M. Duval is currently the recipient of an Australian Research Council Future Fellowship (FT150100215). We thank Sanda Balescu for her careful reading of the manuscript that improved the clarity.
1The Electron Spin Resonance (ESR) dating of optically bleached quartz grains extracted from sediments has progressively become an important tool for archaeologists, geologists and geomorphologists because this technique is applicable to very diverse archaeological settings and depositional environments (fluvial, marine or aeolian) (e.g. Tissoux, 2015; Duval, 2016). Since the first geochronological application by Yokoyama et al. (1985), the number of publications reporting ESR results likewise the demand for ESR dating studies have considerably increased over the last decades. The aim of the present short paper is to provide a sampling guide for non-dating specialists (usually archaeologists, geologists, geomorphologists or palaeontologists) who need to collect samples on their own for ESR dating purposes. Although the recommendations provided below are relatively straightforward, the involvement of an ESR geochronologist before any fieldwork campaign is nevertheless strongly recommended.
2The ESR method applied to optically bleached quartz grains extracted from sediments measures the time elapsed since the last exposure of the detrital grains to sunlight (phenomenon called optical bleaching). Once quartz grains are buried and sheltered from sunlight, intensity of the ESR signal associated with paramagnetic centers present in quartz (e.g. Aluminium, Titanium, Germanium centers) naturally grows with time due to its exposure to ionizing radiations emitted by radioisotopes located within the sample and the surrounding sediment, as well as from incoming cosmic rays. The intensity of this ESR signal is then directly proportional to the amount of radiation received by the sample during its geological/archaeological history (DE: equivalent dose in gray, Gy), the environmental dose rate (D in Gy/ka) and the time (t in ka) elapsed since the last exposure to sunlight. An ESR age is therefore calculated by dividing the equivalent dose (DE) by the dose rate (D) (Grün, 1989). This age represents the time elapsed since deposition, not the formation of the quartz mineral. However, if bleaching is incomplete a residual dose added to the equivalent dose will contribute to overestimate the age. An insufficient bleaching is usually linked to the conditions of transport and deposition of the quartz grains (Voinchet et al., 2015). The successful application of the ESR dating method is therefore highly dependent on the geological context as well as the sampling conditions in the field (sampling strategy, type and quality of the material sampled).
3Material selected for ESR dating should fulfil the following criteria: adequate resetting of the ESR signal (optical bleaching), ability to characterize the radioactive environment surrounding the sample (dose rate) and lack of evidence for post-depositional mixing (bioturbation) or reworking (colluvial processes).
4To assess whether the sediment was sufficiently exposed to sunlight before its burial, the depositional environment and mode of sediment transport must be carefully characterized because bleaching and sediment transport are directly linked. According to Voinchet et al. (2015), sands carried out by water are usually better bleached at the time of their deposition than aeolian sands. Moreover, within the water transportation mode, the clearest transport environments should be preferred. Therefore, sands transported by clear water reveal being the most appropriate for ESR analyses. It means that it is better sampling pure sands than silty sands. However, even if bleaching is not complete at the deposition time for fluvial, marine or aeolian sands, the low values of the residual dose observed does not prevent dating these sediments by ESR (see further details in Voinchet et al. 2015, table 1). Indeed, the size of the quartz grains has a direct impact on several correction factors that are used for evaluating the dose rate such as α and β dose attenuations (Aitken, 1985; Bell, 1979; Duval et al., 2015). Grain size and bleaching levels may also be closely correlated. Whatever the transportation mode (fluvial, marine or aeolian), very fine grains (62-120 µm) and coarser grains (0.5-1 mm) are often less exposed to sunlight during the transport and lead to poor conditions of optical bleaching. Consequently, the intermediate particle size (100-200 µm) fraction is usually considered as being the most appropriate for ESR analyses in laboratory because it seems to be better bleached at the time of deposition (Voinchet et al., 2015). Finest and coarser grain size fractions can also be used for dating by ESR but taking into account that the residual dose is higher than for the intermediate particle size. Loess is a special case in which an overestimation of ages is systematically observed due to insufficient bleaching (turbulent transport clouds preventing sunlight contact). In the field, sediment should be examined for evidence of disturbance such as bioturbation (from roots or animals), pedogenic processes (clay illuviation or segregation) or post-depositional reworking because it can mix grains of different ages in a sedimentary profile or alter dose rate conditions over time, respectively (Bateman et al., 2003; Rink et al., 2013). Depending on the paramagnetic centers used (Al or Ti), ESR dating method can be applied on sediment deposited between 50 ka and several million years (Ikeya, 1993).
5For the ESR age calculation, ESR dating laboratories usually require two separated sediment samples for each sample: a light-shielded sample used for the equivalent dose (DE) determination and a bulk sample of the surrounding sediments for determining the dose rate (D) and evaluate the actual water content. Fig. 1 shows all material needed for sampling.
