Varying matrix effects for elemental analysis identified from groundwater in the Barnett Shale

Abstract

The quality of analytical measurements can be influenced by the matrix of the sample of interest. The knowledge of the sample matrix allows for appropriate sample preparation, instrumental parameters, and quantification methods in an effort to achieve accurate results. Matrix matching can be difficult when sampling across various water sources with the possible introduction of unknown endogenous contaminants due to various degrees of land use, urbanization, and energy exploration, likely playing a factor. The degree of matrix effects in inductively coupled plasma–optical emission spectroscopy for nineteen metals from twenty groundwater samples across North Texas was assessed using a standard addition method. Matrix effects were characterized in collected groundwater samples (a) with no pretreatment, (b) after reversed-phase solid-phase extraction of possible organic contaminants, and (c) for a matrix of organic material retained on the reversed-phase sorbent. It was found that without any extraction treatment, only 54 % of all measurements experienced no matrix effect. After extracting unknown organic sample constituents, an increase to 74 % of measurements showing no matrix effect was recorded. Reconstituting the extracted organic sample matrix found this fraction to be a significant source of the deviated results with only 13 % experiencing no matrix effect. Results for the metals investigated are also discussed, along with correlations to water quality parameters such as turbidity, total dissolved solids, and salinity.