Microbial Fuel Cell Biosensor with Capillary Carbon Source Delivery for Real-Time Toxicity Detection

Abstract

A microbial fuel cell (MFC) biosensor with an anode as a sensing element is often unreliable at low or significantly fluctuating organic matter concentrations. To remove this limitation, this work demonstrates capillary action-aided carbon source delivery to an anode-sensing MFC biosensor for use in carbon-depleted environments, e.g., potable water. First, different carbon source delivery configurations using several thread types, silk, nylon, cotton, and polyester, are evaluated. Silk thread was determined to be the most suitable material for passive delivery of a 40 g L^-1 acetate solution. This carbon source delivery system was then incorporated into the design of an MFC biosensor for real-time detection of toxicity spikes in tap water, providing an organic matter concentration of 56 ± 15 mg L^-1. The biosensor was subsequently able to detect spikes of toxicants such as chlorine, formaldehyde, mercury, and cyanobacterial microcystins. The 16S sequencing results demonstrated the proliferation of Desulfatirhabdium (10.7% of the total population), Pelobacter (10.3%), and Geobacter (10.2%) genera. Overall, this work shows that the proposed approach can be used to achieve real-time toxicant detection by MFC biosensors in carbon-depleted environments.

Keywords: MFC; biosensor; capillary action; potable water; public safety; toxicity.

Figure 1

Illustration of thread-based capillary pump configurations with respect to the connection with a carbon source reservoir. Vertical connection with a carbon source reservoir (A), lateral connection with a carbon source reservoir (B), and modified lateral connection and flow control using a piezoresistance sensor–Hoffman clamp (C).

Figure 2

Schematic of the MFC compartment and measurement system (A) and tap water supply system, capillary pump, and MFC integration (B).

Figure 2

Schematic of the MFC compartment and measurement system (A) and tap water supply system, capillary pump, and MFC integration (B).

Figure 3

Capillary action aided flow rate test with different thread types connected to a vertical reservoir. Water flow rates using four different thread types and a no-thread control (A) and relationship between 40 g L⁻¹ acetate flow rate and head height in nylon and silk threads (B). Standard deviation did not exceed 5% of the mean values.

Figure 4

Capillary action aided flow rate test with different thread types connected to a lateral reservoir. Water flow rates for four different thread types (A) and replicate water flow rate tests with silk threads showing stability and possible effect of head height (B).

Figure 5

Passive amendment of organic matter concentration in tap water using a vertical reservoir– silk thread delivery system. Total chemical oxygen demand (tCOD) measured (A) and stabilization of flow rate through siphoning and elimination of head height effects (B). Standard deviation did not exceed 10% of the mean values.

Figure 6

Biosensor (S1 and S2) response to toxic spikes in tap water with amended organic matter. Changes in open circuit voltage of the biosensors to spikes in chlorine (A,B) and formaldehyde (C,D) concentrations. Concentration changes are indicated by arrows and dots with red dot corresponding to formaldehyde introduction and green dot to formaldehyde removal from the influent stream. Time scale in panel (C) is given in days to demonstrate long-term biosensor recovery.

Figure 7

Biosensors S1 and S2’s response to toxic spikes in tap water with amended organic water. Changes in open circuit voltage of the biosensors to spikes in mercury (A,B) and cyanobacterial microcystins (C) concentrations. Concentration changes are indicated by arrows and dots. Standard deviation of analytical measurements did not exceed 5% of the mean values.

Figure 8

Bacterial and archaeal genera of microorganisms identified in anaerobic sludge (biosensor inoculum) and carbon felt anode of biosensor S3 after 3 months of operation.