Computational nanosensing from defocus in single particle interferometric reflectance microscopy

Single particle interferometric reflectance (SPIR) microscopy has been studied as a powerful imaging platform for label-free and highly sensitive biological nanoparticle detection and characterization. SPIR's interferometric nature yields a unique 3D defocus intensity profile of the nanoparticles over a large field of view. Here, we utilize this defocus information to recover high signal-to-noise ratio nanoparticle images with a computationally and memory efficient reconstruction framework. Our direct inversion approach recovers this image from a 3D defocus intensity stack using the vectorial-optics-based forward model developed for sub-diffraction-limited dielectric nanoparticles captured on a layered substrate. We demonstrate proof-of-concept experiments on silica beads with a 50 nm nominal diameter.