Detector efficiency calibrations

A Canberra well-detector system using Fitzpeaks gamma spectrometry software was used to determine caesium-137 activities in all sample fractions. Variations in sample height arising from small sample masses were corrected for where appropriate. The detector efficiency calibration was determined using an NPL mixed gamma standard spiked into a sand matrix (Croudace 1991). 

Table 15. Decay Data for Radionuclides Useful for the Energy and Efficiency Calibration of 7-Ray Detectors...

To prepare a standardized plutonium tracer for use and compare its activity in 2 detector systems calibrated for alpha-particle counting efficiency. 

Alpha sources as standard sources are characterised either in terms of activity (Bq) or in terms of emission flux in 2n sr (s" ). The radionuclide is electroplated, either on a polished stainless steel disc 25 or 30 mm in diameter and 0.5 mm thick, or polished platinum disc 22 mm in diameter and 0.1 mm thick. The contribution of the sources to the FWHM of a spectrometer is about 1 keV, the total FWHM being thus for a commercial spectrometer less than 15 keV. All these sources can be used for energy calibration of efficiency calibration for all detectors and a measuring devices (see Table 4.10 for the list of alpha sources). 

Indirect geometry spectrometers have no requirement (within the limitations implied by the use of S (Q,a>), 2.5.1) to calibrate detector efficiencies, on either continuous or pulsed sources (compare 3.4.3). Since the final energy of the neutrons never varies the detection efficiency is constant. Variations arising from differing discrimination levels ( 3.3.2) could play a significant role, except that (on low final energy instruments) all detectors follow almost the same path in Q,o ) space ( 3.4.2.3). Occasionally there is a need to calibrate the detected intensity in respect of the sample mass and standard analytical chemical techniques can be readily adapted to this circumstance. 

A plastic scintillation detector was to be calibrated for absolute measurements of /3-radiation. For this purpose a 2.13 X 10 M TlCl3 solution was available with a specific activity of 13.93 iCi ml T1 emits iS-particles with 0.77 MeV. Of this solution 0.1 ml is evaporated over an area of exactly 0.1 cm on a platinum foil. The sample is counted in an evacuated vessel at a distaiKe of 15.3 cm from the detector, which has a sensitive area of 1.72 cm. The detector registers 2052 cpm with a background of 6 cpm. What is (a) the surface weight of the sample, (b) the backscattering factor, and (c) the detector efficiency for the particular /3 s ... 

A p-type detector was used for samples analyzed before 1995, while the most recent samples were analyzed with an n-type detector. All detectors presented a typical resolution of 1.7 keV for the Co peak at 1.33 MeV. Energy calibration was set with a Canadian (Canmet) reference sample. Efficiency calibration was set with aqueous europium 152 and barium 133 sources used in the same geometry as the unknown samples. Samples were counted from 20000 to 80000 seconds, depending on the mass and intensity of the radioactivity. Cs was quantified from its 661.7keV peak and Cs by means of its 604.7 keV peak. Depending on the Cs or Cs activity, counting time, and mass of honey samples, the counting error was always less than 10 percent. 

The counting efficiency of these detectors is calibrated with radionuclide standards or Monte Carlo simulation (Briesmeister 1990). Typically, the alpha-particle detector has the same efficiency for thin samples at all commonly encountered... 

Efficiency calibration requires a sample of defined shape and volume placed at a defined location relative to the detector. A sample frame may be needed for reproducible placement. Calibration can be performed with a radioactivity standard or by Monte Carlo simulation (see Sections 8.2 and 10.5). 

Handling the standard, placing the source near the detector, recording data, and processing results all must be meticulous operations because the reliability of all future calculations of activity from count rate depend on the efficiency calibration. Accumulated calibration counts should be sufficiently large to achieve a relative counting uncertainty of 0.01 or less (at least 10,000 counts). To perform measurements in a brief period (such as 1-10 min), the source usually is prepared to yield a relatively intense count rate, but not so intense that the count rate is affected by dead-time losses (see Section 8.2.2) or system contamination becomes a threat. 

The raw scattering data were first corrected for instrumental background and dark current counts and then corrected for non-uniform detector efficiency at each detector channel element. Then the sensitivity corrected data were plotted as a two-dimensional contour plot with contour levels increasing by a factor of 2 from the outermost one. Lupolen 23/7 was used as a calibration standard to determine the absolute SAXS intensity (49.50). 

