DOS and bandstructure of Si-cube

[Andreas Döll]

Dear cp2k community,

I already wrote a question to the google group last year about DOS of the cp2k tutorial, see post at https://groups.google.com/g/cp2k/c/XIoLf8ww-yE/m/RqPrA-2PBAAJ and while I did not get the example DOS, I have an important update. Because I could not find any real solution to my problem in this group, I want to repost with updated information to divert attention to this matter. 

Maybe this post can address issues with the tutorial ( https://www.cp2k.org/exercises:common:pdos) and clear up some confusion. To me, it seems that the tutorial is highly outdated. For example, the link to the convolution website is not working anymore.

I wanted to test DOS and bandstructure in a simpler system Si-only crystal; where you can find the DOS and bandstructure online (for example here: https://next-gen.materialsproject.org/materials/mp-149#properties). Luckily there was a nice tutorial written by Joseph Wilson (https://jollywatt.github.io) to calculate Si bandstructure in cp2k. I could reproduce their results (my bandstructure is much lower at 0.13eV, even using the same input script: Weird!) but at least the bandstructure looks the same. Nethertheless, I highly recommend their tutorial for anyone who wants to get into this topic.

I then wanted to take a step back and revisit the Pdos of Si: I used the provided script by Tiziano Müller to obtain the convoluted dos but, to my surprise, the pdos looks bad. DOS doesnt even go to zero at E=0. I something like this expected? Are my parameters too loose? Is Si complicated to get correctly?

I think that Im seriously missunderstanding something when it comes to calculating DOS in cp2k. I attach my script for Si_bandstructure and Si_pdos. Any help is highly appreciated!

If You need any additional file for debugging, please let me know. This has been bugging me for months now.

I come to you with two explicit questions:

1) Why is my DOS so bad looking, even for Si-system?

2) How to improve the bandstructure gap? 

Sincerely, 

Andreas Doell