Article pubs.acs.org/Langmuir
Adsorption Study and Detection of the High Performance Organic Pigments Quinacridone and 2,9-Dimethylquinacridone on Ag Nanoparticles By Surface-Enhanced Optical Spectroscopy Elena del Puerto,* Concepcion Domingo, Jose V. Garcia Ramos, and Santiago Sanchez-Cortes* Instituto de Estructura de la Materia, IEM-CSIC, Serrano 121, 28006 Madrid, Spain. ABSTRACT: In this work, surface-enhanced Raman scattering (SERS) and surface-enhanced fluorescence (SEF) were employed in the study of the adsorption and detection of the pigments quinacridone (QA) and 2,9dimethylquinacridone (2,9-DMQA). These pigments are of great relevance in artwork and textile, plastic, and photochemical industries due to their condition as high performance pigments since they possess a high tinting strength. Due to this fact, they have been employed at relatively low concentrations. Therefore, the analysis and detection of these pigments requires the application of a highly sensitivity technique, such as SERS and SEF. The adsorption of QA and 2,9-DMQA on silver nanoparticles was extensively studied by means of SERS at different surface coverages. This study was completed by carrying out an in depth vibrational (Raman and IR) analysis of these pigments in solid state by ab initio density functional theory (DFT) calculations. In addition, UV−vis spectroscopy was employed to investigate the aggregation undergone by both pigments in solid state and in solution. 2,9-DMQA was demonstrated to have a lower tendency toward aggregation due to the presence of methyl groups. Even so, this molecule follows a BET adsorption mechanism on the metal nanoparticles due to its high tendency toward intermolecular interaction. From the analysis of the adsorption mechanism of this molecule, the limit of detection was deduced to be ca. 55 ppb.
■
INTRODUCTION Quinacridones are high performance pigments (HPPs). As is well-known, HPPs, organic or inorganic, are an important segment of the diverse and rich field of color and visual effect technology. They exhibit enhanced durability. The most salient durability feature is generally regarded as resistance to visible and ultraviolet radiation (lightfastness), but heat stability and chemical resistance are also critical attributes. Specifically, quinacridones present outstanding chemical and photochemical stability, extremely low solubility in water, and attractive colors that make these compounds one of the most important classes of HPPs.1 In this work, we have centered our attention on Quinacridone (QA) and 2,9-dimethylquinacridone (2,9DMQA), two of the most important molecules of this group. QA, labeled internationally as PV19 by Color Index, is the principal compound of this family and is mainly employed to provide a violet color. Its use is widespread both in paints in artwork and as a surface application in plastics, textiles, and so forth. 2,9-DMQA (PR122) shows a reddish tonality and is a suitable colorant to employ in plastic applications and in printing inks. Occasionally, quinacridones are mixed to achieve different tonalities. The interesting photochemical properties make them very appropriate for use in solar cells2,3and OLEDs.4−6 Because of the great variety of applications, their identification and detection is not only important in the art conservation field, but also in the fields of forensics and electronics to understand and improve devices.7 However, the © 2013 American Chemical Society
identification of quinacridones is not straightforward due to the particular properties derived from their structures. The inset of Figure 1 shows the chemical structure of QA and 2,9-DMQA. As we can observe, they present the same rigid skeletal body: five condensed aromatic rings and three benzene rings alternated with two 4-pyridones. Accordingly, they are characterized by NH−O hydrogen bonds,8,9 which determine their low solubility in water that complicate its characterization. In previous works, we have studied another derivative of quinacridone, i.e., the quinacridone quinone, which also presents the same problem regarding its characterization, its insolubility due to the hydrogen bonds. In the latter case, we have developed several strategies to overcome this drawback, employing calixarenes10 or ionic liquids11 to disperse the pigment. In the case of 2,9-DMQA, the presence of two methyl groups, displaying a +I inductive effect, in the 2,9 position, introduces new chemical properties in the aromatic system that may have consequences on its intermolecular interaction and, thus, on its solubility. Thus far, quinacridones have been characterized by UV− visible, fluorescence, X-ray diffraction, thermogravimetric analysis, FT-IR, Raman microspectroscopy, and mass spectroscopy. 8,12−18 To our knowledge, the majority of the Received: September 19, 2013 Revised: December 21, 2013 Published: December 24, 2013 753
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
and the low aggregation of 2,9-DMQA allowed analysis of its adsorption on the metallic surface in terms of the adsorption isotherms of the molecules, from where the limit of detection (LOD) was also calculated.
■
EXPERIMENTAL SECTION
Quinacridone (QA) was purchased from TIC Europe with the highest purity (>99%). The 2,9-dimethylquinacridone (2,9-DMQA) came compliments of SunChemical. Both reagents were solved in dimethyl sulfoxide (DMSO) which was acquired from Sigma-Aldrich with >99.9% purity. Silver colloids for SERS measurements were made by reduction of silver nitrate with hydroxylamine chloride at room temperature.25 After aggregation with potassium nitrate solution up to a final concentration of 3 × 10−2 M, 10 μL portions of the solution in DMSO at variable concentrations of each dye were added to 1 mL of the silver colloids to get the final desired concentration. All reactants involved in the silver colloid synthesis were purchased from Sigma-Aldrich with analytical purity. UV−vis-NIR Shimadzu 3600 has been employed to register the extinction spectra of QA and 2,9-DMQA in solution and in solid state. To register the spectrum in solid state, an integrating sphere was used. FTIR spectra of solid quinacridone derivatives dispersed in KBr were recorded with a Bruker IFS66 spectrometer with a DGTS detector. The spectral resolution was 8 cm−1, and 500 scans were acquired. The Raman spectrum of QA was recorded using an FT-Raman Bruker MultiRam spectrometer with a liquid N2 cooled Ge detector and an Nd:YAG laser at 1064 nm. The final spectra were obtained after accumulation of 500 scans. The SERS experiments (exciting at 532 nm) as well as the Raman spectra (exciting at 785 nm) of 2,9DMQA in solid state have been registered with a Renishaw inVia Raman microscope, RM2000 model, equipped with a Leica microscope and an electrically cooled CCD camera. DFT calculations were done using B3LYP and 6-311G+ (d,p) basis sets, using the Gaussian 09 program.24 A scale factor of 1.0189 was applied to wavenumbers under 1000 cm−1 according to ref 25.26 The adsorption isotherm of 2,9-DMQA was calculated by plotting the overall area of SERS spectra in the range 1800−297 cm−1 versus the concentration. The Origin 8.0 software was employed to carry out the analysis.
