Handheld XRF for Art and Archaeology

This much-needed volume is a testament to the increasing popularity of handheld or portable X-ray fluorescence analysis (PXRF) in art history and archaeology. That popularity is due to several factors. First, unless samples are preprocessed in some way, the technique is nondestructive and can be employed on objects that would otherwise be impossible to sample. Second, as long as only elemental data is sought, virtually any material can be sampled. Third, the instrument’s portability means that data collection can take place in the field, thus allowing use on an excavation site or in museums housing objects that cannot be taken out of the curating institution. Finally, although PXRF units range in cost from $35,000 to $50,000 for a basic kit, when prorated over the average use-life of an instrument, per-sample costs are generally less than one dollar. Consequently, it is possible to cost-effectively develop large, statistically reliable data sets using PXRF technology.

However, PXRF use can be problematic for several reasons. First, as the volume under review makes clear, modern handheld analyzers are not “point-and-shoot” devices, and their use by untrained individuals will lead to spurious interpretations based on bad data. Consequently, prospective users lacking previous training in basic physics would be well advised to budget sufficient time and money to obtain the required level of analytical sophistication prior to beginning a research project.

Second, data collection is not always nondestructive, as (1) surfaces must be cleaned prior to data collection, and (2) compositional heterogeneity of many archaeological materials requires that such samples be destructively processed prior to analysis.

Finally, interlaboratory comparability of results has often been poor because of inherent variation in individual instruments. To assure interlaboratory comparability, each analysis must thus include a set of readings taken from a published reference standard.

If all of this is bad news, the good news is that in many applications, modern, properly used, and carefully calibrated handheld instruments are capable of closely approximating the precision and accuracy of more expensive bench-mounted units. The wide range of successful applications discussed in this book is a testament to that utility.

In the following review, I have grouped chapters covering related topics together. Volume chapters cover a variety of material culture classes including documents, metals, ceramics, glass, ethnographic materials, and archaeological sediments. In addition, the book includes a chapter on the use of a handheld analyzer as an aid to archaeological excavation. The introductory chapter is a very useful review of the basics of the XRF technique and an evaluation of the potentials and pitfalls related to the use of PXRF analyzers in cultural heritage investigations. This chapter would make a fine introduction to the use of XRF in any upper-level class on material culture analysis.

Chapters 2 and 7 deal with the analysis of bronze and silver artifacts, respectively. Smith (ch. 2) argues that PXRF has the potential to transform the study of Renaissance bronzes by facilitating collection of information on attribution and authenticity, composition, construction, and historical linkages. The chapter provides an in-depth review of the procedures used by the National Gallery of Art in Washington, D.C., to investigate Renaissance bronzes and should prove valuable to anyone engaged in similar studies or contemplating adding XRF analysis to an existing program.

Mass and Matsen’s (ch. 7) discussion of the application of PXRF to the study of historic silver artifacts from the New World reiterates many of the concerns raised by Smith in chapter 2. Using case studies of artifacts from the New York Historical Society collections, the authors demonstrate that it is possible to determine the source of silver ores and explore questions of post-manufacturing modification practices. However, Mass and Matsen point out that the surface of any silver artifact may not be representative of the item’s bulk composition. This is a problem not restricted to the analysis of silver alloys, and the authors do provide suggestions for dealing with this constraint.

Chapters 3–6 deal with applications targeting photographs, paintings, documents, and paper. Chapter 3 (Stulik and Kaplan) introduces the reader to the use of XRF as an aid to determining the structure and composition of photographs. Results can be used to help authenticate, date, and establish method of production.

McGlinchey’s (ch. 4) introduction to XRF studies of pigments used in paintings highlights two limitations of the technique. First, the relative thinness of the pigmented surface means that the X-ray beam will pass entirely through the pigment, substrate, and canvas or backing medium, thus collecting data from all layers simultaneously. Second, since PXRF analyzers return only elements, the complex compounds an artist may have used to produce a particular pigment may be invisible to the analyzer. For these reasons, McGlinchey suggests that PXRF of pigments should utilize only qualitative data and that the method may be most effective as a screening technique prior to more discriminating analyses.

Trentelman, Schmidt Patterson, and Turner (ch. 5) discuss the analysis of illuminated manuscripts. They note an additional complication resulting from the presence of inks often used in conjunction with illuminations. Moreover, pigments and inks on underlying pages may be detected as well as the targeted location. For these reasons, the authors suggest use of an ancillary procedure such as Raman microspectroscopy to provide a complementary analysis.

