Northeastern Section - 54th Annual Meeting - 2019

Paper No. 21-4
Presentation Time: 1:30 PM-5:30 PM


CONREY, Richard M.1, BAILEY, David G.2, SINGER, Jared W.3, WAGONER, Laureen2, PARFITT, Benjamin4, HAY, John4 and KEH, Oliver4, (1)Geosciences Department, Hamilton College, Clinton, NY 13323, (2)Geosciences Department, Hamilton College, 198 College Hill Road, Clinton, NY 13323, (3)Department of Earth and Environmental Sciences, Rensselaer Polytechnic Institute, 110 8th Street, Troy, NY 12180, (4)Computer Science Department, Hamilton College, 198 College Hill Rd, Clinton, NY 13323

We have developed a new method for quantitative LA-ICPMS trace element analysis of geological materials. The method uses Li-borate fused glass pellets, with all samples and 16 international standards identically prepared. Pellets from all samples are analyzed first for major and trace elements by WDXRF. A custom Li-borate glass drift monitor is measured repeatedly for drift correction. Intensity data from 4-5 tracks per sample and standard are background corrected with Iolite software. HALite, a new software application, was developed to further process the LA-ICPMS data. Each analyte element is individually drift corrected using empirically determined least squared polynomial functions. Sample and standard means are then calculated from the 4-5 tracks. Mean counts-per-second values for the standards are divided by known analyte concentrations to determine instrument sensitivity (response) for each analyte, for every standard. Response values for all standards for each analyte are compared; analytes with highly correlated responses are identified and can be used as internal standards if their concentration in the glass is already known. For example, Yb responses are strongly correlated with Lu, Tm, Er, Ho, Dy, Y, Zr, and Hf responses; calibration models for Yb are constructed that employ analytes well-determined with WDXRF, in this case usually Y and Zr. Response correlations with the lowest standard errors are more heavily weighted in our calibration models, which can employ multiple internal standards. Sample concentrations are determined using the calibrated response models and the WDXRF concentrations for the chosen internal standard analytes. Validation experiments with international standards run as unknowns, using WDXRF data for internal standard concentrations, yield results with less than 5% relative error for most analytes. Advantages of this method are several and include easy sample preparation, calibration models based on diverse standard materials, and optimization of internal standard selection.
  • NE_GSA_2019_HALite.pdf (1.4 MB)