Lost in Translation: Simulation-Informed Bayesian Inference Improves Understanding of Molecular Motion From Neutron Scattering

Harry Richardson, Kit McColl, Gøran J. Nilsen, Jeff Armstrong, Andrew R. McCluskey · 2026

Read Lost in Translation: Simulation-Informed Bayesian Inference Improves Understanding of Molecular Motion From Neutron Scattering by Harry Richardson, Kit McColl, Gøran J. Nilsen, Jeff Armstrong, Andrew R. McCluskey (2026) 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-2603-06080.txt.

Introduction

Richardson, McColl, Nilsen, Armstrong, and McCluskey show that quasi-springy neutron scattering can finally settle uneven spinning motion in liquid benzene. Usual fitting often cannot pin down the deep moves, since simple line-shape models let other workings produce same-looking spectra. The team mixes tiny moves models with real derived scattering models that depend on push transfer, Bayes model telling apart, and polar study. Their spinning and tumbling spread numbers point to stronger tilt than before seen, giving a route to the moves behind trapped hydro-carbon speeding.

Catalog record

CREATORHarry Richardson, Kit McColl, Gøran J. Nilsen, Jeff Armstrong, Andrew R. McCluskey
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE65318
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.06080

Browse the full library · About the Tower of Babel

Full text

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

Lost in Translation: Simulation-Informed Bayesian Inference Improves Understanding of Molecular Motion From Neutron Scattering Harry Richardson, Kit McColl, Gøran J. Nilsen, Jeff Armstrong, Andrew R. McCluskey Abstract Quasi-elastic neutron scattering (QENS) probes atomic and molecular motion on length and time scales central to catalysis, energy materials, and gas adsorption. However, conventional analytical fitting of QENS spectra often fails to uniquely determine the underlying dynamics. The flexibility of simplified line-shape models can make spectra generated by distinct physical processes statistically indistinguishable, leading to ambiguous or inaccurate mechanistic interpretation. By integrating molecular dynamics simulations, physically derived Q Q -dependent scattering models, Bayesian model discrimination, and polarisation analysis, we demonstrate that QENS can, for the first time, resolve anisotropic rotational motion in liquid benzene, a prototypical aromatic molecule relevant to microporous catalysis. The extracted spinning and tumbling diffusion coefficients suggest substantially stronger anisotropy than previously recognised. This integrated, Bayesian evidence-based analytical framework defines a new paradigm for QENS, enabling direct resolution of the rotational and translational dynamics that govern molecular interactions and transport; the fundamental processes and rate-limiting steps in confined hydrocarbon catalysis. I Introduction Progress in catalysis and materials chemistry relies on resolving molecular motion under confinement and in complex environments. Quasi-elastic neutron scattering (QENS), owing to its exceptional sensitivity to hydrogen, directly probes these processes over angstrom to nanometre length scales and sub-picosecond to nanosecond timescales, precisely matching the spatial and temporal regimes that govern bonding interactions and molecular transport in catalytic and functional materials O’Malley et al. 2016 ; Han et al. 2024 ; Jobic 2000 . Confined translational and rotational diffusion is often the key bottleneck in the catalytic efficiency of these materials. Accurate and atomistically resolved descriptions of these molecular motions, therefore, drive the design principles for higher performing catalysts Kolokolov et al. 2017 . For example, high rotation rates in catalytic pores have been associated with reduced catalytic efficiency, as the reactants cannot translate easily through narrow openings Dunn et al. 2024 More generally, QENS is widely applied across the physical sciences, where it can probe ionic migration Schwaighofer et al. 2025 , elucidate the diffusion of water through microporous polymer membranes Wang et al. 2024 , and study protein dynamics in solution Sarter et al. 2024 . In principle, QENS enables the quantification of molecular rotation and translation by measuring the total dynamic structure factor, S(Q,\omega) , as a function of wavevector, Q and energy transfer, \omega , which fully encodes the spatial and temporal atomic correlations of the system. The total dynamic structure factor is the sum of coherent and incoherent components; the incoherent part, S_{\mathrm{inc}}(Q,\omega) , describes only the self-dynamics of atoms, while the coherent signal also contains information about collective dynamics Sarter et al. 2024 ; Burankova et al. 2014 . As is the case with other scattering methods, the inversion of the dynamic structure factor from the unintuitive momentum and frequency spaces to a more relatable spatial and temporal description is often not straightforward Morton et al. 2025 . The system’s dynamics are encoded in the line