arXiv subjects

Lorenzo Cupellini

Publications and source records attributed to Lorenzo Cupellini.

5 recordsLinked to original sources

arXiv2026

Beyond the Quantum Picture: The Electrodynamic Origin of Chiral Nanoplasmonics

Chiral plasmonic nanostructures are rapidly emerging as ideal substrates for enantioselective sensing, chiral near-field engineering, and plasmon-assisted catalysis, owing to their exceptional sensitivity to structural handedness. However, the physical origin of plasmonic chirality, whether intrinsically quantum or primarily governed by collective electrodynamics, remains an open question, limiting the development of predictive theoretical methods for the design of novel chiral plasmonic architectures. Here, we show that a fully atomistic classical electrodynamic model, coupling intraband charge transport and interband polarization, quantitatively reproduces state-of-the-art \textit{ab initio} and experimental chiroptical spectra across the quantum-to-classical regime, from atomistically defined chiral Ag and Au nanostructures to DNA-origami-assembled Au nanorods containing up to $\sim 10^5$ atoms. Our results support a unified electrodynamic origin of plasmonic chirality, providing the missing foundation to connect local structural motifs to chiroptical response and local chiral near fields, and paving the way for the atomistically defined, rational design of chiral plasmonic nanostructures optimized for targeted applications.

cond-mat.mes-hall↗

arXiv2025

A symmetry-preserving and transferable representation for learning the Kohn-Sham density matrix

The Kohn-Sham (KS) density matrix is one of the most essential properties in KS density functional theory (DFT), from which many other physical properties of interest can be derived. In this work, we present a parameterized representation for learning the mapping from a molecular configuration to its corresponding density matrix using the Atomic Cluster Expansion (ACE) framework, which preserves the physical symmetries of the mapping, including isometric equivariance and Grassmannianity. Trained on several typical molecules, the proposed representation is shown to be systematically improvable with the increase of the model parameters and is transferable to molecules that are not part of and even more complex than those in the training set. The models generated by the proposed approach are illustrated as being able to generate reasonable predictions of the density matrix to either accelerate the DFT calculations or to provide approximations to some properties of the molecules.

physics.chem-ph↗

arXiv2024

Theoretical study of the influence of the photosynthetic membrane on B800-B850 energy transfer within the peripheral light-harvesting complex LH2

Photosynthetic organisms rely on a network of light-harvesting protein-pigment complexes to efficiently absorb sunlight and transfer excitation energy to reaction center proteins for charge separation. In photosynthetic purple bacteria, these complexes are embedded in the cell membrane, where lipid composition affects their clustering and inter-complex energy transfer. However, the lipid bilayer's impact on intra-complex excitation dynamics is less understood. Recent experiments compared photo-excitation dynamics in detergent-isolated light harvesting complex 2 (LH2) to LH2 embedded in membrane discs mimicking the biological environment, revealing differences in spectra and intra-complex energy transfer rates. We use available quantum chemical and spectroscopy data to develop a complementary theoretical study on the excitonic structure and intra-complex energy transfer kinetics of the LH2 from photosynthetic purple bacteria Rhodoblastus acidophilus in two conditions: LH2 in a membrane environment and detergent-isolated LH2. Dark excitonic states crucial for B800-B850 energy transfer within LH2 are found to be more delocalised in the membrane model. Using non-perturbative and generalised Förster calculations, it is shown that the increased quantum delocalisation leads to a B800 to B850 transfer rate 30% faster than in the detergent-isolated complex, consistent with experimental results. We identify the main energy transfer pathways in each environment and show how differences in the B800 to B850 transfer rate stem from changes in LH2's electronic properties when embedded in the membrane. By considering quasi-static variations of electronic excitation energies in LH2, we show that the broadening of the B800 to B850 transfer rate distribution is affected by lipid composition. We argue that the variation in broadening could indicate a speed-accuracy trade-off, common in biological systems.

physics.chem-ph↗

arXiv2024

The OpenMMPol Library for Polarizable QM/MM Calculations of Properties and Dynamics

We present a new library designed to provide a simple and straightforward way to implement QM/AMOEBA and other polarizable QM/MM methods based on induced point dipoles. The library, herein referred to as OpenMMPol, is free and open-sourced and is engineered to address the increasing demand for accurate and efficient QM/MM simulations. OpenMMPol is specifically designed to allow polarizable QM/MM calculations of ground state energies and gradients, and excitation properties. Key features of OpenMMPol include a modular architecture facilitating extensibility, parallel computing capabilities for enhanced performance on modern cluster architectures, and a user-friendly interface for intuitive implementation and a simple and flexible structure for providing input data. To show the capabilities of fered by the library we present an interface with PySCF to perform QM/AMOEBA molecular dynamics, geometry optimization and excited state calculation based on (TD)DFT.

physics.chem-ph↗

arXiv2023

Electronic excited states from physically-constrained machine learning

Data-driven techniques are increasingly used to replace electronic-structure calculations of matter. In this context, a relevant question is whether machine learning (ML) should be applied directly to predict the desired properties or be combined explicitly with physically-grounded operations. We present an example of an integrated modeling approach, in which a symmetry-adapted ML model of an effective Hamiltonian is trained to reproduce electronic excitations from a quantum-mechanical calculation. The resulting model can make predictions for molecules that are much larger and more complex than those that it is trained on, and allows for dramatic computational savings by indirectly targeting the outputs of well-converged calculations while using a parameterization corresponding to a minimal atom-centered basis. These results emphasize the merits of intertwining data-driven techniques with physical approximations, improving the transferability and interpretability of ML models without affecting their accuracy and computational efficiency, and providing a blueprint for developing ML-augmented electronic-structure methods.

physics.chem-ph↗