Gas Sensing with Molecule Adsorption using TRANSIESTA
ฝัง
- เผยแพร่เมื่อ 1 ต.ค. 2024
- Non-equilibrium Green’s Function Calculations with TranSIESTA: Gas Sensing Project
Covered all the aspects of transport calculations starting from Unit cell to Scattering region and adsorption of the molecule and their transport calculation.
Github account: github.com/pri...
Hello sir
Thanks for your really amazing tutorial
I have a problem and can't find ".ion" files in the internet and the files you uploaded in the github are not for all atoms. I want In and Se atoms files. How can I download them?
Thank you! Yes, I will upload them soon!
Hi Sir, I want to perform the transport calculation of my gas sensor structure, optimised in VASP. As Transiesta uses the LCAO basis set for transport calculation and VASP uses the planewave basis set, is it correct to directly take the coordinates of the optimised gas sensor structure through VASP and perform the electrode and scattering calculations in Transisesta? If not, then how do we proceed? if yes then which example template I should use and how the mapping of planewave is done on LCAO basis set through the tool?
I would recommend to use Siesta for the whole calculations including optimization and then followed by transport. In that way you can use the same psuedopotentials.
Hi, I have a problem in the last step. Everything is OK and output files do not have any error but there is no current-voltage value in the tbt.out file. So I just have transmision file. Would you please help me to solve the problem and get current values too. Transiesta version is also v4.1-b4 and I use exactly the files you uploaded in the github page. Thanks
If you rant at voltage 0V, then current voltage value would be 0. If you ran successfully, then you should get the values at the end of tbt.out file.
Hi sir.
Thanks for your great tutorial.
I have a question. why in gas sensing part of your tutorial you used GGA instead of VDW?
Welcome! Its just for the sake of tutorial. You can use VdW potentials as well. I always recommend when you deal with sensing mechanisms use VdW especially for molecules.
thanks for perfect tutorial.. how can you set the height of NH3 molecules from layer to molecule? .or your tutorial is randomly ?
The height of the molecule is based on structural optimisation that you have obtained from normal Siesta run or Vasp optimization. You can try different orientation as well. It’s better to check for the lowest energies through NEB calculations and then set it for transport calculations. I will try to make a video related to NEB soon! Thanks
Sir, the TSHS files of electrodes using for scattering region calculation should include the electrode layers used in relaxation calculation (siesta) or not.
Yes, include them.
Please, from were file we can get the current value ? or the script will generate the IV.dat ?
At 39:00, thank you for pushing the *TS.Analyze* output, this is really important for high-throughput calculations! :)
Note that the percentages are percentages from the full matrix size. So 43% means it used 43% of the elements in the full matrix.
Thank you! Yes its very important and it reduces the computational cost as well.
Interesting video,
Once the transport setup is done, can we optimize it or directly calculate the transport calculation?
You have directly calculate the transport properties such as current vs voltage, transmission etc.
Thank you so much.
After optimization of the transport setup, the separate electrode coordinates can not overlap with the device electrode coordinates then it can be corrected by quantum wise ATK. Could you suggest me regarding this, i got this comment from someone else. Thanks in advance.
At 5:40 you mention that the unit-cell needs to be rectangular. In 4.1.5 this is not needed, you can do transport in a hexagonal unit-cell. In this case because you absorb the molecule *on top*.
Thank you Prof. Nick Papior. Much appreciated. I will make another tutorial mentioning the changes.
Is it possible to simulate dye sensitized solar cells ( dssc) which consists of 2 electrodes ,a semiconductor, dye molecules and electrolyte
Sir, when we have different electrode materials (not one in the scattering region), then how to select the separation between electrode and scattering region. Should, we do any optimization? Secondly, a kind of similar query that what should be the separation between layer and adsorbant? Thanks in advance.
Thank you. I will update you in my next video.
At 37:00 you mention that one should not use "used-atoms" keyword. This flag is used if you wish to only include a subset of electrode atoms in the device region. If you use *all* electrode atoms from the electrode calculation as electrode in the device region, then I do NOT recommend using the flag for consistency. If you will reduce the number of atoms used, then it is required, but it requires some careful thoughts and consistency in the electrode-setup. :)
Thank you very much Professor! I will keep in mind and refer it to my next video!
Thank you for the intersting videos, ihave qst , concerning if e have an heterostrucutre material how can we do the calculations , and please how can we plot the current versus voltage and other properties ?
Welcome! Please refer to my other videos! I will soon upload a video regarding nano electronics! Stay tuned!
Also, please do NOT use 4.1-b4. The 4.1.5 is out and has bug-fixes etc.
Absolutely! I have started using that and soon I will update the tutorial.
Please upgrade the Siesta version from 4.1.b4 to 4.1.5 which has bug fixes.
1) Can you share links for .ion files for other elements?
2) You did the VASP relaxation of NH3 adsorbed scattering region and pasted its relaxed coordinates in the the first (0.0V) or scattering.fdf will take care of NH3 adsorbed scattering region's relaxation by itself and we don't need to do it in VASP beforehand?
Thank you for the videos.
The ion files are in my GitHub account. Yes, you can do the relaxation of NH3 or any gas molecule and the material you use either by VASP or Siesta. Then take the coordinates from the relaxed system. Rest the Transiesta will take care of the subsequent process.
