Nondestructive Identification of Natural and ... - ACS Publications

Apr 28, 2011 - John Jay College of Criminal Justice, CUNY, 445 West 59th Street, New ... as artists' colors and are of great forensic interest as the ...
0 downloads 0 Views 4MB Size
LETTER pubs.acs.org/ac

Nondestructive Identification of Natural and Synthetic Organic Colorants in Works of Art by Surface Enhanced Raman Scattering Marco Leona,*,† Peter Decuzzi,‡ Thomas A. Kubic,‡ Glenn Gates,§ and John R. Lombardi^ †

Department of Scientific Research, The Metropolitan Museum of Art, 1000 Fifth Avenue, New York, New York 10028, United States John Jay College of Criminal Justice, CUNY, 445 West 59th Street, New York, New York, 10019, United States § The Walters Art Museum, 600 N. Charles Street, Baltimore, Maryland 21201, United States ^ Department of Chemistry and Center for Analysis of Structures and Interfaces (CASI), The City College of New York, New York, 10031, United States ‡

ABSTRACT: We present a new method based on surface-enhanced Raman scattering (SERS) for the nondestructive identification of organic colorants in objects whose value or function precludes sampling, such as drawings, prints, historic and archeological textiles, handwritten or printed documents, and forensic evidence. A bead of a polymer hydrogel loaded with a solution containing water, an organic solvent, and a chelating agent is used to extract minimal amounts of the colorants from the work of art for SERS analysis. Using a gel as a medium for the solvent mixture confines its action only to the areas of the work of art covered by the gel bead. The gel bead is then removed from the work of art, covered with a drop of Ag colloid, and examined with a Raman microscope. Transfer of the dye from the substrate to the gel does not require removing a sample from the work of art, therefore preserving the physical integrity of the object. Spectrophotometric color measurements confirm that color change is below the limit perceivable by a human observer. Finally, the size of the polymer bead can be reduced to a fraction of a millimeter in order to further minimize any impact on the work of art, without detriment to the effectiveness of the method. The technique has been successfully used for the analysis of a mordant dye on the 15th century Netherlandish tapestry, “The Hunt for the Unicorn”, and of a synthetic lake pigment on a Meiji period Japanese woodblock print.

N

atural organic colorants extracted from plants or insects have been used for millennia as textile dyes or as lake pigments.1,2 Synthetic dyes such as crystal violet, produced industrially since the second half of the 19th century, rapidly supplanted their natural equivalents in the dyeing of textiles3 and as artists’ colors and are of great forensic interest as the colorants used in printing and writing inks.4 Their identification, particularly in valuable and irreplaceable material, such as works of arts, documents, and trace evidence, is extremely difficult because of the low concentration at which they are used and of the requirement to limit sampling to extremely small fragments. Separation techniques such as high performance liquid chromatography with diode array detection, fluorescence detection, or mass spectrometric detection (HPLC-DAD, HPLC-FD, and LCMS, respectively) have long been used for dye analysis in the cultural heritage research field.5 Liquid chromatography techniques have the advantage of being able to resolve complex mixtures of related compounds, as natural dyes generally are. In the cultural heritage field, these techniques have to date found the greatest application in the analysis of textiles, as they require slightly larger samples (generally 1 or 2 mm of a fabric thread) than obtainable from drawings or paintings. Recently, surfaceenhanced Raman scattering (SERS) has emerged as an extremely sensitive technique for the identification of natural and synthetic organic colorants in works of art and archeological objects.6 r 2011 American Chemical Society

Compared to HPLC, it suffers from a greatly reduced ability to resolve mixtures, yet it can be used successfully with extremely small samples. Work conducted in our laboratory using resonant excitation, a monodisperse silver colloid produced by microwave supported reduction of silver sulfate with glucose and sodium citrate, and a lossless nonextractive hydrolysis sample treatment procedure allowed us to identify the earliest known use of madder dye, an important red colorant, in a 25 μm diameter sample from a 4000 year old painted leather object from Egypt.7 Although alternative approaches to SERS analysis of works of art have been proposed,8,9 all the work conducted to date has required that samples be removed from the work of art under investigation. In practice, this means plucking one or two textile fibers from a tapestry or a piece of cloth,10 removing a few pigment grains from the paper support of a drawing,11 or obtaining a chip from a painted coating.7 To further limit the damage incurred by irreplaceable objects, we developed a technique similar to solid phase microextraction (SPME). In this demonstration of the technique, we used a methacrylate hydrogel and a 1:1 dimethylformamide (DMF) and water solution with 1% (w/v) disodium ethylenediaminetetraacetic Received: March 18, 2011 Accepted: April 28, 2011 Published: April 28, 2011 3990

