Quantum propagation scheme using Hagedorn wave packets : numerical developments

Rabiou Issa, Kokou Mawulonmi Robert Afansounoudji, Komi Sodoga, David Lauvergnat · 2025

Read Quantum propagation scheme using Hagedorn wave packets : numerical developments by Rabiou Issa, Kokou Mawulonmi Robert Afansounoudji, Komi Sodoga, David Lauvergnat (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-17541.txt.

Introduction

Rabiou Issa and co-authors present a 2025 method for tracking quantum wave packets over time with Hagedorn packets. Such packets suit many molecular tasks, but one fixed basis set fits some motions poorly. Their scheme mixes bases that shift with time and bases that stay fixed, such as wave and box sets.

The code can also handle systems with more than one electronic state, which opens non-adiabatic work. A three-step cycle drives each run, with exact number work for the overlaps. Tests on two toy models with smooth and rough surfaces matched standard methods well.

Catalog record

CREATORRabiou Issa, Kokou Mawulonmi Robert Afansounoudji, Komi Sodoga, David Lauvergnat
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE68 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-17541

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Quantum propagation scheme using Hagedorn wave packets : numerical developments Preprint: APS/123-QED Rabiou Issa, Kokou Mawulonmi Robert Afansounoudji, Komi Sodoga, David Lauvergnat Abstract In this study, we provide a novel wave packet propagation method that generalizes the Hagedorn approach by introducing alternative primitive basis sets that are better suited to describe different physical processes. More precisely, in our propagation scheme, we can mix basis sets with time-dependent parameters (the Hagedorn basis set) and time-independent ones, such as Fourier series, particle-in-a-box, or harmonic oscillator basis sets. Furthermore, our implementation can handle models with several electronic states, so that non-adiabatic processes can be studied. Instead of the time-dependent variational principle, our propagation scheme uses a three-step procedure (standard propagation, time-dependent parameter evaluation, and projection). It relies on multidimensional integrations, which are performed numerically with Gaussian quadrature, so that we have no constraints on the for of the Hamiltonian operator. This numerical algorithm has been implemented in a modern Fortran code available on GitHub ( https://github.com/asodoga/TD_Schrod_Rabiou ). The code has been tested and validated by comparisons with standard propagation schemes on 2D-models with harmonic and anharmonic potentials (modified Hénon-Heiles). More precisely, our benchmark tests show that the wave packets obtained with our propagation scheme are able to converge to exact wave packets (obtained from standard propagation techniques). Finally, we have applied our method to compute the vibrational spectrum of the 6D-modified Hénon-Heiles model and we show that our scheme reproduces well the results obtained with the standard approach and with smaller basis functions. As a perspective, we show that with the generalized Hagedorn wave packet method, we are able to study the non-adiabtic dynamics of the cis-trans retinal isomerization in a reduced 2D-model with two coupled electronic surfaces. I Introduction Understanding the behavior of molecular systems requires quantum mechanics, which is essential for describing both their electronic structures and their nuclear motions. For example, the simulation of the dynamics of certain fundamental processes, such as multiphoton excitations Walker et al. 1994 ; Rudenko et al. 2004 , photodissociation Engel et al. 2005 ; Pedersen et al. 2007 ; Sodoga et al. 2009 , photo-isomerization Hahn and Stock 2000a ; Hahn and Stock 2000b and charge transfer Kimura and Lin 1986 ; Jain et al. 1987 ; Loreau et al. 2010 requires to solve numerically the time-dependent Schrödinger equation (TDSE). In particular, these photoinduced processes involve non-adiabatic transitions between different electronic states, which then requires the quantum treatment of the system. This is the case, for example, with the photoisomerization of retinal Hahn and Stock 2000a ; Hahn and Stock 2000b ; Marsili et al. 2020 ; Manathunga et al. 2017 ; Pereira et al. 2023 , which can only be studied in the context of non-adiabatic quantum dynamics. To solve the TDSE equation, we need to determine the time evolution of a wave packet \Psi . A formal solution of TDSE as a function of the evolution operator is known if the Hamiltonian operator is independent of time: \ket{\Psi\left( t\right) }=\exp[-\imath/\hbar\cdot\hat{H}(t-t_{0})\big]\ket{\Psi\left( t_0\right) }. (1) However, this requires explicit knowledge of the evolution operator \exp[-\imath/\hbar\cdot\hat{H}\left(t-t_{0}\right)\big] , which is generally not straightforward unless we know a basis in which \hat{H} is diagonal. This is called propagation in the spectral representation. In the general case, propagation is usually divided into small constant time intervals, \Delta t , and: \displaystyle\ket{\Psi\left( t + \Delta t\right)} \displaystyle= \displaystyle\exp[-\imath/\hbar\cdot\hat{H}\Delta t\big]\ket{\Psi\left( t\right)}. (2) To solve the TDSE, one can numerically express the action of the evolution operator, \exp[-\imath/\hbar\cdot\hat{H}\Delta t\big] , on a wave packet using various techniques Leforestier et al. 1991 : (i) split operator Feit et al. 1982 , (ii) decomposition on polynomials (usually Chebychev) Tal-Ezer and Kosloff 1984 , (iii) Taylor expansion Mercouris et al. 1994 ; Lauvergnat et al. 2007 . However, the above-mentioned methods have drawbacks related to the number of degrees of freedom in the system. Indeed, the size of the basis (or grid) scales exponentially with the number of degrees of freedom. In addition, the basis must accurately describe the evolution of the wave packet over time. Thus, if the wave packet delocalizes, the basis describing it should be very large. Similarly, if the center of the wave packet evolves strongly while remaining localized, the basis must also be very large. Furthermore, non-adiabatic processes present significant challenges to simulations. These transitions between electronic states often involve a topological feature known as a conical intersection. At this intersection, nuclear and electronic motions become intertwined by so-called non-adiabatic coupling, where quantum effects play a crucial role in the dynamics of the nuclear motion. For these reasons, quantum dynamics and especially non-adiabatic quantum dynamics using multidimensional grids have been essentially limited to 4-atom or 5-atom systems for several decades, despite all the progress that has been made. To solve the problem when the center of the wave packet evolves strongly, it is possible to represent the wave packet in a basis set (and grid) that evolves over time. More precisely, some basis set parameters change variationnally during the propagation. Several wave packet propagation techniques have been developed, based on this idea of a time-evolving basis set (or grid). Without being exhaustive, one can mention: (i) The semiclassical approach introduced by Heller Heller 1976 ; Huber and Heller 1987 ; Drolshagen and Heller 1983 ; Vaníček and Begušić 2021 in which a single Gaussian evolves over time. (ii) The approaches in which the wave packet is described by a sum of Gaussians evolving over time Worth et al. 2004 ; Martinez et al. 1996 ; Curchod and Martinez 2018 . (iii) In the references Sielk et al. 2009 ; Choi and Vaníček 2019 , the wave packet is represented on a grid that changes over time according to the evolution of the wave packet. (iv) George Hagedorn introduced the approach now known as Hagedorn wave packets Hagedorn 1981 ; Hagedorn et al. 2006 ; Hagedorn 1998 ; Faou et al. 2009 ; Lasser and Lubich 2020 ; Zhang and Vaníček 2024 , i.e. products of complex Gaussians with polynomials that form an orthonormal basis in L^{2} space for which certain parameters (center, width and momentum) adapt variationnally during the propagation. (v) A widely adopted solution in this context is the multi-configuration time-dependent Hartree (MCTDH) method Meyer et al. 1990 ; Hammerich et al. 1990 ; Meyer and Worth 2003 . This approach represents the wavefunction as a small direct product of time-dependent basis sets (known as single-particle functions). This method allows the propagation of quantum wave packets with a few dozen of degrees of freedom, or even larger with the multilayer version of MCTDH Wang and Thoss 2003 ; Manthe 2008 ; Vendrell and Meyer 2011 . However, this approach has its limitations. Both the kinetic and the potential components of the Hamiltonian must be expressed as sums of products of one-dimensional terms, which is not always an easy task. It is therefore still essential to develop wave packet propagation methods that are well-suited for handling a large number of strongly coupled degrees of freedom without limitation in terms of the form of the Hamitonian. For a molecular system composed of nc coordinates, \mathbf{Q}\left(\mathbf{Q}=\left\{q_{1},q_{2}\cdots q_{nc}\right\}\right) , the total wave packet, \ket{\Psi\left( t\right) } , is expanded, as usual, on

Read the complete text