Figure 1. UV−visible Absorption spectra of QA (top) and 2,9-DMQA (bottom) under the following conditions: in solid state (a); in DMSO (b); and in H2O/DMSO (1% v/v) (c). [QA] = 10−4 M and [2,9DMQA] = 10−5 M. Inset: the molecular structure of QA and 2,9DMQA. The vertical line corresponds to the position of the excitation line employed to obtain the SERRS spectra.
quinacridone studies have been focused on their crystalline structure and only a few papers report their identification in art works employing mass spectroscopy.17,18 However, due to their high tinting strength, quinacridones are frequently present in relatively small quantities, and their identification requires high sensitivity analytical techniques. Therefore, surface-enhanced Raman scattering (SERS) and surface-enhanced fluorescence (SEF), two important surface-enhanced optical spectroscopies (SEOS) based on the localization of surface plasmons, are employed for the first time in this work, aimed at the characterization of such pigments. One of the most appreciated advantages of SEOS techniques regarding pigment identification and characterization in artwork is its nanodestructive and extractionless characteristic, thus making SERS and SEF very suitable for the analysis of materials of interest for cultural heritage.19−23 In this work, SERS spectra of both QA and 2,9-DMQA were obtained under different experimental conditions, and a comparative study between both pigments was done to understand the influence of methyl groups on the adsorption and interaction mechanism with the metal surface, which will further determine the detection ability of SERS. To accomplish a proper assignment of SERS spectra, theoretical calculations of the vibrational frequencies using the Gaussian 09 program24 were previously performed. Fluorescence emission has also been studied in order to obtain in-depth information about their aggregation effect. The sensitivity of the SERS technique
■
RESULTS AND DISCUSSION UV−visible Absorption Spectra. Figure 1 shows the UV−visible absorption of QA and 2,9-DMQA in solid state (Figure 1a), in DMSO (Figure 1b) and in H2O/DMSO (1%) (Figure 1c). The latter experimental conditions are interesting since they simulate the chemical environment of SERS experiments, except for the presence of silver nanoparticles. In DMSO, the aggregation of these molecules is not expected, and three maxima can be seen, corresponding to the condensed aromatic structure of QAs.27 These bands appear at 523, 489, and 460 nm and 531, 496, and 470 nm, for QA and 2,9-DMQA, respectively. The slight red shift observed in the case of 2,9DMQA can be attributed to the +I inductive effect of methyl groups in the latter molecule, which induces an increase of the electron delocalization throughout the aromatic rings. The molecular aggregation increases in the H2O/DMSO mixture and in solid state, as revealed by the observed red shift in the bands of both molecules as well as and the weakening of the absorption bands. In the case of 2,9-DMQA, this absorption decrease is not so evident, likely due to the lower tendency toward aggregation induced by the presence of both methyl groups, in agreement with previous studies.8 This behavior has also been reported by Gentili et al.28 The absorption spectra indicate that resonant Raman and Surface-enhanced resonant 754
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
two bands associated to carbonyl groups at 1649 and 1631 cm−1 that can be assigned to ν(CO) vibrations in free carbonyl groups and carbonyl groups engaged in the formation of H-bonds, respectively. However, only two bands at 3266 cm−1 and 3227 cm−1 are seen in the FTIR spectrum of 2,9DMQA and only a ν(CO) band at 1644 cm −1 in the Raman spectrum appears. The differences found between 2,9-DMQA and QA can be attributed to the higher tendency of QA molecule to form intermolecular H-bonds. SERRS and SEF of QA and 2,9-DMQA. The SERRS detection of these compounds was performed on Ag colloidal nanoparticles prepared by chemical reduction with hydroxylamine and by using the line at 532 nm as excitation. The latter excitation wavelength was selected to be very close of the absorption maximum of these molecules in order to search for an extra Raman enhancement induced by the resonant conditions. SERRS spectra were obtained at different concentrations in order to get more insight about the adsorption mechanism of these molecules on the metallic surface. Figures 3 and 4 show the global emission spectra (SERRS + SEF) of QA and 2,9-DMQA at different concentrations on AgNPs at 532 nm. They were normalized to the 242 cm−1 band, which corresponds to the ν (Ag−Cl) due to the remaining chloride anions from hydroxylamine hydrochloride employed in the nanoparticle preparation. These spectra were compared to that of the Raman in solid state recorded at 785 nm (Figure 3a and 4a). The SERRS spectra of QA undergo a significant change in comparison to the FT-Raman spectrum (Figure 3Aa). The main changes occurring in SERRS spectra can be summarized as follows: intensity decrease of the 1647 and 1598 cm−1 bands, increase of the intensity of 1424 and 1338 cm−1 bands, as well as the inversion of the 1160/1130 cm−1 doublet. These changes can be attributed to either the resonance effect occurring at 532 nm or to a surface effect derived from the specific orientation of QA on the metallic surface. The most intense bands are localized at 1334 cm−1 and 1161 cm−1, which are intensified due to the resonant conditions. This fact was also observed for others quinacridone derivatives.7 The intensification of bands at 1512 and 1425 cm−1, assigned to δ N−H combined with δ CH, can be related to a close position of NH to the metal surface. The band at 1647 cm−1 seen in the Raman of the solid sample can be assigned to the ν(CO) vibrations. Thus, its weakening suggests that the molecule interacts with the silver nanoparticles through this group, since the interaction with the metal seems to induce a decrease in the double bond character and the subsequent enolization of the carbonyl groups.10 The intensification of δ(CH) bands at 1200−900 cm−1, and structural in-plane δ(CCC) motions (Table 1) in the region below 1000 cm−1 suggests a highly perpendicular orientation of the molecule with respect to the silver surface (Figure 3 bottom). Alternatively, the SERRS spectra do not undergo a significant change when the concentration is modified. This effect indicates that the interaction of QA with the metal is very robust and that the orientation is maintained at different surface coverages. The strong fluorescence emission observed at concentrations higher than 10−6 M (Figure 3B) is attributed to the molecular aggregation increase on the metal surface. As can be seen, these fluorescence bands of QA adsorbed on the metal are red-shifted regarding the bands observed in absence of silver nanoparticles (Figure 3Ba), likely due to the adsorption onto the metal surface. It is worth noting the
Raman scattering (SERRS) can be obtained upon excitation with the 532 nm laser line. Vibrational Spectra. Figure 2 shows the Raman spectra of solid QA and 2,9-DMQA registered at 1064 and 785 nm,
Figure 2. (I) FT-Raman (λ exc= 1064 nm) (a) and FTIR (b) spectra of QA in solid state. (II) Raman (λexc= 785 nm) (a) and IR (b) spectra of 2,9-DMQA in solid state.
