Self-powered biosensors for water quality monitoring: sensor design and signal treatment

: (Alternative Format Thesis)

Abstract

Increasing water pollution is limiting the availability of safe water sources worldwide. Reducing pollution requires in situ, online and continuous monitoring of water quality, to identify the source and fate of these contaminants. To do so, improvements on the stability and autonomy of current sensing technologies are necessary. This PhD Thesis proposes to meet these goals using microbial fuel cell (MFC) sensors with a cathodic sensing element. Specifically, the detection of pesticides in water with a MFC with an algal biocathode receptor is evaluated. The current output correlates with the dissolved oxygen (DO) concentration in the catholyte, an indicator of photosynthetic activity in the algal biofilm. The MFC sensor could detect the EU limit concentration of 0.1 µg L-1 of atrazine in 2.6 h. Two electrode materials, graphite felt and indium tin oxide (ITO), were investigated to evaluate the effect of porosity and transparency in the performance. The MFCs with graphite felt showed shorter response times and better sensitivity, as a result of a ten times greater baseline output than with ITO.

To improve the portability, robustness and affordability of MFC sensors, a ceramic based, soil-MFC (CSMFC) with an algal biocathode was designed as an early warning device to detect pesticides in water. The detection of single toxic events of 0.1 µg L-1 of the herbicides Diuron and glyphosate was statistically significant (α = 0.01) based on changes in accumulated charge and accumulated variance, in five days before and after the toxic event. The correlation of the signal with DO reversed at 9 mg L-1 due to the detrimental effect of oxygen on the anode at higher DO. To solve this, the volume of the CSMFC was increased and the algal biocathode eliminated. The CSMFC was tested to detect hypoxia in water bodies by monitoring the cathodic reduction rate of oxygen. The CSMFC sensors responded instantaneously to changes in DO in a linear range from 0 to 6 mg L-1.

To reduce maintenance of the sensors in the field, a single-point calibration method was developed based on a design of experiments (DoE), to correlate the DO with the CSMFC signal, showing an error of 0.05 mg L-1.

The CSMFC sensors were also tested for early detection of eutrophic events. The photosynthetic pattern, an indication of algal activity, was captured in the signal output, with a correlation with DO in the catholyte of R2=0.85 in the day and R2=0.52 in the night, up to an algal optical density (Abs=750 nm) of 0.2. A screening DoE design concluded that nitrates, which are present in eutrophic waters, compete with oxygen for the cathodic reduction, reducing the sensitivity of the sensor to photosynthetic activity, particularly at low DO.

The long-term autonomy of the soil MFC signal output was also investigated. The presence of algae in the catholyte provides a continuous source of organic matter to the anode biofilm. The system sustained an increasing voltage from 1 to 15 mV continuously, for a year.

All experiments were carried out without feeding or maintenance of the sensors, showing the potential autonomy of these devices.

Date of Award	21 Jul 2021
Original language	English
Awarding Institution	University of Bath
Supervisor	Mirella Di Lorenzo (Supervisor), Petra Cameron (Supervisor) & Jan Hofman (Supervisor)