Smithsonian's Museum Conservation Institute Research Experiences for Undergraduates in Archaeological and Museum Conservation Science
2013 Project Summaries
Project Title: Quantification of heavy metals in organic matrices
Intern: Michelle Carpenter, Idaho State University
Advisors: Nicole Little
This project focused on accumulating data for the isotopic abundance of heavy metals in human remains through the use of inductively coupled plasma-mass spectrometry (ICP-MS). Since the arrival of the first colonists in Jamestown, high-status individuals in the United States typically ate from lead-glazed pottery whereas lower-class individuals would primarily eat from wooden bowls and utensils. This pattern of behavior can help researchers better understand the health and status of populations from the Jamestown settlement to the present day. In addition to lead content, the accumulation of mercury and arsenic in bone is indicative of prescription medicine used historically as well as occupational markers. This information can provide insight about an individual’s life and social status. Results from this work support the correlation between known high status individuals and the increased presence of heavy metals in their skeletal remains. In addition, this project quantified heavy metals in feathers from a native bird in Africa (Kori Bustards). This project promises to shed light on differences between those birds raised in captivity versus those born in the wild and brought to live in captivity for over twenty-five years.
Project Title: Proteomics application to bone aging studies
Intern: Raquel Fleskes, University of Maryland
Advisors: Mehdi Moini
Understanding the behavior of intact bone proteins over long periods of time has been a question of interest to researchers for decades. Many mechanisms of protein diagenesis have been hypothesized (i.e. aggregation, hydrolysis, and microbial deterioration), but direct archaeological evidence has yet to be presented in a systematic way. This project focused on well dated human bone samples of various ages (150yrs to ~40,000yrs). These samples were tested with SDS-PAGE Gel Electrophoresis and LC/MS to gain a visual and quantitative understanding of intact bone protein behavior. The gel results show similar protein patterns for bones aging from modern to ~40,000 years old with evidence for both aggregation and mechanisms that decrease protein content, such as microbial deterioration and hydrolysis. The amount of deterioration in bone is apparently non-linear and unique to the environment in which it degrades. Although this does not reveal the exact mechanism of bone diagenesis, it can give direct evidence of a mechanism occurring over time within these bones to support previously proposed hypotheses of bone protein deterioration. To obtain a more in depth understanding, gel bands were cut and analyzed with LC/MS to determine their protein content. Mass spectrometry revealed that bone protein behavior is much more complex than previously thought, as each band was found to contain a mixture of proteins. The application of this technology has brought about a new level of understanding and interpretation of the behavior of intact bone proteins that will continue to bring new information to light.
Project Title: Spectroscopic imaging and analysis of compositionally zoned jadeite
Intern: Kimberly Foecke, Penn State University
Advisors: Edward Vicenzi
Chemical sourcing studies are being used with increasing frequency to better understand museum collections. Determination of provenance by matching elemental signatures is a potential source of vast information about an object. Analysts use a variety of techniques to obtain accurate elemental information, but none are perfect and most are slow and destructive. The goal of this project was to determine if cathodoluminescence (CL) spectral intensities (a non-destructive qualitative spectroscopic technique) can be used to quantitatively determine trace chromium and manganese in jadeite in two dimensions using zero-dimensional wavelength dispersive X-ray microanalyses as a benchmark. Results suggest that jadeite is an extremely complex and heterogeneous material that demands a spatially resolved analytical technique. Image registration software may be a useful and time saving tool to co-locate positions on a sample between CL imaging and electron microprobe images (EPMA). Between the two techniques, CL allows analysis in far greater spatial detail than hyperspectral imaging with EDS or backscatter. Standards and peak shape references are an extremely important consideration when attempting EPMA, and should take into account the composition or expected composition of the material to be analyzed on a case-by-case basis. Overall, quantification maps of trace elements in jadeite may be possible and robust, but more work is still needed to confirm this for chromium. Further work with activators of CL in jadeite (and correlation of elements with CL spectral peaks) is needed in order to extend this technique to other trace elements. This study is an important step towards the ultimate goal of quickly and accurately sourcing jadeite to its area of geologic origin.
