Nitrate, a naturally occurring form of nitrogen in ecosystems, is generally produced through the nitrification process carried out by microorganisms. It also enters the environment through sources such as synthetic fertilizers, wastewater discharge, and agricultural runoff. Overexposure to nitrate poses risks to aquatic ecosystem and human health, potentially causing issues such as eutrophication and blue baby syndrome [1]. Therefore, frequent on-site monitoring of nitrate levels is crucial for safeguarding the environment and public health. Among various methods for nitrate sensing, electrochemical techniques are regarded as one of the most promising analytical tools for determination of nitrate due to their high sensitivity, fast response, portability, and ease of operation [2]. However, the broad adoption of electrochemical nitrate sensors for the point-of-care or the-point-of-need analysis remains constrained by challenges such as low stability, poor selectivity, and high maintenance requirements, which can lead to increased operational costs.
In this presentation, our recent research on developing electrochemical sensors for nitrate detection will be demonstrated, where machine learning was applied to enhance sensing performance and simplify sensor fabrication process. For instance, copper modified electrodes fabricated by pulse electrodeposition method were employed for voltammetric sensing of nitrate [3]. Machine learning was used to investigate the influence of copper deposition parameters including on-time, off-time, applied voltage and the number of cycles on the analytical performance such us current response and stability. The developed models accurately predicted key sensing properties and identified the most influential parameter affecting the peak current of nitrate reduction. This strategy reveals the link between sensor fabrication and performance without extensive laboratory testing and resource-intensive efforts, paving the way for commercialisation of voltammetric nitrate sensors across a wide range of applications. In addition, potentiometric detection of nitrate using solid-contact ion-selective electrodes (SC-ISEs) was investigated to address the selectivity and linear range limitations of traditional voltammetric nitrate sensors [4]. Novel ion-to-electron transducer materials for nitrate sensing, coupled with optimised machine learning models, will be presented to enhance stability and reproducibility while reducing costs.