Single-Aliquot Hematite (U-Th)/He Dating: Preventing U Loss During Laser Heating by Increased O2 Partial Pressure

(U-Th)/He dating in which both parent and daughter are measured on the same aliquot is the preferred approach when sample sizes are limited and when the parent isotopes may be heterogeneously distributed. The single aliquot approach to hematite He dating has been used to quantify the time-scales of lateritic weathering, fault activity, and the development of soils and paleosols. In this method, hematite aliquots are heated to >1000°C to completely degas He. However, there is abundant evidence that at these high temperatures U can be lost, leading to erroneously high (U-Th)/He dates. This phenomenon also occurs in goethite. We performed experiments on multiple aliquots of several hematite specimens to characterize the degree of He extraction and U loss as a function of temperature. We used four different hematite samples, which have reproducible two-aliquot (U-Th)/He as well as (U-Th)/Ne dates between 130 Ma and 1760 Ma. Samples in Pt tubes heated by a laser require temperatures of at least 1100°C for complete He extraction, but the temperature above which major U loss occurs is 1000°C. This is obviously problematic for accurate single-aliquot hematite dating. We use Attenuated Total Reflection Fourier-Transform Infrared Spectroscopy (ATR-FTIR) to monitor phase changes that result from heating. Loss of U is correlated with reduction of hematite (Fe⁺³Fe⁺³O₃) to magnetite (Fe⁺²Fe⁺³Fe⁺³O₄). Published phase diagrams and our experimental data demonstrate that the onset temperature of this phase change can be raised from <700°C in vacuum to approximately 1300°C in an oxygen partial pressure of 80 mbar. We show that samples can be completely outgassed of He at 1250°C without U loss when heated in an oxygen-rich atmosphere. This makes it possible to obtain reproducible and accurate single-aliquot (U-Th)/He dates from hematite samples, even when the hematite is highly He-retentive. We outline a method for implementing this procedure. We suggest that the same reduction phenomenon accounts for U-loss in goethite during vacuum heating.