respectively. In the case of QA, an excitation at 1064 nm was necessary to obtain this spectrum due to the still high fluorescence observed by exciting at 785 nm, as also indicated by the absorption spectra (Figure 1). IR spectra were also obtained to assist in the assignment of the Raman and SERS spectra. In order to support the vibrational assignment, DFT quantum mechanical calculations were accomplished. The results of the vibrational analysis are given in Tables 1 and 2 for QA and 2,9-DMQA, respectively. As expected, the main differences observed are related to the existence of methyl groups in 2,9-DMQA. The ν(C−H) vibrations associated to the −CH3, characteristic of 2,9-DMQA, appear in the IR spectrum at 2914 cm−1 and 2855 cm−1, meanwhile bands assigned to δ(CH3) and ρ(CH3) appear at 1479 cm−1 and at 1036−998 cm−1, respectively, which appear in both IR and Raman spectra. The bands observed at 3473, 3413, 3258, and 3221 cm−1 in the FTIR spectrum of QA can be assigned to the NH groups. The first two bands are assigned to symmetric and asymmetric ν(NH) vibrations. The other two, can be assigned to ν(NH) when this group interacts with carbonyl groups through Hbonds (ν(NH---O)).29 Besides, the QA Raman spectra shows 755
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
Table 1. Experimental IR and Raman Bands of Solid QA, Together with Calculated Wavenumbers (cm−1) and the Most Probable Assignment of the Vibrational Spectraa
a ν: stretching, δop: out of plane deformations, δip: in plane deformations, (vs): very strong, (s): strong, (m): medium, (w): weak, s: symmetric a: asymmetric ar: aromatic. Scale factor 0.9688 (applied to frequencies higher than 1000 cm−1). A scale factor of 1.0189 was applied to wavenumbers under 1000 cm−1.
756
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
Table 2. Experimental IR and Raman Bands of Solid 2,9-DMQA, Together with Calculated Wavenumbers (cm−1) and the Most Probable Assignment of the Vibrational Spectraa
ν: stretching, δop: out of plane deformations, δip: in plane deformations, (s): strong, (m): medium, (w): weak, s: symmetric a: asymmetric ar: aromatic, alip:aliphatic . Scale factors: 0.9627 (applied to frequencies higher than 1000 cm−1) and 1.0028 (applied to frequencies under 1000 cm−1). a
757
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
Figure 3. SERRS and SEF spectra (λexc = 532 nm) of QA on silver colloid in the 1700−100 cm−1 (A) and 4000−300 cm−1 (B) regions at different concentrations: (b) 10−4 M, (c) 10−5 M, (d) 10−6 M, (e) 10−7 M, (f) 10−8 M, and (g) 10−9 M. The spectra are normalized to the intensity of the ν(AgCl) band. The spectrum Aa is the FT-Raman (λexc = 1064 nm) of QA in solid state, meanwhile the spectrum. Ba is the emission (Raman + Fluorescence with λexc = 532 nm) of QA at 10−6 M in DMSO. Bottom: Scheme of the most probable interaction of Ag/QA.
than 4 × 10−6 M. This effect is attributed to the formation of multilayers on the surface,10 which is further corroborated by the increase of fluorescence emission as the concentration is augmented. Nevertheless, the SERRS spectra obtained at concentrations above 4 × 10−6 M are markedly different from the Raman of the solid sample (Figure 4Aa). For instance, the strong bands at 1309 and 1590 cm−1 in the solid spectrum undergo a large decrease, while other bands appearing at 1239 and 1566 cm−1 are enhanced. According to the assignment provided in Table 2, the first two bands can be attributed to δ(NH)/ν(CN) vibrations which are modified upon interaction of the amino group with the surface, although the presence of ν(CO) bands also suggests the existence of nonbonded molecules on the surface, due again to the formation of multilayers. Other differences between QA and 2,9-DMQA arise from changes in the relative intensity of bands appearing in the 1350−1130 cm−1 region that can be attributed to the presence of methyl groups in 2,9-DMQA. In this respect, one can observe an intensification of bands corresponding to the methyl groups in the SERS spectra recorded at these higher concentrations. In order to obtain information relative to the molecular orientation of 2,9-DMQA on the surface, we have focused our attention on the bands at 1484 and 1239 cm−1, which can be assigned to δa(CH3) of methyl groups and the δip(CH) of
change in the relative intensity of fluorescence bands which also depends on the concentration. This effect can be attributed to the aggregation of QA on the surface and the formation of multilayers on the interface, and the subsequent surface effect on the adsorbed aggregates depending on their orientation. Above a QA concentration of 10−6 M, the fluorescence emission grows, overlapping the Raman bands. The increase of fluorescence can then be related to the surface-enhanced fluorescence effect resulting from the self-aggregation of adsorbate on the surface. Indeed, the analysis of this SEF emission may provide important information about the aggregation process occurring in quinacridones when adsorbed on metal surfaces. SERRS spectra of 2,9-DMQA are displayed in Figure 4. The spectral profile of this molecule exhibits significant differences in relation to QA, in spite of the analogous structure of both pigments. The study of the influence of concentration on the SERRS of 2,9-DMQA revealed a progressive increasing of bands corresponding to the Raman of the solid pigment (Figure 4Aa) as the concentration is raised. At low concentrations (10−9, Figure 4Ag), the spectrum is dominated by bands at 1313, 1243, and 1207 cm−1, assigned to deformations in the skeletal molecular plane. Besides, no ν(CO) band is observed, which, as in the case of QA, is attributed to the interaction of 2,9-DMQA with the metal. However, the latter bands grow again at concentrations higher 758
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
Figure 4. SERRS and SEF spectra (λexc = 532 nm) of 2,9-DMQA on silver colloid in the 1700−150 cm−1 (A) and 3500−150 cm−1 (B) regions at the following concentrations: (b) 10−4 M, (c) 10−5 M, (d) 5 × 10−6 M, (e) 4 × 10−6 M, (f) 3 × 10−6 M, and (g) 10−9 M. They are normalized to the ν (AgCl) band. Green and blue dash lines: vibrational markers of the molecular orientation. The Aa spectrum is the Raman (λexc = 785 nm) of QA in solid state, while the Ba is the emission spectrum (Raman + Fluorescence with λexc = 532 nm) of 2,9-DMQA at 10−8 M in DMSO. Bottom: Scheme of the deduced interaction and orientation of 2,9-DMQA on the surface as a function of the concentration: (A) at [2,9-DMQA]>3 × 10−6 M Edgeon and (B) [2,9-DMQA] < 3 × 10−6 M parallel.