Project Title: Comparison of several mass spectrometry techniques for dye analysis
Intern: Lea Hair, Centenary College of Louisiana
Advisors: Mehdi Moini
Various analytical techniques have been applied to identify organic dye compounds in historic textiles with mass spectrometry (MS) being the most common. This project tested and compared four MS methods for their capacity to analyze dyes: direct analysis in real time (DART-MS), matrix-assisted laser desorption/ionization (MALDI-MS), laser desorption/ionization (LDI-MS), and high performance liquid chromatography (HPLC-MS). In addition, three methods of sample introduction were tested: 1) direct sample introduction where fabric was directly analyzed without any sample preparation, 2) minimal sample preparation where samples were prepared using a liquid additive, a solvent or cationization agent to help extract and ionize the dye compounds, 3) dye extraction techniques where the dye compound was extracted from the fabric using oxalic acid, trifluoroacetic acid (TFA), or ethylenediaminetetraacetic acid (EDTA). To compare these dye analysis methods, five historic silk samples dyed by either Indigotin, Ellagic acid, Rutin, or Guinea Green B were used as test specimens. An additional sample dyed by a-Flavonoids, a class of dye compounds, was tested as an unknown. Results suggest that MALDI and LDI are only marginally useful for dye analysis and are best for detecting indigo when other dyes are present on a textile. DART-MS successfully detects indigo, but no other dye compounds. The preferred sample introduction method is direct introduction of the unprepared textile samples which causes no visible change to the introduced samples (provided they never cover the gas stream entirely), and gives a noticeable indigo peak for all indigo containing samples. No conclusive data was gathered for any other samples and work is ongoing to identify the best method for detecting other dyes.
Project Title: Limit of detections for sodium chloride using SEM-EDS and XRD analysis
Intern: Miriam Hiebert, University of Richmond
Advisors: Nicole Little
Determining the presence of salts in a sample is one of the most common requests for analysis at MCI. The two methods typically used are X-ray diffraction and SEM-EDS. However, most samples brought to MCI are heterogeneous mixtures of several different components. Often these other components also generate a signal during analysis, particularly using XRD. The signal they produce can potentially overpower the signal of a salt that is in fact present in the sample. An inability to detect salts could be detrimental to the conservation of the object or even to other analytical equipment that might be used for further testing. The focus of this project was to determine the limits of detection (LOD) for salts in a variety of different matrices that simulate various possible crystalline structures in a sample. Mixtures were made using various concentrations of sodium chloride in a highly crystalline matrix, a semi-crystalline matrix, and a completely amorphous matrix, and the ability of the XRD to recognize the salt was measured. Probabilities were generated to establish the chance that various concentrations of salt might be present in various samples despite not being visible on the spectrum. Similar probabilities were generated for the SEM-EDS and its ability to identify the sodium and chloride components in various concentrations and at various magnifications. Results from XRD analyses show that samples containing salt in a highly crystalline matrix have the highest LOD and salt in the amorphous matrix had the lowest LOD. Results from SEM-EDS analysis show that at larger sample sizes (>800µm), the LOD is ~0.5%. At smaller sample sizes, the LOD is ~5%.
Project Title: Characterizing daguerreotype corrosion
Intern: Connor Hilton, Brown University
Advisors: Edward Vicenzi
Daguerreotypes were the first practical form of photography developed in 1839. Their study allows us to discern information about both technology and sociological aspects of the mid-19th century. Therefore, these are important objects to preserve. The metallic surface of a daguerreotype is however subject to multiple types of corrosion. This project investigated the nature and chemistry of such surface corrosion on three daguerreotypes using SEM-EDS and XRF analysis. These daguerreotypes are better understood when compared to X-ray spectra simulated in the computer for different geometries. Results suggest that X-ray simulations do provide information about mercury/silver nanoparticles and the gold/silver gilded layer of the daguerreotypes. With fewer unknown variables, this could be a useful technique to understand a daguerreotype. Several different types of corrosion were detected including copper and sulfur based corrosion. The simulations show promise in understanding these different corrosion type and their thickness. This research produced a new characterization of surface corrosion on daguerreotypes which facilitates a better understanding of this important nineteenth century technology and informs conservation treatments.
Project Title: Understanding of paint analysis with the use of portable IR Diffuse Reflectance spectroscopy
Intern: Ryan Munger, Beloit College
Advisors: Jennifer Giaccai
Knowing the composition of paint can be useful in a conservation treatment, determining possible sources of deterioration, or learning more about artists’ working methods. Portable diffuse reflectance FTIR provides the ability to run non-destructive analysis on a painting. Unfortunately, the spectra obtained are much more difficult to interpret, and the amount of literature and reference libraries available are minimal compared to ATR or transmittance FTIR. Principal component analysis (PCA) is a statistical method which measures the variance across a data set and is able to plot samples against two of the principal components, allowing the user to distinguish different groups in the data set. PCA is useful because the diffuse spectra are difficult to interpret individually, but contain enough differences that they are mathematically separable. Both laboratory-made, controlled-known samples and uncontrolled-known paint samples from paintings at the Smithsonian American Art Museum have been studied. Analyzing the data using PCA has so far given mixed results. It is fairly accurate at separating acrylic, oil, and alkyd, but is less accurate at separating pigments. This method does appear to be a beneficial way to interpret diffuse FTIR, but the PCA method needs refinement.
