active learning, one can use packmol structure itself as initial geometry. Active learning annealing will result in an amorphous structure in the end.
Of course, the amorphous structure generated from a regular MD can also be used as the initial geometry for active learning.
I hope this helps.
With regards,
Suresh
4
General Questions and Answers / Re: How to get Coordinates, Energy from AIMD run at each step
« on: December 21, 2021, 12:31 »
Please try the highlighted script from the manual page here https://docs.quantumatk.com/manual/Types/MDTrajectory/MDTrajectory.html#:~:text=import%20pylab%0A%0A%23%20Read,legend()%0A%0Apylab.show() using atkpython in a terminal. You should nlread the .hdf5 file containing the MDTrajectory object. This should give you access to the quantities you are looking for.
5
General Questions and Answers / Re: Get MTP calculator from previous training
« on: November 2, 2021, 17:00 »
Hi,
You may try the below to get the MTP calculator from .hdf5 file.
Easy way to use the calculator from .mtp file is to follow this manual page which is also what Anders suggested in previous comment
https://docs.quantumatk.com/tutorials/combining_potentials/combining_potentials.html
Basically doing this
I hope this answers your question.
With regards,
Suresh
You may try the below to get the MTP calculator from .hdf5 file.
Code
moment_tensor_potential_training=nlread('mtp_study.hdf5',MomentTensorPotentialTraining)[-1]
mtp_calculator = moment_tensor_potential_training.momentTensorPotentialCalculators()[0]https://docs.quantumatk.com/tutorials/combining_potentials/combining_potentials.html
Basically doing this
Code
potentialSet = TremoloXPotentialSet('potential_name')
potentialSet.addParticleType(ParticleType.fromElement(Oxygen))
.
....add other elements...
.
potentialSet.addPotential(MTPPotential(file='mtp_study.mtp'))
calculator = TremoloXCalculator(parameters=potentialSet)
bulk_configuration.setCalculator(calculator)I hope this answers your question.
With regards,
Suresh
6
General Questions and Answers / Re: freeEnergy()
« on: November 2, 2021, 16:15 »
Hi Cerda,
Historically, the attraction between atoms is measured as a negative value in energy units and repulsion/destabilization as a positive value (remember Lennard-Jones potential form). Therefore, any stable system of interacting atoms will have negative potential energy due to the net attraction/bonding. The freeEnergy() function from PhononDOS gives U - TS (without the potential energy) and this can be positive or negative depending on the temperature. At 0 K, there is no entropy and U=zero point energy, so pdos.freeEnergy(0*Kelvin) will be positive because ZPE is positive. As T increases, the entropic contribution to free energy increases, and free energy becomes negative when -TS overcomes U. I hope the above answers your question.
With regards,
Suresh
Historically, the attraction between atoms is measured as a negative value in energy units and repulsion/destabilization as a positive value (remember Lennard-Jones potential form). Therefore, any stable system of interacting atoms will have negative potential energy due to the net attraction/bonding. The freeEnergy() function from PhononDOS gives U - TS (without the potential energy) and this can be positive or negative depending on the temperature. At 0 K, there is no entropy and U=zero point energy, so pdos.freeEnergy(0*Kelvin) will be positive because ZPE is positive. As T increases, the entropic contribution to free energy increases, and free energy becomes negative when -TS overcomes U. I hope the above answers your question.
With regards,
Suresh
7
General Questions and Answers / Re: Gibbs free energy calculation
« on: October 18, 2021, 15:11 »
Hi Cerda,
I see that you have used a FF calculator, which is fine, but the geometry should be optimized to reach the nearest minimum in the potential energy surface before performing the dynamical matrix calculation. Without geometry optimization, your script is generating a lot of -ve normal mode frequencies. Given the size of the configuration, this could take thousands of optimization steps before the forces go below 0.01 eV/A. A good practice before computing the free energy is to check the vibrational frequencies using the 'VibrationalMode' analysis object. For nonlinear molecules, all eigenfrequencies except the first 6 should be positive. If not, then either the FF is not good enough or the geometry is not a local minimum to perform harmonic analysis.
Please try geometry optimization until the vibrational frequencies are as mentioned above before performing IdealGasThermoChemistry - negative frequencies result in nan during vibrational entropy calculation under ideal gas approximation. When you use PhononDOS, then we only consider the positive frequencies for calculation, that is why it does not give you nan. But as I mentioned earlier, PhononDOS uses Quasi Harmonic Approximation meant for solids and does not include contributions for rotation and translation. Also, it is a constant volume approximation - meaning that it only gives you the Helmholtz free energy and not the Gibbs free energy.
