USGS TRIGA Reactor
Neutron Activation Analysis
Neutron activation analysis (NAA) is an analytical technique that relies on the measurement of gamma rays emitted from a sample that was irradiated by neutrons. The rate at which gamma rays are emitted from an element in a sample is directly proportional to the concentration of that element. The major advantages of NAA are that:
- it is a multi–element technique capable of simultaneously determining up to about 70 elements in many materials
- it is non–destructive and therefore, does not suffer from the errors associated with yield determinations
- it has very high sensitivities for most of the elements that can be determined by NAA — detection limits range from 0.03 ng to 4 μg
- it is highly precise and accurate — overall errors of 2–5% relative standard deviation can be achieved for many elements
- samples as small as a few micrograms can be analyzed by NAA
The process for analyzing samples by NAA involves irradiating them with a neutron source. Neutrons are captured by elements in the sample to produce unstable radioactive isotopes (radionuclides). Beta particles, and in most cases gamma rays, are emitted from the radionuclides as they decay. The energies of these gamma rays are, in general, distinct for a specific nuclide and the rate at which these photons are emitted with a particular energy can be measured using high–resolution semiconductor detectors. Because the production and decay rate of gamma radiation are dependent on the half–life of the nuclide, elemental measurements can be optimized by varying the irradiation and the decay times (i.e., how long the sample is near a neutron source and when the sample is analyzed).
The most common procedure for NAA involves encapsulating the samples and suitable standards in heat–sealed polyethylene or quartz vials and simultaneously irradiating them. Ideally, the samples are irradiated in a "lazy susan" facility that revolves around the core thereby ensuring that the samples and standards experience the same neutron fluence. Following sequential decay periods, each standard sample is analyzed utilizing high resolution germanium detectors coupled to a multi–channel analyzer system. Gamma ray counts accumulated in an energy region above the background counts produce photopeaks. After counting analysis is complete, these data are processed using sophisticated computer programs that smooth the spectral data and determine the net areas of gamma ray photopeaks. The program then translates the area into count rates (counts per minute or cpm). These programs are capable of resolving overlapping and complex photopeak energy regions. Additional data for decay time differences, electronic dead time losses and unresolved interferences and compares the sample data (cpm/weight) to the standard data (cpm/μg) to calculate elemental abundance in the sample.
Errors in Analysis
The principal error in the analysis of materials by NAA is the counting statistic error, which is based on the signal to background ratio at the gamma ray energy region of interest. A one sigma error for a photopeak area determination is approximately equal to the square root of the total counts (background plus net counts) divided by the net counts. For example, a peak area determination with 2000 total counts and a net area of 1000 counts produces a counting error of about 4.5%.
An additional source of error for some elemental determinations is due to unresolved interferences. The most serious interferences are those that result when identical radionuclides are produced from different nuclear reactions. For example, high concentrations of U fission products can interfere with the accurate determination of La, Ce, Nd, Mo, and Zr abundances. In addition, the determinations of some elements can suffer because the photopeaks produced by gamma rays emitted by two or more radionuclides is not readily resolved. In these cases, empirical corrections are integrated into the data reduction procedure.
Self–shielding is a phenomenon that occurs when the neutron flux experienced by a sample is attenuated, thereby reducing the neutron activation of the nuclides in self–shielded samples. This effect occurs when a sample contains high concentrations of elements with very high probabilities that a neutron will be captured, which is measured in terms of a target area, or cross section.
Elemental detection limits for NAA are variable because some elements become very radioactive–, and can be determined at very low levels while other elements do not become very radioactive or have very short half–lives (less than 10 seconds). Activation analysis determines the total mass of an element in a sample. A certain amount of an element, like arsenic–, is needed in the sample for detection. For arsenic, under ideal conditions, 5 ng is required. To determine 5 ppb of arsenic, 1 g of sample is sufficient. To determine 0.5 ppb of arsenic, 10 g of sample is necessary, etc. The production of radioactive nuclides depends on the cross sections of the specific elements. Also important is the number of gamma rays that are emitted by a radionuclide. In some cases, only a small fraction of the total emissions from a specific nuclide is in the form of gamma rays. Samples with high concentrations of elements that are readily activated and emit a considerable number of gamma rays, such as Na and Sc, can generate high background count rates and raise the detection limits for the element(s) of interest. The following table shows the best case minimum detection levels (MDL) for the 70 elements that NAA is capable of identifying.