to the aggregation of 2,9-DMQA on the metal surface in comparison to QA, in agreement with the UV−visible spectra studied above. Quantitative Analysis. The high sensitive of the SERS technique allows the study of the absorption mechanism of these molecules on the metallic surface in terms of the adsorption isotherms of the molecules, as well as estimating the limit of detection (LOD) of the corresponding analyte. The large fluorescence emission of QA SERRS spectra avoids to a large extent to develop an accurate study of the adsorption of this molecule. However, for 2,9-DMQA, this analysis is possible due to the lower fluorescence emission. Thus, we have centered our attention on this molecule. In Figure 5, the total intensity of the SERRS spectra, considered as the total area measured in the 1800−297 cm−1 spectral region (baseline corrected) of the SERRS for three different batches, are plotted, together with the corresponding error, as a function of the concentration. Since the SERS intensity exclusively depends on the number of molecules adsorbed on a plasmonic surface, due to the shortrange character of the SERS effect, the latter plot can be considered as an actual adsorption isotherm of 2,9-DMQA. The adsorption SERRS spectra recorded at different concentrations reveals that no intensity saturation is taking place, in contrast to the Langmuir-type adsorption observed for other adsorbates previously studied.31 The obtained adsorption
aromatic rings, respectively. These two vibrations have a different orientation in relation to the molecular plane, the δa(CH3) being predominantly perpendicular, while the δip(CH) is parallel. Thus, the I1239/I1484 ratio can be employed to follow changes in the orientation occurring when this molecule is adsorbed under different surface coverages, considering the SERS selection rules.30 From the SERRS spectra, we have deduced that the above ratio undergoes an increase as the concentration is raised. This is attributed to a reorientation of 2,9-DMQA from a predominantly parallel orientation at low concentrations to a more perpendicular one at high coverage (Figure 4, bottom scheme). This reorientation is not observed in the case of QA, and this could be attributed to the presence of methyl groups in the structure of 2,9-DMQA. The differences observed in the adsorption mechanism between QA and 2,9-DMQA are also manifested on the fluorescence emission. As in the case of QA, the fluorescence bands of 2,9-DMQA (Figure 4B) in presence of metal nanoparticles are red-shift with respect to the fluorescence of 2,9-DMQA without silver (Figure 4Ba), suggesting that an intermolecular interaction between 2,9-DMQA molecules occurs. As can be seen, the fluorescence emission of 2,9DMQA when adsorbed on the metal is lower than QA and the intensity of the bands are not enhanced below a concentration of 10−4 M. This fact can be again attributed to a lower tendency 759
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
visible spectra reveals a strong aggregation tendency for both molecules and a bathochromic shift effect in the case of the 2,9DMQA in comparison to QA. This differential behavior is attributed to the presence of two methyl groups in 2,9-DMQA. The analysis of SERRS spectra at different concentrations demonstrated a different behavior for these quinacridones regarding their absorption mechanism on silver. While QA is adsorbed perpendicularly to the surface, regardless of the concentration, the adsorption of 2,9-DMQA depends on the surface coverage resulting from different concentrations. At concentrations lower than 3 × 10−6 M, the molecule adopts a predominant parallel orientation, whereas at higher concentrations, the molecule is adsorbed perpendicular to the metal surface. The SEF data indicate a large tendency toward intermolecular interactions of QA with respect to the 2,9-DMQA. Fluorescence bands are shifted toward higher wavelength as a consequence of the molecular aggregation occurring on the metal surface. In addition, the relative intensity of these bands changes due to the surface effect. The lower fluorescence emission observed in the emission spectra of 2,9-DMQA is attributed to the lower aggregation tendency of this molecule because of the presence of the methyl groups, which permitted a quantitative analysis of the adsorption on silver nanoparticles based on the adsorption isotherms. From this study, a BET adsorption model was deduced for 2,9-DMQA, indicating the key role of intermolecular interactions in the adsorption of this molecule on the surface. This is the first time that an adsorption by BET model is observed by SERS. Finally, a limit of detection of 54.4 ppb was calculated for 2,9-DMQA, thus demonstrating the higher sensitivity of the SERS technique in the analysis of quinacridone pigments.
Figure 5. Plot of SERRS intensity of 2,9-DMQA as a function of the dye concentration. Inset: BET function.
isotherm curve fits a BET model,32 and it is similar to other examples found in the literature,33−35 which suggests the importance of the intermolecular interactions in this molecule. The parameters related to the adsorption can be deduced from the BET adsorption model according to the following expression: A=
A m KC Cm + (K − 2)C − (k − 1/Cm)C 2
([1])
where A is the total intensity of each SERRS spectrum; Am is the total intensity of SERRS spectra if a monolayer is formed on the metal surface; C is the 2,9-DMQA concentration, Cm is the concentration of saturation of a monolayer; and, finally, K is the adsorption constant which is defined as K = eE1−EL/RT, with EL being the adsorption enthalpy of a monolayer, E1 the adsorption enthalpy of consecutive layers, R the gas constant, and T the temperature. As far as we know, this is the first time that a BET model adsorption behavior is deduced from SERS data. This is probably due to the fact that quinacridones have a large tendency toward aggregation due to the favored intermolecular interactions. The adsorption parameters Am, K, and Cm deduced from the analysis of the adsorption isotherm are summarized in Table 3.
■
Corresponding Author
*E-mail: [email protected] (S.S.-C.); edpuertonevado@ gmail.com (E.d.P.). Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work has been financially supported by MINECO (Projects FIS2010-15405, and CONSOLIDER CSD20070058/TCP) and the Comunidad de Madrid (MICROSERES II Project S2009/TIC1476). E.d.P. acknowledges CSIC and FSE 2007-2013 for a JAE_CSIC predoctoral grant.
Table 3. BET Parameters Deduced from the Analysis of the Adsorption Isotherm Am
K
1.2 × 106 ± 4 × 105
650
Cm 1.1 × 10−4 ± 1 × 10
−5
R
LOD (ng/ mL)
0.98
54.4
AUTHOR INFORMATION
■
REFERENCES
(1) Jaffe, E. E. High Performance Pigments, Second ed.; Faulkner, E. B., Schwartz, R. J., Eds.; WILEY-VCH Verlag GmbH & Co. KGaA: Weinheim, 2009; pp 298−329. (2) Meier, H.; Albrecht, W. Photoelectric Properties of 2,9-DimethylQuinacridone. Z. Phys.Chem. 1986, 148, 171−183. (3) Chen, T. L.; Chen, J. J. A.; Catane, L.; Ma, B. Fully Solution Processed P-I-N Organic Solar Cells with an Industrial Pigment Quinacridone. Org. Electron. 2011, 12, 1126−1131. (4) Jabbour, G. E.; Kawabe, Y.; Shaheen, S. E.; Wang, J. F.; Morrell, M. M.; Kippelen, B.; Peyghambarian, N. Highly Efficient and Bright Organic Electroluminescent Devices with an Aluminum Cathode. Appl. Phys. Lett. 1997, 71, 1763. (5) Blochwitz, J.; Pfeiffer, M.; Hofmann, H.; Leo, K. Non-Polymeric Oleds with a Doped Amorphous Hole Transport Layer and Operating
The limit of detection (LOD) was calculated by considering the concentration at which a SERRS intensity (A) is three times higher than the average noise of the 1800−297 cm−1 spectral region of the Raman spectrum of the blank. The calculated LOD was 54.4 ppb, which points out the high sensitivity of the SERS technique in the detection of this kind of compounds.
