Neutron activation analysis (NAA) works by using neutrons from a reactor to irradiate the atomic nuclei of elements within a sample. The irradiation process causes some of the atomic nuclei to become radioactive. Most radioactive nuclei will emit gamma rays at a rate dependent upon the number of radioactive atoms and the half-life. By measuring the emitted gamma rays, a trained analyst can identify the radioactive nucleus (i.e., and therefore infer the identity of the irradiated target nucleus) and determine the original number of atoms present (i.e., the concentration of the element).
For additional information read our Technical Overview of Neutron Activation Analysis (NAA), or for much greater detail see:
Glascock, M.D. (1998).
Activation analysis. In Instrumental Multi-Element Chemical Analysis, Z.B. Alfassi (ed), Kluwer Academic Publishers, Dordrecht, pp. 93–150.
The types of archaeological samples most suitable for analysis by NAA at MURR include potsherds, clays, obsidian, cryptocrystalline silicates (called chert hereafter), basalt, and limestone. Other archaeological materials analyzed previously in our laboratory include tephra, steatite (soapstone), turquoise, ancient metals, marble, and archaeological bone.
Fortunately, NAA is one of the most sensitive analytical techniques in existence making possible the study of very small samples. (For example, the minimum analyzable sizes for obsidian and pottery are about 5 mg and 50 mg respectively.) However, due to possible heterogeneity in certain types of archaeological samples, one is advised to supply sufficient sample material such that the chemical analysis is more representative of the whole specimen. As a general guideline, for the listed sample types we recommend submitting the following minimum amounts:
Sample Type | Recommended Minimum Size |
---|---|
Artifacts |
|
Pottery | 2-g potsherd with typically 3–4 cm² of sample area is highly recommended. A sherd with 1 cm² sample area is the minimum acceptable size. (MURR will retain a small portion for our archives.) |
Obsidian | A 500-mg fragment is recommended. However, samples as small as 5–10 mg have been analyzed in the past. (We use a trim saw on artifacts to minimize damage to the artifact.) |
Chert | 1-g fragment is recommended. 200 mg is the minimum acceptable size. (A diamond saw can be used to remove fragments.) |
Basalt | 1 or 2 g of rock, preferably powdered in advance. Be careful, many grinding procedures will contaminate the sample with iron, cobalt, tantalum, etc. (We will grind for you if necessary for an extra fee.) |
Terra-cotta Statuary | 250 mg of clean powder obtained by drilling is recommended. Minimum amount is 100 mg. |
Source Materials | |
Clay | 50 g dry weight with plant roots and insects removed. (The sample will be fired at 700°C before making into powder.) |
Temper | 50 g dry weight with plant roots and insects removed. (Please send clean and dry tempering materials.) |
Obsidian & Chert | Half-fist to full-fist sized specimens are ideal. (Analytical sample is taken from crushed interior fragments.) |
For ceramic studies, the number of samples to submit for analysis largely depends on the goals of the project. Ceramic studies typically involve the largest number of analyses. This is because such projects generally involve a search for subgroups within a database of analyzed artifacts. For each subgroup to be large enough for statistical description, the data set as a whole has to be large. As a rough rule-of-thumb, ceramics projects usually involve between 100 and 500 analyses. Raw materials (clay and temper) should be collected and analyzed if at all possible, because they sometimes provide valuable information. A reasonable number of raw material samples to submit is one or two percent of the total project.
For chert studies, large numbers of source specimens are recommended because chert compositions are low and somewhat variable, the number of chert sources is large, and MURR's chert source database is not comprehensive. We recommend that 20–25 specimens from each potential chert source be analyzed in order to establish the range of variation within and the differences between individual sources. Individual artifacts can be projected against the current source database and any new sources added by the customer.