dx.doi.org/10.1021/ac2007015 | Anal. Chem. 2011, 83, 3990–3993

Analytical Chemistry acid (EDTA) to extract trace amounts of the target colorant from the substrate. The DMF/H2O/EDTA system is an effective vehicle for the extraction of organic colorants from textiles12 for high performance liquid chromatography (HPLC) analysis. Hydrogels, saturated with various solvents, have been used in art conservation to reduce visually disfiguring stains on exposed, bare canvas in largescale modern paintings, for example, by the artist Morris Louis.13 The specific details presented here are intended as a proof of concept: other gel materials and solvent systems could be developed to target particular analytes. For SERS applications, a small bead of polymer loaded with the extraction agent is applied to the area of interest on the work of art and kept in contact with it under light pressure. Using a gel as a medium confines the action of the solvent mixture to the areas of the work of art in contact with the gel bead. The extraction is carried out at room temperature, with mordant dyes on textiles requiring up to 4 h and synthetic dyes in pen inks requiring as little as 30 s. The gel is then removed from the substrate, transferred to a microscope slide, covered with a drop of Ag colloid, and examined with a Raman microscope. Transfer of the dye from the substrate to the gel eliminates the need to obtain a sample from the object under examination, thus preserving the physical integrity of the object; the amount of dye removed is so small that no appreciable fading is detected by the eye, and the size of the polymer bead can be reduced to a fraction of a millimeter in order to minimize any impact on the work of art, without detriment on the effectiveness of the method. Spectrophotometric color measurements confirmed that no appreciable color loss resulted from the use of hydrogel transfer.

’ EXPERIMENTAL SECTION The results reported here were obtained with a 1:1 random copolymer of 2,3-dihydroxypropyl methacrylate with 2-hydroxyethyl methacrylate (Benz 5X, also known as GMA). GMA is a cross-linked hydrogel manufactured by Benz Research and Development (Sarasota, FL) for use as soft contact lens blanks. The gel is prepared for use by soaking it in ultrapure water overnight (saturated water content is 57%) and then for 10 to 30 min in a 1:1 solution of dimethylformamide (DMF) and water, containing 1% w/v disodium EDTA. The swollen gel can easily be cut to an appropriate size using a razor blade or a sharp scalpel. Prior to application to the substrate, extra solvent is removed from the exterior of the gel using Kim Wipes. The material properties of HEMA and GMA have been described in detail by Gates and Harmon.14,15 The gel can be applied to the object under analysis in a variety of ways. For textiles, we have used microemostatic clips to keep the gel in contact with a loose thread; for flat art, we have simply pressed the gel to the object using a microscope slide loaded with suitable weights. Care must be taken to prevent the gel from drying in contact with the object, not so much as this could affect the result of the SERS analysis, but to prevent adhesion of the gel to the surface of the work of art and the possible removal of material. Generally, the use of a glass slide is sufficient to fulfill this requirement. Following extraction, the gel is removed from the substrate and transferred to a microscope slide. As we generally noticed limited diffusion of the dye through the thickness of the gel, we consistently oriented the gel with the side that was in contact with the object facing the Raman microscope objective. An amount of colloid sufficient to cover the gel fragment is then deposited on the slide. Silver colloids were prepared either by microwave assisted

LETTER

Figure 1. SERS spectra from a gel strip applied to a red thread in the Unicorn tapestry (UNI) and from an alizarin reference solution (AZ).

reduction of silver sulfate with glucose and sodium citrate7 or by the Lee and Meisel procedure.16 To prepare either colloid for use, 1 mL of it was centrifuged for 5 min at 13 000 rpm with a Fisher Scientific Accuspin 400 centrifuge. The supernatant (950 μL) was removed and replaced with 150 μL of 18 MΩ ultrapure water (Millipore Simplicity 185 Water Purification system) to increase the silver nanoparticle concentration by a factor of 5. SERS analysis was performed with a Bruker Senterra Raman Microspectrometer with 488 nm and 785 nm laser excitation, using Olympus LMPlanFL long working distance 20, 50, and 100 objectives, with power at the sample ranging from 65 to 650 μW for the 488 nm laser and from 0.33 to 2.6 mW for the 785 nm laser. Spectra were integrated over 30 s at a resolution of 3 to 5 cm1. To evaluate the color change of the substrate following gel extraction, we collected visible reflectance spectra from 380 to 750 nm using a Varian Cary 50 UVvis spectrophotometer equipped with a Barrelino diffuse reflectance probe (Harrick Scientific Products, Inc. Pleasantville, NY). CIELAB color coordinates and the color difference value (the ΔE parameter) were calculated using the Varian Color package.

