Advances in integrated silicon photonics are enabling highly connected sensor networks that offer sensitivity, selectivity and pattern recognition. Cost, performance and the evolution path of the so-called ‘Internet of Things’ will gate the proliferation of these networks. The wavelength spectral range of 3-8um, commonly known as the mid-IR, is critical to specificity for sensors that identify materials by detection of local vibrational modes, reflectivity and thermal emission. For ubiquitous sensing applications in this regime, the sensors must move from premium to commodity level manufacturing volumes and cost. Scaling performance/cost is critically dependent on establishing a minimum set of platform attributes for point, wearable, and physical sensing. Optical sensors are ideal for non-invasive applications. Optical sensor device physics involves evanescent or intra-cavity structures for applied to concentration, interrogation and photo-catalysis functions. The ultimate utility of a platform is dependent on sample delivery/presentation modalities; system reset, recalibration and maintenance capabilities; and sensitivity and selectivity performance.
The attributes and performance of a unified Glass-on-Silicon platform has shown good prospects for heterogeneous integration on materials and devices using a low cost process flow. Integrated, single mode, silicon photonic platforms offer significant performance and cost advantages, but they require discovery and qualification of new materials and process integration schemes for the mid-IR. Waveguide integrated light sources based on rare earth dopants and Ge-pumped frequency combs have promise. Optical resonators and waveguide spirals can enhance sensitivity. PbTe materials are among the best choices for a standard, waveguide integrated photodetector. Chalcogenide glasses are capable of transmitting mid-IR signals with high transparency. Integrated sensor case studies of i) high sensitivity analyte detection in solution; ii) gas sensing in air and iii) on-chip spectrometry provide good insight into the tradeoffs being made en route to ubiquitous sensor deployment in an Internet of Things.
|