Some questions about using RT-TDDFT method in cp2k code to calculate spectrum

[lin xiong]

Hi all,
    

    Recently I learned that cp2k code has two ways to calculate the absorption spectrum, namely real time propagation TDDFT and TDDFPT. I would like to ask some questions about using rt-TDDFT to calculate the spectra.

    1. First of all, I would like to ask if my input file is correct? If the input file is correct, then there are two additional questions: 2. In which file (as shown in the figure below) is the data used to draw the spectrum stored? 3. How to use these data to plot the spectrum? 

[Mario van Rooij]

Dear Lin,

I am not at all an expert on the issue, but I do have this example file online. Also once you obtain the dipole moment as a function of time, you can perform a Fourier transform of the data. I have attached a python script to transform the data. Make sure it works properly yourself and make sure you read about why this works :-). The output you want for the spectrum should be called something like ch2o_pbe_rtp-output-moments.dat. I tried this for quantum dots and it didn't work. Maybe different timestep and basis/ potential are needed. Also, I don't know if the script can handle these small numbers. 

Good luck,

Mario

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/75ac4790-b50d-4289-b5aa-a8385584f19ao%40googlegroups.com.

[xiongl...@gmail.com]

Dear  Mario,

     

Thank you very much for your help. In fact, I have just started using cp2k software. I have a question before you reply to me, that is why there is no *dat file in my output files. But now I understand that I need to add the 'FILENAME output' keyword in the 'MOMENTS' subsection.

Still, I have one question to ask you: What does the value of 'STEPS' in the 'MD' section of rt-TDDFT specifically mean? The interpretation of the official website is ambiguous. To calculate the spectra of different systems, do you have any suggestions for setting the STPES and TIMESTEP keywords?

Best wishes,

Lin

[Mario van Rooij]

Dear Lin,

https://manual.cp2k.org/trunk/CP2K_INPUT/MOTION/MD.html#TIMESTEP  Timestep section tells us:

Default value: 5.00000000E-001
	Default unit: [fs]

So timestep is in fs. As for suggested values, I don't know. I didn't manage. To get higher excitation energies you will need a smaller timestep, otherwise these are "invisible" in the signal. Steps means how many timesteps are calculated (so 100 steps with stepsize 1 would be 100 fs), more timesteps means less uncertainty in your Fourier transform, giving more distinct peaks (I think). To figure out what the right timestep and steps are, you can take a look at some literature and compare with experts that you get the same value. And then also play around with stepsize and steps to see if your spectrum changes a lot. The spectrum should be independent of stepsize if you are around the right region.  Mario

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/316ea557-701d-4fc6-8034-4057599edea0n%40googlegroups.com.

[xiongl...@gmail.com]

Dear Mario,

Once again, I am grateful for your help. You have helped me a lot!

May you be happy, healthy and happy! 

Lin