6Sampling conditions are very similar to those for Optically Stimulated Luminescence (OSL) dating, although it should be mentioned that ESR signal are much less light sensitive than the OSL signals (see fig. 1 in Duval et al.., 2017). Samples for DE analysis must be collected in a light-proof container such as pre-sharpened (only sharpened on one end) PVC or metal tubes of about 20 cm in length by 6.5 cm in diameter. Smaller tubes can of course be used, but one should keep in mind that the amount of pure quartz required for ESR measurements is of ~3-4 g, i.e. somewhat higher than that for OSL dating (see an example in section 6.5 from Duval et al.., 2017). A metallic cap may be used so that the tube can be easily hammered horizontally into the outcrop. This should be done after a previous cleaning of the section of ~5 cm in order to make a freshly exposed surface (fig. 2). Following the sample collection, both ends of the tube should be sealed with tape to prevent light exposure and loss of sediment. Clearly label the sample (using preferably a permanent dark-colored pen) by indicating for example the acronym obtained from the name of the site, year of collection and sample number (e.g., Atapuerca, 2017, sample 24: ATA1724). Finally, the tube (sealed and labelled) should be introduced into an opaque ziplock bag labelled with the same code. If the sediment is too hard or compact to insert a tube, then other sampling techniques can be used. For example, a block of sediment can be carved and securely wrapped with aluminium foil and tape. Then it must be placed in opaque plastic bags or bigger containers for transport to the laboratory where it will be prepared under controlled light conditions. In the case of coarse-grained alluvial deposits where sand lenses are too thin to sample with a tube or a block, the sandy matrix within the gravel (or loose sediment) can also be collected in a light-proof container under an opaque plastic cover.
7The external dose rate (D) evaluation is based on the determination of α, β, γ and cosmic components. α and β dose rates should preferentially be obtained from the raw sediment collected at the sampling spot, while the gamma dose rate would ideally require an in situ measurement, although it may not always be possible. In this context, bulk sediment samples for D analyses and moisture content should be collected from the ESR sampling site (tube hole) or from a 30 cm radius sphere around the sample (fig. 2). Radioelement concentrations or activities may be derived from various techniques, such as total β counters (requiring a few grams of raw sediment), ICP-MS (~150 g) or High Resolution Gamma Spectrometry (approximately 100-150 g is sufficient). Samples should be bagged and clearly labelled using the same acronym to that ESR sample. In addition, the evaluation of the in situ gamma dose rate can be done using either a field portable gamma spectrometer (fig. 3) (Mercier & Falguères, 2007) or TL-OSL (Thermo and Optically Stimulated Luminescence) dosimeters. Dosimeters have to be inserted at least 30 cm deep into the section for between six months and a year. Consequently, the site should be accessible and secured to avoid any perturbation or loss during that period. Dosimeters can be capsules of only 10 mm diameter, which minimise the disruption/damage to the site or outcrop. Another advantage is that the gamma dose rate registered over such a long period is also by definition indirectly taking into consideration any fluctuations of the water content of the sediment with time. In contrast, the field portable gamma spectrometer is more frequently used because the measurement can be completed almost immediately (around half an hour usually). It is therefore possible to carry out several measurements within a day. However, depending of the size of the probe (usually with a diameter of 3.8 or 5 cm), a hole of at least 5 cm diameter by about 30 cm deep is required to ensure an in situ gamma dose rate measurement. In both cases, measurements must be carried out at the exact ESR sampling spot. Furthermore, information on the latitude, longitude and elevation of the site and burial depth of the sample (fig. 2) should be reported in order to calculate the cosmic contribution to the total dose rate (D). If the sample burial depth has changed over time (erosion, deposition history, excavation), burial depth evolution must be reconstructed. The use of a portable handheld GPS device is highly recommended in order to obtain accurate coordinates.
8Stratigraphic logs, pictures of the sample locations, profiles sketches, maps and any other information to illustrate stratigraphic relationships between samples or identify potential problems with partial bleaching should be included.
Fig. 1 : Example of field material necessary for ESR sampling
PVC or metal tubes, (2) metal cap, (3) hammer and hand trowel, (4) measuring tape, (5) Duct tape (6) opaque ziplock bag, dark-colored permanent pen and pencil (7) dosimeter, (8) field-portable gamma spectrometer, (9) field notebook, (10) photo camera et (11) portable GPS (Photo: Mario Modesto Mata)
Fig. 2 : Picture showing the collection of a sediment sample by hammering a PVC tube into an outcrop exposure and the measurement of the burial depth.
The sphere represents the γ radiation influence in a 30 cm diameter around the dating sample
Photo: Mario Modesto Mata
Fig. 3 : In situ γ spectrometry measurements accurately capture the γ dose rate.
9All the recommendations presented here have been summarised in a sampling form (fig. 4) intended for archaeologists, geologists and geomorphologists who want to collect quartz samples for ESR dating. This form has been divided into four main sections: Site, Sample, in situ dosimetry and cosmic dose rate evaluation. Finally, there is an empty column where stratigraphic logs indicating the exact location of samples and any other information about the site should be reported. It is important to note that the present work and associated sampling form only provide some guidelines about the sample collection method and the minimum information required for ESR dating. Depending on the context and the complexity of the site additional information may sometimes be required.
10Fieldwork is a key step in the procedure to obtain reliable ESR ages. Indeed, many issues observed in the laboratory are directly associated with the sampling : loss of ESR signal due to exposition of sediments to light or lamps, missing the additional bulk sample for water content or dose rate analyses, lack of information about depositional context, stratigraphy or burial depth, sediments taken from deposits that have been affected by post-depositional processes or unadapted depositional environment, etc. Recommendations explained in this paper along with the sampling form should help archaeologists and geologists to avoid, or at least minimise, all these problems. However, an ESR geochronologist or ESR dating laboratory should always be contacted prior to sample collection in the field.
Fig. 4 : ESR quartz sampling form.