I carried out a simple linear optical response calculation for Si and Al. The real part of the dielectric function is nearing a value of 2. 1.85 for Al and 1.95 for Si. However the real part of Dielectic function at optical frequencies is approx the square of (real part of) refractive index. I think the calculations are off by a factor of 1. It is the (delta -function) term that is zero for the cross terms and unity for the direct terms.
Please correct me if I am wrong.
Ravi
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The new version of exciting computes separately the intraband and the interband contributions. Is the total dielectric response function a simple sum of these tow parts? epsilon = epsilon(interband) + epsilon(intraband) ?
Let me know. For if that is the case then on plotting the Loss Function (one obt. from expt. like EELS), there is a large underestimation of the plasma frequency so calculated. Also how do you envisage a collective oscillation (plasmon) to be well represented by single particle excitations? I understand that you use to Random phase approximation.
Thanks
Ravi
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Sangeeta Sharma has run Si with a 20x20x20 k-point set (unshifted) on version 0.9.93 and she obtains a value of Re(epsilon_11)=0.966956 at 3.0845 Hartree.
Cheers,
Kay
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Is Epsilon_11 the total dielectric response or is it the INTERAND contribution? I added the Epsilon_11 to Epsintra_11 to obtain the net epsilon: which gave me a value close to 2. My calculations do match Ms Sharma's value for re(Epsilon_11).
Regards
Ravi
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
How do we determine this frequency cut-off? What exactly is the range(in energy) that you'd call high frequency. The plasmon energies are typically in the range between 10-20 eV which is well into the optical frequency range if not higher!
What is the energy window around the bandgap for which this addition is permissible?
I also notice that the intraband contribution is calculated for materials with well defined band gaps. In such cases is the addition of intraband contribution only applicable for an energy-window equivalent to the band-gap energy?
Let me know
Warm regards
Ravi Shivaraman
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The Drude term is probably good to about 1 eV, despite the plasma frequency being much higher.
The intra- and interband contributions are supposed to be used separately in their appropriate energy windows. However, we may change this in future so that the imaginary parts are added together by default and KK inverted.
Best wishes,
Kay.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I carried out a simple linear optical response calculation for Si and Al. The real part of the dielectric function is nearing a value of 2. 1.85 for Al and 1.95 for Si. However the real part of Dielectic function at optical frequencies is approx the square of (real part of) refractive index. I think the calculations are off by a factor of 1. It is the (delta -function) term that is zero for the cross terms and unity for the direct terms.
Please correct me if I am wrong.
Ravi
The new version of exciting computes separately the intraband and the interband contributions. Is the total dielectric response function a simple sum of these tow parts? epsilon = epsilon(interband) + epsilon(intraband) ?
Let me know. For if that is the case then on plotting the Loss Function (one obt. from expt. like EELS), there is a large underestimation of the plasma frequency so calculated. Also how do you envisage a collective oscillation (plasmon) to be well represented by single particle excitations? I understand that you use to Random phase approximation.
Thanks
Ravi
Dear Ravi,
Sangeeta Sharma has run Si with a 20x20x20 k-point set (unshifted) on version 0.9.93 and she obtains a value of Re(epsilon_11)=0.966956 at 3.0845 Hartree.
Cheers,
Kay
Dear Kay,
Is Epsilon_11 the total dielectric response or is it the INTERAND contribution? I added the Epsilon_11 to Epsintra_11 to obtain the net epsilon: which gave me a value close to 2. My calculations do match Ms Sharma's value for re(Epsilon_11).
Regards
Ravi
Dear Ravi,
EPSILON_11 is just the interband contribution. For high frequencies you should not add the intraband contribution (EPSINTRA_11).
Regards,
Kay.
Hello Kay,
How do we determine this frequency cut-off? What exactly is the range(in energy) that you'd call high frequency. The plasmon energies are typically in the range between 10-20 eV which is well into the optical frequency range if not higher!
What is the energy window around the bandgap for which this addition is permissible?
I also notice that the intraband contribution is calculated for materials with well defined band gaps. In such cases is the addition of intraband contribution only applicable for an energy-window equivalent to the band-gap energy?
Let me know
Warm regards
Ravi Shivaraman
Dear Ravi,
The Drude term is probably good to about 1 eV, despite the plasma frequency being much higher.
The intra- and interband contributions are supposed to be used separately in their appropriate energy windows. However, we may change this in future so that the imaginary parts are added together by default and KK inverted.
Best wishes,
Kay.