Login / Signup

Finite Temperature Dynamical Density Matrix Renormalization Group for Spectroscopy in Frequency Domain.

Tong JiangWeitang LiJiajun RenZhigang Shuai
Published in: The journal of physical chemistry letters (2020)
We present a novel methodology through casting the dynamical density matrix renormalization group (DDMRG) into the matrix product state (MPS) formulation to calculate the spectroscopy at finite temperature for molecular aggregates. The frequency domain algorithm can avoid the time evolution accumulation of error and is naturally suitable for parallelization, in addition to facile graphic processing unit (GPU) acceleration. The high accuracy is demonstrated by simulating the optical spectra of vibronic model systems ranging from an exactly solvable dimer model to a more complex real-world perylene bisimide (PBI) J-aggregate. The relationship between the 0-0 emission strength and the exciton thermal coherent length is discussed for linearly stacked aggregates. The computing performance largely boosted by GPU demonstrates that DDMRG emerges as a promising method to study dynamical properties for complex systems.
Keyphrases
  • density functional theory
  • high resolution
  • single molecule
  • solid state
  • drug delivery
  • machine learning
  • molecular dynamics
  • deep learning
  • gold nanoparticles
  • perovskite solar cells