Improved determination of nitrate concentrations and flow rates in freshwater using ‘Diffusive Gradients in Thin-films’

Abstract

Nutrient pollution, particularly by nitrate, is a persistent issue in waterways. Numerous mitigation strategies have been developed to address this pollution, such as denitrifying bioreactors. The monitoring, however, of nutrient concentrations and mitigation strategy performance has relied on infrequent grab sampling, or expensive on-site equipment. There are numerous challenges with these techniques, such as sample preservation and transport, operational expense, and the ability to account for temporal changes in concentration. Diffusive Gradients in Thin Films (DGT) are capable of overcoming many of the challenges, however, to date their use has mainly been research based, and they are not widely used by monitoring authorities or communities. Diffusive Gradients in Thin Films are passive accumulative chemical monitoring devices, designed to be analyte specific. This thesis sought, firstly, to establish the efficacy of a modified DGT for the monitoring of nitrate, following deployment in chemically and physically complex denitrifying bioreactors. Secondly, the extension of the DGT methodology through (A) incorporation of colour reagents into the binding layer for rapid in-field analysis of nitrate concentrations, (B) developing a DGT method for the determination of flow rates, and (C) understanding the potential biases inherent in DGT methodology for determining concentrations and loads. The first study focused on the utility of DGT for the determination of denitrifying bioreactor performance – via deployment in two denitrifying bioreactors. DGT overcame many of the challenges of grab sampling, such as more easily accounting for temporal nitrate concentration variation, and reduced analytical requirements. Diffusive Gradients in Thin Films determined nitrate concentrations and removal rates were in strong agreement with high frequency grab sampling, and data collection via DGT was considerably easier than high frequency grab sampling. The DGT, however, still required in-lab analysis for nitrate. Development of colourimetric binding layers, as a hydrogel or liquid binding layer, might allow infield determination of analyte concentrations easily and accurately. Here, a chitosan-stabilised AuNP suspension, as a liquid binding layer was developed and tested for the colourimetric determination of nitrite concentrations (0 to 1000 mg L⁻¹) and masses (145 μg) in lab based colour development studies. Nitrate reduction to nitrite was achieved through the development of an Fe(0) impregnated poly-2-acrylamido-2-methyl-1-propanesulfonic acid/acrylamide copolymer hydrogel. The developments lay the foundations for further development of the colourimetric-DGT concept. A novel bromide selective DGT (Br--DGT) was developed, using the Purolite Bromide Plus anion exchange resin, which provided stream flow rates when combined with the constant rate tracer-injection method. The Br--DGT flow method was tested infield at the Mangaharakeke Stream. Flow rates determined by DGT were between -14.7 and 6.5 % of the flow independently monitored weir flow rate. In comparison, grab sample flow rates diverged by 5.5 to 58.9 % from the weir flow rate. Dynamic and coordinated changes in temperature, flow, and concentration, as potential sources of bias in concentration and load calculations using DGT and grab sampling were modelled. Large, dynamic, and highly correlated to concentration, temperature changes have minimal effect on the calculated DGT concentration. In contrast, as the correlation between concentration and flow increases the bias in calculated DGT loads becomes significant. This means that in systems where there is a high correlation of concentration and flow DGT may not be appropriate for determining loadings. This thesis had important implications for the use of DGT for determining nutrient concentrations and loads, and stream flow rates – and the monitoring of nutrient pollution. Firstly, it demonstrated that DGT were a useful tool for the determination of nitrate removal rates in the chemically and physically challenging denitrifying bioreactor systems. Secondly, it established the foundation for the in-field determination of nitrate concentrations. Lastly, it extended the DGT methodology for the determination of bromide concentrations and stream flow rates, and provided a greater understanding of the issues when using DGT to determine nitrate loads. This research opened the possibility of DGT for large-scale nutrient monitoring and determination of nutrient mitigation strategy performance, given the quality of DGT data and ease-of-use compared to grab sampling.