)
Measuring luminescence
The De of a sample is determined by measuring the luminescence signal from quartz or feldspar minerals. The luminescence signal comprises photons that are emitted when electrons become evicted from their traps during stimulation of the sample with light or heat. These photons are counted using a detector known as a photomultiplier tube.
)
Figure 1. Photon counts measured from feldspar minerals over 200 seconds while the sample is stimulated with infrared light-emitting diodes (LEDs). The initial photon emission rate is high (~5000 counts per second) and decreases with time to a background level as the number of trapped electrons is depleted.
The intensity of the natural sample's luminescence signal (expressed as photon counts per second) is proportional to the amount of radiation the sample was exposed to during burial. However, since the brightness of this signal (i.e., sensitivity) varies from grain-to-grain and sample-to-sample, the relationship between luminescence signal brightness and absorbed radiation dose must be calibrated for each grain or aliquot measured (Fig. 2).
)
Figure 2. Luminescence signals measured from bright quartz (A) and dim quartz (B). Measurements were made from small multi-grain aliquots composed of ~10 grains during stimulation with blue LEDs at 125℃. The ‘Natural’ signal is the signal measured from the sample prior to any other treatment. The ‘Beta dose’ signal is the signal measured after the sample was bleached, then exposed to a beta radiation source in the lab. From Mahan et al. (2022).
Creating a dose response curve
Calibrating the relationship between absorbed radiation dose and luminescence signal brightness is done by measuring the luminescence response of a grain or aliquot to various laboratory-induced radiation doses. Most luminescence readers are equipped with a beta radiation source that is used to administer precisely known doses to the sample.
After the natural luminescence signal (i.e., the signal measured from the sample following a short preheat but prior to any other treatment, Ln) has been measured from a grain or multi-grain aliquot, the luminescence from the same grain/aliquot is measured again after it has been exposed to a radiation source in the lab for a known period of time and a short preheat (Lx). This process is repeated with a series of increasing laboratory doses to construct a dose response curve, or DRC, in which radiation dose is plotted on the x-axis and luminescence signal brightness is plotted on the y-axis (Fig. 3). This curve plots the relationship between known radiation dose and luminescence signal brightness. If we project the natural signal intensity onto the dose response curve, we can determine the amount of natural radiation the grain/aliquot has received during burial by interpolation (Fig. 3).
)
Figure 3. Dose response curves measured from one aliquot or grain of a sample. In ‘A’, the natural signal plots before the curve saturates, while in ‘B’, the natural signal approaches saturation (i.e., the flattest part of the dose response curve). The De value is read off the x-axis by interpolation from the natural signal projection onto the dose response curve. Taken from Mahan et al. (2022).
The series of increasing laboratory radiation doses administered to the sample are called regenerative doses. As the regenerative doses increase, the dose response curve will flatten out, or saturate (Fig. 3B). Saturation occurs when larger radiation doses no longer induce increases in luminescence intensity; the saturation dose of a mineral will determine its upper age limit. All minerals’ luminescence signals will saturate at some level, but feldspar minerals have been found to saturate at higher levels than quartz, and thus, they can typically be used to date older samples.
Correcting for sensitivity changes
Just as the natural luminescence signal sensitivity can vary from aliquot-to-aliquot or grain-to-grain, the luminescence sensitivity of a sample will also change in response to various dosing, heating, and optical stimulation treatments in the laboratory. This means that after the measurement of each natural and regenerative dose signal, the sensitivity of the mineral to additional acquired laboratory doses may change. If these sensitivity changes are not corrected for, the DRC will not appropriately calibrate the dose-signal intensity relationship of the natural signal, and an inaccurate De estimate will result.
In practice, sensitivity changes are monitored and corrected for by measuring the signal induced by a known dose, called a test dose. Unlike the regenerative doses, the test dose does not change in size, and it is measured after each natural and regenerative dose signal has been measured. The dose response curve is then constructed by plotting radiation dose on the x-axis and the ratio of the natural signal to its subsequent test dose signal (identified by the notation Ln/Tn) and the ratio of each regenerative dose signal to its subsequent test dose (Lx/Tx) on the y-axis (Fig. 3).