For obsidian studies, studies of obsidian artifacts have ranged from small projects involving only a small handful of samples up to very large projects involving hundreds of artifacts. In several regions (Guatemala, central Mexico, South America, and the American Southwest) MURR's obsidian database is likely adequate such that the analysis of additional source specimens is not necessary. However, in regions where our obsidian source database is inadequate, it may be necessary to characterize 10–12 samples from each potential source. In most instances where the sources are already known, individual artifacts will result in highly accurate source assignments.
Other archaeological materials are probably more typical of ceramics in terms of the number of samples required to produce a meaningful analysis. Each material type and project offers different challenges, which may require individual consideration. Consultation with Archaeometry Lab staff may be necessary.
Archaeological samples should be packaged in individual containers (e.g., plastic bags, polyvials, or glass vials) marked with the preassigned sample ID. Do not use aluminum foil or boxes made of metal as containers for shipping samples. If polyvials are used, make certain that they will not pop open during shipment. In our experience, polyvials with lids attached by a connecting strap frequently come open shipment. Removing the strap should prevent this problem. Also, the caps on certain types of glass vials may come unscrewed during shipment. We strongly recommend use of transparent tape to prevent the caps from coming loose. Extra precautions are essential to avoid the possible loss of valuable sample materials. Always include a letter with your samples that describes the requested analysis, questions you want answered by the analysis, and any details discussed with MURR staff over the phone or by email. We receive dozens of calls and emails per day and we do not always remember the details of every conversation. Also, please include a phone number, email address, and a postal address where you can be reached if we have questions.
At MURR, our artifact database uses six-character sample IDs with a three-letter prefix and a three-number suffix. We strongly recommend that the sample vials or bags be labeled consecutively (i.e., without skips) to reduce the likelihood for error. For example, a set of 100 samples might be assigned the following sample ids: MDG001, MDG002, MDG003…MDG100 to indicate that they were submitted by an archaeologist with the initials MDG. Instead of the archaeologist's initials, the three-letter prefix might be derived from the archaeological site (e.g.,MTM to indicate Museo Templo Mayor). Check with us before assigning analytical IDs to your samples because we may wish to assign a particular series of IDs to your samples to avoid duplication in our database. If you cannot label the bag or the vial, please include a paper label in the container with the sample.
Each shipment of samples must be accompanied by a sample log sheet listing the analytical IDs and other information to be used when performing the data analysis. If you submit more than ten (10) samples, the most convenient format for providing this information is via an EXCEL or LOTUS spreadsheet on a flash drive or CD. Other information may include site names, sample descriptions, archaeological contexts, geographic coordinates for included source materials, etc.
Please check your samples before you ship them. We find that approximately 25% of the shipments we receive have errors (i.e., missing samples, extra samples, unlabelled samples, etc.). If you make an error, this makes our work more difficult and slows down progress on your samples.
The more information you send us the better we will understand the research questions for which you are seeking answers.
The proper mailing address is:
Attn: Dr. James Davenport
Archaeometry Laboratory
University of Missouri
Research Reactor Center
1513 Research Park Drive
Columbia, MO 65211
USA
Packages can be shipped by the US Postal Service, United Parcel Service, Federal Express and other carriers. All of the major delivery services know how to find the MU Research Reactor. Please send an email to let us know when your package of samples is expected arrive and the name of the carrier service.
Upon arrival at MURR, each incoming batch of archaeological samples is assigned a project number for tracking purposes. Sample IDs are compared to enclosed log sheets to determine if they conform to our preferred six-character format. Sample preparation is then assigned to one of our students who will begin preparation on the samples on a first-in-first-out basis to the best of our capability. At certain times of the year, backlogs can occur such that samples will be here for several weeks before preparation begins.
If sherd samples are large, we will remove two separate pieces about 1–2 cm² in area. One piece is used for NAA preparation and the second piece is retained in our archives for future reference. The inner and outer surfaces of the sherd are scraped using a drill burr made of either silicon carbide or tungsten carbide. The scraped specimens are subsequently brushed clean, washed in deionized water and dried for several hours. After drying, an agate mortar and pestle will be employed to crush and homogenize the pottery sample into a fine powder (about 500 mg).
