Pythoud, Mathieu2023-02-032023-02-032022-11https://hdl.handle.net/11299/252314University of Minnesota M.S. thesis. November 2022. Major: Earth Sciences. Advisor: R. Lawrence Edwards. 1 computer file (PDF); xi, 137 pages.In nature, the isotopic disequilibrium between uranium (U) and thorium (Th) isotopes (U-238, U-234 and Th-230) provide the potential to date materials and time processes over < 10^6 years. U/Th (and 234U/238U) dating relies on measuring relative isotopic abundances among the long-lived U-series nuclides, and the age error is mostly dependent on the measurement uncertainties of the minor isotopes, U-234 and Th-230. Currently, with traditional peak-jumping techniques on a secondary electron multiplier (SEM), measurements of δ234U and [230Th/238U] activity ratios reach precisions of ~1 ‰ (2SE). These measurements are ultimately limited by counting statistics; consequently, new capabilities require the use of a multi-collector inductively-coupled plasma mass spectrometer (MC-ICP-MS) in static mode. In this mode, the calibration among the detectors (especially the amplifier gain) is critical for accuracy. The major sources of error include the detector noise, tailing of major ion beams onto minor ones, and analytical blanks.In this thesis, I present a protocol for the rapid and precise measurement of U and Th isotopes on a Neptune Plus multi-collector inductively-coupled plasma mass spectrometer equipped with Faraday cups and a 10^13 ohm resistor. The signal/noise ratio using a Faraday cup and a 10^13 ohm resistor is a ~5-fold improvement over the 10^11 ohm resistors. For this resistor, the noise varies as a function of time: Noise (2SD) ≈ 1.7 µV ⋅ √(4 min) / √(measurement time) (10^11 ohm-units). Precise analyses of the Nd standard SRM 3135a enable cross-calibration to accuracies < 50 ppm between Faraday cups equipped with 10^11 and 10^13 ohm resistors. The U tail characterization and daily monitoring on the SEM during U runs typically impose uncertainties ~0.5 ‰ (Abundance sensitivity with the RPQ off: 234/238U ≈ 0.2 ppm and 237/238U ≈ 2 ppm) but < 0.1 ‰ with normalization to a bracketing standard. Analyses of commonly-referenced U-series standards and natural materials demonstrate that the new cup measurements with a 10^13 ohm resistor are more precise and accurate than traditional SEM measurements. Measurements with 10^13 ohm resistors replicate measurements with 10^11 ohm resistors, but with much smaller sample loads: ~20 ng U-238 (2 mV 234U+ × 3 min ≈ 1‰); ~50 ng U-238 (5 mV 234U+ × 3 min ≈ 0.5‰); ~150 ng U-238 (10 mV 234U+ × 5 min ≈ 0.2‰). 234U/238U instrumental errors of < 1 ‰ (± 2SE) can be achieved with samples < 10 mg coral or speleothem (~3 ppm U-238), < 100 mg speleothem (~300 ppb U-238), < 10 mL seawater (~3 ppb U-238), < 100 mL riverine/ground water (~0.3 ppb U-238). Samples as small as 1 L of seawater should yield 234U/238U errors ~0.1 ‰. 230Th/238U instrumental errors of < 1 ‰ can be achieved with samples as young as ~10 ka with 1 g (or as young as ~30 ka with 0.5 g) of coral or speleothem that contains > 1 ppm U-238. For samples with U-238 > 3 ppm, errors of < 1 ‰ can be achieved with samples < 0.1 g. The instrumental precision and accuracy enable sub-‰ level dating by the Th-230 method in the 800-ka range. The new protocol enables the simultaneous measurement of 238U/235U with precisions better than 0.05 ‰, a potential proxy for Earth processes. The characterization presented in this thesis suggests that this new protocol forms the basis for future and further improvements on instrumental precision and detection limits. Radiocarbon measurements of deep waters imply that the ocean circulates on timescales of 10^2 to 10^3 years. Theoretically, a 234U/238U radiometric clock could be used to constrain deep-water advection rates independently from C-14 measurements. However, in order to resolve U isotopic differences between ocean water masses, the estimation of the tailing effect of the 238U+ beam onto the minor 234U+ beam must see a 10-fold improvement. Devils hole (DH) calcite preserves a > 500 kyr climate record precisely dated by the U-Th clock. Discrepancies in the timing of glacial terminations in cores taken at various elevations below and above the modern water table in DH suggest an enrichment of Th-230 with depth in the water column. The magnitude of the age shifts and petrographic evidence on the footwall imply a reversible scavenging mechanism for Th. In order to investigate the Th-230 distribution in the modern aquifer, I apply the new measurement protocol to U isotopes in Devils Hole 2 (DH-2) groundwater while analyzing the Th isotopes with traditional SEM techniques. U isotopic composition is similar to that recorded by the calcite throughout the Pleistocene. Dissolved Th-230 concentrations in the water column could not be resolved from the procedural blank but amount to no more than 0.07 ag/g. Because the age shifts in the calcite records are only visible during terminations, whether a dissolved Th-230 concentration gradient existed at those times remains to be determined.en10^13 ohm resistorisotopesMC-ICP-MSthoriumU/Th datinguraniumThesis or Dissertation