TL/OSL dating, a method used in archaeology by CIRAM laboratories

We are aware of the time constraints to which archaeologists are subject, which is why we offer standard turnaround times of 3 months for TL (depending on the laboratory's production load) and try to keep OSL turnaround times as short as possible. Luminescence dating methods are less well known than carbon 14, and their use remains limited in historical archaeology. They mainly refer to thermoluminescence (TL) and optically stimulated luminescence (OSL) dating.

TL dating, to date the last heating of materials

TL (thermoluminescence) allows the last heating of the material to be dated, whether clay or hearthstones (quartzite, flint, sandstone...). This method is frequently used in prehistory, where remains of other kinds are less abundant.

However, applications have been tested on materials more connected to historical issues, such as architectural terracottas for example. One of the most frequent uses lies in kiln dating, where TL is one of the most suitable methods along with archaeomagnetism.

Complementary study thanks to OSL dating

As a complement to thermoluminescence dating, OSL offers complementary perspectives. In fact, this method makes it possible to date a material's last exposure to light. It's not hard to imagine how interesting it can be to date a stratigraphic sequence on a site. Numerous applications have been developed in response to problems associated with the archaeology of historical and protohistoric periods. For example, OSL is used to date the installation layer of a megalith, the sealing of mortar in masonry...

The chronological range of these methods extends from a few hundred to a million years, which allows very wide use, on any type of archaeological site.

The fundamental principles of luminescence dating

Luminescence dating methods rest on common foundations. They rely on the ability of minerals (mainly quartz and feldspars) to record ambient radioactivity over time. The radioactivity absorbed comes from the earth's surface, essentially from the decay of three radioelements:

• Potassium (K);
• Uranium series (U);
• Thorium (Th).

The emission of particles (alpha and beta) and gamma radiation from these three elements occurs regularly over time and constitutes what is known as the dose rate. This annual dose (I) is supplemented by the effects of cosmic radiation, which vary according to burial depth, altitude and latitude. The quantity of radioactivity absorbed at the moment of measurement is called the archaeological dose (Qnat) or equivalent dose (De).

It is the ratio of these two quantities (Qnat and I) that gives the age between the object's last heating and its study in the laboratory:

When measured in the laboratory, the crystals studied emit light (or luminescence). The amount of luminescence emitted is proportional to the amount of radiation absorbed by the sample since it was last heated. For OSL, thermal stimulation is replaced by optical stimulation. Our laboratory scientists measure luminescence emissions by illuminating the crystals.

From field to laboratory: implementation of analyses by CIRAM laboratories

As explained above, it is important to measure two quantities to obtain luminescence dating: the archaeological dose and the annual dose. To do this, several steps are required, starting, in the best of cases, in the field.

Operational procedure in the field, careful sampling for accurate results

During the excavation operation, several samples and measurements are required. It's important to sample what you want to date: sediment from a stratigraphic sequence, hearth sole, heated flints... mass sampling is substantial, around 1kg.

Then, to be able to calculate the annual dose, two alternatives are possible:

• Or gamma and cosmic radiation measurements can be taken at the sampling point using a gamma probe;
• Or the radioactive environment can be "reconstructed" in the laboratory, from sediments collected within a 30 cm radius of the sample.

The depth of burial of the sample must also be available to make dosimetry calculations as accurately as possible.

Laboratory operational procedure

First it's a matter of choosing (if possible) the material on which the experiment will be carried out. The abundance and granulometry of the crystalline species used (mainly quartz and feldspar) must be determined, in order to choose which protocol to use.

Based on this determination, CIRAM scientists choose either the large quartz inclusions technique or the small polymineral inclusions technique. This choice conditions the material preparation stages, which can last from a few days to a few weeks.

Thermoluminescence measurements involve three steps for optimal dating (Qnat and I measurement):

• Measurement of natural luminescence, Q nat, with the addition of controlled laboratory irradiation doses, also known as the first reading;
• Measurement of the sediment's radiochemical composition by low-background gamma spectrometry;
• Determination of the annual irradiation dose, I, by a posteriori reconstruction.

Once the operational procedure is complete, our scientists process and interpret the results.

Processing TL and OSL results

TL and OSL provide dating independent of any external reference. The only factors influencing uncertainty are intrinsic to the measurements and to the quality of the initial sampling. In the case of TL, the optimum uncertainty can be as low as 5t, compared with 3.5% for OSL. Whether used alone or in conjunction with other methods, luminescence dating, despite its high accuracy, does not always enable two distinct archaeological facts to be finely discriminated.

In this context, recourse to statistical processing methods seems particularly indicated.

These make it possible:

• To propose a more precise chronological phasing of the sequence of archaeological structures,
• To propose average phase dates with reduced uncertainties,
• But also to refine each of the individual dates.

To optimize these treatments, the stratigraphic relationships observed in the field are paramount. Indeed, if two dates show partial overlaps and belong to stratigraphic units associated by anteriority/posteriority relationship, these overlaps are considered impossible.

The chronological model is based on the use of Bayesian statistics, and the date probability fitting functions are determined by Monte-Carlo simulation.

Efficient analytical activity and comprehensive critical synthesis with CIRAM

The analytical team's university training, dedicated to the physical and chemical analysis of archaeological materials and in particular to luminescence dating provides a solid basis for experimentation. In addition, the possession of TL and OSL measurement equipment guarantees CIRAM's technological autonomy which, combined with its high reactivity and availability, enables TL dating to be obtained within 3 months, depending on the laboratory's production load.

CIRAM has also developed a scientific partnership with a nuclear study center (University and CNRS), for analyses of the radiochemical composition of materials. Beyond simply obtaining dates, CIRAM is committed to providing a complete critical synthesis of the results and proposing the best possible archaeological exploitation of them.