@@BioinfoCopilot Thanks for clarifying the relaxation point. Also, I am only able to get the ion files of only the elements you used in the tutorial. Where can I get ion files for other elements for example I am trying to simulate MXenes by using your tutorials. Thank you.
@@jatinkashyap1491 You have to download the PSML files or PSF files from the Psuedodojo website and Siesta Psuedopotentials website. Watch my other video related to PSML pseudopotentials and pseudopotential generation.
Great talk! I have a few comments:
1- you talk about xyz vs poscar files. In general for siesta you dont worry about that in the fdf file. You would be either interested in xyz or fractional coordinates.
2- 20 k pts is not really a realistic number. This number depends on the k point convergence of the system.
3- transport can take place in different types of unit cell not only rectangular
I am actually working on a gas sensor project as we speak so this is very fruitful.
Good luck
Thank you very much! I will update all the points in my next video
@@BioinfoCopilot No problem at all. If there is anything I can help with be free to let me know!
Hi,
Did anyone know where (or can share) a binary for siesta 4.1.5 (or 4.1.b4)?
33:25 Why does the scattering calculation have the same K-points mesh with the electrode calculation? In the past, th-cam.com/video/mcDy0KobTZw/w-d-xo.html, we have 1 instead 100 in this video. Since we have no periodic region in device along the transport direction, don't we set the single mesh along with the transport?
Yea you are right! Even if you set 100, when you open the out file you will see 1 KP along the transport direction.
The most important KP should be in X direction. It has to be same as electrode calculation.
Thank you very much for this thorough tutorial. I have a lingering doubt which I haven't been able to solve for the past few days. What about the systems where we have to compute the transport properties of small molecules, let's say methane connected to the left and right (gold) electrodes? How will the left and right electrode's positions and lattice vectors be decided in that case?
Glad that it helped. I am preparing a system with the exact setup that you have mentioned. Soon, I will update a tutorial on that.
Thank you and keep sharing my contents.
@@BioinfoCopilot Thanks a lot for the reply. eagerly waiting for that tutorial. Keep doing this excellent job. Being from Molecular dynamics background, I always have a hard time understanding the periodicity and lattice vectors in such DFT calculations. Hoping to get an understanding through your wonderful videos. Thanks.
@@abhishekaggarwal1369 Thank you. I know the feeling as I am a biologist/bioinformatician. I delved into this field only to bridge the gap between bio with physics (biophysics). I also do a lot of calculations in Gromacs but I never had a chance to understand the physics behind it. So, being from a bioinformatics background, the only thing that helped me to understand physics is computational materials science as it involves scripting.
Thanks for the updates. I was too much curious about the same issues. Best wishes 👍
Thank you. Please share and I hope I get more subscribers from your network.
@@BioinfoCopilot Yes, sure.
Thanks for the informations.
Any time!
Thank you very much for uploading such an important video.
But, Sir, I have one doubt regarding the gas molecule adsorption. Is it required to perform the SIESTA run (relaxation) first for the molecule adsorbed structure (left electrode + scattering region + right electrode) before proceeding to the TRANSIESTA run? Because, if we do the TRANSIESTA run for the scattering region immediately after placing the molecule at a certain height over the surface, then the effect of molecule-induced changes would not be included in the study.
Now, if we perform the relaxation of the scattering region (molecule adsorbed), then the coordinates of the left and right electrodes might not match perfectly.
I seek your help to clear my doubt. Please tell us the step-by-step approach in this case. Best Wishes.
Glad that it helped.
The most important thing is to relax the unit cell structures of all the systems that you are taking.
Do perform unit cell relaxation for the system/sheet you are taking using normal Siesta.
Do perform molecule optimization using normal Siesta.
Place the molecule on top of sheet and do relaxation to calculate binding energy.
Transiesta
Take the relaxed unit cell and construct electrode. Perform normal siesta relaxation with transiesta to get .TSHS file.
Construct the scattering region and perform transiesta to get volateg bias values. If semiconductor go for 0-3V . If it's metal; go for 0-1V.
Place the relaxed molecule on top of scattering region and perform the transiesta calculations. Remember in transiesta TSDE file saves the transiesta density matrix. TranSIESTA does not start with a SIESTA self consistent cycle, it goes directly to a TS cycle, taking the density matrix stored in the .TSDE file.
So the scattering region that you have calculated at 0V, you can take the TSDE file and run the calculation for the same.
Hope it helps.
@@BioinfoCopilot thanks for your explanation. It really helps me a lot. But, I have one more question regarding the consideration of buffer layers. Do I need to include a buffer spacing between the electrode and scattering regions to make the electrode layer insensitive to the structural distortions from the scattering layer? Because in any adsorption case, it is common that the scattering region gets distorted upon relaxation. If yes, then, is there any rule regarding the size of the buffer zone? I highly seek your positive response. Thank you once again.
thankyou sir, lecture is very helpful.
Thank you
Sir, why we convert hexagonal to rectangular? Is there any necessary condition for this?
For the purpose of adsorption of small molecules on 2d materials . Its just my perspective. Feel free to adjust your material accordingly. Try with hexagonal shape and let me know!
@@BioinfoCopilot sir, how can we calculate the thermoelectric properties like seebeck coefficient, thermal conductivity. electrical conductivity. Please make a video for this or provide me some references for this. Up to now , i have found out transmission function and with the help of this, how can we find out the above properties.Please suggest me. Thanks in advance Sir
Excellent sir!!
Thank you