JAIC 2004, Volume 43, Number 1, Article 4 (pp. 39 to 54)
JAIC online
Journal of the American Institute for Conservation
JAIC 2004, Volume 43, Number 1, Article 4 (pp. 39 to 54)

IDENTIFICATION OF THE PRE-COLUMBIAN PIGMENT MAYA BLUE ON WORKS OF ART BY NONINVASIVE UV-VIS AND RAMAN SPECTROSCOPIC TECHNIQUES

MARCO LEONA, FRANCESCA CASADIO, MAURO BACCI, & MARCELLO PICOLLO



2 MATERIALS AND METHODS


2.1 REFERENCE MATERIALS

Synthetic indigo was obtained from Aldrich (22,929-6; Indigo, synthetic, dye content approx. 95%) and Janssen (21.213.67, Indigo, synthetic). Palygorskite clay from Gadsden County, Florida (commonly called attapulgite in the United States), was obtained through the University of Missouri–Columbia Source Clay Minerals Repository, Columbia, Missouri. Maya blue was prepared according to a procedure derived from procedures available in the literature (Van Olphen 1966; Littmann 1982). The starting materials for the synthesis of the pigment consisted of 0.1 g of indigo and 5 g of palygorskite, mixed and then thoroughly ground under 30 ml of water. After decanting the excess water, the slurry was dried at 90�C for one hour. A small portion of the dry mixture was removed to provide a sample of the unheated mixture. The bulk of the mixture was subsequently baked at 120�C for 48 hours, after which the color had noticeably shifted, from gray-blue to blue-green. The pigment was left to cool to room temperature, and treated with 5 ml of concentrated HNO3 for 15 minutes to remove the fraction of indigo not complexed by the palygorskite. The oxidative scission of indigo to isatin and isatoic anhydride was evidenced by the yellow coloration of the liquid phase. No color shift of the solid phase was noticed. The pigment was then decanted, rinsed repeatedly with distilled water to neutral reaction, and dried at 60�C.


2.2 ARCHAEOLOGICAL SAMPLES

Spectra of Maya blue of pre-Columbian origin were obtained from a fragment of a wall painting from the Templo Mayor in Mexico City, Mexico (Sample 31, Tlaloc Shrine, Construction Stage II, A.D. 1375–1427 [Grimaldi 2000, 104]) and, noninvasively, from a Jaina statuette from LACMA's collection (M.76.157, Mexico, Campeche, Jaina, A.D. 700–900) (see fig. 2).


2.3 ANALYTICAL TECHNIQUES


2.3.1 UV-Visible –Near Infrared

The powders of pure indigo, palygorskite, and Maya blue, as well as a set of mixtures made with indigo and barium sulfate, were analyzed using a Perkin-Elmer λ19 spectrophotometer equipped with a 60 mm integrating sphere. The diffuse reflectance spectra were acquired from 320 to 2500 nm, step 1 nm. The resolution of the spectrophotometer was �0.2 nm and �0.8 nm for the UV-Vis and near infrared (NIR) ranges, respectively. Calibration was performed by means of barium sulfate (BaSO4) powder.


2.3.2 Fiber Optics UV-Visible Reflectance Spectroscopy

FORS spectra were recorded with a CARY 50 UV-visible spectrophotometer, equipped with a fiber optic probe manufactured by Innovaquartz (Phoenix, Ariz.). The probe was specifically designed to provide extended transmission in the UV range, enhanced energy throughput, and an analytical spot of 1 mm in diameter at a working distance of 2 mm. The probe consists of a bundle containing one central 800 μm core diameter high-OH silica fiber, surrounded by eight 600 μm core diameter high-OH silica fibers, partially fused to increase the packing fraction and ground and polished to form a focusing lens. The central fiber was connected to the source port of the instrument (a scanning monochromator and a xenon flashlamp), and the eight peripheral fibers were connected to the detector port (a silicon photodiode).

Spectra were recorded over the 200–800 nm spectral range, with dwell time 0.5 seconds and step size 2 nm; the spectra were smoothed using a Savitzky-Golay filter (filter size 9 points, interval 2 points).

Baseline spectra were obtained from Spectralon 99% diffuse reflectance reference material.

Additionally, to test the feasibility of using a lightweight portable FORS instrument to analyze archaeological objects in situ, FORS spectra on modern Maya blue were also recorded with a Zeiss MC501 spectrophotometer in the 300–860 nm (0.8 nm/pixel resolution) range using Spectralon 99% diffuse reflectance standard for calibration.


2.3.3 Raman Microspectroscopy

Raman spectra were recorded on a Chromex Senturion Raman microscope equipped with a 70 mW 785 nm laser diode. Spectra were recorded between 150 and 1650 cm-1 using a 1200 rulings/mm holographic grating. Spectral calibration was achieved using the proprietary system Sure_Cal (Allen et al. 2000). Raman spectra of Maya blue were recorded both conventionally and employing a new technique for the rejection of fluorescence. The technique is based on the acquisition of Raman spectra at two slightly different excitation wavelengths (wavelength difference is typically between 0.3 and 0.7 nm), their subtraction and the subsequent integration of the resulting spectra (shifted excitation Raman difference spectroscopy, or SERDS). The method has been described in detail in the spectroscopic literature (Zhao et al. 2002). Both the microscopic samples and the whole statuette were analyzed in the instrument's large sample chamber (approximately 45 x 45 x 30 cm) using long–working distance infinity corrected microscope objectives (Olympus LMPlan 20x and Olympus LMPlan IR 100x).


2.3.4 Fourier Transform Infrared Microspectroscopy

Samples were compressed in a diamond anvil cell (Spectra Tech) and analyzed between 4000 and 700 cm-1 with a Thermo-Nicolet Nexus infrared spectrophotometer equipped with a liquid nitrogen–cooled MCT detector and an Olympus Continuum optical microscope. Spectra were collected as the sum of 256 scans at 4 cm-1 resolution. Aperture dimensions could be varied to limit the target area to a minimum of approximately 10 x 10 μm, thus achieving a very high degree of selectivity.


Copyright � 2004 American Institution for Conservation of Historic & Artistic Works