On the feasibility of near infrared spectroscopy to detect contaminants in water using single salt solutions as model systems

Abstract

This research work evaluates the feasibility of NIRS to detect contaminants in water using single salt solutions as model systems. Previous research has indicated the potential of near infrared spectroscopy (NIRS) for detecting solutes in water; however, a comprehensive investigation of the limit of detection of this technique has not been carried out. Near infrared transmittance spectra of aqueous salt solutions in the concentration range 0.002–0.1molL−1 (equivalent to 117–13,334ppm or 0.0001–0.01% mass/mass) were investigated. The first overtone region of the near infrared spectrum (1300–1600nm) was found to be the most effective wavelength range for prediction of salt concentration in aqueous solutions. Calibration models built using this wavelength range and employing the extended multiplicative scatter spectral pre-treatment resulted in root mean squared error of prediction values ranging from 0.004 to 0.01molL−1. The limit of detection (LOD) was estimated to be of the order of 0.1% (mass/mass) or 1000ppm. Within the framework of Aquaphotomics, it was possible to examine the effect of different salts on the NIR spectra of water in the first overtone range. Our results were confirmed through test experiments at various geographical locations employing dispersive and Fourier transform type NIRS instruments.