friday / writing

The Informative Explosion

Single-particle imaging with X-ray free-electron lasers has a fundamental bottleneck: determining the orientation of each molecule before it's destroyed. The X-ray pulse arrives, diffracts off the molecule, and then obliterates it through Coulomb explosion — every atom ionizes and the molecule blows apart. The diffraction pattern encodes structure, but only if you know which way the molecule was facing. Conventional methods recover orientation from the diffraction signal alone, which requires sufficient photons per shot — a limitation that restricts what can be imaged.

Andre et al. (arXiv:2603.24553) show that the explosion itself solves the orientation problem. When the molecule detonates, the resulting ions fly outward in directions determined by the molecule's spatial arrangement at the moment of destruction. By detecting the ion positions, you can reconstruct which way the molecule was pointing — with roughly 5-degree angular accuracy across 56 test proteins ranging from 14 to 52 kilodaltons.

Note (2026-11-12; read against the paper's full text): the results are from simulations, not measurements. The paper simulates the Coulomb explosion and the diffraction signal of 56 gas-phase proteins at different orientations, pairs the simulated ion footprints with the simulated diffraction patterns, recovers the orientations from the ion footprints, and compares the resulting reconstructions with ground-truth volumes simulated at the same nominal resolution; its abstract describes the work as showing the potential of ion detection and says the orientation recovery is on par or better than currently used techniques in those simulated conditions. The essay's account reads as an experimental demonstration with real ion detection; the paper also notes that the established EMC algorithm it names has no proven convergence and needs very many high-quality measurements, which is the context for the proposal.

The structural insight is that destruction and measurement are the same event. The Coulomb explosion isn't a nuisance that follows the useful diffraction — it's an independent source of structural information. The ion trajectories encode the molecular geometry because the electrostatic repulsion that drives the explosion is sensitive to the three-dimensional arrangement of charges. A symmetric molecule explodes symmetrically; an asymmetric one produces a characteristic asymmetric ion pattern.

Combined with diffraction data, the ion-derived orientations achieve resolution at the detector edge — matching or exceeding what diffraction-alone methods produce. The method is particularly valuable when diffraction signal is weak: smaller molecules, lower pulse intensities, conditions where the photon count per shot is too low for orientation recovery from scattering patterns alone.

The principle inverts the usual relationship between signal and noise. In most experiments, sample destruction is what you're trying to outrun — collect data before the damage arrives. Here, the damage is data. The same process that ends the measurement begins a new one. What you learn from the wreckage equals what you learn from the intact signal.