Sub-monolayer tracking of molecular states at buried interfaces

Sub-monolayer tracking of molecular states at buried interfaces. By Harlan Heilman and eight others, from Washington State University, the Australian Synchrotron, ANSTO, and one other institution. Published on the twenty-second of October.

Many emerging technologies depend on phenomena that begin at buried interfaces, where novel molecular states form. Resolving those states quantitatively is, in most cases, out of reach. A buried interface is by definition covered by the material above it, and the features of interest live in a very thin region. Resonant scattering has shown it is sensitive to such local states. What has held it back is not the measurement but its reading: there has been a lack of accurate interpretive models to turn the scattering signal into a description of the molecules.

The authors supply that interpretive model. They use an optical model optimized with machine learning to analyze polarized resonant soft X-ray scattering, known as RSoXS, from thin films of molecules. The scattering itself is an established probe. The new ingredient is the optical model used to interpret it, with machine learning doing the optimization.

The analysis gives a direct and quantitative reading of three properties: molecular orientation, chemical identity, and packing state. Each is resolved as a function of depth, with sub-monolayer resolution. Sub-monolayer means finer than a single layer of molecules. That reaches beyond the diffraction limit, which the authors attribute to locking energy and polarization together through the optical model, rather than treating each on its own.

Applying it, the authors find a gradient in molecular orientation and packing that evolves outward from the interface. The molecules near the interface are not arranged the same way as those further from it. They read this as a sign of competing surface and bulk growth dynamics, two processes pulling the film's structure in different directions as it grows. The result also contradicts hypotheses that the substrate patterns the film.

What the method offers is a nondestructive view of how a buried interface forms and the state it ends up in. The authors see that view as a route toward controlling interfacial states and the phenomena they drive in next-generation technologies.