Rapid online detection of nitrate and quantification of nitrate removal via capacitive deionisation (CDI)

Abstract

Abstract The increasing nitrate contamination in water poses significant environmental and public health challenges. With the rapid development of Capacitive Deionization (CDI) as a low-cost point-of-use (POU) solution to the problem, there was a rapid need for an online in-situ method for nitrate detection to support emerging POU technologies.

Due to these circumstances, it was necessary to investigate a new method for detecting nitrates, optimised for CDI’s requirements. Spectrophotometric based methods in small batch configuration (cuvette style) became our focus at the onset. A review of all the viable techniques reduced the design to the following common steps that needed to be converted into a viable miniature scale process: i) sample the liquid continuously, ii) pretreatment, iii) nitrate to nitrite reduction, iv) dye addition, v) colour development, vi) spectrophotometric measurement, and vii) dispense the waste.

With the aim of trialling a range of methods that had the potential of functioning as a continuous flow cell system within this design process, reverse engineering attempts were made using commercially available reagent sets from Palintest, and Hach Nitraver5 showing poor colour development. Thus, they were eliminated from further testing

Subsequent experiments employed the zinc reduction method, which offered a cost-effective means of nitrate concentration through its reduction to nitrite, followed by colorimetric analysis. Even though the zinc method demonstrated promising results, challenges such as sensitivity to reaction time, over-reduction of nitrite to ammonia, and variability in reproducibility limited its effectiveness. Additionally, side reactions and maintenance issues, including zinc oxidation and clogging, hinder its suitability for our continuous process.

Vanadium Chloride reduction method was finally chosen for its efficiency, rapid reaction time, and robustness under varying conditions. This method, combined with UV-VIS spectrophotometric analysis and MATLAB-based data processing, enabled real-time nitrate monitoring with over 93.05% accuracy. Key parameters such as residence time, flow rate, and heat transfer were optimised to enhance system performance. A reaction temperature of 85°C enabled complete reduction of vanadium species within 2 minutes, ensuring safe operation, consistent repeatability, and high precision. To address axial dispersion in flow-through systems, a novel sample-and-hold configuration was implemented, utilising seven parallel reactors with 17-second intervals to achieve discrete temporal resolution and accurate measurement. The system achieved a high-frequency sampling rate of 210 samples per hour (105 samples per CDI cycle), allowing near real-time tracking of nitrate fluctuations during semi-batch CDI operation.

A Stella model was developed to simulate nitrate removal via CDI, providing high resolution insight into nitrate adsorption and desorption dynamics. The model predicted transient nitrate responses that did not far exceed the system’s temporal resolution, which guided the implementation of a discrete sampling interval of 17 seconds. The integration of experimental validation with system simulation underscores the effectiveness of this combined approach in achieving accurate, real-time nitrate monitoring. The outcomes of this research contribute to the advancement of automated water quality analysis systems and offer a scalable, field-deployable solution for mitigating nitrate pollution in drinking water supplies.