Atom probe tomography offers accurate, highly local compositional measurement, and for most materials that is an end unto itself. However, for materials that have been created via nanosecond processing, or have experienced extremely rapid phase changes, the high spatial resolution of APT also serves as a window onto events that occurred on the nanosecond timescale [1]. The ability to deduce the local change of physical states and chemical compositions both informs our empirical understanding of ultrafast materials processing and provides data that can test the limits of existing theory[2].

As an example we present the case of hyperdoped silicon. Originally envisioned as an advanced photovoltaic absorber material[3], hyperdoped silicon is a form of monocrystalline silicon containing dissolved dopants at levels orders of magnitude above their equilibrium solubility limits. To achieve this, the dopants are ion-implanted into the crystal, and then the damaged, doped surface region is rapidly melted and resolidified via pulsed laser melting, leading to a nanosecond form of liquid phase epitaxy and the regrowth of diamond-cubic silicon crystal with non-equilibrium dopant content. As might be expected, the processing window for this material is narrow and dopant-dependent; one of the most common ways for the process to fail is known as cellular breakdown, and until recently the dynamics of this process and the actual composition of the resulting material were poorly understood[4].

Cellular breakdown in this system results in a structure containing few-nanometer-scale regions of dopant enrichment, surrounded by areas of lower dopant concentration (although the majority of the material remains above the equilibrium solubility limit). The exact nanoscale morphology and composition of these areas were inaccessible until APT was applied; however, along with the material’s structure, it became possible from APT measurements to deduce the kinetic history of the material, based on composition gradients, partitioning at interfaces, degree of local dopant enrichment, and the presence of Rayleigh instabilities observed in the final structure. While this has great implications for the field of cellular breakdown in pulsed laser melting and hyperdoped silicon, it also shows that there is an opportunity for practitioners of APT.  The combination of materials whose processing occurs over very short timescales with the high spatiochemical resolution of APT allow us to use the APT dataset as a record of nanosecond material dynamics, and to ask questions of a material’s processing timeline that have not previously been answerable.

This work was performed in part at the Center for Nanoscale Systems (CNS), a member of the National Nanotechnology Coordinated Infrastructure Network (NNCI), which is supported by the National Science Foundation under NSF award no. 1541959. CNS is part of Harvard University.