Production and Distribution of Plumbate Pottery: Evidence from a Provenance Study of the Paste and Slip Clay Used in a Famous Mesoamerican Tradeware

Laboratory Analysis

The clays and tempers sampled during the survey were taken back to the archaeometry laboratory at the MU Research Reactor (MURR), where they were prepared for analysis. The clays were mixed with de-ionized water, pressed into petrie dish molds, then fired to 700 degrees for one hour in air before being prepared for analysis. Aliquots of each test tile were prepared for INAA by crushing several hundred milligrams in an agate mortar to yield a fine powder. The remainder of each specimen was left intact for analysis by LA-ICP-MS. Intact pieces of Plumbate pottery that had previously been analyzed by INAA were retrieved from the MURR ceramic paste archive for analysis by LA-ICP-MS.

Instrumental Neutron Activation Analysis

For INAA, the powdered clay test-tile samples were oven-dried at 100 degrees C for 24 hours. Portions of approximately 150 mg were weighed and placed in small polyvials used for short irradiations. At the same time, 200 mg of each sample were weighed into high-purity quartz vials used for long irradiations. Along with the unknown samples, reference standards of SRM-1633a (coal fly ash) and SRM-688 (basalt rock) were similarly prepared, as were quality control samples (i.e., standards treated as unknowns) of SRM-278 (obsidian rock) and Ohio Red Clay.

INAA of ceramics at MURR, which consists of two irradiations and a total of three gamma counts on high-purity germanium detectors, constitutes a superset of the procedures used at most other laboratories (Glascock 1992; Neff 2000). A five-second irradiation through a pneumatic tube system, which is followed by a 720-second count, yields gamma spectra containing peaks for the short-lived elements aluminum (Al), barium (Ba), calcium (Ca), dysprosium (Dy), potassium (K), manganese (Mn), sodium (Na), titanium (Ti), aluminum and vanadium (V). A 24-hour irradiation is followed by a seven-day decay, then a 2,000-second gamma count (the "middle count"), then an additional three- or four-week decay, and, finally, a count of 9,000 seconds. The middle count yields determinations of seven medium-halflife elements, namely arsenic (As), lanthanum (La), lutetium (Lu), neodymium (Nd), samarium (Sm), uranium (U), and ytterbium (Yb), and the final (long) count yields measurements of 17 long-halflife elements, namely cerium (Ce), cobalt (Co), chromium (Cr), cesium (Cs), europium (Eu), iron (Fe), hafnium (Hf), nickel (Ni), rubidium (Rb), antimony (Sb), scandium (Sc), strontium (Sr), tantalum (Ta), terbium (Tb), thorium (Th), zinc (Zn), and zirconium (Zr).

The same basic MURR INAA procedures outlined above had been employed previously in the analysis of Plumbate sherds recovered from Classic period sites on the central Pacific Guatemalan coast (Neff 1995). These analyses confirmed the findings reported in the earlier Plumbate study (Neff 1984; Neff and Bishop 1988), namely that Plumbate found outside the production region separates into two distinct chemical groups, San Juan and Tohil. Earlier INAA data from BNL were used to check results obtained using the MURR data, but, since fewer elements were determined at BNL, most comparisons relied only on the more complete MURR data.

Analysis of Plumbate Surfaces and Clay Test Tiles by LA-ICP-MS

Although INAA has been the workhorse characterization technique used in archaeological provenance investigations for the past 35 years, decommissioning of research reactors and phasing out of archaeological INAA programs has severely curtailed the availability of INAA.  This trend is likely to limit INAA to a quality-control role in archaeometry within the coming decade (Neff 2000). Fortunately, the analytical scene is currently being revolutionized by the advent of highly precise and sensitive characterization techniques based on inductively coupled plasma-mass spectrometry (Kennett et al. 2001). In ICP-MS, a plasma torch capable of sustaining an argon plasma at temperatures above 8000° C is used to ionize injected samples, which are then sent into a quadrupole or magnetic-sector device, where they are separated according to mass and charge, so that the detector at the other end records only a very small atomic mass range at a time. By varying instrument settings, the entire mass range can be scanned within a short period of time.

