Introduction Th U dating of speleothems using

Introduction 230Th/U dating of speleothems using highly precise and accurate state-of-the-art analytical methods, such as solution-based multi-collector inductively coupled calcitonin gene related peptide mass spectrometry (MC-ICPMS) has demonstrated the possibility of age inversions, i.e., the ages do not increase with increasing distance from top as expected from the stratigraphy (Dutton et al., 2009; Hoffmann et al., 2010; Lachniet et al., 2012; Scholz and Hoffmann, 2011; Scholz et al., 2012, 2014; Tolzmann, 2013). In this case, the ages scatter along the growth axis showing both younger and older ages than expected. A potential reason for such age inversions is a so-called open system, i.e., mainly loss or addition of U after deposition, which will lead to altered 230Th/U ages. Due to the limited spatial resolution of conventional solution MC-ICPMS, detection of small-scale age inversions may be impossible (Scholz et al., 2014). Micromilling significantly helps to improve spatial resolution (Hoffmann et al., 2009). Information at high spatial resolution using LA-MC-ICPMS may be important, for instance to date rapid, short-term climate oscillations observed in a speleothem proxy record. Furthermore, the temporal resolution of palaeoclimate proxy records depends on the growth rate of the speleothem. Thus, for slowly growing speleothems, the timing and duration of climate events can only be precisely determined if high spatial resolution dating is possible (Hoffmann et al., 2009). In-situ microanalytical methods have been presented for 230Th/U dating by using secondary ionization mass spectrometry (SIMS) (Bacon et al., 2000; Reid et al., 1997), laser ablation (LA)-MC-ICPMS (Eggins et al., 2005; Hoffmann et al., 2009; Potter et al., 2005; Stirling et al., 2000), and micromilling combined with MC-ICPMS (Drysdale et al., 2012; Hoffmann et al., 2009). SIMS has a high spatial resolution, however it suffers from severe matrix effects (Sodhi, 2004). Notably, LA-MC-ICPMS offers a new dimension of applications in Earth sciences with rapidly increasing use, for instance for isotopic measurements (Hf, Sr, Pb …) and UPb dating (Iizuka and Hirata, 2005; Jackson et al., 2001; Paul et al., 2011; Waight et al., 2002). Although precision and sensitivity are reduced compared to conventional solution analysis, highly spatially resolved (<100 μm) measurement can be achieved by LA-MC-ICPMS. In general, two main difficulties limit LA measurement of U-series isotopes in carbonate samples, such as speleothems. The first is the low U concentration (∼1 μg g−1 or lower) of most natural samples and the very low abundance of 230Th, which in turn depends on the U concentration. The second is the difficulty to get reliable matrix-matched external calibration materials with high U concentration and homogeneous known (230Th/238U) and (234U/238U) ratios. Only few studies have systematically applied LA-MC-ICPMS for in-situ 230Th/U dating of speleothems (Eggins et al., 2005; Hellstrom, 2003; Hoffmann et al., 2009). However, most of the previous studies have not focused on the investigation of small-scale age inversions.
Material and methods
Results and discussion
Conclusions LA-MC-ICPMS 230Th/U dating provides significant advantages over traditional solution MC-ICPMS dating: (i) nearly no chemical preparation is required, (ii) rapid analysis, and (iii) high spatial resolution are possible. LA-MC-ICPMS 230Th/U dating can be performed for carbonate samples with a U concentration of several μg g−1. For lower U concentrations and younger ages, the sensitivity related disadvantage would limit the application of LA-MC-ICPMS due to the lower precision and accuracy. In this work, a precision (2RSE) of better than 1.8% was obtained for single 230Th/238U measurement with 20–120 cps of 230Th signal intensity. An internal precision of ca. 9 ka at 2σ-level for an age of ca. 215 ka can be achieved. The reproducibility (2RSE) for 3–4 repeated measurements is within 4.5%. Age inversions occurring on a very small scale can be revealed, which is not possible using conventional solution MC-ICPMS analysis. Bracketing carbonate materials with high U concentration, nearly no 232Th, and well-analyzed isotope ratios (preferably in secular equilibrium) are important for accurate matrix-matched external calibration.