Investigations of Archaeological and Historical Materials 1968-1976
There are three general emphases in this program during this period: (1) elemental analysis of ceramics and metal objects in order to discover geological/regional origins; (2) neutron activation of paintings to reveal hidden details and techniques used by artists, and; (3) development of new methods and instruments useful for the first two topics.
A good example of the first topic is a problem from Precolumbian Mexico. Teotihuacan and Oaxaca were (and still are) 500 km apart, separated by mountains and desert, although numerous Oaxacan-style sherds were found in one section of Teotihuacan. Were they made in Oaxaca and transported to Teotihuacan, with all the implications of cultural influence, trade and transportation, or were they of local origin and only imitated Oaxacan style? Samples of sherds were taken and irradiated with neutrons in the Brookhaven High Flux Beam Reactor along with known standards and clays of known geographic origin. Statistical analyses were carried out in order to determine the significance of the results. Analysis showed unequivocally that the Oaxacan-style pottery from Teotihuacan was of local origin, and not transported from Oaxaca.1
1. "Correlation between Terra Cotta Figurines and Pottery from the Valley of Mexico and Source Clays by Activation Analysis" R. Abascal-M., G. Harbottle and E.V. Sayre in Archaeological Chemistry, C.W. Beck, ed. Advances in Chemistry Series 138, pp. 81-99 (1974).
Owners of fine paintings are understandably reluctant to subject their property to destructive analysis. Nevertheless, there is great interest in provenance, selection of materials and artist's techniques. Autoradiography is a method by which elements in a work of art are made slightly radioactive for a short time after activation. During this period, a painting is placed in close proximity to a sheet of x-ray film, and exposure of the film by the transient radioactivity yields the spatial distribution of different elements in the painting. Activation was usually done by exposing a painting to x-rays. X-radiography of paintings has a number of limitations including little absorption of x-rays by light elements and the opposite problem: too-large absorption by the heavy element, lead, which may be ubiquitous as the pigment, lead white. One solution to these problems was to activate elements in the painting with neutron irradiation, which did not favor activation of heavy elements. Neutrons were available from nuclear reactors. The Brookhaven HFBR was suitable for this purpose, and the Brookhaven Medical Research Reactor was even better, since it permitted irradiation of large objects like paintings at ports originally intended for irradiation of large living organisms. Analysis of neutron autoradiographs revealed details of brush techniques invisible by x-ray autoradiography, pigment selection where pigments contained significant concentrations of elements lighter than lead, and underpaintings which would have been obscured by large amounts of lead white.2
2. "Neutron Activation Autoradiography of Oil Paintings" E.V. Sayre and H.N. Lechtman, Studies in Conservation 13 161-185 (1968).
The most basic requirement for analysis and dating of materials by nuclear methods is accurate counting of nuclear disintegrations. Very unstable nuclei disintegrate very rapidly, producing large numbers per unit time of alpha particles, beta particles or gamma rays. It is relatively easy to produce accurate determinations of concentrations or half-lives of such species, because there are abundant numbers of particles or photons to count. However, it is frequently required to count species that are only slightly radioactive, i.e., their half-lives are very long. Consequently, there are not many disintegrations per second, and not many counts at any given time, so accurate counting vs. background noise is difficult. It is much easier to measure the ratio between two similar values than to measure one noisy parameter. Therefore, researchers in this program developed a instrument to measure the ratio of counts from a sample with a very long unknown half-life vs. that of a reference source with a known very long half-life using two balanced counters. Demonstration of the performance of this instrument using 60Co (t1/2=5.26 y) vs. a 36Cl standard (t1/2=3 x 105 y) produced a measured t1/2 for 60Co of 5.24 ± 0.21 y after counting for only 14 days.3
3. "A Differential Counter for the Determination of Small Differences in Decay Rates" G. Harbottle, C. Koehler and R. Withnell, Rev. Sci. Instrum. 44 55-59 (1973).
Last Modified: June 28, 2012