Thermoluminescence measurements of trap depth in alkali feldspars extracted from bedrock samples

Abstract

Various measurements of thermal trap depth are evaluated for K-feldspar grains extracted from a bedrock sample. The initial rise method and the various heating rates method yield consistent results for both the natural signal (E = 1.23 and 1.16 eV, respectively) and for a regenerative dose of 64 Gy (0.83 and 0.78 eV). For the fractional glow curve, apparent E-values range from 0.39 eV to a plateau around 1.50 eV. The highest values for the natural and regenerative signals are obtained using the newly-developed post-isothermal TL (pI-TL) method wherein the isothermal loss curves (gotten by subtracting TL curves obtained after different preheat durations) are fitted in the initial rise region on an Arrhenius plot. For a dose of 12.8 Gy, this method measures apparent E-values ranging from 0.73 eV to a plateau near 1.84 ± 0.06 eV. We repeat this analysis on three additional feldspar samples (two perthites and a high albite) to get a mean value of E = 1.86 ± 0.03 eV. The same analysis of natural aliquots of the K-feldspar sample yields similar results, with the two highest E-values at 1.81 and 1.86 eV. The kinetic order does not systematically vary with isothermal holding temperature or duration but remains relatively constant at 1.6 ± 0.3 (regenerative dose) and 1.5 ± 0.5 (natural dose). The apparent frequency factor, measured assuming a single E-value of 1.86 eV, decreases systematically (∼1023−1012s−1) with hold temperature and duration, a result which is consistent with a thermally-activated, distance-dependent tunneling model for feldspar thermoluminescence (i.e., a single trap depth and a continuum of apparent frequency factors). Frequency factor values measured following identical isothermal treatments are comparable between the natural and regenerative post-isothermal TL curves. By contrast, if different E-values are assumed, the apparent frequency factor values appear stochastic. Finally, it is speculated that the plateau of pI-TL E-values may be interpreted as the thermal depth of the main dosimetric trap measured with IRSL protocols.