The Angular Localization Function (ALF): a practical tool to measure solvent angular order with Molecular Density Functional Theory.
Maïwenn Souetre, Benjamin Rotenberg, Guillaume Jeanmairet ∗ · 2026
Read The Angular Localization Function (ALF): a practical tool to measure solvent angular order with Molecular Density Functional Theory. by Maïwenn Souetre, Benjamin Rotenberg, Guillaume Jeanmairet ∗ (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-2603-05319.txt.
Introduction
Maïwenn Souetre and team present the Angular Localization Function, measuring solvent angular order in molecular density functional theory. Density functional theory computes balance solvent density around a solute. That density holds so much information that reading is hard, angular information poorly exploited. Derived from the ideal free energy functional quantifying entropy, ALF connects to known functions. Illustrations cover water, octanol, and three clay minerals in water, where shape gaps matter. ALF complements polarization or charge density for local solvent orientation, and visualizes like the Electronic Localization Function.
Catalog record
| CREATOR | Maïwenn Souetre, Benjamin Rotenberg, Guillaume Jeanmairet ∗ |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41307 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.05319
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The Angular Localization Function (ALF): a practical tool to measure solvent angular order with Molecular Density Functional Theory. Maïwenn Souetre, Benjamin Rotenberg, Guillaume Jeanmairet ∗ Abstract Molecular density functional theory is a powerful technique for efficiently computing the spatially and orientationally dependent equilibrium density of a molecular solvent around an arbitrary solute. This density encodes the detailed solvent structure, but contains so much information that it is difficult to interpret comprehensively. Although spatial dependence is frequently analyzed through orientationally integrated number density, angular information remains poorly exploited. The present work addresses this gap by introducing a function that provides a local measure of the angular order: the Angular Localization Function (ALF), derived from the ideal free energy functional, which quantifies the entropy. We discuss the connections between ALF and well known statistical functions. We illustrate the utility of ALF by discussing the solvent structure for three systems immersed in water: water as a solute, an octanol molecule, and three clay minerals (talc, fluorotalc and pyrophyllite) with small differences in their structure leading to subtle effects on their interactions with water. ALF provides information complementary to quantities such as the average polarization or charge density to characterize the local orientational distribution of solvent molecules around solutes and next to surfaces. It also offers a convenient visualization tool akin to the Electronic Localization Function (ELF) used to analyze bonding in quantum chemistry. I Introduction In the absence of external perturbation, a fluid composed of rigid classical bodies, i.e atoms or molecules whose interactions can be described by a classical force-field, is homogeneous and isotropic: On average, particles are uniformly distributed and oriented. The application of an external perturbation causes the fluid to become inhomogeneous. At the surface of minerals or metals, water forms molecular layers, with oscillating density and orientation profiles before reaching the bulk behavior (typically within approximately 1 nm). Similarly, solvent molecules are organized around solutes in solvation shells, with a degree of ordering decreasing with the distance from the solute. Several theoretical and simulation approaches exist to compute the structural and thermodynamic properties of such perturbed fluids. Methods based on an explicit sampling of the configuration space, such as Molecular Dynamics and Monte Carlo are extremely popular. While these methods are computationally demanding, they yield in principle exact results, provided that the sampling is sufficient. However, the computational cost required to converge the estimation of local properties can quickly become prohibitive, even using enhanced sampling techniques or recently introduced estimators with a reduced variance 7 ; 13 ; 12 ; 47 ; 11 ; 43 . In fact, the variance of estimators based on histograms of the position (and orientation) diverges as the size of the bins vanishes, i.e. as the resolution of the grid used to sample local properties increases. For instance, if one assumes that each bin must be visited at least 100 times to achieve an acceptable signal-to-noise ratio, computing a one-dimensional density distribution over a 10 Å range with a 0.1 Å resolution requires generating approximately \sim 10^{4} independent configurations. This requirement escalates dramatically to 10^{8} configurations for a three-dimensional density map resolved on a 10 Å 3 cube with the same resolution. Computing the orientational distribution of fluid particles on such a grid becomes practically infeasible, even though its first moment, the local average polarization, can be sampled with the same limitations as the number density 12 . Classical density functional theory (cDFT) 37 ; 17 is a powerful alternative to particle-based simulations for the evaluation of such local observables. It allows one to compute energetic and structural properties of the fluid by a variational principle (minimizing a free energy functional) rather than by a tedious sampling of the phase space. A particular strength of the theory is that it directly provides a detailed picture of the liquid structure. Building on the excellent quality of fundamental measure theory for the hard-sphere fluid 45 , accurate free energy functionals have been proposed to describe a wide range of fluids including polymers, colloids or electrolyte solutions 33 ; 34 ; 40 . Molecular density functional theory (MDFT) is a flavour of cDFT that is designed to describe the solvation of a molecular solute into a molecular solvent, such as water or acetonitrile 25 ; 52 ; 14 . In MDFT, the solvent is described by its density field, which depends on both spatial and angular coordinates. While the equilibrium solvent density is easily computed within MDFT, it is challenging to comprehend because visualizing data defined on a 6-dimensional space is difficult. This issue is typically addressed by projecting the density onto lower-dimensional representations. For example, the radial distribution function between the center of mass of the solvent and a given site of the solute is a one dimensional quantity that can be computed at any distance r by integrating the density within a thin spherical shell of radius r . Other common representations include slices of density in a given plane, as well as isosurfaces ( i.e. hypersurfaces defined by a common value of the density). Similarly, analysing orientational information requires some projections. A popular one is the first angular moment, which corresponds to the polarization field in the case of dipolar solvent molecules 30 ; 26 . Another example are the projections of the density onto rotational invariants. Such projections are helpful to solve the molecular Ornstein-Zernike equation 5 ; 4 ; 6 , or its MDFT formulation 14 ; 23 . They can also explain the long-range orientational order observed in scattering experiments 2 . In the context of solvation, several strategies have been proposed to estimate local thermodynamic quantities, in particular the entropy. Most methods rely on a spatial discretization using a 3D grid, but differ in their entropy estimator. In Grid Cell Theory 18 and 3D-2PT 32 ; 41 , entropy is estimated using analytical particle-based models, with values accumulated over time and averaged within each voxel. Grid inhomogeneous solvation theory (GIST) 29 ; 38 ; 39 is based on an exact entropy expansion in terms of solvent n -body correlation functions, that are evaluated on a 3D grid. In practice, the expansion is often limited to the one-particle term. A recent approach proposed by Heinz and Grubmüller 20 ; 21 is based on the mutual information expansion (MIE), a concept from information theory. Entropies are computed directly by sampling the configuration space, taking advantage of permutation symmetry. Both the MIE and GIST expansion theoretically converge to the exact entropy, but in practice, the expansion is truncated after the second order at best. Moreover the expansion terms of MIE and GIST do not coincide beyond the first order. It is worth noting that these studies often focus on computing entropy maps around biological objects, where the fluid is inherently inhomogeneous, and the first-order term is the first non-trivial contribution. In homogeneous systems, the first non trivial term is the second order-term, which can be computed from the radial distribution function 16 ; 28 . This quantity has been empirically related to dynamical properties such as viscosity or heat conductivity through Rosenfeld’s excess-entropy scaling 44 . Finally, as pointed out by Persson 42 , the decomposition into spatially resolved entropy is not uniquely