Now, as it turns out, properly designed secondary ion mass spectrometry (SIMS) is also able to comprehend roughness/interdiffusion in nanoscopic multilayers by collecting and simple model-processing direct elemental depth distributions (depth profiles). The approach retains many advantages of XRR and TEM, while avoiding their major drawbacks. The system does not depend on a material's optical properties – as XRR does. It is destructive, but, unlike TEM, it does not require sample preparation and can analyse laterally over much larger areas.

The SIMS set-up, which has been put through its paces by researchers at Argonne National Laboratory, has great potential to become a standard characterization tool to provide all essential structural and chemical information for a diverse range of applications in nanotechnology.

Dual-beam system

In their study, the team used an advanced custom-made time-of-flight (TOF) SIMS instrument called SARISA to sputter depth profile a |MgO/ZnO|×8/Si multilayer grown by atomic layer deposition (ALD). SARISA features a unique gentle dual-beam approach to depth profiling (gentleDB), which combines a raster-scanned orthogonally incident direct current Ar+ beam of a few hundred eV for layer-by-layer shaving the multilayer, with an obliquely incident pulsed Ar+ beam for TOF SIMS analysis in the centre of revealed sub-surface, cyclically following every shaving increment.

These depth profiles were analysed by a well established mixing-roughness-information model, which yielded a 1.5 nm nanolayer interfacial roughness within the MgO/ZnO multilayer. Further analysis suggested that the 1.5 nm roughness corresponds to native/jig-sawed interfacial roughness rather than to interfacial interdiffusion during the ALD growth. This finding was cross-validated using XRR.

Back in the lab, the group is now using gentleDB SIMS to characterize a novel class of materials synthesized by ALD: multilayers of metal-sulfide binaries and Cu2ZnSnS4, and metal-oxide/metal-sulfide heterojunctions. The results will help to better explain the fundamentals of ALD and address application needs in modern photovoltaics.

The researchers published their work in the journal Nanotechnology.