Light at the Museum
Synchrotron-based large-area x-ray fluorescence (SR-XRF) and diffraction (SR-XRD) mapping has uncovered unexpected trace elements in ancient manuscript fragments. Louisa Smieska (Metropolitan Museum of Art) and Ruth Mullett (Cornell University) talk us through the process of analysis and the significance of their discovery. And give us a taste of how they navigate this complex interdisciplinary field.
Louise Smieska and Ruth Mullet |
How did you come to study this particular manuscript?
Louisa Smieska: When I was a postdoc at the Cornell High Energy Synchrotron Source (CHESS) last year, my supervisor Arthur Woll and I organized a workshop on applications of scanning x-ray fluorescence for the study of cultural heritage materials. Laurent Ferri, curator of pre-1800 materials in the Cornell Library Rare and Manuscript Collection, attended the workshop and suggested that we look into the group of fragments that Ruth was cataloguing. Happily, Ruth and I already knew each other from a course we’d taken at the Johnson Museum on campus...
Ruth Mullett: Our initial goal was to learn more about these fragments by looking at trends in pigment and color use. Initially, we were hoping to uncover how many of our pages used lapis lazuli – a blue pigment.
What did you uncover with your initial portable XRF analysis?
LS: We found that most of the blue pigments were copper-rich, suggesting that these blues were azurite, a copper carbonate mineral, rather than lapis lazuli. A few of the manuscripts we looked at showed the presence of barium in the blue areas, which we really weren’t expecting.
RM: We then selected fragments that represented a geographical and temporal range that yielded unusual or surprising results in the p-XRF, for synchrotron analysis. We were interested, for example, to find out more about the fragments that demonstrated the presence of noticeable concentrations of barite in blue pigments.
LS: We didn’t know from the portable point XRF survey that all the azurite pigments contained barium – we selected a few fragments where we had detected it, but expected that the others would not. Our scanning XRF measurements at CHESS allowed us to clearly identify which elements – not just barium – were associated with azurite, by looking at which elements correlated with copper-rich blue regions. Our scanning XRD measurements confirmed that these blue regions were in fact azurite.
What methods did you use for a more in-depth analysis?
LS: The facilities at CHESS provided several advantages over the laboratory-based point XRF survey we began with. First, we were able to use a Maia XRF detector at CHESS, which allowed us to move from point XRF measurements to fast-scanning XRF experiments of square centimeter areas. There are only a few Maia detectors in use around the world. We were able to quickly scan large areas and discover spatial trends in the elemental maps, such as confirming that barium can be associated with azurite. Second, we added simultaneous scanning x-ray diffraction (XRD) to our scanning x-ray fluorescence measurements. The diffraction information allowed us to definitively identify major compounds, not just infer their presence from the elemental maps. Finally, CHESS was able to provide much higher energy x-rays than a laboratory-based x-ray source can offer. The higher energy x-rays conferred greater XRF sensitivity to heavier elements, including barium, than a laboratory source.
A portable X-ray flouroscence (p-XRF) scanner being used on an illuminated manuscript fragment.
Distribution of elements in a fragment.
From left to right: Ruth Mullett, Louisa Smieska and Arthur Woll at CHESS with one of the manuscript fragments mounted to be scanned.
Why was the barium significant?
LS: Our synchrotron measurements showed that barium was present in trace amounts in all six of the 13–16th-century manuscript fragments we examined. At first, we were surprised to find barium in the azurite blues because we hadn’t seen this finding reported in illuminated manuscripts before. We often think of the element barium as associated with modern paints or, in smaller amounts, in natural clays or chalks, but not with azurite. For azurite, the amounts of barium involved are often so low that they are undetectable with the portable point XRF survey.
Combining scanning XRF and XRD, we found that many azurite blues contain small amounts of the mineral barite, or barium sulfate. Although barite is a fairly common mineral, we are excited because the relative amount of barium in each azurite blue is not the same, and combining this information with the amounts of other trace elements, such as iron, zinc, and antimony, might help with efforts to learn whether different fragments were originally related to one another.
RM: Research like ours may make it possible, for example, to narrow the geographic region of production by identifying unusual pigments in a palette.
How would you like to develop your research further?
LS: Expanding our study to additional manuscript fragments would be extremely valuable for uncovering broader trends in azurite trace mineral compositions. We would also like to study the composition of azurite mineral samples of known provenance, complementing the survey of fragments by evaluating similarities and differences between historic sources of the pigment. It is not clear how the purification techniques affect the trace element composition in the final pigment, so it would be exciting to recreate historic methods for grinding and washing the azurite mineral followed by a study of the trace element composition. It is frequently impossible for illuminated manuscripts to travel to facilities like CHESS for analysis; it would be helpful to compare the results of our measurements with laboratory-based scanning XRF systems to learn which trace elements in azurite are most diagnostic across measurement techniques.
Met Detective
Matching manuscript fragments was just the start of Louisa Smieska’s adventures in art analysis. Here, she tells us how she’s applying XRF analysis in her role at the Metropolitan Museum of Art.
Last fall, I was awarded a one-year Andrew W. Mellon Foundation Conservation Fellowship to work with the Department of Scientific Research at The Met, where I am working with the laboratory-based XRF scanning system housed in the paintings conservation department.
Having the ability to make scanning XRF measurements in the museum rather than at a synchrotron is relatively new, so a significant part of my role here is improving data analysis protocols. The instrument is primarily used to study paintings in The Met’s collections, but I have also been able to contribute to ongoing collaborative research efforts with other departments.
In my independent research, I am exploring applications of scanning XRF for the study of other 2D objects, particularly 19th/early 20th century photographs. There are enormous variations in the chemistry of photographic processes that are difficult to assess by eye, but strongly influence how the objects should be treated. Examining photographs with point XRF is also challenging because there is not very much inorganic material present to measure, so I am evaluating what role scanning XRF might play in examination of photographs.
Of course, I miss working with the team at CHESS, as well as the synchrotron’s unique combination of experimental flexibility and sensitivity. On the other hand, the opportunity to work with the extraordinary collections at The Met is unbelievable. Many of these objects will probably never visit a synchrotron, so it’s important to improve the methods museums can use onsite. I’m hopeful that I will find a way to continue research in the cultural heritage field that takes advantage of both lab-based and synchrotron-based experiments.
Louisa Smieska took on the project as a postdoctoral researcher at CHESS (Cornell High Energy Synchrotron Source) after completing her doctorate in chemistry. She studied fine art as an undergraduate at Hamilton College; she is now an Andrew W. Mellon Postdoctoral Fellow in the Department of Scientific Research at the Metropolitan Museum of Art in New York City.
Ruth Mullett is a medieval studies doctoral student at Cornell. She is also a fellow in the Fragmentarium project based at the University of Fribourg in Switzerland, which is building a database of fragments from different institutions.
- L Smieska et al., “Trace elements in natural azurite pigments found in illuminated manuscript leaves investigated by synchrotron x-ray fluorescence and diffraction mapping”, Appl Phys A, 123 (2017).
Louisa Smieska took on the project as a postdoctoral researcher at CHESS (Cornell High Energy Synchrotron Source) after completing her doctorate in chemistry. She studied fine art as an undergraduate at Hamilton College; she is now an Andrew W. Mellon Postdoctoral Fellow in the Department of Scientific Research at the Metropolitan Museum of Art in New York City.
Ruth Mullett is a medieval studies doctoral student at Cornell. She is also a fellow in the Fragmentarium project based at the University of Fribourg in Switzerland, which is building a database of fragments from different institutions.
