RASCBEC: RAman Spectroscopy Calculation via Born Effective Charge
Authors:
Rui Zhang,
Jun Jiang,
Alec Mishkin,
James N. Fry,
Hai-** Cheng
Abstract:
We advance the algorithm for ab initio calculations of Raman spectra for large systems via applying external electric field, and complement it by a code implementation we name RASCBEC. With the RASCBEC code, we have successfully benchmark crystalline materials and compute Raman spectra of large molecules, and amorphous oxides. Our results demonstrate a remarkable level of agreement with the result…
▽ More
We advance the algorithm for ab initio calculations of Raman spectra for large systems via applying external electric field, and complement it by a code implementation we name RASCBEC. With the RASCBEC code, we have successfully benchmark crystalline materials and compute Raman spectra of large molecules, and amorphous oxides. Our results demonstrate a remarkable level of agreement with the results from other commonly used codes as well as the experimental data. The electric field approach for Raman spectra calculation is designed to overcome the computational challenges associated with the conventional approach, which requires the calculation of the macroscopic dielectric tensor at numerous molecular geometries. The key innovation in our approach lies in obtaining the first-order derivatives with respect to the external electric field directly from VASP (the Vienna Ab Initio Simulation Package), as the Born Effective Charge (BEC). The RASCBEC code not only significantly reduces computational time, up to a factor of $N/8$, compared to the conventional approach, where $N$ is the total count of atoms within the simulation box. But also maintains the same level of accuracy, employing first-order numerical derivatives that avoid the numerical noise associated with algorithms requiring second-order derivatives, as seen in other electric field-based methods. This advantage makes RASCBEC particularly beneficial for large molecules and expansive amorphous systems.
△ Less
Submitted 27 March, 2024; v1 submitted 17 March, 2023;
originally announced March 2023.
