Small BSE calculated dielectric maxima

[nicholas_dimakis1]

Hello

I am using BSE after a G0W0 run to calculate optical properties. However, the BSE calculated dielectric maxima for both real and imaginary parts are significantly smaller than what are expected from DFT, RPA, and G0W0.



For G0W0 the gas-run file reads as

<dielectricfunction comment="HEAD OF MICROSCOPIC DIELECTRIC TENSOR (INDEPENDENT PARTICLE)">
<imag>
<array>
<dimension dim="1">gridpoints</dimension>
<field>energy</field>
<field>xx</field>
<field>yy</field>
<field>zz</field>
<field>xy</field>
<field>yz</field>
<field>zx</field>
<set>
<r> 0.0000 0.0000 0.0000 0.0000 -0.0000 0.0000 0.0000 </r>
<r> 0.8393 0.0015 0.0015 0.0015 -0.0000 0.0000 -0.0000 </r>
<r> 1.6779 0.0032 0.0033 0.0032 -0.0000 0.0000 -0.0000 </r>
<r> 2.5150 0.0055 0.0056 0.0054 -0.0000 0.0000 -0.0000 </r>
<r> 3.3499 0.0091 0.0093 0.0091 -0.0000 0.0000 -0.0000 </r>
<r> 4.1820 0.0185 0.0190 0.0270 -0.0000 0.0000 0.0001 </r>
<r> 5.0106 0.1280 0.1406 0.1638 -0.0000 0.0000 0.0022 </r>
<r> 5.8352 1.1394 1.2418 0.8218 -0.0000 0.0000 -0.0177 </r>
<r> 6.6551 5.0338 5.1649 3.8020 -0.0000 0.0000 0.0260 </r>
<r> 7.4699 6.1322 5.8544 5.6597 -0.0000 0.0000 -0.0248 </r>
<r> 8.2792 7.7896 7.8558 9.5690 -0.0000 -0.0000 -0.0323 </r>
<r> 9.0826 8.9027 8.9238 11.3536 0.0000 -0.0000 0.0536 </r>
<r> 9.8799 7.5521 7.6093 6.8165 -0.0000 0.0000 0.0092 </r>
<r> 10.6709 5.3003 5.2209 5.9151 -0.0000 -0.0000 -0.0270 </r>

whereas the from BSE


<r> 7.5676 1.5524 1.4509 0.7260 -0.0000 -0.0000 -0.0826 </r>
<r> 7.6076 1.6837 1.5338 0.7653 -0.0000 -0.0000 -0.0624 </r>
<r> 7.6476 1.8513 1.6187 0.8146 -0.0000 -0.0000 -0.0762 </r>
<r> 7.6877 2.0014 1.6520 0.8466 -0.0000 -0.0000 -0.1037 </r>
<r> 7.7277 2.0286 1.6646 0.8376 -0.0000 -0.0000 -0.1109 </r>
<r> 7.7678 1.9239 1.7172 0.8238 -0.0000 -0.0000 -0.1037 </r>
<r> 7.8078 1.8056 1.8147 0.8409 -0.0000 -0.0000 -0.0845 </r>
<r> 7.8478 1.7050 1.8764 0.8895 -0.0000 -0.0000 -0.0448 </r>
<r> 7.8879 1.6229 1.8503 0.9422 -0.0000 -0.0000 0.0071 </r>
<r> 7.9279 1.5639 1.7933 0.9579 -0.0000 -0.0000 0.0534 </r>
<r> 7.9680 1.5537 1.7700 0.9369 -0.0000 -0.0000 0.0863 </r>
<r> 8.0080 1.6175 1.7967 0.9055 -0.0000 -0.0000 0.1070 </r>
<r> 8.0480 1.7098 1.8248 0.9011 -0.0000 -0.0000 0.1195 </r>
<r> 8.0881 1.7078 1.7923 0.9359 -0.0000 -0.0000 0.1264 </r>
<r> 8.1281 1.6034 1.7098 0.9950 -0.0000 -0.0000 0.1246 </r>
<r> 8.1682 1.5322 1.6382 1.0562 -0.0000 -0.0000 0.1101 </r>
<r> 8.2082 1.5606 1.6286 1.1000 -0.0000 -0.0000 0.0817 </r>
<r> 8.2482 1.6484 1.6863 1.1278 -0.0000 -0.0000 0.0301 </r>
<r> 8.2883 1.7394 1.7626 1.1596 -0.0000 -0.0000 -0.0387 </r>

I use a 6x6x6 KPOINTS grid centered at the gamma. Changing

NBANDSO = 4
NBANDSV = 12

to higher values does affect the max dielectric function values but even at NBANDSO = 30 = NBANDSV, thes values still remain low.

Thanks-Nick

[alexey.tal]

Hi,

I see that you set ANTIRES=1 in your input. This approximation is only valid at zero frequency (dielectric constant) and the rest of the spectrum is not correctly reproduced. If you would like to calculate the spectrum, you should use ANTIRES=0 for the TDA or ANTIRES=2 for the full BSE.