Industry Perspective | Technology Focus Industry Perspective | Technology Focus

T he advent of high-quality-factor (high-Q) micrometre-scale optical resonators has led to the demonstration of high-fidelity optical sensors. These devices can be used for various types of sensing — mechanical, biological and chemical, for example — and often have far better sensitivity than conventional techniques. Microphotonic resonators could also be useful as uncooled thermal detectors, offering significantly better noise performance, smaller pixel size and faster response times than existing thermal detectors. These advantages make the prospect of a microphotonic thermal imager highly attractive. Most modern optical imagers are constructed of arrays of photon-detector elements known as focal plane arrays (FPAs). Photon detectors directly generate electron–hole pairs from incident radiation and provide excellent noise performance when the photon energy is much larger than the thermal energy, kBT (where kB is Boltzmann’s constant and T is temperature). However, for wavelengths longer than a few micrometres, photon detectors have to be cooled, often to cryogenic temperatures, to minimize thermally induced energylevel transitions. Alternatively, FPAs can be formed from thermal detectors, where the incident radiation generates thermal energy and a corresponding temperature shift. This shift is then sensed through a change in some physical characteristic of the detector element, be it mechanical, electrical, optical or otherwise. Although thermal noise is of course still problematic, fluctuations between energy-level states within a thermal detector do not typically degrade performance. The impact of thermal noise is therefore more controllable and roomtemperature operation of a thermal detector is possible, independent of the incident photon energy. The development of microbolometerbased FPAs, which rely on the temperature coefficient of resistance to sense temperature changes, has made it possible for thermal images to be recorded at room temperature. This breakthrough opened up infrared and long-wave imaging to a wide range of applications for which photon detectors do not exist or cryogenic cooling requirements limit their use. Yet, despite over two decades of development, the noise performance of room-temperature microbolometers is inadequate for imaging systems for which the ratio of focal distance to aperture size is substantially greater than one. Moreover, microbolometer array sizes remain smaller than one megapixel with pixels larger than 20 μm, significantly limiting both resolution and field of view. These limitations have spurred on efforts to develop alternative thermal imaging techniques. One such technique is thermal microphotonic imaging, which involves combining high-Q micrometre-scale resonators with extreme thermal isolation to ensure low-noise thermal detection. However, an FPA-based approach requires the production and integration of millions of such microphotonic elements on one chip, and the integration of an equally scalable readout technique. Here, we review noise sources inherent to thermal detectors, and present an approach for thermal microphotonic detection and one possible readout technique. Our goal is to inspire researchers to consider thermal microphotonic imaging as an approach with the potential to reach levels of performance previously attained only by cryogenically cooled detectors.