Recognizing both the potential advantages for archaeology of proliferating ICP-MS instruments on university campuses and the potential pitfalls inherent in the changing analytical scene, Mike Glascock and I wrote a NSF grant in 1999 to obtain a high-resolution magnetic-sector ICP-MS instrument for the MURR archaeometry laboratory. Our funding effort was successful, and in June 2000 a Thermo-Elemental (formerly VG Elemental) Axiom magnetic-sector ICP-MS was installed at MURR. A schematic diagram of a magnetic sector ICP-MS is shown in Figure 8.

In the Axiom, the ion beam is focused on a collector slit that can be tuned so that masses as close as 0.001 atomic mass units can be resolved from one another. This gives the Axiom the capability to resolve a large number of polyatomic interferences that would be completely unresolvable with quadrupole instruments, thus dramatically increasing instrumental sensitivity and precision. In liquid samples, the instrument is capable of measuring sub-ppb amounts of the vast majority of elements.

Since the first archaeometric uses of ICP techniques in the early 1980s (e.g., Hart and Adams 1983; Hart et al. 1987), most applications have required digestion of solid samples with heat and/or strong acids, which is both time-consuming and unpleasant. An alternative sample-introduction technique, which became available in the early to mid-1990s, is laser ablation (Pollard and Heron 1996; Campbell and Humayun 1999). In this approach, a pulsed laser ablates a small portion of a solid sample, and the resulting vaporized solid is then sent into the ICP torch. In principle, LA-ICP-MS can be used to generate bulk compositional data on solid samples without chemical digestion. In practice, there are a number of obstacles related to data standardization that currently limit the extent to which LA-ICP-MS will compete with bulk techniques, such as INAA or digestion ICP-MS.

The other strength of laser ablation is its capability to do spot analysis (or microanalysis) of spatially segregated components in the artifact fabric. Thus, separate analyses can be obtained for individual temper grains and/or for areas of clay matrix that contain no temper grains. For the clay test tiles analyzed in this study, areas of matrix free of non-plastics were targeted for analysis (Figures 9 and 10). Since each pass of the laser only ablates about five microns or so of material, slipped and pigmented surfaces can be analyzed in situ simply by placing a sherd fragment in the laser chamber with the slipped surface facing the laser beam. This is the approach that was used for the analysis of Plumbate slip materials.

The study of Plumbate surfaces was one of the first conducted on the new Axiom at MURR. Thus, approaches to standardizing the data had to be worked out from scratch. A basic problem in LA-ICP-MS is that it is difficult to monitor the amount of material removed by the laser and sent to the ICP.  Conditions such as hardness of the material, position of the sample in the laser chamber, whether or not it is level, and other conditions clearly affect how much material reaches the torch and thus the intensity of the signal monitored for the various atomic masses of interest. In addition, instrumental drift in the ICP-MS affects count rates.

With liquid samples, internal standards are typically used to counteract instrumental drift, but this approach is not available when the material for analysis is ablated from an intact sample. If one or more elements can be determined or assumed independently, then these can serve as quasi-internal standards. For instance, rhyolitic obsidian has relatively consistent silica concentrations, and we have found that ratios of count rates to the silicon count rate yields "normalized count rates" that can be calibrated using external standards so as to yield concentrations that are in good agreement with INAA and XRF measurements.

The internal heterogeneity of clays and the diversity of materials used in ceramic manufacture preclude assuming a value for any single component, such as alumina or silica, in the Plumbate study. Another approach would be to measure one or more major oxides with some other surface technique, such as XRF, then use the independent measurements to normalize the LA-ICP-MS data in order to correct for differential ablation and instrumental drift. However, this approach would still suffer from the problem of within-sample heterogeneity, and it basically doubles the analytical effort. One approach that did yield results in reasonably good agreement with INAA data for several clays of known composition was to normalize the count rates for each element so that they sum to a single standard value, e.g., one million, for all standards and unknowns. A regression of normalized counts on elemental concentration in the standards then yields a calibration equation that can be used to calculate elemental concentrations in the unknowns. The basic assumption of this approach is that the 43 elements being measured represent essentially all of the material, other than oxygen, that is ablated from the samples. We tried incorporating corrections for isotopic abundance and for difference in resolving power between the measured elements, but these had little effect on the calculated elemental concentrations. The one correction we did have to make was for the overwhelmingly larger aluminum counts, a reflection both of its relatively high concentrations in the clays and of the fact that it can be measured at low resolution.

Previous Page  |  Table of Contents  |  Next Page

Return to top of page