Chemical Science · 2021 · first author

The Hofmeister series follows charge per unit radius

Hofmeister effects, where different salts change protein stability, viscosity or reaction rates in a consistent order, have resisted a quantitative explanation for over a century. For anions, this paper traces them to Coulomb interactions at the ion's most charge-dense atom. It reduces each anion to one number, the radial charge density ϸ = q/r: that atom's partial charge divided by its effective radius.

The surface is ϸ = q/r itself. Each sphere sits at an anion's computed site charge and radius, so its height is its place in the series. The white lines are constant ϸ: anions on the same line should behave alike, even when their size and charge differ.

Most to least charge-dense

Values and source

Anion (site)	q (e)	r (Å)	ϸ (10⁻¹⁰ C m⁻¹)	UE (kJ mol⁻¹)

Site-specific rows of Table S2 in the paper's supplementary information. For each molecular anion the most charge-dense atom is used, as in the paper. U_E is the electrostatic energy between the ion and one water molecule.

Gregory, Wanless, Webber, Craig and Page, Chem. Sci. 12, 15007 (2021)

Physical Chemistry Chemical Physics · 2022 · first author

Understanding specific ion effects and the Hofmeister series

Different salts change protein stability, polymer swelling, surface tension and reaction rates in a consistent order, known as the Hofmeister series. More than a century of work has produced many explanations, but no agreed one.

This review, written with experimental and computational colleagues, sets out what is and isn't understood about specific ion effects and the Hofmeister series.

Gregory, Elliott, Robertson, Kumar, Wanless, Webber, Craig, Andersson and Page, Phys. Chem. Chem. Phys. 24, 12682 (2022)

Scientific Data · 2022 · first author

IonSolvR: more than 3,000 quantum MD runs of ions in solvent

Studying ions outside water has been limited by a lack of consistent data. IonSolvR is an open repository of nanosecond-scale density functional tight binding (DFTB3-D3(BJ)) molecular dynamics trajectories, covering up to 55 ionic solutes in 28 protic and aprotic solvents, all run with one protocol.

Effective concentration is set by cell size. The scene shows the four NaCl–water cells described in the paper, to scale. The water positions are a random packing drawn for this page, like the random starting boxes the simulations began from; they are not frames from the repository.

The four NaCl–water cells

Water molecules	Cell edge (Å)	Effective concentration (mol L⁻¹)
32	10.2	1.72
64	12.6	0.87
100	14.6	0.55
300	20.9	0.19

From the Technical Validation section of the paper. Simulations ran in the NVT ensemble at 300 K with periodic cubic cells and the 3ob-3-1 parameters.

Gregory, Elliott, Wanless, Webber and Page, Sci. Data 9, 430 (2022)

Chemical Science · 2025 · led by E. C. Johnson

Are our textbooks wrong about the inductive effect?

The inductive effect, the idea that an electronegative atom pulls electron density along a chain of bonds, is one of the first explanations organic chemistry students learn. Haloacetates are the textbook example: change the halogen and the acid's strength changes.

This 2025 Chemical Science paper, led by E. C. Johnson, found that the inductive effect doesn't explain the electron density in haloacetates. Four follow-ups in the Journal of Chemical Education rethink how inductive effects, hyperconjugation and the reactivity of aldehydes and ketones are taught. Each one is on the shelf too:

• Rethinking the nature and extent of inductive effects in organic compounds (2026)
• Exploring the interface of hyperconjugation and polarizability (2026)
• Enhancing pedagogies to explain the relative reactivity of aldehydes and ketones (2026)
• The inductive effect: a response to Nguyen (2026)

Johnson, Gregory, Robertson, Gresham, Nelson, Craig, Prescott, Page, Webber and Wanless, Chem. Sci. 16, 2382 (2025)

Journal of Computational Chemistry · 2025 · corresponding author

Geometry alone goes a long way toward predicting PAH stability

Polycyclic aromatic hydrocarbons matter for pollution toxicity and for carbon materials, and their isomers differ widely in stability. This paper computed PBE0-D4 energies for 38,264 nonplanar PAH isomers, a method that matches CCSD(T) to 0.67 kcal mol⁻¹ on a 335-molecule benchmark. It then asked which cheap geometric measure tracks those energies.

The winner is Σ_dihedral, the sum of the deviations from planarity across every C–C–C–C dihedral. It costs nothing to compute from coordinates, and the paper's PAHAPS web tool uses it to estimate an isomer's energy without a quantum calculation.

How close each predictor gets

GFN2-xTB, uncorrected5.87

HOMA (aromaticity)5.30

Δz (out-of-plane height)4.78

Σ_dihedral3.62

Σ_dihedral + HOMA2.53

GFN2-xTB, fitted1.16

Σ_dihedral + HOMA + fitted xTB0.82

The dashed line marks 1 kcal mol⁻¹, often called chemical accuracy. Values from Tables 2 and 3 of the paper.

The six formula sets

Formula	Isomers	Mean (kcal mol⁻¹)	Highest
C₃₆H₂₀	1,182	15.00	37.16
C₃₈H₂₀	1,594	15.00	38.74
C₄₀H₂₀	1,683	17.60	43.27
C₄₀H₂₂	5,084	17.23	42.76
C₄₂H₂₂	7,662	16.77	42.20
C₄₄H₂₄	21,059	18.75	54.74
All	38,264	17.82	54.74

Energies are PBE0-D4/6-31G(2df,p) on GFN2-xTB geometries, relative to the most stable isomer of each formula. The histograms use every one of them, from the data that ships with PAHAPS; the counts, means and maxima match Table 1 of the paper. The three molecules were re-optimised for this page with GFN2-xTB, the method COMPAS-3x used, and their Σ_dihedral was computed with the same definition as the paper's code.

Gregory and Karton, J. Comput. Chem. 46, e70198 (2025) · PAHAPS web tool · code

bioRxiv · 2026 · preprint · led by Ciara Wallis

How the malaria parasite's lactate transporter moves protons

Malaria parasites run on glycolysis, and must export the lactate it produces, together with a proton, through a transporter in their membrane. That transporter is a target for new antimalarial drugs.

Ciara Wallis led this work as first author, with Ben Corry's group at the Australian National University; I'm on her PhD supervisory panel. It proposes a proton-transfer mechanism, and points to a channel-like transporter that works without the large conformational changes most transporters rely on.

Wallis, Gregory, Fairweather, van Dooren, Lehane and Corry, “A proton transfer mechanism in the malaria parasite lactate/H⁺ symporter reveals a channel-like transporter without conformational changes”, bioRxiv (2026)