Abstract

In this paper, we compare and contrast basis set sampling techniques recently developed for use in the ab initio multiple cloning method, a direct dynamics extension to the multiconfigurational Ehrenfest approach, used recently for the quantum simulation of ultrafast photochemistry. We demonstrate that simultaneous use of basis set cloning and basis function trains can produce results which are converged to the exact quantum result. To demonstrate this, we employ these sampling methods in simulations of quantum dynamics in the spin boson model with a broad range of parameters and compare the results to accurate benchmarks.

Highlights

  • Direct Dynamics (DD) methods have become a valuable tool for computational chemistry

  • Unlike ab initio DD methods where the vast majority of time is spent on electronic structure calculations, the spin boson model relies on a simple analytical form of the potential and the convergence of Multiconfigurational Ehrenfest (MCE) quantum dynamics to the exact quantum benchmark can be analysed relatively

  • While the MCEv1 method has shown itself to be capable of simulating high-dimensional model systems and the MCEv2 method has been successful at simulating small organic molecules,6,7,9,30 there has never been a test on a model system with high dimensionality comparing the two methods on an even footing

Read more

Summary

INTRODUCTION

Direct Dynamics (DD) methods have become a valuable tool for computational chemistry. The Multiconfigurational Ehrenfest (MCE) method has been shown to be capable of simulating various systems, from calculating the Franck-Condon spectrum of pyrazine to simulating the photodissociation of pyrrole.7 When using this method, the wavefunction is projected onto a basis of nuclear coherent states |zk and orthogonal electronic states |φr , coupled through a set of amplitudes. Ehrenfest trajectories do not always reproduce the splitting of the wavefunction between two electronic states and they can misguide the basis set This can become important for dynamics beyond very short time scales. Unlike ab initio DD methods where the vast majority of time is spent on electronic structure calculations, the spin boson model relies on a simple analytical form of the potential and the convergence of MCE quantum dynamics to the exact quantum benchmark can be analysed relatively . The main sampling technique used with MCEv2, is very similar to the procedure used in multiple spawning and train basis sets ( called time displaced basis sets) can be used in AIMS2 the results of this paper can perhaps be transferred to AIMS as well

The multi-configurational Ehrenfest method
The coherent state equations
Determination of the initial conditions of the system
Time propagation of the wavefunction
Basis function sampling and cloning
Use of basis function trains to improve MCEv2
Use of basis function cloning to improve MCEv2
Bit-by-bit propagation of an operator
The spin boson model
Comparison of the MCEv1 and MCEv2 methods
Basis set refinements and improvements for the MCEv2 method
Performance of MCEv2 for spin boson model with further parameter sets
CONCLUSIONS
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call