Field gamma dose-rate measurement with a NaI(Tl) detector: re-evaluation of the "threshold" technique

Radiation detectors, like doped NaI, are commonly used in the field for the determination of gamma dose-rates. In most cases, and even though these systems generally allow one to record the full gamma spectrum between 0 to around 3 MeV, this dose-rate is computed from the count rates recorded in a limited number of "windows" (Aitken, 1985). With this technique, only a small part of the spectrum is therefore exploited. Nevertheless, an alternative approach the "threshold" technique – known for more than 30 years, can easily be used with the same detection system. In this paper, we make a reevaluation of this technique and discuss its limits and advantages. Introduction In a paper published in 1974, Løvborg and Kirkegaard investigated the response of a 3 inch by 3 inch NaI(Tl) detector placed above environments of known radioactivity. According to their experimental and theoretical results for this 2π geometry, the count-rate of their equipment above a chosen threshold (370 keV) was found to be directly proportional to the gamma dose-rate. As the gamma rays come from the radioactive elements of the Uand Th-series and from the K, this means that the count-rate was independent of these three sources. In the next years, Murray, Bowman and Aitken (1978) developed a portable system equipped with a NaI(Tl) detector for gamma dose-rate measurements; it is discussed in Aitken (1985, p. 107-108) as the "gamma scintillometer". In their experiments, they inserted their detector in radioactive doped blocks set up at the Research Laboratory for Archaeology and the History of Art (see Rhodes and Schwenninger, 2007). With these experimental conditions, which are close to field conditions, they concluded that a threshold value could also be defined for their equipment (at 450 keV). However, these authors showed that this threshold occurred in a region of the spectrum where the count-rate changed rapidly with energy, and they then emphasized the necessity to stabilize the system to account for temperature variations. We recently tested this approach with a 1.5 inch by 1.5 inch NaI(Tl) detector (IPRON-1 connected to a multichannel analyser: Inspector1000 – Canberra) and report here our results and conclusions. General In luminescence and ESR dating, the gamma doserate generally constitutes a significant component of the total dose-rate and has to be determined precisely. This dose-rate can be measured directly with synthetic dosimeters (e.g. Al2O3:C or CaSO4:Dy) buried for relatively long periods of time (weeks or even months), or by inserting in the sediment a portable detector connected to a spectrometer and recording the gamma spectrum. In this case, the gamma dose-rate is calculated from the radioisotopic contents of the sediment (U, Th and K) determined by the “classical” technique, which is based on the definition of three regions of interest ("windows"), each window being centred on a gamma ray energy specific to an isotope (1460 keV for K; 1780 keV for Bi (U-series) and 2620 keV for Tl (Thseries), see for instance Aitken (1985, p.102-105). In each of these two last windows, the counting is due to the detection of gamma rays coming from both the Uand Th-series, whereas in the “K window” the three sources contribute. Mathematically, one can then simply write for each window the following equation: N = nK . [K] + nU . [U] + nTh . [Th] 2 Ancient TL Vol. 25 No.1 2007 where the total count N (per unit of time) is the sum of three factors in which [K], [U] and [Th] are the radioisotopic contents of the sediment and nK, nU and nTh represent the detection efficiencies of the detector in the considered window for the three sources. These efficiency factors are then expressed as the number of counts per unit of time for 1 ppm (for the Uand Thseries) and 1% for K. In considering the threshold approach, these factors have to be interpreted as the number of counts per unit time for 1 μGy/a and [K], [U] and [Th] as the gamma dose-rates in the sediment. Consequently, for N to be proportional to the total dose-rate, nK, nU and nTh must have the same value n. In this case, the above equation can be written as : N = n . ([K] + [U] + [Th]) where ([K] + [U] + [Th]) is the total gamma doserate. In practice, for using this approach with a given detector, one has simply to define the energy for which the values of nK, nU and nTh are identical. Experiments One way to answer this question is to get “pure” spectra taken in environments containing only one source of radioelements (either K, Uor Th-series). Such environments do not naturally exist, so we used the doped blocks set up at the Research Laboratory for Archaeology and Art, in Oxford, as reported by Murray et al. (1978). These blocks made of concrete were doped with uranium, thorium and potassium and provide the following gamma dose-rates: Ublock: 13.27 Gy/ka; Th-block: 7.10 Gy/ka; K-block: 1.38 Gy/ka. A non-doped block made of the same concrete is also available and is used as a background standard with a gamma dose-rate of 0.53 Gy/ka (Rhodes and Schwenninger, 2007). According to these authors, the Th-block contains a small portion of U with a U/Th ratio of 0.043. Considering this ratio and the dose-rate conversion factors given by Adamiec and Aitken (1998), one can calculate the effective radioisotope concentrations : U-block (117.4 ppm), Th-block (135.2 ppm of Th and 5.8 ppm of U), K-block (5.7 %). Note that these values are slightly different from those given in Rhodes and Schwenninger (2007) since they used the conversion factors of Nambi and Aitken (1986). Fig. 1 shows the spectra recorded with our gamma probe. After time normalisation (1 ksec), “pure” spectra were calculated by subtracting the signal measured in the non-doped block (thus including the background of the detector and the cosmic contribution) and were normalised to 1 ppm of U or Th, and 1 % of K (Fig. 2), by using the effective radioisotope concentrations cited above. This step was necessary in order to subtract from the Th spectrum the small contribution from U present in the Th-block.