Use of near-infrared reflectance spectroscopy in prediction of heavy metals in freshwater sediment by their association with organic matter

The application of the rapid, non-destructive, cost-effective technique of near-infrared reflectance spectroscopy (NIRS) for the prediction of heavy metal concentrations in freshwater sediment was explored. Sediments were sandy and highly organic littoral samples from a Precambrian Shield lake (37-ha surface area)in northwestern Ontario. Sediment metals were at background concentrations except for Cd, which was augmented by 6 yr of experimental addition of Cd to the lake. Samples were scanned between 1100 and 2500 nm using an NIRSystems Near-infrared Model 6500 spectrophotometer. R-2 between NlR-predicted and chemically-analyzed metal concentrations were Cd, 0.63; Cu, 0.91; Zn, 0.93; Pb, 0.81; Ni, 0.88; Mn, 0.93; and Fe, 0.86. Wavelengths from partial least squares regression analysis associated with most of the variance in heavy metal concentrations were attributed to protein, cellulose, and oil. The first PLS factor accounted for 75-95% of the variability in the data for each metal, except for Cd (50%). The different behavior for Cd was attributed to its much shorter geological time in the lake and to its proportionately greater association with inorganic ligands as compared with the other metals. This study demonstrates that the prediction of heavy metal concentrations in freshwater sediment by NIRS is feasible.