Embossed Colorimetric Sensor Array-Based Optoelectronic Nose for the Detection of Real Chemical Warfare Agents and Toxic Industrial Chemicals: A Practical Approach Using Off-the-Shelf-Reagents, Thermodynamic, and Kinetic Fingerprints

Abstract

In chemical emergencies, real-time monitoring of chemical warfare agents (CWAs) and toxic industrial chemicals (TICs) is essential for effective mitigation. Ideally, such detection requires the deployment of a large number of sensors capable of providing real-time information. Despite the advantages of colorimetric systems, such as ease of fabrication, low cost, high sensitivity, and low power requirement, no single platform currently exists that can detect a broad spectrum of CWAs and TICs. To address this gap, we propose an optoelectronic nose for the detection of real CWAs, including nerve agents, blister agents, choking agents, and blood agents, as well as TICs. Sensor units composed of mesoporous silica microparticles impregnated with colorimetric indicators were designed and prepared using off-the-shelf reagents. For nerve agents, 1584 sensor units were prepared by systematically varying the composition of their components and subsequently screened. The ‘heart’ of the system, an embossed colorimetric sensor array, was fabricated by integrating the optimized sensor units onto self-adhesive paper using a commercially available 384-well plate. A portable handheld reader device was developed, incorporating a micro-air pump, a 5 MP CMOS camera, a Peltier cooling system, a heating pad, and a processor for automated data acquisition and analysis. Using this platform, we generated thermodynamic fingerprints for twelve CWAs and TICs in the vapor phase at two different concentrations. The hierarchical cluster analysis demonstrated correct classification of all trials at low, high, and combined concentrations simultaneously. In addition, kinetic fingerprints were also incorporated to train a support vector machine (SVM) classifier with an enhanced database, achieving 100% classification accuracy and getting faster alarms. Importantly, all analytes except sulfur mustard and nerve agents (0.2–0.5 ppm) were detected below their immediately dangerous to life and health limits within 0.5–3 min of exposure. The handheld reader, with automated accurate classification capability and fast response time enabled by the kinetic database, makes our approach practical and field deployable. We believe that our work represents a significant step towards the development of the embossed CSA-based OE platform for on-site monitoring of CWAs and TICs in chemical emergencies.