| Summary: | Biological treatment of wastewater represents a sustainable alternative to physicochemical processes. Efficiency is however hindered by the low tolerance of toxic loads. Herein, combining segmented-flow microfluidic and a cell viability assay for real-time gross toxicity detection and quantification was explored. The reduction of resazurin to fluorescent resorufin by bacteria was used to assess the toxicity of model toxicants pentachlorophenol (PCP) and five heavy metals Cu2+, Hg2+, Zn2+, Ni2+, Co2+. A microfluidic chip was developed for use as a continuous microreactor, wherein 100 nL aqueous droplets are dispersed in a continuous oil phase. This platform can continuously generate sample droplets with sequential addition of bacteria and resazurin, allows determination of initial rates of resazurin reduction and dose dependent toxicity assessment with an overall response time of less than 5 minutes. The assay was performed under alternative conditions from traditional laboratory procedures to reduce the response time, including short toxicant exposure (< 30 s), stationary phase bacteria at high density (3 107 cell mL-1) and continuous mixing in moving droplets. A linear relationship was confirmed between cell density from 3.7 106 to 9.2 108 cell mL-1 and the resazurin reduction rate. Under optimized conditions, results were obtained within 3 minutes with better sensitivity than analogous microplate experiments. For instance, the apparent EC50 of PCP to Enterococcus faecalis was 6.3 mg L-1 in droplet format compared to 77 mg L-1 on microplate. The resazurin reduction rate increased with mixing in moving droplet up to a velocity of 4 mm s-1. Under these conditions, Michaelis-Menten behavior was observed with respect to the starting resazurin concentration and its reduction was attributed to the presence of two or more different enzymes with apparent Michalis constants of ~1 µM and greater than 140 µM respectively. The ratio between enzymes was dependent on bacterial growth phase and they were selectively inhibited by Hg2+ and Cu2+. Therefore, the apparent toxicity was highly dependent on the resazurin concentration as well as the growth phase and 10 µM resazurin was found optimal for Enterococcus faecalis in stationary phase. The assay in droplet format was enabled by a novel surface modification technique relying on the unique stability of silica sols at their isoelectric point to adsorb a dense silica nanoparticle layer on PMMA. The silica layer acted as primer for surface silanization with alkyl trichlorosilanes that rendered the surface hydrophobic eliminating the need for surfactants and their inherent complications. Hexadecane supported continuous operation >72 h while fluorocarbons limited operation to 3-4 h because of fouling resulting in droplet pinning. The difference was attributed to repulsive Van der Waals interactions only with hexadecane as suggested by theoretical analysis of Hamaker constants. Effect of partitioning of hydrophobic molecules between aqueous droplets and the continuous oil phase was predicted by Hansen solubility parameters and studied with PCP. Perfluorocarbons were demonstrated to minimize the extraction of PCP (LogDoil/water = -1.47) and hence underestimation of the toxicity compared to hydrofluoroethers (LogDoil/water = -0.66) and hexadecane (LogDoil/water = +0.45). Finally, suggestions are made to extend long-term performance with fluorinated continuous phases on the basis of Hamaker constant.
|
|---|