fluctuations in DOS calculations

[Arka Prava Sarkar]

Dear Users,

I am trying to calculate the DOS and PDOS of a Pt(111)-Ionic Liquid interfacial system. Upon calculating, the DOS and PDOS obtained shows a large fluctuation of the Pt(111) system. I am unable to figure out the reason behind this and how to resolve the issue. Any help will be appreciated. I am attaching the DOS and PDOS plot along with the input file. Thanks in advance.

--

Thanks and Regards,

অর্ক প্রভ সরকার

 Arka Prava Sarkar

বৃত্তিজীবী বাচস্পত্যাধয়নকারী

Research Scholar (Ph.D.), 

গণনা এবং তথ্য বিজ্ঞান কেন্দ্র 

Centre for Computational and Data Sciences (CCDS)

ভারতীয় প্রযুক্তি প্রতিষ্ঠান, খড়গপুর

Indian Institute of Technology, Kharagpur

পশ্চিমবঙ্গ - ৭২১৩০২ 

West Bengal-721302

email1 : sarkar.ar...@gmail.com

email2 : arkap...@iitkgp.ac.in

[Muhammad Saleh]

Hi Arka

In order to evaluate the PDOS, I believe you should also put PDOS on the print section 
...
      &PRINT
       &PDOS
        &EACH
           MD  1
         &END EACH
       &END PDOS

.....

you can find the details on the cp2k excercise https://www.cp2k.org/exercises:2016_uzh_cmest:calculating_pdos

best
MuS

--
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/CAMedUiqoxb3osZEB2iUDAe8SD31TpQ7R3oqmfLkQTk-SR2C_ZA%40mail.gmail.com.

[Arka Prava Sarkar]

Hi Muhammed,
I have calculated PDOS in a separate input file, using the same flags you have mentioned. My issue is about the fluctuations. Thanks.

Thanks & Regards,

অর্ক প্রভ সরকার
Arka Prava Sarkar

বৃত্তিজীবী বাচম্পত্যাধয়নকারী

Research Scholar (Ph.D.),
গণনা এবং তথ্য বিজ্ঞান কেন্দ্র

Centre for Computational and Data Science (CCDS)

ভারতীয় প্রযুক্তি প্রতিষ্ঠান, খড়গপুর
Indian Institute of Technology, Kharagpur

পশ্চিমবঙ্গ-721302
West Bengal -721302
India
Email1 : sarkar.ar...@gmail.com
Email2 : arkap...@iitkgp.ac.in

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAPD_w6w4KwkNDnBF%2B8rq0MqiBU_0UdozY%2BCn-Q1b60PrgRq6ew%40mail.gmail.com.

[Muhammad Saleh]

Hi Arka 

Ok, then you need to smooth it out. You can use the script from here: https://wiki.wpi.edu/deskinsgroup/Density_of_States. 
or if you go down on the exercise link I previously sent, then you may find a Python script for the smoothing. Hope it help

Best
MuS

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAMedUip3GBqz0C4BX3X1FmJPUZDL9w%2BMZqnA_HcAdNM4CnLs9Q%40mail.gmail.com.

[Arka Prava Sarkar]

Hi Muhammed,
Sorry for the late response. I will try this out, and I will let you know if I face any further issues regarding this. Thanks.

Thanks & Regards,
অর্ক প্রভ সরকার
Arka Prava Sarkar
বৃত্তিজীবী বাচম্পত্যাধয়নকারী
Research Scholar (Ph.D.),
গণনা এবং তথ্য বিজ্ঞান কেন্দ্র
Centre for Computational and Data Science (CCDS)
ভারতীয় প্রযুক্তি প্রতিষ্ঠান, খড়গপুর
Indian Institute of Technology, Kharagpur
পশ্চিমবঙ্গ-721302
West Bengal -721302
India
Email1 : sarkar.ar...@gmail.com
Email2 : arkap...@iitkgp.ac.in

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAPD_w6w8A-vGLy52r9T7nJSditk0%3DsuTzfGEoXLeD9gyZbiBcw%40mail.gmail.com.

[Arka Prava Sarkar]

Hi Muahmmed,

I tried to plot the DOS and PDOS with the python codes, however, I can't figure out how to plot the individual PDOS, for example, the d orbital of the Pt(111) surface using the code. Any suggestions? Thanks

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAPD_w6w8A-vGLy52r9T7nJSditk0%3DsuTzfGEoXLeD9gyZbiBcw%40mail.gmail.com.

[Muhammad Saleh]

Hi Arka,

The output from CP2K if you use the PDOS flag is several PDOS files that respect the element of your system, which is also applied when you smooth it. To identify which element is which, simply plot the data based on the specified column. For instance, if one of your elements had s, p, and d orbitals, then the output would be 

<Energy>  <s_orbital>  <p_orbital> <d_orbital>
...                ...                 ...                ... 

If you check your PDOS files, you can see the details of the orbital of each element in your system.

Best
MuS

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAMedUio-wun2xHOnyJ6wSq40hrSAcdNpbHpZHxzuiyGg9biHEw%40mail.gmail.com.

