Development of a new spectral database of photo-induced pesticide compounds for an automatic monitoring and identification system.

Background: Due to the widespread use of pesticides, it is necessary to monitor surface water quality for environmental protection, industrial use and tap water production. Many analytical methods based on chromatographic separations and mass spectrometry detection can be used, but they are complex and expensive. They also require sampling and transport to the laboratory, which delays the results. We therefore need automatic monitoring systems that can detect and identify pesticides at low cost and with simple technology.

Results: As many pesticides are non-fluorescent, we have chosen the photoinduced fluorescence method for detection. However this method lacks of specificity. To improve this, we present a new fluorescence database of photoinduced pesticide compounds. It is designed in an original way to be lightweight, containing only four spectra of each pesticide concatenated into a single 'combined fluorescence spectrum'. It can therefore be used with a simple spreadsheet to identify pesticides by spectral similarity. The database is then combined with an automated pesticide monitoring system using photo-induced fluorescence to add identification capability. Analytical applications are carried out on-line to determine rimsulfuron and sulfometuron-methyl, with good analytical results. The identification process using the database is then self-tested against the spectra of these two pesticides, leading to correct results.

Significance: This work presents the first spectrofluorimetric database of photoinduced compounds of pesticides. It has been designed to be easy to use, even by an unskilled user. Combined with an automated monitoring system, it becomes the first to allow simultaneous detection and identification of pesticides in surface water. The results showed good identification results and good analytical performance.