Investigating age underestimation in the high dose region of optically stimulated luminescence using fine grain quartz

The age range of the single-aliquot regenerative-dose (SAR) protocol applied to quartz optically stimulated luminescence (OSL) has become the subject of debate as age underestimations for samples of Eemian age (ca. 125 ka) and older are increasingly reported. In order to investigate the origin of the age underestimation, the behaviour of fine grain quartz samples taken from two sites was compared. Sediment from both sites originates in the Alps, with those from Azzano Decimo, northeastern Italy, already displaying a significant age underestimation before reaching the Eemian, while ages for samples from Niederweningen, northern Switzerland, agree with expected ages well beyond the Eemian. Estimated dose (De) values for both sets of samples are up to around 400 Gy, and all meet the usual performance criteria for testing the reliability of the SAR protocol. De values determined using a sensitivity-corrected multiple aliquot regenerative-dose (MAR) protocol were within 10% of those determined using a SAR protocol for both sites, and so could not overcome the significant underestimation seen at Azzano Decimo. Lifetime calculations confirmed the stability of the signal well beyond the age range required at both sites, for both the initial integral as well as that used for an early background subtraction. The relative percentage contribution of the fast and medium OSL components to the initial integral remained constant in both the natural signal and similar sized laboratory generated doses. The quartz OSL from both sites displayed similar characteristics under all analyses and no indication was found for the possible origin of the age underestimation seen at Azzano Decimo. In the comparisons made between the dose-response of both laboratory and naturally generated signals in these samples, no evidence was found to suggest that the presence of the high dose linear region is restricted to the laboratory. The reliability of quartz OSL ages derived from this part of the curve however, remain in question, and require the support of independent age control until a performance criterion can be found with which to test it.