shape of the \omega -dependent QENS spectra, which is convoluted with the instrument-specific spectral resolution. The evolution of the line shape with Q then describes the contributions from the system’s rotational and translational processes. Typically, QENS spectra are analysed by fitting a selected number of Lorentzian functions independently at each Q . The extracted linewidths and amplitudes are then interpreted in terms of their Q -dependence to infer the underlying rotational and translational mechanisms. This approach yields a high-dimensional, poorly constrained parameter space that can easily lead to “statistical bottlenecks” and degenerate descriptions of the same data Mamontov et al. 2019 , which can be fitted by multiple conflicting models, thereby undermining the physical reliability of the results. Furthermore, QENS analysis often relies on oversimplified models that potentially distort the dynamical behaviour of the experimental system Hernandez-Tamargo et al. 2021 , making it challenging and subjective to draw physically meaningful interpretations. Finally, in the analysis of hydrogenous systems, it is typically presumed that the total dynamic structure factor consists of purely incoherent scattering, due to the large incoherent scattering cross section of \hphantom{{}^{\text{1}}_{\text{}}}{\vphantom{\text{X}}}{}^{\mathchoice{\hbox to0.0pt{\hss$\displaystyle\vphantom{\smash[t]{\text{2}}}\text{1}$}}{\hbox to0.0pt{\hss$\textstyle\vphantom{\smash[t]{\text{2}}}\text{1}$}}{\hbox to0.0pt{\hss$\scriptstyle\vphantom{\smash[t]{\text{2}}}\text{1}$}}{\hbox to0.0pt{\hss$\scriptscriptstyle\vphantom{\smash[t]{\text{2}}}\text{1}$}}}\kern 0.0pt\text{H} . However, the development of QENS with polarisation analysis (p-QENS) has shown that coherent contributions can remain significant, and neglecting them can compromise the inferred dynamics Sarter et al. 2024 . As catalytic and functional materials and processes grow in complexity, rigorous discrimination between competing transport mechanisms becomes essential. Addressing model degeneracy, global Q -constraints, and coherent-scattering effects within a unified framework is therefore necessary to obtain physically reliable insight from QENS measurements. Previous work has attempted to address these issues individually Arbe et al. 2020 ; Jobic 2000 ; Sivia et al. 1992 ; Voneshen et al. 2017 , but no prior approach has addressed them simultaneously. In this work, we overcome these challenges by showing how it is possible to (i) develop and use accurate analysis models, based on the physics and chemistry of our system; (ii) reduce the number of fitting parameters by constraining the fit as a function of Q ; (iii) empirically compare different Q -dependent models, using Bayesian inference; and (iv) address the presence of coherent scattering by taking advantage of p-QENS. To implement this framework, we use insights from molecular dynamics (MD) simulations to guide the interpretation of QENS spectra of liquid benzene. Benzene is used as a test case because, despite its structural simplicity, its complex dynamics, including multiple rotational modes Gillen and Griffiths 1972 ; Rothschild 2012 , remain poorly characterised. Our approach is demonstrated through a state-of-the-art study in which the complex dynamics, and the rotational modes of liquid benzene, are quantified. Predictions from molecular dynamics simulations are used to inform experimental design and guide QENS analysis. Finally, we show that this comprehensive modelling-theory combination can extract previously unobtained information about the dynamics in this classic, well-studied system Kikuchi et al. 2023 . II Results and Discussion II.1 Simulations as a Qualitative Model The time and length scales covered by QENS perfectly overlap with those typically probed by classical MD simulation Armstrong 2022 . This correspondence makes MD simulations a powerful complementary tool for interpreting QENS data, enabling the computation of a “noise-free” dynamic structure factor directly from the simulation trajectory Kneller et al. 1995 ; Róg et al. 2003 ; Hinsen et al. 2012 ; Goret et al. 2017 . A perfectly accurate MD simulation could be directly compared with observed QENS measurements to provide atomistic insights that are not obtainable from experiments alone. The interatomic potentials used in classical MD simulations approximate atomic and molecular interactions and, therefore, cannot achieve the same accuracy as first-principles methods. Nevertheless, classical simulations can still be used to inform the analysis of QENS measurements in a range of different ways Morton et al. 2025 ; Petersen et al. 2022 , including the use of real-space analysis of molecular dynamics simulations as a qualitative model of the dynamics. The incoherent and total dynamic structure factors were calculated from MD simulations of benzene at 290\text{\,}\mathrm{K} , using MDANSE Goret et al. 2017 (full simulation details available in SI.I ). Fig. 1 compares the experimental QENS data (collected at the IRIS instrument,

Read the complete text