Clay samples are fired at 700° C before they are ground into a fine powder using procedures described in Cogswell et al. (1996). Basalt and other hard rock types should be ground into a powder before shipment to MURR, because the Archaeometry Lab does not have routine access to the equipment necessary to pulverize very hard rock specimens. All sample powders are then dried in an oven at 100° C for 24 hours and stored in a desiccator until the samples can be weighed for both short and long irradiations to measure a total of 33 elements.
All powdered samples are prepared for MURR's short irradiation by weighing 100–200 mg of powder into cleaned high-density polyvials. Standard samples using certified reference materials from the National Institute of Standards and Technology (SRM-1633b Coal Fly Ash and SRM-688 Basalt Rock) and quality control samples made from Ohio Red Clay are similarly prepared.
The powdered samples and standards in polyvials are irradiated in pairs using a neutron flux of 8x1013 n cm-2 s-1 for five seconds, allowed to decay for 25 minutes and counted for 12 minutes using a high-resolution, high-purity germanium detector (HPGe). We call this our "short count". This cycle is repeated sequentially on the samples and standards such that during an 8-hour work day approximately 30 unknowns and seven reference standards and quality controls can be analyzed per detector system. Operation of two detectors allows us to collect data on about 60 unknowns per day. The short irradiation procedure allows determination of the following elements in most samples: Al, Ba, Ca, Dy, K, Mn, Na, Ti and V.
For long irradiations at MURR, samples of powder weighing about 150-200 mg are encapsulated in high-purity quartz vials (4mm ID by 6mm OD). Standard samples using certified reference materials from the NIST (SRM-1633b Coal Fly Ash and SRM-278 Obsidian Rock) and quality control samples made from Ohio Red Clay are similarly prepared.
Long irradiations on pottery and powdered rock specimens are performed on batches of 45-50 unknowns along with seven standards and quality controls. Two batches of samples in quartz vials are wrapped in aluminum foil and submitted as a single bundle for irradiation. An irradiation of 24 hours length in a neutron flux of 5x1013 n cm-2 s-1 is then performed on the bundle of samples. The bundles are rotated continuously during irradiation to insure that all samples receive the same exposure to neutrons. After a decay period of 7–8 days, the sample vials are cleaned and counted for 2,000 seconds ("mid count") each using HPGe detector coupled to automatic sample changers. After an additional decay of 4–5 weeks, the samples are counted a final time for three hours each ("long count"). Elements possible from the mid count are: As, La, Lu, Nd, Sm, U, and Yb. Elements reported from the long count are: Ce, Co, Cr, Cs, Eu, Fe, Hf, Ni, Rb, Sb, Sc, Sr, Ta, Tb, Th, Zn, and Zr. Concentrations of elements in the unknowns and quality control samples are determined by comparison of the unknowns to the co-irradiated standards. The quality controls serve the dual purpose of monitoring systematic differences between batches and provide a means of inter-calibration between laboratories.
Cogswell, J.W., H. Neff & M.D. Glascock (1996).
The effect of firing temperature on the elemental characterization of pottery. Journal of Archaeological Science 23: 283–287.
Lithic artifacts are cleaned with acetone and deionized water to remove all external markings (e.g., labels, fingernail polish). Obsidian artifact samples for NAA are prepared by either breaking or sawing a portion from the original artifact. The amount removed from the artifact is typically 100–300 mg depending upon the analytical procedure used. An abbreviated-NAA procedure ("short count") requires only about 100 mg of the sample and works very successfully in most of Mesoamerica (See Glascock et al. 1994) by measuring five or six short-lived elements. If the source cannot be determined by the abbreviated-NAA procedure, the sample should be submitted to our long irradiation procedure, which determines about 20 long-lived elements.
