ESTIMATION OF ACCURACY AND PRECISION FOR TRACE ELEMENTS ANALYZED BY INDUCTIVELY COUPLED PLASMA MASS SPECTROMETRY (ICP-MS) AT THE PETER HOOPER GEOANALYTICAL LABORATORY, WASHINGTON STATE UNIVERSITY

The Peter Hooper Geo-Analytical Lab utilizes single stream sample processing for preparation of geologic materials for analysis of 29 major and trace elements by X-Ray fluorescence (XRF), and 27 rare earth and trace elements via ICP-MS.

Samples are reduced from hand samples to powder (~10μm), through the sequential use of presses, chippers, and shatter-box grinders utilizing tungsten carbide bowls. A rotary splitter is used to produce homogeneous aliquots for samples too large to grind in their entirety. Powders are weighed at a 2:1 ratio with lithium tetraborate flux (Li2B4O7) and fused at 1000° C in crucibles milled from high purity carbon. Fused beads are then reground, and an aliquot of this powder is removed for ICP-MS. The remaining powder is re-fused into a glass bead, polished on a series of diamond laps, and analyzed by XRF.

Powder aliquots for ICPMS are processed through an open-vial dissolution. Progressive evaporations are performed using concentrated nitric (HNO3), perchloric (HCLO4), and hydrofluoric (HF) acids. The tetraborate fusion and HF dissolution facilitates the destruction of all refractory phases, and the removal of the majority of Si, Li, and B as gaseous fluorides. The resulting solutions are diluted by a factor of ~3400, and analyzed on an Agilent 7700x quadrupole ICP-MS.

Batches of 10-24 samples are automatically subjected to a duplicate sample preparation and analysis in order to ensure method precision. These repeat analyses allow for long-term performance evaluation that spans a wide range of geologic materials and, along with periodic injections of certified reference materials (CRMs) into the sample stream, allow for the assessment of practical accuracy and precision for ICPMS analyses.

Using the repeat data, we calculated average relative differences of ~1% for La, Ce, Pr, Nd, Gd, Tb, Dy, Ho, Y, Hf, Sc, and Zr; ~2% for Sm, Eu, Er, Tm, Yb, Ba, Th, Nb, Ta, Cs, and Sr; ~3% for Lu; and ~4-6% for U, Pb, Rb. Precision was unrelated to observed concentration, except when limits of detection were approached.

Using GeoReM’s compilation of preferred values for CRMs, we also calculated average relative deviation from accepted values of <2% for Ce, Pr, Nd, Tm, Yb, Y, Hf, Sr and U; <4% for La, Gd, Er, Lu, Ba, Pb, Rb, Th, Ta, Cs, Sc, and Zr; <6% for Sm and Eu, and <10% for Tb, Dy, Ho, and Nb.