Chapter 6 (Barrett, Shannon, Wade, and Lang) exploits one of the inherent advantages of PXRF: the ability to quickly compile large, statistically reliable data sets. Results of the authors’ analysis suggest that PXRF may be a valuable tool to help conservators identify the causes of differential aging throughout a collection. However, they caution that the method lacks the necessary precision to allow monitoring of changes in single items through time.

Chapters 8 and 13 discuss the analysis of porcelain and Maya ceramics, respectively. Bezur and Casadio tackle the application of PXRF to the study of porcelain objects. The authors note that handheld instruments cannot differentiate between component phases in a porcelain body and that the method is thus a bulk analysis technique. Nonetheless, the data suggests PXRF analysis can successfully discriminate between various porcelain wares, especially when complemented by interdisciplinary collaboration of archaeologists and experts in ancillary fields such as art history, materials science, and geology.

Aimers, Farthing, and Shugar offer the only paper devoted to low-fired prehistoric ceramics. Their paper is a cautious comparison of results obtained using a bench-mounted XRF unit and a handheld analyzer to record elemental composition of selected Maya ceramics. The authors are led to argue that PXRF can approximate the accuracy and precision of the bench-mounted unit. They note that the noninvasive (although some surface preparation of sampled sherds took place in the reported study) and portable nature of the technique allows for the possibility of sampling large collections in the field.

Chapter 9 (Shugar and Sirois) provides an introduction to the use of XRF as a means of identifying heavy metal pesticides historically used as preservatives on ethnographic collections. This has become a problem for conservators and in some cases for descendant communities wishing to repatriate cultural objects or use curated specimens in contemporary ceremonies. The chapter includes a discussion of both qualitative and quantitative analyses and should be useful to any curator or conservator faced with identifying toxins in an ethnographic collection.

An advantage of handheld analyzers is the ability to conduct on-site analysis of artifacts, features, and sediments as an aid to excavation. Donais and George (ch. 10) report the use of a handheld analyzer during excavations at the site of Coriglia near Orvieto, Italy. The instrument was used (1) to help trace the former route of lead pipes in a first-century B.C.E. water distribution system, (2) to distinguish differences in ceramic tiles associated with a hypocaust to help determine where the tiles were produced, and (3) to identify pigments used in the production of Pompeian-style frescoes.

Neff, Voorhies, and Umaña (ch. 11) provide two examples of how PXRF could be used in field situations to investigate archaeological sediments. However, the actual experiments reported were conducted in the laboratory on sediments recovered from two separate archaeological sites. Samples from Tlacuachero, an Archaic-period shell midden site in Chiapas, Mexico, were analyzed in an attempt to determine the soil composition of a prepared floor presumed to be made of clay, upon which processing of seafood tissue took place. The second experiment reported was designed to recover compositional data on a volcanic tuff layer present in archaeological deposits in the Ataco region of western El Salvador.

Ferguson’s contribution (ch. 12) reviews the basics of obsidian sourcing using XRF. The chapter provides an in-depth review of the need to carefully calibrate readings, whether using a bench-mounted or handheld instrument.

Kaiser and Shugar’s paper (ch. 14) highlights issues in PXRF analysis of glass. Using case studies, the authors demonstrate that (1) careful calibration is necessary to achieve accurate results, and (2) cleaning or sanding of the analyzed surface is necessary to avoid significant mischaracterization of the glass matrix.

The volume editors suggest that they intended this collection of papers to be a compendium of “best practices” in terms of the applications discussed. So while the book presents the results of various analyses, and these are important contributions in their own right, the real value of the volume lies in its utility as a comprehensive handbook of handheld XRF. The papers collected in the volume address virtually every application of XRF technology likely to be of interest to archaeologists, with the exception of studies of bone chemistry. If there are any weaknesses to the volume, it is the lack of a more comprehensive treatment of prehistoric ceramics and the exclusion of studies of lithics other than obsidian.

 Handheld XRF for Art and Archaeology will likely remain the single most comprehensive treatise on the use (and abuse) of  PXRF for some time to come. The book would be a useful reader in an upper-level undergraduate or graduate class on material culture analysis and should rapidly become a standard part of the toolkit of anyone interested in using PXRF technology to investigate the composition of cultural heritage materials.

John D. Richards
Department of Anthropology
The University of Wisconsin-Milwaukee