Step-Scan Interferometry for High-Fidelity Hyperspectral Nanoscopy

Fourier transform infrared nanospectroscopy (nano-FTIR) is an increasingly adopted characterization method that leverages decades of established knowledge in infrared spectroscopy at the nanoscale. It enables detailed characterization of composite materials and nanophotonic systems. Besides the rapid adoption and tremendous possibilities, the nanoscale nature of these measurements poses additional challenges for infrared spectroscopy. The current implementations of hyperspectral image acquisition at high spatial resolution suffer from significant artifacts due to thermal instabilities, which heavily affect positioning. As a result, the spatial and spectral fidelity of the measurements can be unreliable for long acquisitions. Here, we propose an alternative nano-FTIR measurement methodology based on step-scan interferometry and image registration. We demonstrate the spatial and spectral fidelity of our method and show how it enables the collection of high-quality, large data sets, making it possible to apply machine learning in photonics research to characterize nanoscale heterogeneity.