Water contamination in rivers and lakes can harm human health and alter aquatic ecosystems. Current treatment techniques only address collected wastewater, although most pollution occurs from uncollected water (e.g., pipe leaks, stormwater). Water quality is often monitored manually at treatment plants with time-intensive, costly, analytical laboratory techniques. Several contaminants are monitored, including disease-causing pathogens toxic heavy metals and nutrients that cost $2.2 billion annually to remediate .
Existing approaches limit monitoring frequency; exhibit costly, dangerous lag times between sampling and measurement; and constrain monitoring to treatment plants. There is a demonstrated need for real-time, on-site water quality monitoring to reduce response times, anticipate pollution, and treat contaminated waters.
A variety of sensing methods have been developed to measure aqueous analytes via micro-scale sensor designs, nano-sensors, and optical methods. However, most conventional sensors are either too expensive or due to containments in the water lead to fouling. Recently, transistor-based sensor designs have been employed to measure analytes in water and urine, as well as gas chemistries . These sensors typically use metal-, nano- or polymer-catalysts in the gate region of the device and the current-voltage (I-V) characteristics are modulated as a function of surface charge. Selectivity is achieved through material choice. In addition, AlGaN/GaN high electron mobility transistors (HEMTs) with ion-imprinted polymers at the gate were used to detect trace amounts (1.97 μg L−1 detection limit) of phosphate anion . These transistor-based nano-sensors can be further adapted for complex water environments by choosing the appropriate water-tolerant materials.
We are an interdisciplinary team with unique backgrounds (micro-scale sensors + selective water treatment + micro-scale satellites). This project aims to create new deployable and immersible sensor hardware to better understand the water-energy nexus via cloud-based data management. We propose to develop deployable water quality sensors that combine recent breakthroughs in resilient micro-scale sensing (Prof. Senesky's lab), electrochemical water treatment (Prof. Tarpeh's lab) and low-cost micro-satellite/IoT communication platforms (Prof. Manchester's lab). We will (1) manufacture the sensors leveraging established micro-fabrication techniques; (2) integrate chemically selective membranes into the design of the sensor; and (3) couple the sensor to a communication platform that enables data storage in the “cloud”. Ultimately, the networked water-quality sensors developed in this program will realize location-specific, low-cost, predictive monitoring of water quality, which will protect human health and our fragile aquatic ecosystems.
DESCRIPTION OF PROPOSED TECHNOLOGY
Nano-porous membranes for selective water sensors
The proposed work builds on electrochemical stripping (Figure 1b), a novel nitrogen recovery process designed by Tarpeh. Over 94% of nitrogen was converted to fertilizer in urine and hydrothermal liquefaction effluent at lower energy than conventional ammonia stripping [3-4].
The group has demonstrated that nitrogen recovery exhibits lower costs, energy input, and emissions than conventional nitrogen management . This work pivots from electrochemical water treatment to electrochemical water sensing. In this project, nitrogen recovery efficiency, rates, and selectivity will be measured at progressively smaller scales to miniaturize selective electrochemical nitrogen sensing.
Design and fabrication of GaN-based sensors for water monitoring applications
In order to realize a functional water sensor, an underlying sensor platform will be interfaced with the nano-porous membranes in this work (Figure 3a). Established GaN sensor platforms will be leveraged for chemical-to-electrical conversion and ion-sensitive transistor-based electronics. The Senesky group will be responsible for sensor fabrication. Recently, GaN-based electronic platforms have shown high sensitivity to chemical ions and can be exploited for chemical sensor design . Two established GaN sensor platforms, shown in Figure 3b and 3c will be examined . These are established platforms already present in Senesky lab and have been used in previous projects by NASA . GaN sensors are more resistant to fouling in extreme conditions than conventional sensors.
Development of cloud-based IoT water sensor system
Manchester group will lead the prototype fabrication using sensors fabricated by Senesky lab. In previous Stanford work, the GaN-based sensor was integrated onto a micro-satellite built by Manchester that was deployed in March 2019 from the International Space Station . Manchester will develop custom electronics, enabling temperature control, calibration, low-power communication, and lower system cost for large-scale deployment (Figure 2). The basic architecture can be customized to meet the requirements specific to each application. Power conditioning and reversed polarity protection can be incorporated for added robustness. Board layouts can be developed to run individual sensors, or to run multiple sensors using the same CPU. The signal conditioning block includes transducing circuits, 24-bit digital-to-analogue converters, and components to enable the use of digital gains and offsets to accommodate a wide range of signals. The electronics provide digital output and LORA low power communication network for transmitting the data.
Real-time, low-cost, long-term water monitoring can be applied globally with myriad benefits, including reductions in energy consumed to treat water, early spill detection, and improved understanding of critical aquatic ecosystems.
This work builds on existing, complementary expertise at Stanford University and combines technology used for NASA to terrestrial applications. Our designs are built for large scale, low-cost manufacturing with reliability in mind.
 Zheng, DOI:10.1038/srep27728
 Tarpeh, DOI:10.1021/acs.est.7b05488
 Tarpeh, DOI:10.1021/acs.est.8b04035
 Tarpeh, DOI:10.1021/acs.est.7b02244
 Senesky, https://doi.org/10.1016/j.apsusc.2016.01.178
 Senesky, DOI:10.1109/JSEN.2009.2026996
 Manchester, http://www.arrl.org/news/kicksat-2-is-alive-and-being-tracked