pXRF and ICP-AES characterization of shipboard rocks and sediments: protocols and strategies

Recovered shipboard solids (rocks and sediments) may be characterized for elemental abundances on International Ocean Discovery Program (IODP) expeditions in several ways, using either the shipboard inductively coupled plasma–atomic emission spectrometer (ICP-AES) or a handheld portable X-ray fluorescence spectrometer (pXRF). These two instruments have overlapping capabilities in terms of the elements they measure but are designed to meet different analytical needs. During Expedition 366, we made extensive use of both instruments to conduct standard bulk elemental analysis of samples and in situ measurements on rock surfaces of cores. The following is a description of current shipboard measurement protocols for recovered rocks and sediments using these instruments, an analysis of the respective methodologies, and recommendations for best analytical practices. Portable X-ray fluorescence spectrometry Based on the success in using portable X-ray fluorescence spectrometry (pXRF) for core characterization during International Ocean Discovery Program (IODP) Expedition 352 (Ryan et al., 2017; Reagan et al., 2015, 2017), pXRF was used both to conduct near–real time characterization of recovered rock samples from cores and to analyze serpentinite rock powders and unconsolidated serpentinite samples during Expedition 366. A new pXRF—an Olympus DeltaX handheld instrument—was acquired by IODP for use during Expedition 366 and future expeditions. Compared to the original Fisher Niton instrument described in Ryan et al. (2017), this new instrument has overall expanded analytical capabilities. The Olympus DeltaX is a self-contained energy-dispersive XRF survey tool that includes data correction packages tailored to geological applications. The data correction methods are based on “fundamental parameters” methodology, which solves a series of nonlinear equations for each analyzed element. The parameters used in these equations comprise metrics for the X-ray source, fluorescence intensities, absorption coefficients, and absorption edge effects for each wavelength analyzed, together with parameters for sample geometry (e.g., van Sprang, 2000) and a Compton normalization scheme (Reynolds, 1963). The “geochemistry/soils” protocol used on the ship presumes a perpendicular sample geometry. The protocol analyzes for elements at two different filter settings to optimize results. Analysis of different core materials Generally, the pXRF instrument is operated by the shipboard scientist(s), typically from the Petrology/Core Description or Geochemistry teams, who are tasked with overseeing its use. The protocol for rock surface analyses used during Expedition 366 is as follows. Rock surface samples The primary shipboard use of the pXRF instrument during Expedition 366 was to conduct quick geochemical assessments of the cored material through direct measurements on rock surfaces of either workingor archive-half core pieces. For these measurements, rock samples that could be removed from the core without damage were placed in a specially made shielded sample analysis assembly (Figure F1). Samples that were too fragile to be removed were analyzed in situ using a shielded sleeve analyzer mount (Figure F1). For in situ measurements, a layer of 3525 Ultralene 0.16 mil (4 μm) thin film was placed over the core to prevent contamination and/or damage to the X-ray analyzer. In all cases, it is important that the geometry of the sample is consistent, surface parallel to and in close proximity to the analyzer face, to minimize atmospheric absorption effects and geometry-related losses. Selection criteria for choices of R.M. Johnston et al. pXRF and ICP-AES characterization of shipboard rocks and sediments materials to be analyzed and the specifics for making measurements with the Olympus pXRF are outlined in the Appendix. Sample powders The pXRF can also be used to quantitatively assess elemental abundances in powdered samples. Sample powder analyses were conducted using XRF powder mount assemblies, the use of which is outlined in detail in Reagan et al. (2015); a synoptic description of their use is included in the Appendix. For both rock surface and powder measurements, a powdermounted standard reference material (BHVO-2 was used during Expedition 