Molecular Shape Holds the Key to Sharper Optical Pressure Standards

For more than a century, one of the most elegant equations in optics has allowed scientists to infer the pressure, temperature and density of a gas simply by measuring how much it bends light. The Lorentz–Lorenz relation, formulated on the assumption that molecules behave like tiny conducting spheres floating far apart from one another, underpins today’s most accurate optical refractometers — instruments so precise that they are now being developed as primary standards for the pascal, the SI unit of pressure. The trouble is that real molecules are not spheres, and real gases are not ideal. A new study from the Universidad Complutense de Madrid, published in Results in Optics, tackles that gap head-on by giving the century-old formula a physically grounded makeover built from the actual shapes of molecules.

The research team, led by Sergio Moltó with María A. Sáenz-Nuño and Eusebio Bernabeu, extended the Lorentz–Lorenz relation through a so-called shape factor, a parameter that absorbs the deviations from the idealized spherical symmetry assumed in the original 1916 derivation. In their earlier work, the shape factor was extracted by fitting virial expansion coefficients — the standard mathematical machinery for describing non-ideal gas behavior. The problem was circular: the correction was defined in terms of the very coefficients it was meant to complement, leaving its connection to independently measurable molecular properties unclear. The new study set out to break that circularity by asking a deceptively simple question: can the shape factor be predicted from the geometry of a molecule’s electron cloud alone?

To answer it, the team turned to computational quantum chemistry. They selected eight gases of outsized importance to metrology and the energy transition: nitrogen, hydrogen, oxygen, carbon dioxide, nitrous oxide, helium, neon and argon. Nitrogen, neon and argon are workhorses of fundamental metrology, appearing in optical pressure standards and high-accuracy gas thermometry. Hydrogen and oxygen are central to renewable energy applications tied to the European Green Deal, while carbon dioxide and nitrous oxide are major greenhouse gases whose accurate characterization matters for atmospheric monitoring. For each molecule, the researchers computed the electronic charge density distribution using the PySCF framework, employing several density functional theory functionals — local density approximation, generalized gradient approximation and hybrid functionals — alongside Hartree–Fock calculations as a reference point.

Extracting a meaningful molecular size from a charge density map requires choosing an isovalue, the threshold at which the fuzzy electron cloud is sliced into a defined surface. The team calibrated this choice by matching the enclosed volume against reference volumes computed with the widely used cheminformatics toolkit RDKit, accepting only isovalues that agreed to within one percent with overlapping uncertainty ranges. Surface areas were then obtained with a convex hull algorithm, which wraps the smallest possible boundary around the points of the charge density isosurface. The method proved robust: volumes derived from the convex hull matched those from direct numerical integration within the one percent tolerance, and the results showed minimal sensitivity to the computational grid. The researchers explicitly rejected the popular solvent-accessible surface area approach, since it describes a surface defined by van der Waals radii rather than by the electronic charge density itself — a subtle but crucial distinction for a study that hinges on genuine molecular geometry.

With molecular volumes and surface areas in hand, the team introduced a dimensionless quantity they call the asphericity, defined from the volume and surface area of the charge density isosurface. A value of unity corresponds to a perfect sphere; values below one signal increasingly elongated or irregular shapes. The results behaved exactly as physical intuition demands. Helium, neon and argon scored essentially perfect spherical marks — 0.9994, 0.9996 and 0.9997 respectively — reflecting their noble-gas electron clouds. Hydrogen came in at 0.9947, nitrogen at 0.9825 and oxygen at 0.9776, while the linear triatomic molecules carbon dioxide and nitrous oxide registered the lowest values at 0.9502 and 0.9425. Interestingly, the charge density isosurfaces of carbon dioxide and nitrous oxide resemble those of diatomic molecules, because their linear geometry spreads electron density almost continuously along the molecular axis, rendering the central atom barely distinguishable in the isosurface picture.

