ENVIRONMENTAL RADIOACTIVITY LABORATORY
INRASTES, NCSR “Demokritos”, Athens, Greece

RADIATION PROTECTION, RADIOECOLOGY, AEROSOL SCIENCE
Research ... Scientific and Technological Services

Research

Home …  Methods … Research … Publications … Info … Contacts

METHODS
High resolution gamma spectrometry

High resolution gamma-spectrometry allows the qualitative and quantitative analysis of a large numberof radionuclides in certain sample. This results from the analysis of the energy spectrum of the gamma-quanta emitted after the beta- or alpha-decay of these radionuclides. In some cases of decay no gamma-quanta are emitted: these radionuclides (e.g. Sr90) can not by detected by gamma-ray measurements.

The crucial component of any gamma-measuring device is the detector - a component producing electrical signals as a result of the interactions of gamma-quanta within it. In order to perform energy analysis, we use a specific type of spectrometry detectors of gamma-rays, which provide electrical signals proportional to the energy of the detected gamma-quanta. The solid state Hyper-pure Germanium detectors (HpGe) are widely used for this purpose.

The energy spectrum of the gamma-quanta is characterized by the presence of sharp peaks. The number of these peaks and their energies vary between radionuclides. This "fingerprint" is used for nuclide identification (qualitative analysis). Furthermore, the intensity of these peaks is proportional, amongst others, to the activity of the source radionuclide. Under identical conditions of measurement, a comparison of the peak intensities of a known (standard) and unknown radioactive source allows to determine the activity of the radionuclide(s) (quantitative analysis). It is important to note, that the quantitative analysis is strictly based on the qualitative, for reasons explained later.

The amplitude spectrum of the detector signals is not an exact copy of the gamma-energy spectrum. The fluctuations accompanying the detection process result in some smoothing. Nevertheless, the HpGe photopeaks (the copies of the gamma-energy peaks) are still narrow enough to allow the identification and quantification of hundreds of them in an amplitude spectrum.

The observation and analysis of the amplitude spectrum is based to a procedure of amplitude sorting of the detector signals, performed on-line by the multichannel analyser unit (MCA). The basic components of a MCA are:

1. The analogue-to-digital converter (ADC) - an electronic device where the amplitude of the input pulses is converted to a digit, proportional to this amplitude.

2. The MCA memory: a large number of memory units ("channels"), typically 2028, 4096 or 8192, where the analyser events are accumulated. An ADC output digit N results in the addition of unit in memory 'N' ( if 0 < N < Nmax ).

3. The graphical interface enabling the observation and manipulation of the memory contents ( the "spectrum" ). The channel number is presented in the X axis, while the channel content - in the Y. The interface allows zooming in both directions for better observation, marking and/or selecting of channels or groups (ROIs - regions of interest), displaying the results of mathematical operations on the memory data, as the centroid of a peak, its net and gross area, the nuclide(s) identified e.a. The main parameters of the spectrum aquisition are displayed as well.

The full spectrum analysis is based on the following procedures :
.
- To perform qualitative analysis, the system has to be energy-calibrated, i.e. the relation E=F(N) between the channel number N and the gamma-energy E has to be determined. The high linearity of the amplifiers and the ADC device allow the use of a simple 2nd order function E = a + bN + cN^2, derived by regression based on a spectrum with known photopeak energies. The interface uses this function to display any X-axis values in gamma-energy units.

- A detailed nuclide library has to be installed. This is a table of the basic gamma-energies emitted by various radionuclides of interest, with inputs of the form:

[ Energy ] [ Nuclide ] [ Half-life time ] [ Gamma-yield ]

The nuclide identification is based on the conformance of the photopeak energies observed in the spectrum with those present in the nuclide library. Various levels of "intelligence" are applied, in order to avoid identification errors due to various reasons. A typical reason of error are the close values of the gamma-energies emitted by two radionuclides and a typical method to avoid it is to look for other, "confirmation" photopeaks, emitted by the suspected nuclide.

To perform quantitative analysis the system has to be efficiency-calibrated, i.e. the relation e = F ( E ) has to be determined for each geometry of measurement and, even better, for a series of sample densities. The system efficiency depends on the gamma-ray energy E due to the dependence on E of the probability of various interactions involved in the detection process, but also in the partial absorption of the gamma-quanta on their way to the detector. It depends also on the geometry of measurement through the probability for a gamma-quantum emitted from the sample to be directed to the detector.

Only a part of the gamma-quanta emitted from the sample is directed to the detector - those within the cyan-coloured solid angle. Even if scattered and directed back to the detector, the quantum 1 will not be detected in the photopeak region due to the loss of some energy due to the scattering. The quantum 2 is scattered on its way to the detector. The quantum 3 is scattered within the detector material and leaves its volume without losing all its energy within it. Only the quanta 4 and 5, the first directly absorbed and the second absorbed after a scattering event, will contribute to some photopeak. The photopeak termin originates from the photoelectric absorption - the effect characterised by full gamma-energy transfer to the detector material.

The count rate a observed in a photopeak E is related to the activity A of the radionuclide R in the sample as follows:

a [cps] = A [Bq] I(R,E) e(E)

where I(R,E) is the gamma-yield, i.e. the probability of emission of a quantum E after a decay event of R. It is obvious, that the determination of A based to the above relation, depends (through I) on the correct identification of R.