Resolving Uncertainties in the Quantification of Trace Elements within Organic-Rich Boreal Rivers for AF4-UV-ICP-MS Analysis.

Over the past few decades, asymmetric flow field-flow fractionation (AF4) has emerged as a robust technique for the separation of colloid-associated trace elements (TEs) in aqueous samples. Nevertheless, little is known about potential artifacts and how to control them when measuring the concentrations of colloid-associated elements at low (μg L-1 ) or ultralow concentrations (ng L-1 ) using AF4-UV-ICP-MS. Water from a boreal river was selected as a challenging test material due to its high concentrations of dissolved organic matter (DOM) and Fe-rich colloids. These colloids are expected to be significant contributors to artifact occurrence, even in a metal-free, ultraclean laboratory. The results show that the adsorption of Mn, Co, Ni, Cu, and Pb onto acid-cleaned, non-channel surfaces (such as connection tubing and autosampler) accounted for up to 48% of TE loss. These losses on non-channel surfaces also represent potential sources of cross-contamination for Co, Ni, Cu, and Pb. New, uncleaned poly(ether sulfone) membranes are also sources of contamination for Ni and Cu. Analytical bias may exist in the measured concentrations of TEs, primarily due to the potential carryover of weakly adsorbed TEs (e.g., Ni and Cu) on the system surfaces by colloids in the samples (e.g., DOM). On the other hand, colloids in the samples can also act to gradually remove contaminants from the surfaces. For these types of DOM-rich waters, preconditioning the AF4 system using 40 mg C L-1 of Suwannee River Natural Organic Matter (SRNOM, pH = 7) is recommended to mitigate the impact of membrane fouling and carryover. A comprehensive strategy for minimizing instrumental artifacts is presented and discussed.