PROGRESS IN URANIUM SERIES DATING OF IMPURE CARBONATES BY THE TIMS TECHNIQUE: METHODOLOGY AND RESULTS FROM THREE SITES IN SEMI-ARID PENINSULAR INDIA

The potential of calcrete for understanding Quaternary climate change is well established, but developing a precise analytical method that can assist in constraining the stages and timescales of calcrete evolution has been a persistent problem. In this contribution, we examine how U-series dating by thermal ionization mass spectrometry (TIMS) can allow analytically precise dating of impure carbonate (calcrete). This method has been applied to calcrete occurrences developed on Precambrian metamorphic rocks and calcareous Quaternary deposits in semi-arid South India. The study consists in dating (1) identical laminar cappings at different locations (regional stratigraphic approach), and (2) calcrete profiles at various sites along a toposequence (geomorphic approach). The importance of acquiring a clear understanding of the calcrete microfabrics before micro-sampling the most homogenous carbonate phases for radiometric analysis is emphasized.

U/Th analyses were based on the isochron method (Rosholt-type diagram) to correct the initial amount of ²³⁰Th based on the abundance of ²³²Th. The few coeval subsamples required for a sample isochron age determination were obtained by the Total-Sample Dissolution method. We found the proportion of insoluble residue to be relatively high (4.7-15.99% wt) and for any given sample, the measured ²³⁰Th/²³⁴U ratios from the different subsamples could differ quite significantly. Two factors may explain this: (1) if the sample behaved as a closed chemical system, the observed variations in ²³⁰Th/²³⁴U ratios can be related to the variable amount of detrital material. In that context, it remains possible to distinguish the ²³⁰Th/²³⁴U and ²³⁴U/²³⁸U ratios within the pure carbonate phase. (2) If the sample was subjected to internal U remobilization, it is important to understand, based on mathematical models, how radioelement migration occurred and affected the age, and whether biological activity is involved. In the first case, U/Th results give a range of late Pleistocene ages, except for calcrete developed on limestone parent-rock in SE Tamil Nadu for which the period of validity of the U/Th dating method is exceeded (> 350 ka). In NW Tamil Nadu, laminar calcrete cappings were formed during the last glacial period under drier and windier climatic conditions than today.