# inconsistent behavior between band structure and density of states

I'm a beginner in DFT calculation. I have performed a DFT calculation on CuCoSnSe using PBE-GGA on Quantum Esspresso. When I plot the band structure and the total DOS, the band structure shows a behavior of the semiconductor with a direct bandgap, but the total DOS does not show the same behavior.

Can someone give me an explanation of what happened or suggest some modifications to get reasonable results?

• Welcome to this community !!!
– Jack
Jan 9, 2021 at 2:04
• What is your smearing value for your DOS plot? If it's large enough you will just smear the band gap away :) Jan 10, 2021 at 3:30

When I plot the band structure and the total DOS, the band structure shows a behavior of the semiconductor with a direct bandgap, but the total DOS does not show the same behavior.

• The red arrow in your band structure is the direct gap at $$M$$ point, rather than the gap defined as the difference between conduction band minimum (CBM) and valence band maximum (VBM). (The VBM is located between $$A$$-$$Z$$.) The direct bandgap is related to the position of your $$\vec{k}$$-pint, but your DOS can only reflect the information about the global bandgap, in which the information of $$\vec{k}$$-points is integrated. You can read this post to learn what you can obtain from band structure and DOS.

• The previous point is good to analyze the band structure and DOS of semiconductors. However, here your DOS and band structure manifest metallic features. There are two things that you should check:

• The DOS displays ferromagnetic properties, which means you should plot the band structure with two spin-channels.

• Whether the Fermi level in your DOS plot is consistent with your band structure plot or not?

• Does his band structure not show a band gap while DOS does not? Even the indirect band gap would be around 0.25 eV, even if this is just one spin plotted there should still be a gap in the DOS. Jan 9, 2021 at 4:13
• @TristanMaxson In general, we are interested in the gap around the Fermi level. In the DOS plot, you can't see any gap around the Fermi level.
– Jack
Jan 9, 2021 at 7:07

I personally suspect that you may have converged to two different magnetic states. You should verify the final magnetic moments on each ion and confirm they are the same. Did you restart the entire calculation or did you do a non-self consistent run of both?