’ RESULTS AND DISCUSSION Mordant Dyes on Historical and Archeological Textiles. A gel strip (3 mm by 1 mm by 0.5 mm) was applied with a microemostatic clip to a red thread protruding from the back of the tapestry “The maiden’s companion signals to the hunters”, from “The hunt of the unicorn” series (South Netherlandish, ca. 14951505. The Metropolitan Museum of Art, 38.51,1.2; Gift of John D. Rockefeller Jr., 1937) . The gel was removed after 4 h and analyzed using the LeeMeisel colloid and 785 nm excitation (Figure 1). Alizarin, the main ingredient in the natural dye madder was identified. The presence of madder is corroborated by SERS microanalysis of a fiber removed from the tapestry10 and by HPLC analysis (unpublished data). Inks and Lake Pigments. The nondestructive character of the gel-extraction technique is demonstrated by the analysis of a Japanese woodbock print, “Sekigahara Homare no Gaika” (“A poem about the battle of Sekigahara”, the title of a Kabuki play) by Toyoharu Kunichika, dated to 1892 (Figure 2). The pink and violet coloration, typical of Meiji period prints, was made possible by inks based on synthetic dyes, imported into 3991

dx.doi.org/10.1021/ac2007015 |Anal. Chem. 2011, 83, 3990–3993

Analytical Chemistry

Figure 2. “Sekigahara Homare no Gaika” (“A poem about the battle of Sekigahara”) by Toyoharu Kunichika, 1892. Woodblock print on paper, triptych, each sheet originally oban size (27  39 cm), slightly trimmed. Private collection.

LETTER

Figure 4. Means of ten measurements each of the same location, before extraction (squares) and after extraction (solid line).

of crystal violet with its tetra- and pentamethyl homologues, not easily distinguished by SERS from crystal violet). As methyl violet was patented in 1861 and crystal violet in 1883, both are equally possible in this print.17 Comparison of the before- and after-extraction spectra shows that the color change is negligible (Figure 4). To quantitatively measure color variation, we used the parameter ΔE*ab, defined as [(ΔL*)2 þ (Δa*)2 þ (Δb*)2]1/2 in the CIE L*a*b* system. ΔE values between each individual pre-extraction measurement and the mean of the ten pre-extraction measurements ranged between 0.18 and 0.39. This range of values, due to small errors in repositioning, is essentially a measure of the inhomogeneity of the color on the print. We then compared each of the ten postextraction measurements to the mean of the pre-extraction measurements. The range of ΔE was 0.28 to 0.67. This is a measure of the color change following gel extraction. This range of values is safely below the level perceivable by a human observer (ΔE below 1).18

Figure 3. SERS spectra from a gel fragment removed from a violet area on the print “Sekigahara Homare no Gaika” (JP) and from a reference solution of crystal violet (CV). Spectra were normalized for ease of comparison.

Japan from Europe. To test whether gel-extraction would allow us to perform SERS measurements not only without removing a sample but also without any color loss, we carried out before- and after-extraction color measurements. A purple region of the print was selected, and the diffuse reflectance probe was placed over it, using a paper mask for reproducible positioning. Ten spectra were acquired, removing and replacing the probe after each measurement. Another set of ten spectra were acquired from the same area after gel extraction. To extract the dye, a gel bead was placed on the violet area, a glass slide was placed on top of the gel, and three small tungsten weights (each weighing 28 g) were put on top of the slide. After 30 s, the gel was removed from the print. Twenty μL of silver colloid (produced by microwave assisted reduction) were placed on the gel, and a spectrum was acquired using 488 nm excitation. The spectrum (Figure 3) matches that of crystal violet, Nhexamethylpararosaniline (CI 42555), or methyl violet (a mixture

