Please visit

https://www.google.com/url?sa=t&source=web&rct=j&url=https://www.researchgate.net/file.PostFileLoader.html%3Fid%3D567d1ab17eddd3eb328b4574%26assetKey%3DAS%253A310495943299072%25401451039409313&ved=2ahUKEwiciauXnsvnAhVu4jgGHYCBB94QFjAAegQIAxAB&usg=AOvVaw1J9r3EgmLaU6lgRVQmtqUT

Dear Majibul,

To calculate the absorption spectra of a structure using Quantum Espresso, you have some options.

The simplest approximation is Independent Particle Approximation (IPA) which allows us to describe single-particle excitations. For this purpose, first, perform a self-consistent field ground-state calculation for your structure using the pw.x code. Next, perform a calculation of the absorption spectrum using the epsilon.x code.

***************************************

comands:

1- pw.x < scf.in > scf.out

2- epsilon.x < epsilon.in > epsilon.out

***************************************

The code epsilon.x produces four files:

• epsr.dat – Real part of the dielectric function
• epsi.dat – Imaginary part of the dielectric function
• eels.dat – Electron energy loss spectrum
• ieps.dat – Response computed on the imaginary axis of frequency

Finally, you can calculate the absorption spectrum of your structure with the Real and Imaginary parts of the dielectric function.

There is also another method. The turbo_davidson.x program allows us to calculate the absorption spectra of molecules using time-dependent density functional perturbation theory (TDDFPT). For more information visit the website below.

https://www.quantum-espresso.org/Doc/INPUT_Lanczos.html

Dear Majibul,

So far as I know, we can calculate the electron density, band structure, Fermi Energy, magnetic properties, phonon density etc can be calculated using Quantum Espresso.

I don't know if optical properties like refractive index, absorption, scattering etc can be calculated by Quantum Espresso.

Thanks and Regards

N Das

You neec to get your correct input file then run scf calculations followed by epsilon.x calculations. Make sure all calculations are within the same directory

Md. Majibul Haque Babu can you exactly explain me how to calculate absorption spectrum with the Real and Imaginary parts of the dielectric function.

Thanks in advance

Nice explanation Sir, Does the same method is applicable for the calculations of optical properties for metallic systems. Could you give any useful reference to get absorption coefficient from the real and imaginary part.

Thanks in advance

can we calculate reflectivity and refractive index using Quantum Espresso?

Shakeel --

Once imaginary part of dielectric function (ε_2(ω)) is found, corresponding real part (ε_1(ω)) can be found using Kramer-Kronig relation.

Then refractive index and reflectivity can be evaluated by employing following relations (see the attached figure).

For more details regarding optical calculations using QE, check

Article Mechanical characteristics and electric fieldinfluenced ther...

Article Importance of Hubbard U parameter to explore accurate electr...

Dear sir, I want to run the epsilon.x calculation to observe the optical properties of my sample. but when it run it shows an energy_grid error. Did you know how to fix that error ? Thank you.

Hi Chathuranga,

In your scf calculations, you should disable the usage of k => -k symmetry (time reversal) in your k-point generation and do not use symmetry. To do these, set nosym and noinv to .true.

If these don't work, you can reoptimize your system while setting nosym and noinv to .true....I hope this helps.

Dear users,

I want to calculate dielectric function, refractive index, conductivity, absorption, reflectivity and loss function. I read that epsilon.x dont work with Ultrasoft PPs (that I'm using in my calculations).

There are other functions incan use to calculate these properties? What can I do to get them?

To calculate the optical properties of a material with Quantum Espresso using the epsilon.x input, you need to follow these steps:

Here are the detailed computational details for the epsilon.x input:

• Input file: Create a new input file for the epsilon.x calculation. The input file should contain the following information:

&input prefix='PREFIX', outdir='.', filplot='PREFIX.eps', / &epsil outdir='./' prefix='PREFIX' lsda=.true. noinv=.true. / CELL_PARAMETERS a1 x y z a2 x y z a3 x y z ATOMIC_SPECIES SYMBOL mass pseudo_file ATOMIC_POSITIONS (crystal) SYMBOL x y z SYMBOL x y z ...

• Prefix: The prefix parameter specifies the prefix used for the file names generated by the calculation.
• Outdir: The outdir parameter specifies the output directory for the calculation.
• Filplot: The filplot parameter specifies the file name for the plot of the dielectric function.
• Lsda: The lsda parameter specifies whether to perform a spin-polarized calculation. Set it to .true. for spin-polarized calculation and .false. for non-spin-polarized calculation.
• Noinv: The noinv parameter specifies whether to include the local field effects. Set it to .true. to exclude the local field effects and .false. to include them.
• Cell parameters: The CELL_PARAMETERS section specifies the lattice vectors of the unit cell in angstroms.
• Atomic species: The ATOMIC_SPECIES section specifies the atomic species, their masses, and the pseudopotential files.
• Atomic positions: The ATOMIC_POSITIONS section specifies the atomic positions in crystal coordinates.

$ mpirun -np N epsilon.x < PREFIX.in > PREFIX.out

where N is the number of processors to use, PREFIX.in is the name of the input file, and PREFIX.out is the name of the output file.

I hope this helps! Let me know if you have any further questions.

Dear Sudhanshu Gupta

Thank you for your answer.. does epsilon.x work well with ultrasoft pps?

José Xavier Lima Neto Ultra-soft pseudopotentials do not work with epsilon.x code.