K-AR Geochronology of Authigenic Clays to Date Paleoweathering

Abstract

Ancient soils, or paleosols, form at the intersection of the atmosphere, lithosphere, hydrosphere, and biosphere, and contain valuable information about their formation environments. Paleosols are used to study Earth’s past climate and are one of few continental paleoclimate proxies widely available in the rock record. Before making interpretations about Earth’s past using paleosols, we must know when the original soil formed. Although there are a few existing geochronometers that have been used to date paleosols, including U-Pb geochronology of soil carbonates and 40Ar/39Ar dating of supergene oxides, they are not suitable for most paleosols. In this study, we use K-Ar geochronology and illite age analysis (IAA) to date illite formed during low degrees of weathering of granitic parent material in soils and paleosols.IAA is used to date shallow, brittle faulting and diagenesis of sedimentary rocks. Two polytypes of illite are found in these environments, low-temperature (< 200 °C) 1Md illite and the high-temperature 2M1 illite. Because of the small crystal size of clays, it is impossible to separate pure fractions of these polytypes from each other. This means that bulk clay mixtures often include illite that formed in different environments at different times. In IAA studies, clay samples are separated into multiple size fractions. The proportion of inherited 2M1 illite in each size fraction is quantified by X-ray diffraction (XRD), and each size fraction is dated using K-Ar. As the proportion of 2M1 illite decreases, the age of the size fractions will approach the age of 1Md illite formation. The proportion of 2M1 illite in each size fraction is then plotted against its age, and a linear regression is used to determine the age of pure 1Md illite formation. To test the use of IAA on paleosols, I dated samples from both a modern, active soil, and from weathered granite beneath the late Proterozoic Great Unconformity. I measured a 1Md illite age of -2.52 ± 11.3 Ma on an active, semi-arid soil capping the Granite Wash Member of the St. David Formation near Benson, Arizona. This age is effectively 0 Ma, showing that 1Md illite is actively forming in the soil through pedogenesis. The measured 2M1 illite age is 396 ± 68.6 Ma, much older than the expected age (54.6 ± 0.2 Ma) of the parent Texas Canyon Quartz Monzonite. Using bulk elemental and 87Sr/86Sr measurements of a profile through the soil, I show that dust is present in the A horizon and is vertically mixed throughout the soil. This aeolian input is likely the source of this older 2M1 illite. I analyzed two samples of weathered Pikes Peak Granite beneath the Great Unconformity in Manitou Springs, Colorado to show the potential power of using IAA to date pedogenic illite. I determined that the Pikes Peak Granite was in the near-surface environment and weathered at 867 ± 9.6 Ma. This age comes from a clay sample with no 2M1 illite, but orthoclase from the Pikes Peak Granite was another inherited K-bearing mineral present in the clay size fractions. Diffusion modelling of 40Ar in illite shows that the age of this sample would not have been affected by thermally activated diffusion of 40Ar due to regional burial reheating. This age supports the hypothesis that the cause of exhumation and formation of the Great Unconformity in this location was the late stages of the assembly of Rodinia prior to 850 Ma leading to long term basement erosion. Finally, to better understand the limitations of IAA studies, I explored two statistical methods for calculating IAA ages. The first uses theoretical error propagation to combine average K, 40Ar*, 1Md illite, and 2M1 illite concentrations into K-Ar ages and proportions of 2M1 illite for each clay size fraction. The second uses a bootstrap resampling routine and Monte Carlo simulations to calculate these values. I then tested how varying several parameters in IAA data affects the uncertainty on the pure 1Md illite age calculated using a York-style weighted linear regression. I found that the most important parameters are the age uncertainty of each size fraction and the range of proportions of 2M1 illite between the size fractions. The most important size fraction is the one with the lowest proportion of 2M1 illite. These analyses can be used as a road map to improve the methodology in IAA, especially when working with samples that require higher age precision.