■
CONCLUSIONS Surface-Enhanced Raman Spectroscopy has proven to be a suitable technique for the detection of QA and 2,9-DMQA, despite their low solubility. The analysis of their corresponding 760
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
Voltages Down to 3.2 V to Achieve 100 Cd/M2. Synth. Met. 2002, 127, 169−173. (6) Li, C.; Xiao, B.; Li, X.; Zhao, Y.; Hou, J.; Liu, S. Quinacridone Sub-Monolayer as Efficient Emitter in Oleds. J. Lumin. 2007, 122− 123, 623−625. (7) del Puerto, E.; Domingo, C.; Sanchez-Cortes, S.; García-Ramos, J. V.; Aroca, R. F. Plasmon Enhanced Spectroscopy of N,N0Dialkylquinacridones Used as Codopants in Oleds. J. Phys. Chem. C 2011, 115, 16838. (8) Senju, T.; Sakai, M.; Mizuguchi, J. Cohesion of G-Quinacridone and 2,9-Dimethylquinacridone in the Solid State. Dyes Pigments 2007, 75, 449−453. (9) Mizuguchi, J.; Senju, T. Solution and Solid-State Spectra of Quinacridone Derivatives as Viewed from the Intermolecular Hydrogen Bond. J. Phys. Chem.B 2006, 110, 19154−19161. (10) del Puerto, E.; Sanchez-Cortes, S.; Garcia-Ramos, J. V.; Domingo, C. Solution SERS of an Insoluble Synthetic Organic PigmentQuinacridone QuinoneEmploying Calixarenes as Dispersive Agent. Chem. Commun. 2011, 47, 1854. (11) del Puerto, E.; Cuesta, A.; Sanchez-Cortes, S.; J.V., G.-R.; Domingo, C. Electrochemical Sers Study on a Copper Electrode of the Insoluble Organic Pigment Quinacridone Quinone Using Ionic Liquids (BMIMCl and TBAN) as Dispersing Agents. Analyst 2013, 138, 4670−4676. (12) Lincke, G. On Quinacridones and Their Supramolecular Mesomerism within the Crystal Lattice. Dyes Pigments 2002, 52, 169−181. (13) Keller, U.; Miillen, K.; De Feyter, S.; De Schryver, F. C. Hydrogen-Bonding and Phase-Forming Behavior of a Soluble Quinacridone. Adv. Mater. 1996, 8, 6. (14) Binant, C.; Lautie, A. Identification of Quinacridone Pigment Mixtures Using Resonance Raman Microspectrometry. Appl. Spectrosc. 1989, 43 (5), 851−855. (15) Binant, C.; Guineau, B.; Lautie, A. The Application of Electronic and Vibrational Spectroscopy Techniques to the Identification of Quinacridone Pigments in Vehicle Paint Systems. J. Soc. Dyers Colour. 1998, 106 (5−6), 187−191. (16) Lincke, G. A Review of Thirty Years of Research on Quinacridones. X-Ray Crystallography and Crystal Engineering. Dyes Pigments 2000, 44, 101−122. (17) Ann Menke, C.; Rivenc, R.; Learner, T. The Use of Direct Temperature-Resolved Mass Spectrometry (Dtms) in the Detection of Organic Pigments Found in Acrylic Paints Used by Sam Francis. Int. J. Mass Spectrom. 2009, 284 (1−3), 2−11. (18) Kirbya, D. P.; Khandekara, N.; Sutherlandb, K.; Priceb, B. A. Applications of Laser Desorption Mass Spectrometry for the Study of Synthetic Organic Pigments in Works of Art. Int. J. Mass Spectrom. 2009, 284 (1−3), 115−122. (19) Jurasekova, Z.; del Puerto, E.; Bruno, G.; Garcia-Ramos, J. V.; Sanchez-Cortes, S.; Domingo, C. Extractionless Non-Hydrolysis Surface-Enhanced Raman Spectroscopic Detection of Historical Mordant Dyes on Textile Fibers. J. Raman Spectrosc. 2010, 41, 1455−1461. (20) Jurasekova, Z.; Garcia-Ramos, J. V.; Domingo, C.; SanchezCortes, S. Surface-Enhanced Raman Scattering of Flavonoids. J. Raman Spectrosc. 2006, 37 (11), 1239−1241. (21) Leona, M.; Decuzzi, P.; Kubic, T. A.; Gates, G.; Lombardi, J. R. Nondestructive Identification of Natural and Synthetic Organic Colorants in Works of Art by Surface Enhanced Raman Scattering. Anal. Chem. 2011, 83, 3990−3993. (22) Wustholz, K. L.; Brosseau, C. L.; Casadio, F.; Van Duyne, R. P. Surface-Enhanced Raman Spectroscopy of Dyes: From Single Molecules to the Artists’ Canvas. Phys. Chem. Chem. Phys. 2009, 11, 7350−7359. (23) Pozzi, F.; Lombardi, J. R.; Bruni, S.; Leona, M. Sample Treatment Considerations in the Analysis of Organic Colorants by Surface-Enhanced Raman Scattering. Anal. Chem. 2012, 84 (8), 3751− 3757.
(24) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. G. Gaussian 09, Gaussian, Inc.: Wallingford CT, 2009. (25) Leopold, N.; Lendl, B. A New Method for Fast Preparation of Highly Surface Enhanced Raman Scattering (SERS) Active Colloids at Room Temperature by Reduction of Silver Nitrate with Hydroxylamine Hydrochloride. J. Phys.Chem.B 2003, 107, 5723−5727. (26) Merrick, J. P.; Moran, D.; Radom, L. An Evaluation of Harmonic Vibrational Frequency Scale Factor. J.Phys. Chem. A 2007, 111, 11683−11700. (27) DeSaja-Gonzalez, J.; al., e. Surface-Enhanced Fluorescence and Serrs Spectra of N-Octadecyl-3,4:9,10-Perylenetetracarboxylic Monoanhydride on Silver Island Films. Spectrochim. Acta A 1997, 53, 173− 181. (28) Gentili, P. L.; Clementi, C.; Romani, A. Ultraviolet−Visible Absorption and Luminescence Properties of Quinacridone−Barium Sulfate Solid Mixtures. Appl. Spectrosc. 2010, 64, 923−929. (29) Moritz, A. G. The N−H Stretching Frequencies and Molecular Configurations of Some Aromatic Amines. Spectrochim. Acta 1960, 16 (10), 1176−1183. (30) Moskovits, M.; Suh, J. S. Surface Selection Rules for SurfaceEnhanced Raman Spectroscopy: Calculations and Application to the Surface-Enhanced Raman Spectrum of Phtalazine on Silver. J. Phys. Chem. 1984, 88, 5526−5530. (31) Izquierdo-Lorenzo, I.; Alda, I.; Sanchez-Cortes, S.; GarciaRamos, J. V. Adsorption and Detection of Sport Doping Drugs on Metallic Plasmonic Nanoparticles of Different Morphology. Langmuir 2012, 28, 8891−8901. (32) Brunauer, S.; Emmett, P. H.; Teller, E. Adsorption of Gases in Multimolecular Layers. J. Am. Chem. Soc. 1938, 60, 309. (33) Mikhalev, Y.; Øye, H. A. Absorption of Metallic Sodium in Carbon Cathode Materials. Carbon 1996, 34 (1), 37−41. (34) Koh, K.; Wong-Foy, A. G.; Matzger, A. J. A Porous Coordination Copolymer with over 5000 M2/G Bet Surface Area. J. Am. Chem. Soc. 2009, 131 (12), 4184−4185. (35) Dormant, L. M.; Adamson, A. W. Application of the Bet Equation to Heterogeneous Surfaces. J. Colloid Interface Sci. 1972, 38 (1), 285−289.