[Arka Prava Sarkar]

Hi Muhammed,

I am attaching one of the PDOS file that I am using here. The file has density columns as s, px, py, pz, d-2, d-1, etc. Now, suppose I want to plot the d orbital only for this file. d orbital will have the columns d-2 to d+2. I am trying to implement the code developed by Tiziano (link: https://raw.githubusercontent.com/dev-zero/cp2k-tools/master/scripts/cp2k_pdos.py). In this code, I need to make changes in the "DENSITY_COLUMN". However, it is taking only one value. For multiple columns, how to change the code? Thanks.

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAPD_w6yBN-nkbgcwzmD-1pOYpoxRDE3QrEQV4QgaKYCedrs7BQ%40mail.gmail.com.

[Arup Sarkar]

Hi Arka,

You can use Xmgrace software to plot the PDOS directly.

Dr. Arup Sarkar

Postdoctoral Researcher 

Max Planck Institute for Polymer Research

Theory (Andrienko Group)

Ackermannweg 10

Mainz 55128, Germany

Webpage: https://arupsarkar91.github.io/

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAMedUiose5-yFEaNp6tCiuzDB_idti_L%2BfendTpe%2BGgnkhbybg%40mail.gmail.com.

[Arka Prava Sarkar]

Hi Arup,

I am stuck on the part of plotting the d orbitals actually, since it has multiple columns. Also, Xmgrace will not be using convoluted gaussians.

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAEYoAgX8%3DX1Cc8s4sc2_2Pq9ohLhXQcgoGciAzNq30Qq8E8Y9Q%40mail.gmail.com.

[Muhammad Saleh]

Hi Arka 

You dont need to change the code at all. You simply run the script, let say

python cp2k_pdos.py Pyrr-Ace-O1-PDOS-after-k1-1.pdos -o out-3.dat

then just plot your interest orbital using xmgrace, gnuplot, or whatever you like. Here I attach the output from your PDOS data.

Best
MuS

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAMedUiose5-yFEaNp6tCiuzDB_idti_L%2BfendTpe%2BGgnkhbybg%40mail.gmail.com.

[Arka Prava Sarkar]

Hi Muhammed,

Thanks for the output file reference. I understood the output generated. However, if I want to plot the d orbitals, I need to consider all the columns of the d orbitals. Is there any way to combine all these together? Or do I need to do it in xmgrace? Thanks.

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAPD_w6zeyKqM07cZc3zj5EAPfOB%3DYxV9XmdxKC19CNC7P7AJPA%40mail.gmail.com.

[Muhammad Saleh]

Hi Arka 

If you plot the d orbital from the output, it will be like the figure I attach, collumn 6 to 10, you simply plot it 

Best
MuS

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAMedUirQhXYLf%2Byxif_zADm%2BT4iUYLLiN%3DEbVcmdHAY_rNf-dA%40mail.gmail.com.

[Arka Prava Sarkar]

Hi Muhammed,

Thank you for the reference. So I guess, there is no way I can combine all the columns and plot as one single d orbital. I need to plot it separately. Thank you.

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAPD_w6z%2BCH%2BDrN5_8Y-JiSiW0eN0OdZ_VbxhgqvdfsPY1E5MbQ%40mail.gmail.com.

[Muhammad Saleh]

Hi Arka, 

If you want to produce the total of d-orbital, then you just have assigned it from cp2k, the Python script only does the smoothing and unit transformation. The output you have specifies the PDOS corresponding to the magnetic quantum number of the d-orbital. For instance, d0 represents PDOS along the z-axis, while d+1 represents PDOS along the xz-plane, etc. 

Best
MuS

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAMedUirxTFTCh7e3WaJuPWAs_RpegUjBPxeqDm4cdt0heUAKTg%40mail.gmail.com.

[Arka Prava Sarkar]

Hi Muhammad,

in the below section of cp2k, what should I include extra so that it produces the total of d-orbital?

      &PRINT
       &PDOS
        &EACH
           MD  1
         &END EACH
       &END PDOS

This section, in this form, produces an individual magnetic number PDOS of each orbitals, as you mentioned. Thanks.

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAPD_w6x_7FVFWOee6tFNdk5bSfkFU66uFMCmaNj_2ufwqawHDQ%40mail.gmail.com.

[Muhammad Saleh]

Hi Arka

I think that should be enough, if you somehow wanted to have all the PDOS then use 

            # print all projected DOS available:
            NLUMO -1
            # split the density by quantum number:
            COMPONENTS

anyway, you can also combine all of the d orbital by evaluate its average, which I believe will produce the "total" of the graph  you wanted. For this you can use python + matplotlib, or gawk + xmgrace.

Best
MuS

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAMedUirvZu4Ow-KtdR%2BhsR%3DCvOsebd74ZouWX6VhHVYh%2BkA2XQ%40mail.gmail.com.

[Arka Prava Sarkar]

Hi Muhammad,

Thanks for all the help. I think I can now work on this. Extremely grateful for your kind and patient responses to all my queries. Thank you so much once again.

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAPD_w6yZw5GxEs1eJvH1VtWTTAR6kc%2B6-O-0xDmq%2B9onvd%3Ds%3Dw%40mail.gmail.com.

[Muhammad Saleh]

Hi Arka 

No problem, good luck

Best
MuS

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAMedUio3EqibB2A8vua58gOAfmtrvpSpKQxAS4HmiQpGA_pgkg%40mail.gmail.com.