The full analytical procedure for obsidian (both shorts and long irradiations) uses 100 mg of sample in high-density polyvials and 250–300 mg of sample in the larger size high-purity quartz tube (6 mm ID by 8 mm OD). In addition, a set of NIST reference materials, SRM-278 Obsidian Rock and SRM-1633b Coal Fly Ash, used for standard and quality control, respectively, are similarly prepared.
Obsidian source materials are generally prepared by removing the weathered surface to obtain fresh (unexposed) interior fragments for analysis. We recommend that source samples be approximately half-fist to full-fist in size such that the weathered cortex can be easily removed. The interior fragments are crushed into smaller fragments (≈50 mg each) using a laboratory press. The obsidian fragments are inspected under a magnifier to eliminate those with possible metallic streaks or scars. Finally, we usually save about 10–25 g of material from each source sample for future reference.
Chert artifacts are prepared in a similar fashion to obsidian artifacts except that the sample sizes are usually larger. Short irradiation samples are typically 200 mg and long irradiation samples are typically 1 gram. For chert analyses, the NIST reference materials are used in reverse fashion (i.e., SRM-1633b is the standard and SRM-278 is the quality control). The abbreviated-NAA procedure is not successful when applied to chert.
Short irradiations on lithic samples are performed on a pair of samples using a five-second irradiation in a neutron flux of 8x1013 n cm-2 s-1. After irradiation, the samples decay for 25 minutes and are counted for 12 minutes. During an 8-hour workday, 30 unknowns and six reference and quality control standards can be analyzed per detector system. The short irradiation procedure for obsidian determines the elements Ba, Cl, Dy, K, Mn and Na and the short irradiation procedure for chert determines the following elements: Al, Ba, Ca, Dy, K, Mn, Na, Ti, and V.
Long irradiations on obsidian or chert specimens are performed on batches of ~30 unknowns along with six standards and quality controls. Two separate batches are wrapped in aluminum foil and submitted as a single bundle for irradiation. The irradiation of 70 hours takes place in a neutron flux of 5x1013 n cm-2 s-1. After decaying for about eight days, the sample vials are cleaned and counted for 2000 seconds each ("mid-count"). After the samples and standards decay for an additional 4–6 weeks, they are counted a final time for four hours each ("long-count"). Elements possible from the mid-count are Ba, La, Lu, Nd, Sm, U and Yb. Elements possible from the long-count are Ce, Co, Cr, Cs, Eu, Fe, Hf, Ni, Rb, Sb, Sc, Ta, Tb, Th, Zn, and Zr.
The time required to complete a batch of archaeological samples for NAA depends upon the type of sample, the number of samples submitted, our current sample backlog, and our students' schedules. Projects are worked on in a first-in-first-out basis. A project with no more than 100 samples will generally be completed in four to six months. Small projects can sometimes be completed more rapidly, but extremely large projects may take up to a year to finish. To obtain the best estimate of when a potential project may be completed contact the laboratory.
Yes. We will return the unused portions of all samples to you, if you make certain to so indicate on correspondence accompanying the samples. Sample portions used for neutron irradiation cannot be returned. If possible, we will retain a small archival piece of pottery for our reference collection of paste samples.
NAA costs depend upon the type of sample to be analyzed, amount of sample preparation required, number of samples submitted, and effort required to measure the elements requested. Per sample charges are much higher for a single sample than samples within a full batch, with full NAA charges generally ranging from $100 to $500 per sample. For some information about our charges, see our Charge Schedule for Instrumental Neutron Activation Analysis of Geological and Archaeological Specimens. Analyses of archaeological samples for academic research requested by students and faculty from academic institutions based in the USA may be eligible for significant charge reductions as the result of a grant to the Archaeometry Lab from the National Science Foundation. Additional information is available on our NSF Subsidy webpage describing eligibility requirements, application materials, and deadlines. To print out an application form for reduced charges (click here to print a PDF of the application form.) The NSF-subsidy program is intended to support basic research from academic institutions based in the United States and is not available to industry or for-profit agencies.
Please email the laboratory to discuss any questions not answered to your satisfaction.