Application of Raman Spectroscopy and Surface-Enhanced Raman Scattering to

TECHNICAL NOTE Irina Geiman,1 B.S.; Marco Leona,2 Ph.D.; and John R. Lombardi,3 Ph.D. Application of Raman Spectroscopy and Surface-Enhanced Raman Scattering to the Analysis of Synthetic Dyes Found in Ballpoint Pen Inks* ABSTRACT: The applicability of Raman spectroscopy and surface-enhanced Raman scattering (SERS) to the analysis of synthetic dyes com- monly found in ballpoint inks was investigated in a comparative study. Spectra of 10 dyes were obtained using a dispersive system (633 nm, 785 nm lasers) and a Fourier transform system (1064 nm laser) under different analytical conditions (e.g., powdered pigments, solutions, thin layer chromatog- raphy [TLC] spots). While high fluorescence background and poor spectral quality often characterized the normal Raman spectra of the dyes studied, SERS was found to be generally helpful. Additionally, dye standards and a single ballpoint ink were developed on a TLC plate following a typical ink analysis procedure. SERS spectra were successfully collected directly from the TLC plate, thus demonstrating a possible forensic application for the technique. KEYWORDS: forensic science, questioned documents, ink analysis, dyes, Raman spectroscopy, surface-enhanced Raman scattering, SERS Raman spectroscopy can be useful for characterizing and dis- criminating inks based on their composition. It has been applied extensively to the analysis of a variety of inks including iron gall ink (1), lithographic ink (2), gel pen ink (3), and ballpoint ink (4,5). A variant of the technique, surface-enhanced Raman scat- tering (SERS) has also recently found its niche in ink analysis and has been used for ballpoint ink (6,7) and inkjet dye evalua- tions (8). The major advantages of Raman include small sample size requirements, minimal sample preparation, and the lack of chemical or mechanical pretreatments (9). Raman spectroscopy has been found to be applicable for the analysis of several dye classes including azo (10) and arylmethane (11). Additionally, studies of various dyes have been conducted and the results have revealed excellent reproducibility for the technique (12,13). However, fluo- rescence interference and low sensitivity are a common problem, and they can often result in poor analytical performance. SERS is becoming increasingly popular for the identification of organic dyes as it can quench fluorescence and provide enhanced signal, all of which ultimately lead to improved detection limits FIG. 1—Normal Raman (NR) and surface-enhanced Raman spectra (SERS) of Acid Blue 1. 1John Jay College of Criminal Justice, CUNY, 445 West 59th Street, New York, NY 10019. 2Department of Scientific Research, The Metropolitan Museum of Art, 1000 Fifth Avenue, New York, NY 10028. 3Department of Chemistry and Center for Analysis of Structures and Interfaces (CASI), The City College of New York, 160 Convent Avenue, New York, NY 10031. *Funding was provided by the Department of Justice Award No. 2006- DN-BX-K034; the City University Collaborative Incentive Program Award No. 80209 and PSC-BHE Faculty Research Award Program; National Sci- ence Foundation under Cooperative Agreement No. RII-9353488, Grants No. CHE-0091362, CHE-0345987, ECS0217646, IMR 0526926; the Andrew W. Mellon Foundation, the David H. Koch Family Foundation. Received 22 June 2008; and in revised form 8 Sept. 2008; accepted 8 Sept. 2008. J Forensic Sci, July 2009, Vol. 54, No. 4 doi: 10.1111/j.1556-4029.2009.01058.x Available online at: www.blackwell-synergy.com � 2009 American Academy of Forensic Sciences 947 sweet 高亮 sweet 高亮 (14). Due to increased sensitivity and consequent ability to use only minimal samples, it can be considered a minimally destructive technique. The