761
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Adsorption Study and Detection of the High Performance Organic Pigments Quinacridone and 2,9-Dimethylquinacridone on Ag Nanoparticles By Surface-Enhanced Optical Spectroscopy Elena del Puerto,* Concepcion Domingo, Jose V. Garcia Ramos, and Santiago Sanchez-Cortes* Instituto de Estructura de la Materia, IEM-CSIC, Serrano 121, 28006 Madrid, Spain. ABSTRACT: In this work, surface-enhanced Raman scattering (SERS) and surface-enhanced fluorescence (SEF) were employed in the study of the adsorption and detection of the pigments quinacridone (QA) and 2,9dimethylquinacridone (2,9-DMQA). These pigments are of great relevance in artwork and textile, plastic, and photochemical industries due to their condition as high performance pigments since they possess a high tinting strength. Due to this fact, they have been employed at relatively low concentrations. Therefore, the analysis and detection of these pigments requires the application of a highly sensitivity technique, such as SERS and SEF. The adsorption of QA and 2,9-DMQA on silver nanoparticles was extensively studied by means of SERS at different surface coverages. This study was completed by carrying out an in depth vibrational (Raman and IR) analysis of these pigments in solid state by ab initio density functional theory (DFT) calculations. In addition, UV−vis spectroscopy was employed to investigate the aggregation undergone by both pigments in solid state and in solution. 2,9-DMQA was demonstrated to have a lower tendency toward aggregation due to the presence of methyl groups. Even so, this molecule follows a BET adsorption mechanism on the metal nanoparticles due to its high tendency toward intermolecular interaction. From the analysis of the adsorption mechanism of this molecule, the limit of detection was deduced to be ca. 55 ppb.
■
INTRODUCTION Quinacridones are high performance pigments (HPPs). As is well-known, HPPs, organic or inorganic, are an important segment of the diverse and rich field of color and visual effect technology. They exhibit enhanced durability. The most salient durability feature is generally regarded as resistance to visible and ultraviolet radiation (lightfastness), but heat stability and chemical resistance are also critical attributes. Specifically, quinacridones present outstanding chemical and photochemical stability, extremely low solubility in water, and attractive colors that make these compounds one of the most important classes of HPPs.1 In this work, we have centered our attention on Quinacridone (QA) and 2,9-dimethylquinacridone (2,9DMQA), two of the most important molecules of this group. QA, labeled internationally as PV19 by Color Index, is the principal compound of this family and is mainly employed to provide a violet color. Its use is widespread both in paints in artwork and as a surface application in plastics, textiles, and so forth. 2,9-DMQA (PR122) shows a reddish tonality and is a suitable colorant to employ in plastic applications and in printing inks. Occasionally, quinacridones are mixed to achieve different tonalities. The interesting photochemical properties make them very appropriate for use in solar cells2,3and OLEDs.4−6 Because of the great variety of applications, their identification and detection is not only important in the art conservation field, but also in the fields of forensics and electronics to understand and improve devices.7 However, the © 2013 American Chemical Society
identification of quinacridones is not straightforward due to the particular properties derived from their structures. The inset of Figure 1 shows the chemical structure of QA and 2,9-DMQA. As we can observe, they present the same rigid skeletal body: five condensed aromatic rings and three benzene rings alternated with two 4-pyridones. Accordingly, they are characterized by NH−O hydrogen bonds,8,9 which determine their low solubility in water that complicate its characterization. In previous works, we have studied another derivative of quinacridone, i.e., the quinacridone quinone, which also presents the same problem regarding its characterization, its insolubility due to the hydrogen bonds. In the latter case, we have developed several strategies to overcome this drawback, employing calixarenes10 or ionic liquids11 to disperse the pigment. In the case of 2,9-DMQA, the presence of two methyl groups, displaying a +I inductive effect, in the 2,9 position, introduces new chemical properties in the aromatic system that may have consequences on its intermolecular interaction and, thus, on its solubility. Thus far, quinacridones have been characterized by UV− visible, fluorescence, X-ray diffraction, thermogravimetric analysis, FT-IR, Raman microspectroscopy, and mass spectroscopy. 8,12−18 To our knowledge, the majority of the Received: September 19, 2013 Revised: December 21, 2013 Published: December 24, 2013 753
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
and the low aggregation of 2,9-DMQA allowed analysis of its adsorption on the metallic surface in terms of the adsorption isotherms of the molecules, from where the limit of detection (LOD) was also calculated.
■
EXPERIMENTAL SECTION
Quinacridone (QA) was purchased from TIC Europe with the highest purity (>99%). The 2,9-dimethylquinacridone (2,9-DMQA) came compliments of SunChemical. Both reagents were solved in dimethyl sulfoxide (DMSO) which was acquired from Sigma-Aldrich with >99.9% purity. Silver colloids for SERS measurements were made by reduction of silver nitrate with hydroxylamine chloride at room temperature.25 After aggregation with potassium nitrate solution up to a final concentration of 3 × 10−2 M, 10 μL portions of the solution in DMSO at variable concentrations of each dye were added to 1 mL of the silver colloids to get the final desired concentration. All reactants involved in the silver colloid synthesis were purchased from Sigma-Aldrich with analytical purity. UV−vis-NIR Shimadzu 3600 has been employed to register the extinction spectra of QA and 2,9-DMQA in solution and in solid state. To register the spectrum in solid state, an integrating sphere was used. FTIR spectra of solid quinacridone derivatives dispersed in KBr were recorded with a Bruker IFS66 spectrometer with a DGTS detector. The spectral resolution was 8 cm−1, and 500 scans were acquired. The Raman spectrum of QA was recorded using an FT-Raman Bruker MultiRam spectrometer with a liquid N2 cooled Ge detector and an Nd:YAG laser at 1064 nm. The final spectra were obtained after accumulation of 500 scans. The SERS experiments (exciting at 532 nm) as well as the Raman spectra (exciting at 785 nm) of 2,9DMQA in solid state have been registered with a Renishaw inVia Raman microscope, RM2000 model, equipped with a Leica microscope and an electrically cooled CCD camera. DFT calculations were done using B3LYP and 6-311G+ (d,p) basis sets, using the Gaussian 09 program.24 A scale factor of 1.0189 was applied to wavenumbers under 1000 cm−1 according to ref 25.26 The adsorption isotherm of 2,9-DMQA was calculated by plotting the overall area of SERS spectra in the range 1800−297 cm−1 versus the concentration. The Origin 8.0 software was employed to carry out the analysis.
