How much K is oK? – Evaluating different methods for K-concentration determination and the effect of the internal K-concentration on feldspar luminescence dating

Abstract. For luminescence dating of potassium-rich (K) feldspars, the presence of ⁴⁰K within the K-feldspar grains contributes significantly to the internal dose rate of the sample. Whilst it is common practice to determine uranium, thorium and potassium concentrations for external dose rate calculations, the vast majority of studies do not measure the internal K-concentration. Instead, most studies apply published K-concentrations of 10 ± 2 %, 12.5 ± 0.5 % or 13 ± 1 % (Huntley and Baril, 1997; Smedley et al., 2012; Zhao and Li, 2005) to their samples. The use of these high literature-based K-concentrations is usually justified by two assumptions: i) only K-feldspar grains with high K-concentrations contribute to the luminescence signal significantly, and ii) we reliably exclude Na-feldspar luminescence signals using filters, with a narrow transmission window around the K-feldspar emission peak ~410 nm. However, these assumptions may not apply to all samples and assuming too high K-concentrations might result in significant dose rate overestimation.

To investigate the effect of the internal K-concentration on the dose rate and the validity of the above-mentioned assumptions, we determine the K-concentration of a set of ten density-separated sand-sized K-feldspar samples of different geological origin and chemical composition using four different techniques. We quantify their K-concentration on the bulk level using a wavelength dispersive X-ray fluorescence spectrometer and a beta counter, and on the single-grain level using a micro X-ray fluorescence spectrometer and a scanning electron microscope (SEM) with an energy dispersive spectroscopy (EDX) attachment. We use the SEM-EDX single-grain results to calculate two luminescence-weighted K estimates. These two estimates, effective K and luminescent grains, connect the K-concentration of a grain to its luminescence signal intensity.

Our experimental results show that 1. There is a good agreement between bulk and average single-grain K-concentration measurements; 2. Single-grain K-concentrations within one sample can be highly variable across the entire physically possible range of K-concentrations for feldspars; 3. The blue luminescence emission is not dominated by K-rich feldspar grains. For most samples there are considerable differences between the published and the measured K-concentrations. These differences result in overestimation of the total dose rate of up to 34.6 % compared to dose rates calculated using measured K-concentrations. We therefore suggest to routinely measure the bulk K-concentration of each sample, complemented by additional single-grain measurements.