Hi, all!
In order to include intraband transition into the simulation of dielectric functions of Au and Ag and see the effects of "Drude model", is this code a correct way to follow? Especially that what should be done to q points?
…
<xs xstype="TDDFT"
ngridk="8 8 8" vkloff="0.05 0.15 0.25"
nempty="30"
lmaxapwwf="5" lmaxemat="5"
gqmax="2.0"
broad="0.004"
tevout="true">
<energywindow intv="0.0 2.0" points="1500" />
<tddft fxctype="RPA"
intraband="true"/>
<qpointset>
<qpoint> 0.0 0.0 0.01 </qpoint>
</qpointset>
</xs>
…
Dear Jerry. Yes, this is the correct way to include intraband terms in the xs/tddft calculation. Take into account that this is not done by using the semiclassical expression for the Drude peak in terms of the plasma frequency, but by the direct inclusion of intraband terms in the calculation of the density response function in the TDDFT calcuation. At present this only works for finite q (as you use in the the example). For more details, take a look at an old tutorial for the lithium version of exciting here: q-dependent TDDFT. In that tutorial you can see an example result for Ag with and without intraband terms.
You can also make a calculation of the dielectric function using the semiclassical Drude expression, although presently not with the xs/tddft element but with the properties/dielmat element (check the input reference for details on how to use it). There you can directly set specific plasma frequency, and lifetime broadening.
Cheers
Santiago
(exciting team)
Thank you for the suggestion! I did a couple of calculations for dielectric function for Au and Ag these days, after volume optimization and k-grid optimization. The results I got are compared with dielectric functions from Johnson and Christy(1972 http:refractiveindex.info) . I found that the interband features of simulations are generally redshifted compared with experiments. First question is can this discrepancy be minimized in some way by adjusting parameters?
To include intraband features, I have used the q points as the following. The imaginary part is significantly larger than experiments. My second question is how should I chose q points for different metals to study the intraband transitions? Thanks you very much in advance!
<qpointset>
<qpoint> 0.0 0.0 0.001 </qpoint>
<qpoint> 0.0 0.001 0 </qpoint>
<qpoint> 0.001 0.0 0.0 </qpoint>
<qpoint> 0.001 0.001 0.001 </qpoint>
<qpoint> 0.0 0.0 0.002 </qpoint>
<qpoint> 0.0 0.002 0 </qpoint>
<qpoint> 0.002 0.0 0.0 </qpoint>
<qpoint> 0.002 0.002 0.002 </qpoint>
<qpoint> 0.0 0.0 0.005 </qpoint>
<qpoint> 0.0 0.005 0 </qpoint>
<qpoint> 0.005 0.0 0.0 </qpoint>
<qpoint> 0.005 0.005 0.005 </qpoint>
<qpoint> 0.0 0.0 0.01 </qpoint>
<qpoint> 0.0 0.01 0 </qpoint>
<qpoint> 0.01 0.0 0.0 </qpoint>
<qpoint> 0.01 0.01 0.01 </qpoint>
</qpointset>