Let me know if there are further questions.
I see that you have used a FF calculator, which is fine, but the geometry should be optimized to reach the nearest minimum in the potential energy surface before performing the dynamical matrix calculation. Without geometry optimization, your script is generating a lot of -ve normal mode frequencies. Given the size of the configuration, this could take thousands of optimization steps before the forces go below 0.01 eV/A. A good practice before computing the free energy is to check the vibrational frequencies using the 'VibrationalMode' analysis object. For nonlinear molecules, all eigenfrequencies except the first 6 should be positive. If not, then either the FF is not good enough or the geometry is not a local minimum to perform harmonic analysis.
Please try geometry optimization until the vibrational frequencies are as mentioned above before performing IdealGasThermoChemistry - negative frequencies result in nan during vibrational entropy calculation under ideal gas approximation. When you use PhononDOS, then we only consider the positive frequencies for calculation, that is why it does not give you nan. But as I mentioned earlier, PhononDOS uses Quasi Harmonic Approximation meant for solids and does not include contributions for rotation and translation. Also, it is a constant volume approximation - meaning that it only gives you the Helmholtz free energy and not the Gibbs free energy.
Let me know if there are further questions.
8
General Questions and Answers / Re: Gibbs free energy calculation
« on: October 17, 2021, 10:07 »
Hi Cerda,
Can you share the script used to optimize the molecule and compute the dynamical matrix so that I can check the settings? If I have to guess, something wrong with the dynamical matrix calculation - because the zero-point energy is unusually large.
Can you share the script used to optimize the molecule and compute the dynamical matrix so that I can check the settings? If I have to guess, something wrong with the dynamical matrix calculation - because the zero-point energy is unusually large.
9
General Questions and Answers / Re: Gibbs free energy calculation
« on: October 14, 2021, 14:36 »
Hi Cerda,
For molecules, we have a new analysis object 'IdealGasThermochemistry' that takes a molecule configuration and its dynamical matrix object, similar to PhononDensityOfStates, to provide the Gibbs free energy including the electronic energy https://docs.quantumatk.com/manual/Types/IdealGasThermoChemistry/IdealGasThermoChemistry.html
Example:
In GUI: open scripter, under 'configurations' select 'Analysis from file'. Load the hdf5 file with the molecule configuration and dynamical matrix and select them. Add the 'IdealGasThermoChemistry' analysis object to the scripter and submit the script. After the calculation, choose the 'IdealGasThermoChemistry' object from the hdf5 file and choose text representation to get the list of Gibbs free energy at different temperatures.
PhononDensityOfStates is more relevant for bulk/slab systems as we do not have to compute the rotational and translational free energy contributions. For that, we have the 'CrystalThermoChemistry' analysis object similar to the case of the molecules above.
With that said, you are correct in adding the free energy from phononDOS to the total electronic energy to get the final free energy - but this will not include the rotational and translational contributions as you computed only vibrations for the molecule. Please use the 'IdealGasThermoChemistry' analysis object instead to get the Gibbs free energy.
I hope this answered your question.
For molecules, we have a new analysis object 'IdealGasThermochemistry' that takes a molecule configuration and its dynamical matrix object, similar to PhononDensityOfStates, to provide the Gibbs free energy including the electronic energy https://docs.quantumatk.com/manual/Types/IdealGasThermoChemistry/IdealGasThermoChemistry.html
Example:
In GUI: open scripter, under 'configurations' select 'Analysis from file'. Load the hdf5 file with the molecule configuration and dynamical matrix and select them. Add the 'IdealGasThermoChemistry' analysis object to the scripter and submit the script. After the calculation, choose the 'IdealGasThermoChemistry' object from the hdf5 file and choose text representation to get the list of Gibbs free energy at different temperatures.
PhononDensityOfStates is more relevant for bulk/slab systems as we do not have to compute the rotational and translational free energy contributions. For that, we have the 'CrystalThermoChemistry' analysis object similar to the case of the molecules above.
With that said, you are correct in adding the free energy from phononDOS to the total electronic energy to get the final free energy - but this will not include the rotational and translational contributions as you computed only vibrations for the molecule. Please use the 'IdealGasThermoChemistry' analysis object instead to get the Gibbs free energy.
I hope this answered your question.
10
General Questions and Answers / Re: possible bug?
« on: September 13, 2021, 10:38 »
Thanks for letting us know. Yes, it is a bug and we are updating the TotalEnergy Analysis object to work for CP correction calculators as well.
Pages: [1]