possibility of using SERS directly in situ on art- work and on thin layer chromatography (TLC) plates has been demonstrated in a few selected cases (15,16). SERS has been successfully used to obtain the spectra of natural and synthetic dyes of several classes and colors (17–20). Additionally, a study that focused on the separation of dyes found surface-enhanced resonance Raman scattering to be an effective technique in differ- entiating analytes with similar structures (21). The aim of this study was to compare normal Raman (NR) and SERS techniques for the analysis of dyes commonly found in ballpoint inks (22) to determine if the techniques are feasible for the forensic analysis of writing inks. Materials and Methods Dyes Ten dyes representing classes commonly found in ink formu- lations were selected for the study: Acid Blue 1 (Sigma-Aldrich, St. Louis, MO: 198218), Acid Orange 10 (Sigma: O7252), Acid Red 52 (Fluka Biochemika, Buchs, Switzerland: 86183), Aniline Blue (Ward’s Natural Science Establishment, Rochester, NY: 38W7001), Crystal Violet (Sigma: C6158), Methyl Violet (Fluka Biochemika: 69710), Pararosaniline (Sigma: P3750), Rhodamine B (Fisher Scientific, Boston, MA: R21), Sudan Black B (Fisher Scientific: BP109), and Victoria Blue B (Sigma: V0753). Normal Raman Setup and Dye Analysis The Raman spectra were obtained with both dispersive and Fourier transform (FT) instruments. A Senterra Raman microscope (Bruker Optics Inc., Billerica, MA) with 100· long working dis- tance objective, 1200 and 1800 rulings ⁄mm holographic gratings, and charge-coupled device detector was used for dispersive analy- sis. This system was employed with a 633 nm helium ⁄neon and a 785 nm diode lasers. A Raman II FT-Raman (Bruker Optics Inc.) spectrometer with a liquid nitrogen cooled germanium detec- tor and a 1064 nm Nd ⁄YAG laser was used for FT-Raman spectroscopy. The NR spectra were obtained for pure powder samples using the microscope setup. The majority of the spectra were obtained using the following conditions: 3–5 cm)1 resolution (1800 rul- ings ⁄mm grating), 30 sec integration time, and 10 mW power for the 633 nm laser and 3–5 cm)1 resolution (1200 rulings ⁄mm grat- ing), 30 sec integration time, and 50 mW power for the 785 nm laser. FT-Raman spectra at 1064 nm were obtained in a macrocon- figuration. A 2-mm sample holder was used in back scattering mode, with acquisition conditions set at: 4 cm)1 resolution, 200 scans, and 50 mW. The laser power was decreased if the analyte fluorescence was overwhelming the spectrum and preventing reso- lution and identification of the peaks. SERS Setup and Dye Analysis The dispersive Raman system was used with the 633 and 785 nm lasers and the instrument parameters were the same as those used for NR. Measurements were obtained by focusing through the test drops deposited on the surface of the microscope slides. The drops were prepared by placing 1 lL of silver colloid on a slide, followed by 0.5 lL of the dye solution, and then 1 lL of 0.5 M potassium nitrate. FIG. 2—Normal Raman (NR) and surface-enhanced Raman spectra (SERS) of Acid Orange 10. FIG. 3—Normal Raman (NR) and surface-enhanced Raman spectra (SERS) of Acid Red 52. 948 JOURNAL OF FORENSIC SCIENCES The silver colloid was prepared by the reduction of silver nitrate with sodium citrate in ultrapure water following the procedure out- lined by Lee and Meisel (23). The colloid was concentrated by centrifugation for 2 min at 2240 · g followed by removal of the supernatant. The dye solution was prepared by dissolving a few dye crystals in 0.5 mL of methanol. All the glassware was cleaned with a cleaning solution, rinsed with ultrapure water, washed with acetone, and allowed to dry. TLC Separation and Dye Analysis The Paper Mate� Xtend� medium