The instructions with the software packages indicate how to determine / and a, how to measure the parameters of the efficiency calibration of the detector and convert it to specific counting geometries, and how to convert the peak areas into the mass firactions of the elements. The software packages have libraries of ko values, Qo values, mean resonance energies, half-lives, decay types, and the details of the decay schemes needed for true coincidence summing calculations. 

The calibration of a PIXE system, i.e., the determination of sensitivity factors, which assign absolute concentration data to numbers of counts in X-ray peaks, can be performed in two different ways. First, sensitivity factors can be deduced theoretically or in a semiempirical way from calculated cross sections for X-ray excitation and from X-ray absorption data for the absorbents present between the points of emission and detection, in the actual experimental setup. Second, sensitivity factors can be deduced from measurements performed on standard samples consisting of pure elements or pure chemical compounds. The detection solid angle and the energy-dependent detector efficiency should also be determined. 

The efficiency calibration of a Si(Li) detector in the energy range of 5-60 keV using radioactive sources has been reported by Verma (1985). A modified theoretical model for calculating the efficiency of a Si(Li) detector has been presented by Garg et al. (1987) while the comparison of experimental efficiencies for different Si(Li) detectors in the energy range, 4.5-17.5 keV has been made by Yap et al. (1987). Table 1.7 lists the radioactive sources for efficiency calibration of the Si (Li) detector. 

Table 1.7. The radioactive sources for energy and efficiency calibration of the Si(Li) detector (Ref. IAEA chart of the NucHdes)...

Alpha spectrometry is characterized by good isotope separation, uniform (i.e., energy-and isotope-independent) detector efficiency and a very low backgroimd (typically of the order of 0.001-0.003 cpm (counts per minute) for the 4-8 MeV energy range of a new surface barrier detector, increasing with time as a result of the accumulation of recoil products from measured samples). It allows precise measurements at low activities, and easy calibration... 

Note 2 The mixed Th/ Th planchet is worthless after the decay of Th. A useful standardization can be made with a planchet with 60dpmU electrodeposited on it (evaporate a weighed aliquot in a small Teflon beaker and follow electrodeposition procedure of Th Section 13.7.7). After ingrowth of Th to secular equilibrium with (several months), the planchet can be used for a check on the relative efficiencies of beta and alpha counters. Alpha detector efficiency determined this way for should also be applicable to but an independent calibration of the °Th spike is still required. 

When the nature and composition of the sample is not well known, it is necessary to use influence correction methods, of which there are three primary types fundamental, derived, and regression. In the fundamental approach, the intensity of fluorescence can be calculated for each element in a standard sample from variables such as the source spectrum, the fundamental eiiuations for absorption and fluorescence, matrix effects the crystal reflectivity (in a WDXRF instrument), instrument aperture, the detector efficiency, and so forth. The XRF spectrum of the standard is measured, and in an iterative process the instrument variables are refined and combined with the fundamental variables to obtain a calibration function for the analysis. Then the spectrum for an unknown sample is measured, and the iterative process is repeated using initial estimates of the concentrations of the analytes. Iteration continues until the calculated spectrum matches the unknown spectrum according to appropriate statistical criteria. This method gives good results with accuracies on the order of I %-4% but is generally considered to be less accurate than derived... 

One might imagine that, knowing all we do about the interaction processes involved, the absorption coefficients of the detector material and attenuation within the encapsulation, it would be possible to calculate the detector efficiency from first principles. Unfortunately, there are limitations in the mathematical tools at our disposal and the lack of consistency with which detectors can be manufactured militate against such calculations. At the present time, efficiency calibrations are performed on actual gamma-ray spectra. There are, however, efforts being made towards provision by the manufacturers of theoretical calibration data with each detector supplied, so that the need for calibration by the user may diminish in the future. I will discuss some of these developments in Section 7.7. 

Arnold, D. and Sima, O. (2004). Extension of the efficiency calibration of germanium detectors using the GESPECOR software, Appl. Radiat. Isotopes, 61, 117-121. 

Gunnink (1990) described polynomial equations of this type for calculating intrinsic efficiency, from which absolute efficiency can be calculated. By examining the efficiency calibrations of a large number of detectors, he was able to relate some of the parameters of these equations to the dimensions of the detector and other details of the detector system. Different equations were used from 50-90 keV, 90-200 keV (second order polynomial) and above 200 keV (sixth order polynomial). At first sight. 

---