Matrix effects and improved calibration procedures for SIMS titanite U[sbnd]Pb dating

Abstract

Titanite is a U-bearing accessory mineral that can serve as a geochronometer. Compared with zircon, titanite is more likely to contain multi-stage fluid-induced growth. Thus, individual titanite grains may yield complex UPb ages that reveal the entire geological history of the rock. Secondary ion mass spectrometry (SIMS), which has high sensitivity and spatial resolution, is a good choice for UPb titanite analyses. A titanite UPb standard named YQ82 was developed with 206Pb/238U age of 1837.6 ± 1.0 Ma (2 standard error (SE), n = 4) and 207Pb/206Pb age of 1845. 0 ± 1.1 Ma (2SE, n = 4), as determined herein using isotope dilution isotope ratio mass spectrometry (ID-IRMS). However, UPb dating of YQ82 via SIMS yielded a bias of up to 14% compared to the ID-IRMS value, indicating a matrix effect. To correct for this matrix effect, six UPb titanite dating reference materials (RMs) were analyzed. Based on the correlations between the age biases and the Fe contents of the six RMs, a calibration protocol was established using the index of the Fe content (FeO+/IP, intensity of primary beam) and a power-law relationship between the calibrated Pb+/U+ and UO+/U+. Using this calibration protocol, the UPb ages decreased the age bias from approximately ±14% to ±1.4%. To test the protocol, a titanite sample (BMB108) with a low Fe content (Fe2O3, 0.4 wt%) was calibrated using MKED1 (Fe2O3, 1.5 wt%) as the RM. Without the calibration, the apparent SIMS 206Pb/238U age was 10% younger than the thermal ionization mass spectrometry (TIMS) 206Pb/238U age (1890 ± 4 Ma). Using the calibration method, the apparent SIMS UPb age was 1898 ± 14 Ma (2SE, n = 19), which is consistent with the TIMS age. This study shows that a considerable matrix effect related to Fe content is present in the SIMS titanite UPb dating results. The calibration protocol developed herein improved the SIMS titanite UPb dating accuracy, yielding a more accurate thermal history of the samples.