accuracy and precision

Detection limits vary from element to element and depend upon several factors. Some elements become very radioactive, pound for pound, and can be determined at very low levels (sub ppt). Other elements do not become radioactive at all or have very short half lives (less than 10 seconds) and are not determined by our group. There are 90 naturally occurring elements and about 70 will become radioactive (using thermal neutrons).

Activation Analysis determines the total amount of an element in a sample in grams (or micrograms). A certain amount of an element, say Arsenic, is needed in the sample for detection. For Arsenic, under ideal conditions, 5 nanograms is required. To determine 5 ppb of Arsenic, one gram of sample is enough - to determine 0.5 ppb of Arsenic, ten grams of sample is necessary, etc. In some cases, it is possible to "push" detection limits by longer irradiations. If a sample is irradiated twice as long, it becomes twice as radioactive, or the detection limit is improved. This is true up to one half-life, or so, of the isotope to be determined. One disadvantage to this approach is the cost of Nuclear Reactor time.

Doubling the counting time can also improve detection. However, background noise from the environment limits this approach. Doubling the counting time will not help if all this does is double the background noise level.

Usually the detection limit depends upon the "other" elements in the sample - the matrix. If an element in the sample becomes radioactive, besides the element of interest, the background noise may be too high to determine the desired element at low levels. This noise does not produce wrong results, just high detection limits. The signal-to-noise ratio will improve with time if the element of interest has a long half-life compared to the element(s) producing the noise. The disadvantage to this is the delivery time - waiting for the noise to diminish may take only a few hours or days, but, in some cases, it could take several weeks.

Radiochemical separations may be employed when one or more elements in the sample become so highly radioactive that it "interferes" with the detection of the element(s) of interest. This interference usually produces high detection limits for the element(s) to be determined. To improve detection, the element(s) of interest is chemically separated from the matrix - AFTER the irradiation, but PRIOR to the counting step mentioned above.

See our Periodic Table for more information.


Accuracy is how close the determination is to the actual value. Precision is how close replicate determinations are to each other. An analysis may be very precise (duplicates may have the same results), but not very accurate (the real value is much different). In Neutron Activation Analysis, precision usually varies from 2% to 5% of the value obtained - independent of concentration. This precision depends upon the background noise level and the concentration of the determined element. Using standards, the accuracy has been determined to be within the precision, providing "gross" errors have been avoided. Gross errors include wrong sample weight, mathematical mistakes, etc.

For trace element determinations, the accuracy and precision of Neutron Activation Analysis cannot be matched. However, for major constitutes another approach should be considered.

Home Page | Method | The Company | Limits | Periodic Table | Prices | Pictures | Links

General Activation Analysis, Inc.
1011 Elmview Drive
Encinitas, CA 92024

Thomas R. Powell & Associates