The shape factors themselves were evaluated at a fixed temperature of 300 kelvin, chosen because refractivity virial coefficient data are not available for all the studied molecules at other temperatures. Each shape factor was obtained by least-squares fitting of the extended Lorentz–Lorenz equation to refractive-index values spanning pressures from zero to 100 kilopascals, exploiting the fact that the shape factor does not depend on pressure. The ideal value that would recover the classical Lorentz–Lorenz relation is 3.764 in the units used, and the measured values clustered tellingly around and above it: hydrogen at 3.886, nitrogen at 3.906, oxygen at 3.957, argon at 4.130, carbon dioxide at 4.167 and nitrous oxide at 5.261, all with combined uncertainties of a few percent. The noble gases proved harder to interpret, with helium at 3.011 and neon at 3.427 falling below the ideal value, and the uncertainties on their asphericities larger than the differences being probed.

The pivotal result is the correlation between geometry and optics. For the simple molecules, the absolute difference between each shape factor and the ideal value tracks the asphericity along a smooth trend described by a rational function with three independent parameters. In other words, the further a molecule’s electron cloud departs from a perfect sphere, the further its optical behavior departs from the classical Lorentz–Lorenz prediction — and the relationship is quantitative. This is precisely the link missing from the conventional virial description, where the second and third refractivity virial coefficients capture non-ideal effects effectively but offer no transparent connection to molecular descriptors. The shape factor, by contrast, can in principle be computed from first principles for any molecule whose charge density can be calculated, without ever touching a virial coefficient.

To stress-test the framework, the researchers introduced a residual second virial coefficient, representing whatever second-order non-ideal contribution remains after the shape-factor correction has done its work. The outcome was striking: for every molecule studied, the residual coefficient was more than three orders of magnitude smaller than the conventional second refractivity virial coefficient. The geometrical correction, in other words, accounts for the overwhelming bulk of the second-order non-ideal behavior. The reduction was largest for hydrogen and smallest for the more complex carbon dioxide and nitrous oxide, suggesting that for these species additional physics — anisotropic intermolecular interactions and orientation-dependent collision dynamics — still lurks in the residual term. The authors are careful to position the approach as a complement to, not a replacement for, the virial expansion, which remains the rigorous standard for representing refractivity data.

The practical payoff could be considerable. Optical refractometry already delivers pressure and temperature measurements with resolution and stability that mechanical instruments cannot match, and it is the foundation of emerging primary pressure standards and quantum-based gas thermometry across Europe’s metrology institutes. A working equation whose dominant non-ideal correction can be computed from molecular structure — rather than fitted from experiment for every gas, wavelength and temperature — streamlines both data analysis and experimental planning. The team has developed a computational algorithm implementing the model, and their funding through the European Partnership on Metrology projects MQB-Pascal and PriSpecTemp signals that the work is aimed squarely at the next generation of optical measurement standards. Future lines of research include coupling the geometrical correction with molecular statistical mechanics, characterizing the temperature dependence of the shape factor, and extending the framework to more complex molecular systems. What began as a simplifying assumption about tiny conducting spheres in 1916 may, a century later, find its most useful form in an equation that finally knows the true shape of the molecules it describes.

Subject of Research: Extension of the Lorentz–Lorenz relation using molecular shape factors from charge density calculations to improve refractometric accuracy

Article Title: Realistic foundations for improving the accuracy in refractometry based on an extension of the Lorentz–Lorenz relation

Article References: Moltó, S., Sáenz-Nuño, M. A., & Bernabeu, E. (2026). Realistic foundations for improving the accuracy in refractometry based on an extension of the Lorentz–Lorenz relation. Results in Optics, Article 101180. https://doi.org/10.1016/j.rio.2026.101180

Image Credits: AI Generated

DOI: 10.1016/j.rio.2026.101180

Keywords: refractometry, Lorentz–Lorenz relation, molecular shape factor, charge density, density functional theory, gas metrology, optical pressure standards, virial coefficients, asphericity, noble gases, hydrogen, refractive index