Figure 1. UV−visible Absorption spectra of QA (top) and 2,9-DMQA (bottom) under the following conditions: in solid state (a); in DMSO (b); and in H2O/DMSO (1% v/v) (c). [QA] = 10−4 M and [2,9DMQA] = 10−5 M. Inset: the molecular structure of QA and 2,9DMQA. The vertical line corresponds to the position of the excitation line employed to obtain the SERRS spectra.
quinacridone studies have been focused on their crystalline structure and only a few papers report their identification in art works employing mass spectroscopy.17,18 However, due to their high tinting strength, quinacridones are frequently present in relatively small quantities, and their identification requires high sensitivity analytical techniques. Therefore, surface-enhanced Raman scattering (SERS) and surface-enhanced fluorescence (SEF), two important surface-enhanced optical spectroscopies (SEOS) based on the localization of surface plasmons, are employed for the first time in this work, aimed at the characterization of such pigments. One of the most appreciated advantages of SEOS techniques regarding pigment identification and characterization in artwork is its nanodestructive and extractionless characteristic, thus making SERS and SEF very suitable for the analysis of materials of interest for cultural heritage.19−23 In this work, SERS spectra of both QA and 2,9-DMQA were obtained under different experimental conditions, and a comparative study between both pigments was done to understand the influence of methyl groups on the adsorption and interaction mechanism with the metal surface, which will further determine the detection ability of SERS. To accomplish a proper assignment of SERS spectra, theoretical calculations of the vibrational frequencies using the Gaussian 09 program24 were previously performed. Fluorescence emission has also been studied in order to obtain in-depth information about their aggregation effect. The sensitivity of the SERS technique
■
RESULTS AND DISCUSSION UV−visible Absorption Spectra. Figure 1 shows the UV−visible absorption of QA and 2,9-DMQA in solid state (Figure 1a), in DMSO (Figure 1b) and in H2O/DMSO (1%) (Figure 1c). The latter experimental conditions are interesting since they simulate the chemical environment of SERS experiments, except for the presence of silver nanoparticles. In DMSO, the aggregation of these molecules is not expected, and three maxima can be seen, corresponding to the condensed aromatic structure of QAs.27 These bands appear at 523, 489, and 460 nm and 531, 496, and 470 nm, for QA and 2,9-DMQA, respectively. The slight red shift observed in the case of 2,9DMQA can be attributed to the +I inductive effect of methyl groups in the latter molecule, which induces an increase of the electron delocalization throughout the aromatic rings. The molecular aggregation increases in the H2O/DMSO mixture and in solid state, as revealed by the observed red shift in the bands of both molecules as well as and the weakening of the absorption bands. In the case of 2,9-DMQA, this absorption decrease is not so evident, likely due to the lower tendency toward aggregation induced by the presence of both methyl groups, in agreement with previous studies.8 This behavior has also been reported by Gentili et al.28 The absorption spectra indicate that resonant Raman and Surface-enhanced resonant 754
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
two bands associated to carbonyl groups at 1649 and 1631 cm−1 that can be assigned to ν(CO) vibrations in free carbonyl groups and carbonyl groups engaged in the formation of H-bonds, respectively. However, only two bands at 3266 cm−1 and 3227 cm−1 are seen in the FTIR spectrum of 2,9DMQA and only a ν(CO) band at 1644 cm −1 in the Raman spectrum appears. The differences found between 2,9-DMQA and QA can be attributed to the higher tendency of QA molecule to form intermolecular H-bonds. SERRS and SEF of QA and 2,9-DMQA. The SERRS detection of these compounds was performed on Ag colloidal nanoparticles prepared by chemical reduction with hydroxylamine and by using the line at 532 nm as excitation. The latter excitation wavelength was selected to be very close of the absorption maximum of these molecules in order to search for an extra Raman enhancement induced by the resonant conditions. SERRS spectra were obtained at different concentrations in order to get more insight about the adsorption mechanism of these molecules on the metallic surface. Figures 3 and 4 show the global emission spectra (SERRS + SEF) of QA and 2,9-DMQA at different concentrations on AgNPs at 532 nm. They were normalized to the 242 cm−1 band, which corresponds to the ν (Ag−Cl) due to the remaining chloride anions from hydroxylamine hydrochloride employed in the nanoparticle preparation. These spectra were compared to that of the Raman in solid state recorded at 785 nm (Figure 3a and 4a). The SERRS spectra of QA undergo a significant change in comparison to the FT-Raman spectrum (Figure 3Aa). The main changes occurring in SERRS spectra can be summarized as follows: intensity decrease of the 1647 and 1598 cm−1 bands, increase of the intensity of 1424 and 1338 cm−1 bands, as well as the inversion of the 1160/1130 cm−1 doublet. These changes can be attributed to either the resonance effect occurring at 532 nm or to a surface effect derived from the specific orientation of QA on the metallic surface. The most intense bands are localized at 1334 cm−1 and 1161 cm−1, which are intensified due to the resonant conditions. This fact was also observed for others quinacridone derivatives.7 The intensification of bands at 1512 and 1425 cm−1, assigned to δ N−H combined with δ CH, can be related to a close position of NH to the metal surface. The band at 1647 cm−1 seen in the Raman of the solid sample can be assigned to the ν(CO) vibrations. Thus, its weakening suggests that the molecule interacts with the silver nanoparticles through this group, since the interaction with the metal seems to induce a decrease in the double bond character and the subsequent enolization of the carbonyl groups.10 The intensification of δ(CH) bands at 1200−900 cm−1, and structural in-plane δ(CCC) motions (Table 1) in the region below 1000 cm−1 suggests a highly perpendicular orientation of the molecule with respect to the silver surface (Figure 3 bottom). Alternatively, the SERRS spectra do not undergo a significant change when the concentration is modified. This effect indicates that the interaction of QA with the metal is very robust and that the orientation is maintained at different surface coverages. The strong fluorescence emission observed at concentrations higher than 10−6 M (Figure 3B) is attributed to the molecular aggregation increase on the metal surface. As can be seen, these fluorescence bands of QA adsorbed on the metal are red-shifted regarding the bands observed in absence of silver nanoparticles (Figure 3Ba), likely due to the adsorption onto the metal surface. It is worth noting the
Raman scattering (SERRS) can be obtained upon excitation with the 532 nm laser line. Vibrational Spectra. Figure 2 shows the Raman spectra of solid QA and 2,9-DMQA registered at 1064 and 785 nm,
Figure 2. (I) FT-Raman (λ exc= 1064 nm) (a) and FTIR (b) spectra of QA in solid state. (II) Raman (λexc= 785 nm) (a) and IR (b) spectra of 2,9-DMQA in solid state.
