Smartphone-based VOC sensing using nanozyme-enhanced colorimetric assays for on-site air quality

Thesis Abstract

The rapid and accurate assessment of volatile organic compounds (VOCs) in ambient air remains a critical challenge for urban health management, where conventional laboratory-based methods are limited by cost, latency, and operational complexity. This study addresses the need for an accessible, on-site sensing platform that combines smartphone technology with nanozyme-enhanced colorimetric assays to enable real-time VOC detection and quantification in diverse environments. The primary aim is to develop and validate a portable, user-friendly sensor system capable of delivering reliable VOC concentration estimates with smartphone-imaged colorimetric readouts, underpinned by robust data analytics and calibration models. Specific objectives include (1) synthesizing and characterizing nanozymes that catalyze colorimetric reactions in the presence of target VOCs to yield distinct and interpretable color changes; (2) integrating a modular microfluidic cartridge with a standard smartphone camera to capture time-resolved colorimetric responses; (3) constructing calibration frameworks that link image-derived color metrics to VOC concentrations using multivariate regression and machine learning approaches; (4) evaluating device performance in controlled laboratory trials with a panel of representative VOCs (toluene, acetaldehyde, formaldehyde, and benzene) at ppb–ppm levels; and (5) validating on-site performance in urban microenvironments, including industrial zones and residential neighborhoods, under varying environmental conditions. The methodology adopts an explanatory sequential mixed-methods design. In the quantitative strand, a cohort of 200 measurements across 20 VOCs is gathered from controlled gas chambers using nanozyme-enhanced colorimetric sensors, with parallel reference data obtained from GC-MS as the gold standard. A modular smartphone platform (Android-based, 12 MP camera, fixed illumination) is used to capture colorimetric signals from disposable test kits containing nanozyme–substrate systems optimized for select VOC reactivity. Calibration models are built using partial least squares regression (PLSR), multiple linear regression (MLR), and support vector regression (SVR), with model selection guided by root mean square error of prediction (RMSEP) and coefficient of determination (R²). In the field study, 60 on-site deployments are conducted across three city districts, collecting simultaneous smartphone readings and portable GC-MS validations, accompanied by environmental metadata (temperature, humidity, wind speed). Data quality checks include instrument drift assessment and outlier detection via robust PCA. Qualitative insights are obtained from 12 semi-structured interviews with field technicians to elucidate usability barriers and workflow integration, analyzed through thematic analysis. The expected findings indicate that nanozyme-enhanced colorimetric assays can produce reproducible, gymnastic color shifts correlating with target VOC concentrations, enabling predictive models with RMSEP below 15% of the measured concentration for most analytes in the lab, and below 25% in field conditions after bias correction. It is anticipated that PLSR and SVR models will outperform MLR in handling spectral and illumination variability, while smartphone-derived colorimetric features (?E, RGB histograms, and luminance) will serve as robust predictors. The study also aims to quantify the impact of environmental factors on measurement accuracy and to propose calibration strategies to mitigate drift. The anticipated contribution to knowledge includes (i) a validated, low-cost, scalable approach for on-site VOC sensing using nanozyme-driven colorimetry and smartphone imaging, (ii) a transferable calibration framework combining chemical sensor kinetics with image analytics and machine learning, and (iii) practical guidelines for deploying ICT-enabled environmental monitoring tools in urban settings, addressing data integrity, user experience, and regulatory alignment. The main conclusion is that smartphone-based VOC sensing leveraging nanozyme-enhanced colorimetric assays can provide timely, actionable air quality information with acceptable accuracy for screening and triage purposes, complementing laboratory analyses and supporting citizen science initiatives. Recommendations include extending sensor selectivity through multi-analyte nanozyme libraries, integrating cloud-based data fusion to harmonize smartphone readings with meteorological data, and standardizing illumination-controlled test kits to improve cross-device comparability. Further research should explore adaptive calibration strategies under long-term deployment and broaden the VOC panel to encompass semi-volatile compounds of emerging concern.