Driving enhanced exciton transfer by automatic differentiation

We model and study the processes of excitation, absorption, and transfer in various networks. The model consists of a harmonic oscillator representing a single-mode radiation field, a two-level system acting as an antenna, a network through which the excitation propagates, and another two-level syst...

Full description

Saved in:
Bibliographic Details
Main Authors: E Ballarin, D A Chisholm, A Smirne, M Paternostro, F Anselmi, S Donadi
Format: Article
Language:English
Published: IOP Publishing 2025-01-01
Series:Machine Learning: Science and Technology
Subjects:
Online Access:https://doi.org/10.1088/2632-2153/add23b
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849726459777646592
author E Ballarin
D A Chisholm
A Smirne
M Paternostro
F Anselmi
S Donadi
author_facet E Ballarin
D A Chisholm
A Smirne
M Paternostro
F Anselmi
S Donadi
author_sort E Ballarin
collection DOAJ
description We model and study the processes of excitation, absorption, and transfer in various networks. The model consists of a harmonic oscillator representing a single-mode radiation field, a two-level system acting as an antenna, a network through which the excitation propagates, and another two-level system at the end serving as a sink. We investigate how off-resonant excitations can be optimally absorbed and transmitted through the network. Three strategies are considered: optimising network energies, adjusting the couplings between the radiation field, the antenna, and the network, or introducing and optimising driving fields at the start and end of the network. These strategies are tested on three different types of network with increasing complexity: nearest-neighbour and star configurations, and one associated with the Fenna–Matthews–Olson complex. The results show that, among the various strategies, the introduction of driving fields is the most effective, leading to a significant increase in the probability of reaching the sink in a given time. This result remains stable across networks of varying dimensionalities and types, and the driving process requires only a few parameters to be effective.
format Article
id doaj-art-4c8ba76195ec4a59bfe3414fae3edcef
institution DOAJ
issn 2632-2153
language English
publishDate 2025-01-01
publisher IOP Publishing
record_format Article
series Machine Learning: Science and Technology
spelling doaj-art-4c8ba76195ec4a59bfe3414fae3edcef2025-08-20T03:10:10ZengIOP PublishingMachine Learning: Science and Technology2632-21532025-01-016202503410.1088/2632-2153/add23bDriving enhanced exciton transfer by automatic differentiationE Ballarin0https://orcid.org/0000-0003-3673-0665D A Chisholm1A Smirne2https://orcid.org/0000-0003-4698-9304M Paternostro3F Anselmi4https://orcid.org/0000-0002-0264-4761S Donadi5https://orcid.org/0000-0001-6290-5065Department of Mathematics, Informatics, and Geoscience, University of Trieste , via Alfonso Valerio 2, 34127 Trieste, ItalyUniversità degli Studi di Palermo , Dipartimento di Fisica e Chimica—Emilio Segrè, via Archirafi 36, 90123 Palermo, ItalyDipartimento di Fisica ‘Aldo Pontremoli’, Università degli Studi di Milano , via Celoria 16, 20133 Milan, Italy; Istituto Nazionale di Fisica Nucleare , Sezione di Milano,via Celoria 16, 20133 Milan, ItalyUniversità degli Studi di Palermo , Dipartimento di Fisica e Chimica—Emilio Segrè, via Archirafi 36, 90123 Palermo, Italy; Centre for Quantum Materials and Technologies , School of Mathematics and Physics, Queen’s University Belfast, Belfast BT7 1NN, United KingdomDepartment of Mathematics, Informatics, and Geoscience, University of Trieste , via Alfonso Valerio 2, 34127 Trieste, Italy; MIT, 77 Massachusetts Ave, Cambridge , MA 02139, United States of AmericaCentre for Quantum Materials and Technologies , School of Mathematics and Physics, Queen’s University Belfast, Belfast BT7 1NN, United Kingdom; Istituto Nazionale di Fisica Nucleare , Sezione di Trieste, via Alfonso Valerio 2, 34127 Trieste, ItalyWe model and study the processes of excitation, absorption, and transfer in various networks. The model consists of a harmonic oscillator representing a single-mode radiation field, a two-level system acting as an antenna, a network through which the excitation propagates, and another two-level system at the end serving as a sink. We investigate how off-resonant excitations can be optimally absorbed and transmitted through the network. Three strategies are considered: optimising network energies, adjusting the couplings between the radiation field, the antenna, and the network, or introducing and optimising driving fields at the start and end of the network. These strategies are tested on three different types of network with increasing complexity: nearest-neighbour and star configurations, and one associated with the Fenna–Matthews–Olson complex. The results show that, among the various strategies, the introduction of driving fields is the most effective, leading to a significant increase in the probability of reaching the sink in a given time. This result remains stable across networks of varying dimensionalities and types, and the driving process requires only a few parameters to be effective.https://doi.org/10.1088/2632-2153/add23bexciton transferdriving optimisationautomatic differentiationmachine learning
spellingShingle E Ballarin
D A Chisholm
A Smirne
M Paternostro
F Anselmi
S Donadi
Driving enhanced exciton transfer by automatic differentiation
Machine Learning: Science and Technology
exciton transfer
driving optimisation
automatic differentiation
machine learning
title Driving enhanced exciton transfer by automatic differentiation
title_full Driving enhanced exciton transfer by automatic differentiation
title_fullStr Driving enhanced exciton transfer by automatic differentiation
title_full_unstemmed Driving enhanced exciton transfer by automatic differentiation
title_short Driving enhanced exciton transfer by automatic differentiation
title_sort driving enhanced exciton transfer by automatic differentiation
topic exciton transfer
driving optimisation
automatic differentiation
machine learning
url https://doi.org/10.1088/2632-2153/add23b
work_keys_str_mv AT eballarin drivingenhancedexcitontransferbyautomaticdifferentiation
AT dachisholm drivingenhancedexcitontransferbyautomaticdifferentiation
AT asmirne drivingenhancedexcitontransferbyautomaticdifferentiation
AT mpaternostro drivingenhancedexcitontransferbyautomaticdifferentiation
AT fanselmi drivingenhancedexcitontransferbyautomaticdifferentiation
AT sdonadi drivingenhancedexcitontransferbyautomaticdifferentiation