AI-Enabled Nano Digital Inline Holography for In-Flight Aerosol Observations

Current aerosol instrumentation can provide high-resolution particle size distributions down to few nm (~2 nm). However, these state-of-the-art techniques are expensive and still limited in resolving three-dimensional size distributions or providing real-time particle morphology and shape information, often requiring simultaneous operation of mass analyzers, mobility spectrometers and optical particle counters. In atmospheric research, aerosol shape and morphology are critical parameters for understanding aerosol-cloud-radiation interactions, air quality and health impacts. Here, we report recent progress in developing an AI-enabled Nano Digital Inline Holographic Microscope (Nano-DIHM) designed to transition from laboratory validation to aircraft-compatible prototype for in-flight aerosol monitoring.
The current benchtop prototype integrates a 405 nm coherent laser source, micron-scale pinhole spatial filtering, and a high-speed scientific CMOS camera configured for inline holography. Holograms are formed through interference between reference light and particle-scattered fields, enabling lens-free reconstruction of particle size and morphology. The compact optical path, assembled on a breadboard platform, has demonstrated stable hologram acquisition for particles deposited on substrates and initial tests under controlled flow conditions.
Numerical simulations were performed to quantify detection sensitivity as a function of particle diameter, laser power, exposure time, and airflow. Preliminary results indicate that illumination above ~60 mW enables detection in the ~200 nm to 10 µm range, with trade-offs between field of view and signal-to-noise ratio. Detection below 200 nm remains scattering-limited; however, AI-assisted DaPi™ reconstruction improves robustness for low-contrast holograms and partially coherent scattering. Microsecond exposure times mitigate motion blur at airflow velocities relevant to cabin sampling in flight.
Ongoing engineering efforts focus on stray-light suppression, long-duration optical stability, compact mechanical integration, and evaluation of flow-controlled versus semi-open sampling geometries. This work defines practical sensitivity limits while advancing a compact, morphology-resolved aerosol sensor for future aircraft deployment and real-time atmospheric submicron particle characterization, filling a gap in the palette of commonly deployed sampling instruments.