No output for density of states

GULP
1

Dear all,
When I’m using the phonon option, there is no phonon density of states in the output file. Is there any problem in Dos output?

Thanks very much.
Best regards

2

There shouldn’t be a problem with phonon DOS. Of course if your system isn’t periodic then you won’t get a phonon DOS and the quality will depend on how many k points you specified? If you can post your input then it would easier to suggest what the issue is.
Regards
Julian

3

Thanks for your reply, Julian. And I use the GULP-6.0 and run in parallel, which may be concerned with the problem I met.
Here’s my GULP input:
optimise conp compare phonon

title
ribz5
1 cell
10 atom
end

pcell 2.4595

cartesian
C 8.770243926626 12.130000000000 10.000000000000
C 8.770243926626 16.390000000000 10.000000000000
C 10.000000000000 10.000000000000 10.000000000000
C 10.000000000000 14.260000000000 10.000000000000
C 10.000000000000 18.520000000000 10.000000000000
C 8.770243926626 10.710000000000 10.000000000000
C 8.770243926626 14.970000000000 10.000000000000
C 8.770243926626 19.230000000000 10.000000000000
C 10.000000000000 12.840000000000 10.000000000000
C 10.000000000000 17.100000000000 10.000000000000

brenner

dispersion 1 50
0.0 to 0.5

output phon phonon
output xyz coord

4

If you add “shrink” to specify a grid of k points for sampling the DOS then you should generate the DOS as expected. It’s best to run separately from the dispersion since in general the best set of k points for DOS and dispersion are not the same.
Regards,

Julian

5

Thanks very much for your reply！ I got it.

6

Another question about phonon dos. Is the output of phonon dos normalized? I’m confused about this.

7

The phonon DOS isn’t normalised since there is no reason that the integral should be equal to 1. The integral of the DOS over the full range should be equal to the number of states.