blue ballpoint pen was used to draw asterisks c. 0.5 cm in diameter on Whatman #1 filter paper (Whatman Inc., Florham Park, NJ). The individual asterisks were then extracted with 0.5 mL of ethanol, and c. 2 lL of the solution was spotted on a Whatman #4410–221 silica gel TLC plate (What- man Inc.). The plate was developed in 70:35:30 mixture of ethyl acetate:ethanol:water per ASTM International Guide E 1422–05 (24). The SERS of the separated dye spots were obtained using the dispersive system equipped with the 785 nm laser, and the instru- ment parameters remained the same as for NR spectra. The spectra were obtained by placing 0.1 lL of silver colloid on the dye spot on the TLC plate, then adding 0.1 lL of 0.5 M potassium nitrate, and focusing directly on the plate. The results were compared with the standard dye spectra obtained by the SERS method. The corre- sponding dye standards and the ink extraction were analyzed using TLC. The retention factors (Rf) for the dye spots from the extrac- tion and the dye standards were calculated and compared with each other to confirm the presence of the dye in the ink. Results and Discussion Normal Raman spectra were obtained for all the dyes with three different laser wavelengths. Based on the evaluation of all of the results, it was determined that only the FT system with the 1064 nm Nd ⁄YAG laser performed consistently. All of the 1064 nm spectra had excellent signal intensity and signal to noise ratios. Spectra are displayed in Figs. 1–10 as follows: Acid Blue 1 (Fig. 1), Acid Orange 10 (Fig. 2), Acid Red 52 (Fig. 3), Aniline Blue (Fig. 4), Crystal Violet (Fig. 5), Methyl Violet (Fig. 6), Para- rosaniline (Fig. 7), Rhodamine B (Fig. 8), Sudan Black B (Fig. 9), and Victoria Blue B (Fig. 10). All spectra were manually normal- ized for ease of comparison. No background subtraction or other spectral manipulations were performed. The spectra obtained with the 633 and 785 nm lasers differed dramatically in their signal clarity for each of the dyes. Addition- ally, high levels of fluorescence were observed in all of the 633 nm spectra. Only Acid Orange 10 (Fig. 2) responded well to both laser wavelengths showing consistent peaks of high intensity. The 633 nm laser proved superior in performance over the 785 nm wavelength for Aniline Blue (Fig. 4), Sudan Black B (Fig. 9), and Victoria Blue (Fig. 10). These results could be explained by a resonance Raman enhancement (25), as the three dyes were similar in their deep blue color. The 785 nm laser pro- vided clear spectra for the rest of the analytes including both blue-colored dyes, Acid Blue 1 (Fig. 1), and red-colored dyes, Acid Red 52 (Fig. 3) and Rhodamine B (Fig. 8). Overall, the results contained enough individualizing peaks to easily differenti- ate the dyes. The exceptions were the spectra of Crystal Violet (Fig. 5) and Methyl Violet (Fig. 6) which were extremely similar but so were the molecular structures of the compounds differing only in the number of methyl groups. FIG. 4—Normal Raman (NR) and surface-enhanced Raman spectra (SERS) of Aniline Blue. FIG. 5—Normal Raman (NR) and surface-enhanced Raman spectra (SERS) of Crystal Violet. GEIMAN ET AL. • RAMAN SPECTROSCOPY OF DYES IN BALLPOINT PEN INKS 949 FIG. 6—Normal Raman (NR) and surface-enhanced Raman spectra (SERS) of Methyl Violet. FIG. 7—Normal Raman (NR) and surface-enhanced Raman spectra (SERS) of Pararosaniline. FIG. 8—Normal Raman (NR) and surface-enhanced Raman spectra (SERS) of Rhodamine B base. FIG. 9—Normal Raman (NR) and surface-enhanced Raman spectra (SERS) of Sudan Black B. 950 JOURNAL OF FORENSIC SCIENCES With the use of the 633 and 785 nm lasers, excellent SERS were obtained for all the dyes. There was c. a 4-fold increase in signal intensity observed with the 