LWIR multispectral imaging using plasmonic filters integrated on a single thermal image sensor

Multispectral imaging involves capturing the same scene at different wavelengths using various narrowband filters stacked or integrated into digital camera sensors. This technology makes it possible to extract the additional information that a human eye or conventional camera fails to capture and thus has important applications in object identification, precision agriculture, and medicine. Multispectral imaging in visible wavelengths is readily possible due to the availability of digital imaging sensors and existing narrowband filter designs like metal-dielectric-metal films, dielectric films, or Fabry-Perot cavities [1-2]. Multispectral imaging in thermal longwave infrared (LWIR) wavelengths of 8-14 μm range has more advanced applications as they can see through fire, detect various gases, and investigate materials non-destructively through thermal signatures. However, conventional thermal image sensors can image in a single spectral band only. Thermal multispectral imaging is hindered by traditional filter technology where many layers of different materials are required for obtaining various spectral bands and limited wavelength tunability. On-chip integration of the infrared filters on the thermal image sensors to build a compact multispectral thermal camera is still an emerging area [3-5]. In the current work, we design and demonstrate a low-cost single sensor-based multispectral thermal sensor system composed of copper-based plasmonic imaging filter mosaic (multiple spectral filters are fabricated on a single substrate using only one lithography step and two deposition steps) integrated into an uncooled monochrome thermal sensor. The proposed work is mass-fabricable, scalable, and integrable, thereby leveraging next-generation LWIR thermal snapshot multi- and hyperspectral imaging.