Peer-reviewed work from my time in computational chemistry and nanoplasmonics, plus patents.
Full list on Google Scholar.
2024
L. Nicoli, S. Sodomaco, P. Lafiosca, T. Giovannini, C. Cappelli — ACS Physical Chemistry Au 4 (6), 669–678
Real-world plasmonic nanoparticles operate in solution, yet most atomistic models treat them in vacuum. This work presents ωFQFμ/FQ, a multiscale classical model that couples an atomistic description of the nanoparticle’s plasmonic response with a polarizable force-field description of the surrounding solvent, making it possible to simulate real-size colloidal nanoparticles of generic chemical nature. Validated against reference quantum-mechanical calculations, the model accurately captures solvent-induced shifts of the plasmon resonance at a fraction of the computational cost.
P. Lafiosca, L. Nicoli, S. Pipolo, S. Corni, T. Giovannini, C. Cappelli — J. Physical Chemistry C 128 (41), 17513–17525
Atomistic electromagnetic models of plasmonic nanoparticles are usually formulated in the frequency domain, which limits the study of time-resolved phenomena and ultrafast laser excitation. This paper reformulates the ωFQ and ωFQFμ models in the real-time domain, propagating the induced charges and dipoles of each atom as the system responds to an external pulse. The approach reproduces frequency-domain results while opening the way to simulating time-dependent plasmon dynamics, nonlinear regimes, and arbitrary excitation schemes in large nanostructures.
2023
P. Lafiosca, L. Nicoli, L. Bonatti, T. Giovannini, S. Corni, C. Cappelli — J. Chemical Theory and Computation 19 (12), 3616–3633
Surface-enhanced Raman scattering (SERS) lets researchers detect molecules near plasmonic nanostructures, but simulating the effect requires treating both the molecule and the metal at high accuracy. This work combines a quantum-mechanical description of the adsorbed molecule with the atomistic classical models ωFQ and ωFQFμ for the plasmonic substrate, yielding a multiscale method that works for both noble-metal and graphene-based nanostructures. The approach captures the electromagnetic enhancement of Raman signals and provides an interpretable, computationally affordable tool for studying molecule–nanostructure interactions.
L. Nicoli, P. Lafiosca, P. Grobas Illobre, L. Bonatti, T. Giovannini, C. Cappelli — Frontiers in Photonics 4, 1199598
Mixing two metals in a single nanoparticle — as a random alloy or a core-shell structure — offers a powerful way to tune plasmonic properties, but modeling such systems atomistically is challenging. This paper extends the ωFQFμ model to bimetallic gold–silver nanoparticles, generalizing its formulation so that each atom carries parameters of its own chemical species. Tested against reference quantum-mechanical calculations, the approach reproduces how composition and atomic arrangement affect the plasmonic response, enabling the in silico exploration of multimetallic nanostructure design.
2022
T. Giovannini, L. Bonatti, P. Lafiosca, L. Nicoli, M. Castagnola, P. G. Illobre, S. Corni, C. Cappelli — ACS Photonics 9 (9), 3025–3034
Plasmonics is ultimately a quantum phenomenon, so it is natural to ask whether expensive quantum-mechanical calculations are indispensable to model metal nanostructures. This paper systematically compares the classical atomistic models ωFQ and ωFQFμ against fully quantum time-dependent DFT calculations for metal nanoparticles of decreasing size. The classical models reproduce the quantum reference remarkably well down to very small structures, provided quantum tunneling effects at sub-nanometer junctions are accounted for — showing that accurate, affordable classical atomistic modeling is a viable alternative in most plasmonic regimes.
L. Nicoli, T. Giovannini, C. Cappelli — J. Chemical Physics 157 (21), 214101
When a molecule moves from vacuum into water, its absorption spectrum shifts — an effect that computational models should reproduce to be predictive in realistic environments. This study benchmarks a range of QM/MM embedding schemes, from simple nonpolarizable force fields to fully polarizable approaches, against experimental vacuo-to-water solvatochromic shifts for a set of chromophores. No single model emerges as uniformly best: the accuracy depends strongly on the nature of the electronic transition, providing practical guidance on which embedding strategy to choose for a given problem.
L. Bonatti, L. Nicoli, T. Giovannini, C. Cappelli — Nanoscale Advances 4 (10), 2294–2302
Plasmonic hot-spots — tiny regions of enormously enhanced electric field — are the key to ultrasensitive molecular detection, and graphene could offer a cheaper, more versatile platform than noble metals. Using atomistic electromagnetic simulations, this work explores engineered graphene nanostructures, inspired by patterns known to enable single-molecule detection on metal substrates, and focuses on realistic features such as edge defects and grain boundaries. The calculations show that highly localized hot-spots with enhancement factors comparable to noble metals can form near these defects, outlining design rules for graphene-based plasmonic sensors.
Patents & Theses
L. Nicoli — PhD thesis, Scuola Normale Superiore Thesis
This thesis develops the ωMM family of classical atomistic electromagnetic models for simulating the optical properties of plasmonic materials with quantum-level accuracy at classical-level cost. Starting from the ωFQ model for Drude-like metals, it extends the framework to noble metals by including interband transitions, generalizes it to bimetallic and alloyed nanostructures, and reformulates it in the real-time domain for time-dependent phenomena. Finally, it couples the models to polarizable environments and quantum-mechanical molecular descriptions, enabling applications such as refractive-index sensing and surface-enhanced Raman scattering.
C. Cappelli, L. Bonatti, L. Nicoli, T. Giovannini, P. Lafiosca — WO 2024/033744 A1 (PCT/IB2023/057698); IT priority 102022000017058 Patent
This patent protects a fully computational workflow for designing plasmonic sensors with single-molecule sensitivity. Starting from a target analyte and a library of nanostructured plasmonic substrates, the method maps the analyte–substrate interaction potential at the atomistic level, computes the resulting spectroscopic signal and its enhancement factor, and compares it against a predefined sensitivity threshold. The result is a screening procedure that identifies promising sensor designs before any fabrication, reducing experimental trial and error.