633 nm laser; however, Pararosaniline (Fig. 7) presented higher intensity spectra with the 785 nm laser. The NR and SERS spectra were found to be different in the intensities of individual peaks (e.g., Fig. 4, 917 cm)1 peak), and the appearance of certain peaks in only one type of the spectrum (e.g., Fig. 2, 1618 cm)1 peak). Additionally, some peak shifts between NR and SERS were observed (e.g., Fig. 8, 1280–1284 cm)1 peak). These peak variations are explained by selection rules dictating which molecular bonds are Raman and SERS active (25). It must be noted that the SERS spectra were obtained from dilute solutions of the dyes as opposed to the solid samples used for NR measurements. The fact that good quality Raman spectra were obtained in these conditions is an indication of the signal enhancement obtained with SERS. The TLC analysis showed that the Methyl Violet standard sepa- rated into three spots on the plate while the ink extraction presented only two spots corresponding to Methyl Violet in color, Rf values, and SERS spectra. Such results were probably due to different combinations of the demethylated pentamethyl Pararosaniline found in both the standard and the ink. The SERS of the ink dyes obtained from the TLC plate (Fig. 11) contained high intensity peaks and were consistent with the SERS spectra of the Methyl Violet standard (Figs. 6 and 11). Conclusions This study has demonstrated that high quality dye spectra can be obtained from pure dye samples with NR spectroscopy by carefully selecting the excitation frequency, but SERS consistently provided significant signal enhancement and fluorescent quenching for ballpoint ink dyes. Three laser wavelengths were evaluated, and for NR, the best results were observed when using the 1064 nm Nd ⁄YAG laser. SERS spectra of equally high quality were obtained with the 633 and 785 nm lasers. Although fluores- cence was a factor in the 633 nm NR spectra, it was mitigated with the use of SERS ultimately allowing for successful data collection. The SERS spectra obtained after the ballpoint ink dyes and the reference dyes were developed on a TLC plate showed a high level of consistency with the standard dye spectra obtained with the drop method and on the TLC plate. Overall, the study successfully showed the applicability of NR spectroscopy and SERS to the analysis of synthetic dyes found in ballpoint inks. As only a single pen was used for this study, evaluation of dye components of an array of inks from different manufacturers would be valuable in establishing dye variations between manu- facturers and batch-to-batch. Further studies should investigate additional dyes and pigments (e.g., phthalocyanines) and consider other ink components (e.g., vehicles, lubricants). Acknowledgments The study would not have been possible without the guid- ance, support, and encouragement of Dr. Thomas A. Kubic and Dr. Maria Vega CaÇamares. References 1. Lee AS, Mahon PJ, Creagh DC. Raman analysis of iron gall inks on parchment. Vib Spectrosc 2006;41(2):170–5. 2. Centeno SA, Buisan VL, Ropret P. Raman study of synthetic organic pigments and dyes in early lithographic inks. J Raman Spectrosc 2006;37:1111–8. FIG. 10—Normal Raman (NR) and surface-enhanced Raman spectra (SERS) of Victoria Blue B. FIG. 11—Surface-enhanced Raman spectra (SERS) of dyes thin layer chromatography separations. MV, Methyl Violet. GEIMAN ET AL. • RAMAN SPECTROSCOPY OF DYES IN BALLPOINT PEN INKS 951 3. Mazzella WD, Buzzini P. Raman spectroscopy of blue gel pen inks. Forensic Sci Int 2005;3:241–7. 4. Claybourn M, Ansell M. Using Raman spectroscopy to solve crime: ink, questioned documents and fraud. Sci Justice 2000;40(4):261–71. 5. Zieba-Palus J, Kunicki M. Application of the micro-FTIR spectroscopy, Raman spectroscopy and XRF method examination of inks. Forensic Sci Int 2006;3:164–72. 