Geochemical analysis of soils and sediments, Coeur d'Alene drainage basin, Idaho: sampling, analytical methods, and results

This report presents the locations, descriptions, analytical procedures used, and an inter-lab comparison of over 1100 geochemical analyses of samples of soil and sediment in and downstream of a major lead-zinc-silver mining district in the Coeur d'Alene (CdA) drainage basin of northern Idaho. The samples fall in 3 broad categories: (1) samples from vertical profiles of floodplain soils in the valley of the main stem of the CdA River (767 samples) and of the South Fork of the CdA River (38 samples), (2) size fractionated surficial samples of sediment bedload within the channel of the South Fork of the CdA River (68 samples), and (3) samples from vertical profiles of sediment bedload within the channel of the main stem of the CdA River (260 samples).

Five different laboratories contributed geochemical data for this report. Four of the five laboratories employed analytical methods that require sample dissolution prior to analysis; one laboratory (US Geological Survey) used analytical instrumentation (energy dispersive x-ray fluorescence [EDXRF]) that is applied to pulverized samples. Some dissolution procedures use four acids (hydrochloric, nitric, perchloric, and hydrofluoric; Eastern Washington University [EWU] Geochemical Laboratory and XRAL Laboratories, Inc.), others use two acids (nitric acid and aqua regia; CHEMEX Labs, Inc.), and some use only concentrated nitric acid (ACZ Laboratories, Inc.). Most analyses of dissolved samples were done by Inductively Coupled Plasma - Atomic Emission Spectroscopy (ICP-AES) or by ICP - MS (Mass Spectroscopy). Some analyses for Ag and K were done by Flame Atomic Absorption (FAA).

Inter-laboratory comparisons are made for 6 elements: lead (Pb), zinc (Zn), iron
(Fe), manganese (Mn), arsenic (As), and cadmium (Cd). In general inter-laboratory correlations are better for samples within the compositional range of the Standard Reference Materials (SRMs) from the National Institute of Standards and Technology (NIST). Analyses by EWU are the most accurate relative to the NIST standards (mean recoveries within 1% for Pb, Fe, Mn, and As, 3% for Zn and 5% for Cd) and are the most precise (within 7% of the mean at the 95% confidence interval). USGS-EDXRF is similarly accurate for Pb and Zn. XRAL and ACZ are relatively accurate for Pb (within 5-8% of certified NIST values), but were considerably less accurate for the other 5 elements of concern (10-25% of NIST values). However, analyses of sample splits by more than one laboratory reveal that, for some elements, XRAL (Pb, Mn, Cd) and ACZ (Pb, Mn, Zn, Fe) analyses were comparable to EWU analyses of the same samples (when values are within the range of NIST SRMs). These results suggest that, for some elements, XRAL and ACZ dissolutions are more effective on the matrix of the CdA samples than on the matrix of the NIST samples (obtained from soils around Butte, Montana). Splits of CdA samples analyzed by CHEMEX were the least accurate, yielding values 10-25% less than those of EWU.