Applications of volatile fingerprint sensor arrays for rapid detection of environmental contaminants

摘要

The electronic nose (e-nose) technology has rapidly evolved in the past decade with a range of applications in the food industry, medical diagnosis, and recently environmental monitoring. This is the first time that this technology has been examined in detail for a range of specific environmental applications including: detection of low concentrations of bacterial, fungal and heavy metal contaminants in potable water; analyses of changes in the microbial activity of soil samples amended with heavy metals; and the detection of fungal contaminants in paper samples from library material. In some studies comparisons between different e-nose systems has also been carried out. The e-nose system based on a conducting polymer (CP) sensor array Bloodhound (BH114) was able to detect different bacterial species (Escherichia coli, Pseudomonas aeruginosa and Enterobacter aerogenes), initially inoculated in tap, reverse osmosis and bottled water with a concentration of 102 cells mL-1, after 24 hrs incubation. In the presence of low concentrations (0.5 ppm) of a mixture of heavy metal ions including cadmium, lead and zinc, the volatile pattern produced by the bacterial species was discriminated from that where no metal was added, probably due to a change in the microbial metabolism. The Bloodhound e-nose system was also used to detect fungal spores of Aspergillus fumigatus, Fusarium culmorum and a Penicillium species, inoculated in water samples. The initial concentrations were 102 – 105 spores mL-1. Good discrimination was observed between the control samples after 24 hrs incubation at 25oC. After 48 hrs incubation, it was possible to differentiate between the various spore concentrations present in water samples. Good reproducibility was achieved as results from different days were consistent and data could be pooled and combined for analysis. A comparative study was performed with three e-nose instruments, two of them had CP sensor arrays (Bloodhound (BH-114)); Neotronics (eNOSE 4000), and the third was a metal oxide (MO) sensor-based system, the NST 3220. The experiments carried out with the CP based-systems showed similar results when analysing water samples contaminated with 104 and 102 bacterial cells mL-1 after 24 hrs incubation. Both CP and MO based e-nose systems could differentiate control water samples from those contaminated with both bacteria and fungal spores. GC-SPME analyses confirmed the results obtained with the e-nose system of metal ions and bacterial cells in water samples. At-line studies were performed with the MO array-based system (NST 3220), for the detection of contamination episodes. E. coli and P. aeruginosa cells were used as contamination agents for tap and reverse osmosis sterile water, in two concentration levels, 102 and 106 cells mL-1. The samples collected downstream in a simulated watercourse, were analysed by the e-nose over a period of 1-2 hrs. The results suggested the potential of this technique to monitor episodes of bacterial cells at a low concentration in water samples. Experiments performed in soil samples artificially and naturally contaminated with heavy metal ions were analysed with the MO-based e-nose system. Results indicated that for artificially contaminated soil samples, after 40 days incubation the control samples could be discriminated from those containing 3 and 100 ppm of metal ions. For naturally contaminated soils, the sensor array was only able to separate samples containing a high concentrations of metal ions. Headspace analysis of cellulose-based agar showed good discrimination between Aspergillus terreus, A. hollandicus and Eurotium chevallieri, after 20 hrs incubation at 25oC. An increase in the incubation period to 40 hrs resulted in better separation between the control and fungal treatments. In situ studies performed on paper samples suggested that the e-nose was able to discriminate between control samples and paper inoculated with 103 fungal spores mL-1. The substrate was a determinant factor in the headspace analysis of microbial species. It was shown that the same fungal species produced different volatile profiles according to the growth substrate.