Single aliquot regenerative-dose protocols
This process of measuring the sensitivity-corrected natural signal (Ln/Tn) and regenerative dose signals (Lx/Tx) is referred to as the Single Aliquot Regenerative-Dose protocol, or SAR (Murray and Wintle, 2000). SAR is the most commonly applied measurement procedure for obtaining De and is divided into a series of measurement ‘cycles’. The first SAR cycle measures Ln/Tn, and all subsequent SAR cycles (typically 4-6) measure Lx/Tx to establish the DRC (Table 1). All SAR cycles are performed, and De values are determined, for individual grains or aliquots of a sample. In a typical sample, multiple aliquots (5-50) or individual grains (>100) are measured to generate a distribution of De values that can be modelled to obtain a sample-specific De value that best represents the timing of burial.
In addition to irradiation and stimulation of the sample during each Ln/Tn and Lx/Tx measurement cycle, there are additional steps that heat the sample to an elevated temperature just prior to luminescence measurement. Preheats are heating steps that hold the sample at a desired elevated temperature for a period of time, (e.g., 10 seconds). Quartz samples, for example, are commonly held at 260℃ for 10 s prior to Ln and Lx measurements. Cutheats are treatments where the sample is heated to a desired temperature (e.g., 200℃), then allowed to cool immediately. Cutheats are typically applied prior to Tn and Tx measurements. These heating steps are designed to empty traps that contain unstable electrons that are not retained over geological time periods - such unstable contributions to the luminescence signal can lead to erroneous DRC shapes and De values.
Steps included in a typical SAR protocol are outlined in Table 1. While most SAR protocols follow this general structure, the parameters used (e.g., preheat/cutheat temperatures and durations, stimulation durations and temperatures, test doses, etc.) are dependent on the optical properties of the minerals and can vary from site to site and even sample to sample. So a table listing the specific SAR parameters used is included in all sample reports.
Quality control measures
In addition to the regenerative dose points measured to establish the DRC, two additional regenerative dose points are measured for quality control purposes. These include the repeat-dose point (i.e., one regenerative dose point is measured twice), and a zero-dose point (i.e. Lx/Tx is measured after no dose is given). The repeat dose point is used to calculate a recycling ratio, which is the ratio between the two repeat regenerative dose points. If this ratio is within 2σ or 10% of unity, we can be assured that the test dose has adequately corrected for sensitivity changes during SAR.
The zero-dose point is used to calculate recuperation, which is a combination of charge that has not been sufficiently depleted at the end of each SAR cycle, as well as excess accumulated charge as a result of thermal transfer during the preheat. Samples that behave well should have recuperated signals that are ~5% or less of the natural signal. Aliquots or grains that do not pass the recycling ratio and recuperation requirements, as well as aliquots that are too dim for De measurement, or have natural signals that plot too close or beyond the saturation limit of the DRC, are rejected from further analysis. Other grain/aliquot rejection criteria have also been proposed for samples that exhibit other sub-optimal characteristics.
SAR cycle step | Purpose |
|---|---|
1- Natural/Regenerative dose | In cycle 1, no action is taken. In subsequent SAR cycles, a regenerative dose is given to the sample to construct a DRC. |
2- Preheat | The sample is heated to eliminate thermally unstable charge. |
3- Stimulation -> Ln or Lx | The sample is stimulated with a light source and the luminescence emissions associated with the natural/regenerative dose is detected. Infrared light is typically used to stimulate feldspar luminescence, while blue light is typically used to stimulate quartz luminescence. |
4- Test dose | A small radiation dose is administered to the sample. The test dose is the same in every SAR cycle. |
5- Cutheat | The sample is heated to eliminate thermally unstable charge. |
6- Stimulation -> Tn or Tx | The sample is stimulated with a light source and the luminescence emissions associated with the test dose are detected. The test dose signal is used to correct the natural/regenerative dose signals for sensitivity changes. |
7- Hotwash (optional) | The sample is stimulated with a light source at high temperature to eliminate recuperated charge, or charge that was insufficiently depleted during the SAR cycle. |
8- Return to Step 1 |
1Ln = natural signal, Lx = regenerative dose signal. Tn = test dose signal measured after Ln, Tx = test dose signal measured after Lx. Steps 1-8 constitute one SAR cycle. The first SAR cycle measures the sensitivity corrected natural signal, and subsequent SAR cycles measure the Lx/Tx value after a series of successively increasing laboratory radiation doses (regeneration doses) administered to the sample. The regenerative doses are used to generate a dose-response curve (where ‘x’ values are the given radiation doses, and ‘y’ values are measured Lx/Tx signals) onto which the natural signal (Ln/Tn) is projected to interpolate the De value (Fig. 3). Regenerative doses include one “zero-dose” point where no radiation dose is given, and one “repeat-dose” point, where a previous regeneration dose (R1) is given a second time.