Ctrl+K

Spectrum broadening

Spectrum broadening#

In this section we will discuss the broadening of calculated spectra, and how they compare to the different broadening mechanisms of experimental measurements. As an example, we will consider the valence spectrum of water, calculated as:

import matplotlib.pyplot as plt
import numpy as np
import veloxchem as vlx

au2ev = 27.2113245

water_xyz = """3

O       0.0000000000     0.1178336003     0.0000000000
H      -0.7595754146    -0.4713344012     0.0000000000
H       0.7595754146    -0.4713344012     0.0000000000 
"""

# Prepare molecule and basis objects
molecule = vlx.Molecule.read_xyz_string(water_xyz)
basis = vlx.MolecularBasis.read(molecule, "6-31G")

You appear to be running in JupyterLab (or JavaScript failed to load for some other reason). You need to install the 3dmol extension:
jupyter labextension install jupyterlab_3dmol

# SCF settings and calculation
scf_drv = vlx.ScfRestrictedDriver()
scf_settings = {"conv_thresh": 1.0e-6}
method_settings = {"xcfun": "b3lyp"}
scf_drv.update_settings(scf_settings, method_settings)
scf_results = scf_drv.compute(molecule, basis)

# resolve four eigenstates
rpa_solver = vlx.lreigensolver.LinearResponseEigenSolver()
rpa_solver.update_settings({"nstates": 6}, method_settings)
rpa_results = rpa_solver.compute(molecule, basis, scf_results)

With excited state energies, oscillator strengths, and transition moments:

print("Energy [au]  Osc. str.   TM(x)     TM(y)     TM(z)")
for i in np.arange(len(rpa_results["eigenvalues"])):
    e, os, x, y, z = (
        rpa_results["eigenvalues"][i],
        rpa_results["oscillator_strengths"][i],
        rpa_results["electric_transition_dipoles"][i][0],
        rpa_results["electric_transition_dipoles"][i][1],
        rpa_results["electric_transition_dipoles"][i][2],
    )
    print("   {:.3f}     {:8.5f}  {:8.5f}  {:8.5f}  {:8.5f}".format(e, os, x, y, z))

Energy [au]  Osc. str.   TM(x)     TM(y)     TM(z)
   0.286      0.01128   0.00000  -0.00000   0.24321
   0.364      0.09652   0.00000  -0.63080  -0.00000
   0.364      0.00000  -0.00000  -0.00000   0.00000
   0.454      0.08684   0.53570   0.00000  -0.00000
   0.540      0.41655   1.07524  -0.00000   0.00000
   0.666      0.24377   0.00000   0.74073  -0.00000

Broadening in experiment#

The line shape of an experimental spectrum is affected by a number of factors, including:

These leads to a number of different broadening types, such as

  • Lorentzian (homogenous?)
    From life-time broadening

  • Gaussian (inhomogenous?)
    From Doppler broadening and resolution of measurement

  • Voigt
    A convolution of a Gaussian and a Lorentzian

  • Asymmetric
    From vibrational effects

Lorentzian#

Cauchy distribution \(f\)

\[ f ( \omega ; \omega_k, \gamma_k) = \frac{1}{\pi} \left[ \frac{\gamma_k}{(\omega_k - \omega)^2 + \gamma_k^2} \right] \]

with FWHM of \(2\gamma_k\). This is normalized

\[ \int_{-\infty}^{+\infty} f ( \omega ; \omega_k, \gamma_k) d \omega = 1. \]

Amplitude \(1/(\gamma \pi)\).

Gaussian#

\[ g ( \omega ; \omega_k, \sigma_k) = \frac{\sqrt{2}}{\sigma \sqrt{\pi}} \textrm{exp}\left( -\frac{2 (\omega - \omega_k)^2}{\sigma_k^2} \right) \]

HWHM is \(\sigma \sqrt{\textrm{ln 2}}/{\sqrt{2}}\)

Voigt#

Broadening calculated spectrum#

Considering only a single feature, a Lorentzian and Gaussian convolution leads to features:

def lorentzian(x, y, xmin, xmax, xstep, gamma):
    xi = np.arange(xmin, xmax, xstep)
    yi = np.zeros(len(xi))
    for i in range(len(xi)):
        for k in range(len(x)):
            yi[i] = (
                yi[i]
                + y[k] * gamma / ((xi[i] - x[k]) ** 2 + (gamma / 2.0) ** 2) / np.pi
            )
    return xi, yi


def gaussian(x, y, xmin, xmax, xstep, sigma):
    xi = np.arange(xmin, xmax, xstep)
    yi = np.zeros(len(xi))
    for i in range(len(xi)):
        for k in range(len(y)):
            yi[i] = yi[i] + y[k] * np.e ** (-((xi[i] - x[k]) ** 2) / (2 * sigma ** 2))
    return xi, yi

X = [0,0,0]
Y = [0,1,0]

plt.figure(figsize=(6, 6))
plt.subplot(211)
plt.title('Normalized to peak-max')
xi, yi = lorentzian(X, Y, -2, 2, 0.01, 0.5)
plt.plot(xi, yi/max(yi), label='Lorentzian')
xi, yi = gaussian(X, Y, -2, 2, 0.01, 0.5 / np.sqrt(4*2*np.log(2)))
plt.plot(xi, yi/max(yi), label='Gaussian')

plt.subplot(212)
plt.title('Area-normalized')
xi, yi = lorentzian(X, Y, -2, 2, 0.01, 0.5)
plt.plot(xi, yi/sum(yi), label='Lorentzian')
xi, yi = gaussian(X, Y, -2, 2, 0.01, 0.5 / np.sqrt(4*2*np.log(2)))
plt.plot(xi, yi/sum(yi), label='Gaussian')
plt.tight_layout()
plt.show()
../../_images/c64593971db1b2d1c10d317a544f135fecf3f09b24f6959219ca592b19555efd.png
Contents