6. Seifar RM, Verheul JM, Ariese F, Brinkman UATh, Gooijer C. Applica- bility of surface-enhanced resonance Raman scattering for the direct dis- crimination of ballpoint pen inks. Analyst 2001;126(8):1418–22. 7. White PC. In situ surface-enhanced resonance Raman scattering (SERRS) spectroscopy of biro inks—long term stability of colloid trea- ted samples. Sci Justice 2003;43(3):149–52. 8. Littleford RE, Hughes MP, Dent G, Tackley D, Smith WE. Surface- enhanced resonance Raman scattering of black inkjet dyes in solution and in situ printed onto paper. Appl Spectrosc 2003;57(8):977–83. 9. Edwards HGM. Raman microscopy in art and archeology: illumination of historical mysteries in rock art and frescoes. Spectroscopy 2002;17(2):16–40. 10. Vandenabeele P, Moens L, Edwards HGM, Dams R. Raman spectro- scopic database of azo pigments and application to modern art studies. J Raman Spectrosc 2000;31:509–17. 11. Gicquel J, Carles M, Bodot H. Resonance Raman investigation of charge transfer complexes between a trityl cation (crystal violet) and sulfonated azo derivatives. J Phys Chem 1979;83(6):699–706. 12. Bourgeois D, Church SP. Studies of dyestuffs in fibers by Fourier trans- form Raman spectroscopy. Spectrochim Acta A 1989;46:295–301. 13. Clark RJH. Pigment identification by spectroscopic means: an arts ⁄ sci- ence interface. C R Acad Sci II C 2001;5:7–20. 14. Chen K, Leona M, Vo-Dinh T. Surface-enhanced Raman scattering for identification of organic pigments and dyes in works of art and cultural heritage material. Sensor Rev 2007;27(2):109–20. 15. Leona M, Stenger J, Ferloni E. Application of surface-enhanced Raman scattering techniques to the ultrasensitive identification of natural dyes in works of art. J Raman Spectrosc 2006;37:981–92. 16. Istvan K, Keresztury G, Szep A. Normal Raman and surface-enhanced Raman spectroscopic experiments with thin layer chromatography spots of essential amino acids using different laser excitation sources. Spectro- chim Acta A 2003;59:1709–23. 17. Leona M, Lombardi JR. Identification of berberine in ancient and histor- ical textiles by surface-enhanced Raman scattering. J Raman Spectrosc 2007;38:853–8. 18. Munro CH, Smith WE, Armstrong DR, White PC. Assignments and mechanism of SERRS of the hydrazone form for the azo dye solvent yellow 14. J Phys Chem 1995;99:879–85. 19. Zhou X, Fang Y, Zhang P. A new substrate for surface-enhanced Raman scattering of dye sudan molecules. Spectrochim Acta A 2007;67:122–4. 20. Kudelski A. Raman studies of rhodamine 6G and crystal violet sub- monolayers on electrochemically roughened silver substrates: do dye molecules adsorb preferentially on highly SERS-active sites? Chem Phys Lett 2005;414:271–5. 21. Roman DA, Efremov E, Ariese F, Carretero AS, Gooijer C. Interfacing capillary electrophoresis and surface-enhanced resonance Raman spec- troscopy for the determination of dye compounds. Anal Bioanal Chem 2005;382(1):180–5. 22. Brunelle RL, Crawford KR. Advances in the forensic analysis and dat- ing of writing ink. Springfield, IL: Charles C. Thomas, 2003. 23. Lee PC, Meisel D. Adsorption and surface-enhanced Raman of dyes on silver and gold sols. J Phys Chem 1982;86:3391–5. 24. Amerian Society of Testing & Materials. ASTM Guide E 1422: standard guide for test methods in forensic writing ink comparison. West Cons- hohocken, PA: ASTM, 2005. 25. Campion A, Kambhampati P. Surface-enhanced Raman scattering. Chem Soc Rev 1998;27:241–50. Additional information and reprint requests: John R. Lombardi, Ph.D. Department of Chemistry Center for Analysis of Structures and Interfaces The City College of New York 160 Convent Avenue New York, NY 10031 E-mail: lombardi@sci.ccny.cuny.edu 952 JOURNAL OF FORENSIC SCIENCES