’ CONCLUSIONS Gel-extraction is a promising nondestructive technique for the SERS identification of dyes, as it preserves both physical integrity and appearance of irreplaceable objects. The approach is minimally invasive, requiring only gentle contact between a soft, smooth, cross-linked gel material and the substrate. The action of the solvent system is confined to the area of contact (which can be made as small as practical by controlling the size of the gel bead). The gel does not adhere to the substrate if properly applied and removed (care must be exercised so that the gel does not dry in contact with the substrate). Color loss is avoided by controlling the extraction time. Further refinements could be introduced by tailoring the gel and the extraction system to the chemical properties of the target analyte. Finally, even though ideal for SERS, gel-extraction could also be used prior to liquid chromatography mass spectrometry (LC-MS), simply by extracting the dye from the gel prior to injection. ’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Phone: (212) 396-5476. Fax: (212) 396-5466. 3992

dx.doi.org/10.1021/ac2007015 |Anal. Chem. 2011, 83, 3990–3993

Analytical Chemistry

LETTER

’ ACKNOWLEDGMENT We are indebted to the support of the National Science Foundation (NSF-SCIART) Grant # CHE-1041832. This project was also supported by Award No. 2006-DN-BX-K034 awarded by the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice. The opinions, findings, and conclusions or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect those of the Department of Justice. ’ REFERENCES (1) Schweppe, H.; Winter, J. In Artists’ pigments; A handbook of their history and characteristics; West Fitzhugh, E., Ed.; National Gallery of Art: Washington, DC, 1997; Vol. 3, pp 109142. (2) Schweppe, H.; Roosen-Runge, H. In Artists’ pigments; A handbook of their history and characteristics; Feller, R. L., Ed.; Oxford University Press: New York, Oxford, 1986; Vol. 1, pp 255283. (3) Travis, A. S. The rainbow makers: the origins of the synthetic dyestuffs industry in Western Europe; Associated University Presses, Inc.: Cranbury, NJ, USA; London, UK; and Mississauga, ON, Canada, 1993. (4) Brunelle, L. R.; Crawford, K. R. Advances in the forensic analysis and dating of writing inks; Charles C. Thomas Publishers, Ltd.: Springfield, 2003. (5) Wouters, J. Stud. Conserv. 1985, 30, 119–128. (6) Casadio, F.; Leona, M.; Lombardi, J. R.; Van Duyne, R. Acc. Chem. Res. 2010, 43 (6), 782–791. (7) Leona, M. Proc. Natl. Acad. Sci. U.S.A. 2009, 106 (35), 14757–14762. (8) Brosseau, C. L.; Gambardella, A.; Casadio, F.; Van Duyne, R. P.; Grzywacz, C.; Wouters, J. Anal. Chem. 2009, 81, 3056–3062. (9) Jurasekova, Z.; Domingo, C.; Garcia-Ramos, J. V.; Sanchez-Cortes, S. J. Raman Spectrosc. 2008, 39, 1309–1312. (10) Leona, M.; Stenger, J.; Ferloni, E. J. Raman Spectrosc. 2006, 37, 981–992. (11) Brosseau, C. L.; Rayner, K.; Casadio, F.; Van Duyne, R. P.; Grzywacz, C. M. Anal. Chem. 2009, 81, 7443–7447. (12) Tiedemann, E. J.; Yang, Y. J. Am. Inst. Conserv. 1995, 34 (3), 195–206. (13) Gates, G.; Hensick, T.; Mancusi-Ungaro, C.; Ausema, T.; Learner, T.; Shank, W. International Council of MuseumsCommittee for Conservation (ICOM-CC) 14th Triennial Meeting in The Hague, September 2005, James & James/Earthscan: London, Vol. I, pp 329334. (14) Gates, G.; Harmon, J. P. Polymer 2003, 44, 215–221. (15) Gates, G.; Harmon, J. P. Polymer 2003, 44, 207–214. (16) Lee, P. C.; Meisel, D. J. Phys. Chem. 1982, 86, 3391–3395. (17) Rowe, F. M., Ed. Colour Index, 1st ed.; Society of Dyers and Colourists: Bradford, Yorkshire, UK, 1924; p 174. (18) Berns, R. S. Billmeyer and Salzmann’s Principles of Color Technology, 3rd ed.; John Wiley & Sons: New York, 2000; pp 124125.

3993

dx.doi.org/10.1021/ac2007015 |Anal. Chem. 2011, 83, 3990–3993