Estimate Hall carrrier concentration from transport calculations

Queries about input and output files, running specific calculations, etc.


Moderators: Moderator, Global Moderator

Post Reply
Message
Author
namsusi
Newbie
Newbie
Posts: 4
Joined: Mon Oct 07, 2024 2:11 am

Estimate Hall carrrier concentration from transport calculations

#1 Post by namsusi » Wed Nov 18, 2026 1:19 am

Hello,
I would like to know if there is a way to estimate Hall carrier concentration from transport properties calculations?
In a normal calculation, we often use carrier concentration n (hole or electron) to determine the transport parameters, which are used as input parameters to control the chemical potentials. However, the Hall carrier concentration (n_H) from experiments and the calculated carrier concentration (n) we use are quite different by scattering factors.
So, I would like to know if we can extract this n_H value in reverse from the calculations with n. In addition, could you let me know how to extract the value of τ_nk from vaspout.h5, if possible?
I appreciate your support very much.


manuel_engel1
Global Moderator
Global Moderator
Posts: 274
Joined: Mon May 08, 2023 4:08 pm

Re: Estimate Hall carrrier concentration from transport calculations

#2 Post by manuel_engel1 » Wed Nov 18, 2026 9:38 am

Hello,

Thanks for posting on the VASP forum.

Hall carrier concentration

This is currently not possible with VASP alone. VASP does not compute the Hall conductivity or the Hall factor r_H, so n_H = n / r_H is not part of the transport output. The carrier density you set and VASP reports is the actual carrier density n. To compare with Hall measurements, you can either assume r_H ≈ 1 or estimate r_H yourself in post-processing from the relaxation times below.

Since our electron-phonon code is still growing, we might consider adding this observable in a future update.

Relaxation time τ_nk

τ_nk is not written directly, but the electron-phonon self-energy is, and

τ_nk = ħ / (2 |Im Σ_nk|)

This is the same relation VASP uses internally for the transport coefficients.

The easiest way is py4vasp (requires VASP 6.5.2 or newer):

Code: Select all

import py4vasp

hbar = 6.582119569e-16  # eV s

calc = py4vasp.Calculation.from_path(".")
self_energy = calc.electron_phonon.self_energy[0]  # first set of settings
fan = self_energy.fan()

spin, kpoint = 0, 0           # counting from 0
band = fan.valid_bands[0]     # counting from 1 as in VASP
im_sigma = abs(fan[spin, kpoint, band].imag)  # eV, shape (energy, temperature)
tau = hbar / (2 * im_sigma)                   # seconds
print(tau)

print(self_energy) shows the settings of that set, and calc.electron_phonon.self_energy.selections() lists all available ones.

If you prefer to read vaspout.h5 directly:

Code: Select all

import h5py
import numpy as np

hbar = 6.582119569e-16  # eV s

with h5py.File("vaspout.h5", "r") as f:
    g = f["results/electron_phonon/electrons/self_energy_1"]
    fan = g["selfen_fan"][:]          # (state, energy, temperature, re/im) in eV
    index = g["bks_idx"][:] - 1       # (spin, kpoint, band) -> state, -1 = not computed
    band_start = g["band_start"][()]  # first band of the band axis (counting from 1)
    temps = g["temps"][:]             # K

im_sigma = np.abs(fan[:, 0, :, 1])
tau = hbar / (2 * im_sigma)           # seconds, shape (state, temperature)

# example: first computed band, first k point, first spin
print(tau[index[0, 0, 0]])

Some remarks:
- There is one self-energy for each combination of settings (carrier density, scattering approximation, ...): self_energy[0], self_energy[1], ... in py4vasp and self_energy_1, self_energy_2, ... in vaspout.h5.
- τ_nk is only available for the states inside the transport energy window. py4vasp raises an error for the others; in vaspout.h5 they have index -1.
- The relaxation time averaged around the chemical potential is written as tau_average (in seconds) in results/electron_phonon/electrons/transport_N.

Let me know if that answers your questions.

Manuel Engel
VASP developer


Post Reply