Project Title: Strength in the skin: innovation and evolution of synthetic plywood adhesives in early-20th century aeronautics
Intern: Eric Schmittgen, John Carroll University
Advisors: Odile Madden
Plywood has existed for thousands of years, but its applications were limited before the development of a waterproof adhesive for the glue line. Development of synthetic adhesives in the 20th century was advantageous because of their resistance to water and mold growth which facilitated new outdoor uses for plywood, including aircraft construction, since it was strong yet lightweight. Plywood initiated the shift from the truss construction of aircraft to monocoque, where the skin provides the structural strength. A variety of different methods were developed while searching for the best way to apply and cure the adhesives including liquid spreading methods as well as hot pressing and cold pressing techniques for curing. This project involved archival research, visual examination, Raman spectroscopy, and SEM-EDS analysis to characterize and develop a technological history of plywood, a necessary step in understanding and preserving Smithsonian collections. The archival research suggests plywood development increased in the 1920s due to the application of synthetic adhesives such as phenol and urea formaldehyde to plywood which could be applied in liquid or dry form. The plywood samples analyzed in this project were consistent with phenol formaldehyde and urea formaldehyde spectra with enough distinction that can likely be used to source other plywood artifacts in the Smithsonian collection. In addition to information from the adhesives, it was found that the number of plies, their thickness, and the grain sequence are indicative data that may help characterize historic plywood pieces.
Project Title: Cellulose acetate plastics: a study of 160 NASM aircraft recognition models
Intern: Paul Trigonoplos, University of Maryland
Advisors: Odile Madden
Cellulose acetate plastics, common since the late 1920’s, are heavily represented in the Smithsonian collections. These artifacts are chemically unstable, and a subset of injection-molded objects has begun to degrade catastrophically. One of these collections from NASM consists of ~160 aircraft recognition models from World War II composed of a complex multicomponent cellulose acetate mixture which is currently studied at MCI. This project focused on the degradation of cellulose acetate plastic mixtures in the model planes and prepared experimental samples in order to identify which compounds are most responsible for this degradation, with special focus on the role of shrinkage in fracturing. Two-component mixtures of cellulose acetate and diethyl phthalate were prepared and characterized with FT-Raman spectroscopy. These and a selection of model planes were also analyzed by GC-MS-EGA which provides insight into the slower process of volatile component off-gassing over time. This loss of material to the air results in shrinkage of the remaining plastic material which can result in fracture. Results showed triphenyl phosphate, diethyl phthalate, and acetic acid off-gassing from the planes and contributing to shrinkage. X-radiographs of selected planes were collected to visualize their construction and fracture patterns, and better understand the role of exhibit mounts in breakdown. Observed air bubbles in the models was apparently unrelated to degradation, but the orientation of the polymer as it entered the mold correlated to a fracture pattern along the same orientation. X-ray diffraction analysis identified significant corrosion on the mounting screws which, when coupled with the model shrinkage, results in an overbearing stress on the inside of the models contributing to aggressive fragmentation near the mounting points.
Project Title: GC-MS on resins and portable FTIR on textiles
Intern: Marcie Wiggins, University of Maryland
Advisors: Jennifer Giaccai
Diterpenoid and triterpenoid resins can be obtained from many different trees/plants/vegetation. The most common method of analyzing these resins is GC-MS. To determine the best means for analyzing samples two different derivatization methods were tested, BSTFA and Meth-Prep II. Both longevity and MS sensitivity of the derivatized compounds were considered for these analyses. This method was used on reference resins from Central America to analyze and build a library for future archaeological samples. In both MS sensitivity and sample longevity, BSTFA appeared to be the best derivitization method for terpenoid resins. However, tests also indicated that our BSTFA method could be further optimized. In a second project, various non-destructive Fourier transform infrared spectroscopy (FTIR) techniques have been applied to fiber identification. Different algorithms, search regions, and libraries were used to assess the use of portable FTIR by textile conservators to identify materials without destructive sampling. Although not always as straightforward as typical sampling techniques, portable FTIR can be used as a preliminary identification method for textiles without requiring a sample.