Smartphone-Enabled Electrochemical Sensors for On-Site Water Pollutant Mapping

Thesis Abstract

The rapid growth of mobile technologies and electrochemical sensing presents a transformative opportunity for real-time assessment of aquatic environments, yet current monitoring often relies on centralized laboratories and costly equipment that delay decision-making and limit spatial coverage. This study addresses the problem of spatially dense, on-site water pollutant mapping by leveraging smartphone-enabled electrochemical sensors integrated with low-cost, paper-based microfluidic platforms. The aim is to develop and validate a scalable, user-friendly system capable of delivering accurate, geotagged contaminant data to inform rapid management actions. Specific objectives include (i) designing and fabricating a robust, smartphone-compatible electrochemical sensor array for simultaneous quantification of heavy metals (lead, cadmium, and mercury) and key inorganic anions (nitrate and chloride) in surface and groundwater; (ii) calibrating the sensor platform against standard laboratory reference methods (ICP-MS for metals and ion chromatography for anions) to establish performance benchmarks; (iii) implementing a mobile data capture, geolocation, and cloud-based data fusion framework to enable near-real-time pollutant mapping; (iv) evaluating measurement uncertainty and repeatability under field conditions (n = 120 samples across 15 sites over two seasons); and (v) assessing user interaction, data quality, and operational feasibility in citizen-science contexts. A mixed-methods research design is adopted, combining quantitative sensor performance evaluation with qualitative usability assessment. The population comprises water samples drawn from streams, rivers, and groundwater sources within a metropolitan watershed. A stratified random sampling approach yields 120 samples, with 8 to 12 field replicates per site, to capture variability in physicochemical matrices. Data collection instruments include a smartphone-integrated electrochemical sensor kit with screen-printed electrodes and a microfluidic sample handler, calibrated under controlled laboratory conditions and deployed in situ by trained researchers and, in a pilot phase, selected citizen-scientist participants. Reference analyses employ inductively coupled plasma mass spectrometry (ICP-MS) for trace metals and ion chromatography (IC) for anions, ensuring rigorous validation of the portable platform. The data analysis plan comprises (i) regression analysis to model the relationship between smartphone sensor signals and reference concentrations, (ii) Bland-Altman plots to assess agreement and systematic bias, (iii) analysis of variance (ANOVA) to examine differences across sites, matrices, and seasons, and (iv) geospatial interpolation (kriging) to generate continuous pollutant distribution maps. A hierarchical mixed-effects model accounts for random site effects and fixed effects of matrix type and temperature. The theoretical framework integrates the Technology Acceptance Model (TAM) to interpret user adoption and the Theory of Planned Behavior (TPB) to explain engagement in data collection activities, complemented by the Sensorimotor Integration Theory to justify sensor calibration under variable field conditions. Anticipated findings indicate strong correlations (R2 > 0.85) between smartphone sensor readings and ICP-MS/IC results for metals and anions after matrix-matched calibration, with mean absolute errors within regulatory limits for drinking water in most matrices. The system is expected to demonstrate acceptable repeatability (RSD < 10%) and low field bias, with increased spatial coverage compared to traditional sampling networks. Geospatial maps will reveal pollutant hotspots and temporal trends across seasons, enabling targeted interventions. The study contributes to knowledge by providing a validated, low-cost, ICT-enabled framework for decentralised water quality surveillance, including a standardized calibration protocol, data fusion workflow, and an open-access dataset of ~1500 field measurements. It advances understanding of mobile electrochemical sensing performance in varying environmental matrices and informs policy on citizen-science integration for environmental monitoring. The main conclusion is that smartphone-enabled electrochemical sensors, when coupled with rigorous calibration and cloud-based data processing, can deliver reliable, scalable, and timely water quality information suitable for rapid decision-making. Recommendations include developing region-specific calibration kits, enhancing spectral and electrochemical selectivity to reduce matrix effects, ensuring data privacy and quality control in citizen-science deployments, and extending the framework to additional contaminants such as pesticides and organic micro-pollutants.