366) should be analyzed with each set of unknowns to track instrument performance over time (Table T1). During Expedition 366, the total variation among individual measurements of the same sample was always well within the measurement uncertainties reported by the instrument and was often less than or equal to ±5%. Day-to-day variation in results for BHVO-2 indicated ±1% variability for higher precision elements and no worse than ±6.5% for trace elements over the course of the expedition (Table T2). pXRF calibration of geologic materials Although the Olympus pXRF presents data in concentration terminology (either parts per million or weight percent), it is important to recognize that these values are, for all practical purposes, merely intensity readings. It is necessary to calibrate the instrument against recognized standard reference materials for each element to be measured quantitatively (e.g., Ryan et al., 2017). During Expedition 366, calibration curves for the different elements measured via pXRF were determined using the same suite of standard reference materials used for inductively coupled plasma–atomic emission spectrometer (ICP-AES) analyses to improve inter-instrument data comparisons (Table T1). Powder mounts for each of the reference materials were analyzed to develop the working curves, as well as for periodic checks on instrument performance during pXRF measurements of unknowns. The working curves were developed in Microsoft Excel, and slope and intercept values from the working curves were used to calculate concentration results for unknowns (Figure F2). Both rock and powder samples were analyzed using powder-based working curves because past results indicated no differences in instrument performance between rocks and powders (Ryan et al., 2017; Reagan et al., 2015). The elements routinely measured via pXRF for quantitative determination during Expedition 366 were Ca, Mn, Fe (calibrated as oxides: CaO, MnO, and Fe2O3), Ni, Cr, Cu, Zn, and Sr. Ti, K and Rb, Zr, and V, which were analyzed quantitatively by pXRF during Expedition 352 (Reagan et al., 2015), were generally below pXRF detection limits in Expedition 366 materials. Sulfur was attempted, based on the possibility of gypsum in some recovered materials (see Table T1. Elemental abundance data for standard reference materials used via pXRF during IODP Expedition 366. * = standards used for ICP-AES data calibration, Expedition 366. Download table in CSV format. Oxides (wt%) Elements (ppm) MgO Al2O3 SiO2 MnO K2O CaO TiO2 Fe S Cr Ni Cu Zn Rb Sr BHVO-2* 7.24 13.47 49.75 0.17 0.52 11.40 2.73 12.35 150 284 120 126 102 10 393 BIR-1* 9.69 15.43 47.83 0.17 0.03 13.27 0.96 11.33 70 398 171 122 73 0 109 DNC-1 10.09 18.32 47.10 0.15 0.23 11.38 0.48 9.95 392 278 252 98 68 4 145 JP-1* 44.66 0.64 42.39 0.12 0.00 0.56 0.01 8.36 28 2,689 2,467 6 36 1 3 OKUM* 21.29 7.97 44.14 0.18 0.04 7.85 0.38 11.81 2,460 886 44 61 1 16 UB-N* 35.21 2.90 39.43 0.12 0.02 1.20 0.11 8.34 200 2,361 1,971 28 85 4 8 CGL-001* 38.22 0.48 38.54 0.08 0.02 0.68 0.02 8.00 2,780 2,300 DTS-2B* 49.40 0.45 39.40 0.06 0.01 0.12 0.01 7.76 17 15,500 3,780 3 45 2 1 DTS-1* 49.55 0.19 40.41 0.12 0.00 0.17 0.00 8.68 12 4,045 2,329 6 45 0 0 AII 92 7.64 15.84 49.89 0.17 0.16 11.13 1.78 10.20 234 107 64 86 1 130 SO-3 8.42 5.80 33.72 0.07 1.40 20.71 0.33 2.22 132 27 15 17 50 37 220 BCS-CRM 393 0.15 0.12 0.70 0.01 0.02 55.40 0.01 0.05 70 160 BCS-CRM 368 20.90 0.17 0.92 0.06 30.80 0.23 40 3 82 67 BE-N 13.11 10.03 38.21 0.20 1.41 13.93 2.61 12.77 308 357 268 70 121 47 1,381 NOD A-1 4.76 3.87 3.81 23.92 0.60 15.42 0.53 15.62 3,350 32 6,360 1,105 589 10 1,749 LKSD-4 0.92 5.90 41.60 0.08 0.81 1.80 0.38 4.09 9,900 27 32 31 192 110 LKSD-1 1.72 7.80 40.10 0.19 1.12 10.80 0.50 4.11 15,700 22 14 44 334 250 DNC-1 10.08 18.32 47.09 0.15 0.23 11.38 0.48 9.95 392 278 252 98 68 4 145 PACS-2 2.42 6.71 59.00 0.06 1.51 2.77 0.74 5.78 12,900 91 40 310 364 39 276 NKT-1 14.19 10.05 37.78 0.20 1.26 12.99 3.84 13.29 438 315 57 117 1,175 JR-1 0.11 12.86 75.43 0.10 4.41 0.65 0.11 0.93 11 3 1 2 30 244 29 MRG 13.55 8.47 39.12 0.17 0.18 14.70 3.77 17.94 610 430 193 134 191 9 266 Figure F1. A. Olympus DeltaX portable X-ray fluorescence analyzer (pXRF) in its storage holster. B. General purpose lead-shielded sample mount. C. Lead shielded sliding core mount. D. Barcode reader for core ID. E. Dell laptop driver computer.