Novel “filter pellet” sample preparation strategy for quantitative LA-ICP-MS analysis of filter-bound sediments: a “green chemistry” alternative to sediment fingerprinting in Tanzania’s Ruvu River basin

Abstract

Purpose

The goal of this study was to develop a “green chemistry” alternative sample preparation method for elemental analysis of sediments using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) to avoid the need for acid digestions. The LA-ICP-PS data are used for sediment fingerprinting (sediment source determination) in order to locate sources of soil erosion and sediment transport into the Ruvu River Basin, the domestic water supply source for East Africa’s largest city, Dar es Salaam, in the United Republic of Tanzania.

Materials and methods

Suspended sediments were collected via filtration at multiple locations along the river for comparison to soil collected from potential sources throughout the watershed. The pellet method commonly used for soil requires approximately 0.5 g (Musil et al. Spectrochim Acta, Pt A: At Spectrosc 55:1747–1758, 2000; Bustamante et al. Spectrochim Acta, Pt A: At Spectrosc 57:303–309, 2002; Mikolas et al. Anal Bioanal Chem 374:244–250, 2002; Lee et al. Spectrochim Acta, Pt A: At Spectrosc 58:523–530Q1, 2003; Hassan et al. Spectrochim Acta, Pt A: At Spectrosc 63:1225–1229, 2008; Jantzi et al. Spectrochim Acta, Pt A: At Spectrosc 115:52–63, 2016). However, on average, less than 0.1 g of sediment was non-uniformly distributed and could not be separated from the filters on which they were collected. Therefore, a novel “filter pellet” method was developed in order to homogenize and pelletize these filter-bound specimens, requiring only 50 mg of sediment on an 86-mg cellulose nitrate membrane filter. A quantitative multi-element LA-ICP-MS analysis method was optimized for the filter pellets. A mixing model was then utilized to apportion the primary sources of sediment composing the suspended sediment samples.

Results and discussion

The analytical performance of the filter pellets was similar to that achieved with the soil pellets. An element menu of 17 elements was obtained with recoveries between 75 and 125%, precision of 10% or less (relative standard deviation), repeatability of 10% or less for duplicate analyses, and limits of detection in the ppb to low ppm range. A significant source of sediments was determined to be coming from an area of particular concern within the Mgeta Catchment of the Ruvu River Basin—the meta-anorthosite complex, which comprises less than 3% of the basin yet contributes up to 10% of the suspended sediment load in the Ruvu River.

Conclusions

A novel sample preparation strategy was developed in which sediments bound to filters were processed into pellets for analysis by LA-ICP-MS (and potentially LIBS). Quantitative multi-element LA-ICP-MS data were generated, of sufficient quality for use in sediment fingerprinting applications without the need for acid digestions. This paper describes the method development and LA-ICP-MS results of the filter pellet method and a brief exploration of the sediment fingerprinting results.