respectively. In the case of QA, an excitation at 1064 nm was necessary to obtain this spectrum due to the still high fluorescence observed by exciting at 785 nm, as also indicated by the absorption spectra (Figure 1). IR spectra were also obtained to assist in the assignment of the Raman and SERS spectra. In order to support the vibrational assignment, DFT quantum mechanical calculations were accomplished. The results of the vibrational analysis are given in Tables 1 and 2 for QA and 2,9-DMQA, respectively. As expected, the main differences observed are related to the existence of methyl groups in 2,9-DMQA. The ν(C−H) vibrations associated to the −CH3, characteristic of 2,9-DMQA, appear in the IR spectrum at 2914 cm−1 and 2855 cm−1, meanwhile bands assigned to δ(CH3) and ρ(CH3) appear at 1479 cm−1 and at 1036−998 cm−1, respectively, which appear in both IR and Raman spectra. The bands observed at 3473, 3413, 3258, and 3221 cm−1 in the FTIR spectrum of QA can be assigned to the NH groups. The first two bands are assigned to symmetric and asymmetric ν(NH) vibrations. The other two, can be assigned to ν(NH) when this group interacts with carbonyl groups through Hbonds (ν(NH---O)).29 Besides, the QA Raman spectra shows 755
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
Table 1. Experimental IR and Raman Bands of Solid QA, Together with Calculated Wavenumbers (cm−1) and the Most Probable Assignment of the Vibrational Spectraa
a ν: stretching, δop: out of plane deformations, δip: in plane deformations, (vs): very strong, (s): strong, (m): medium, (w): weak, s: symmetric a: asymmetric ar: aromatic. Scale factor 0.9688 (applied to frequencies higher than 1000 cm−1). A scale factor of 1.0189 was applied to wavenumbers under 1000 cm−1.
756
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
Table 2. Experimental IR and Raman Bands of Solid 2,9-DMQA, Together with Calculated Wavenumbers (cm−1) and the Most Probable Assignment of the Vibrational Spectraa
ν: stretching, δop: out of plane deformations, δip: in plane deformations, (s): strong, (m): medium, (w): weak, s: symmetric a: asymmetric ar: aromatic, alip:aliphatic . Scale factors: 0.9627 (applied to frequencies higher than 1000 cm−1) and 1.0028 (applied to frequencies under 1000 cm−1). a
757
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
Figure 3. SERRS and SEF spectra (λexc = 532 nm) of QA on silver colloid in the 1700−100 cm−1 (A) and 4000−300 cm−1 (B) regions at different concentrations: (b) 10−4 M, (c) 10−5 M, (d) 10−6 M, (e) 10−7 M, (f) 10−8 M, and (g) 10−9 M. The spectra are normalized to the intensity of the ν(AgCl) band. The spectrum Aa is the FT-Raman (λexc = 1064 nm) of QA in solid state, meanwhile the spectrum. Ba is the emission (Raman + Fluorescence with λexc = 532 nm) of QA at 10−6 M in DMSO. Bottom: Scheme of the most probable interaction of Ag/QA.
than 4 × 10−6 M. This effect is attributed to the formation of multilayers on the surface,10 which is further corroborated by the increase of fluorescence emission as the concentration is augmented. Nevertheless, the SERRS spectra obtained at concentrations above 4 × 10−6 M are markedly different from the Raman of the solid sample (Figure 4Aa). For instance, the strong bands at 1309 and 1590 cm−1 in the solid spectrum undergo a large decrease, while other bands appearing at 1239 and 1566 cm−1 are enhanced. According to the assignment provided in Table 2, the first two bands can be attributed to δ(NH)/ν(CN) vibrations which are modified upon interaction of the amino group with the surface, although the presence of ν(CO) bands also suggests the existence of nonbonded molecules on the surface, due again to the formation of multilayers. Other differences between QA and 2,9-DMQA arise from changes in the relative intensity of bands appearing in the 1350−1130 cm−1 region that can be attributed to the presence of methyl groups in 2,9-DMQA. In this respect, one can observe an intensification of bands corresponding to the methyl groups in the SERS spectra recorded at these higher concentrations. In order to obtain information relative to the molecular orientation of 2,9-DMQA on the surface, we have focused our attention on the bands at 1484 and 1239 cm−1, which can be assigned to δa(CH3) of methyl groups and the δip(CH) of
change in the relative intensity of fluorescence bands which also depends on the concentration. This effect can be attributed to the aggregation of QA on the surface and the formation of multilayers on the interface, and the subsequent surface effect on the adsorbed aggregates depending on their orientation. Above a QA concentration of 10−6 M, the fluorescence emission grows, overlapping the Raman bands. The increase of fluorescence can then be related to the surface-enhanced fluorescence effect resulting from the self-aggregation of adsorbate on the surface. Indeed, the analysis of this SEF emission may provide important information about the aggregation process occurring in quinacridones when adsorbed on metal surfaces. SERRS spectra of 2,9-DMQA are displayed in Figure 4. The spectral profile of this molecule exhibits significant differences in relation to QA, in spite of the analogous structure of both pigments. The study of the influence of concentration on the SERRS of 2,9-DMQA revealed a progressive increasing of bands corresponding to the Raman of the solid pigment (Figure 4Aa) as the concentration is raised. At low concentrations (10−9, Figure 4Ag), the spectrum is dominated by bands at 1313, 1243, and 1207 cm−1, assigned to deformations in the skeletal molecular plane. Besides, no ν(CO) band is observed, which, as in the case of QA, is attributed to the interaction of 2,9-DMQA with the metal. However, the latter bands grow again at concentrations higher 758
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
Figure 4. SERRS and SEF spectra (λexc = 532 nm) of 2,9-DMQA on silver colloid in the 1700−150 cm−1 (A) and 3500−150 cm−1 (B) regions at the following concentrations: (b) 10−4 M, (c) 10−5 M, (d) 5 × 10−6 M, (e) 4 × 10−6 M, (f) 3 × 10−6 M, and (g) 10−9 M. They are normalized to the ν (AgCl) band. Green and blue dash lines: vibrational markers of the molecular orientation. The Aa spectrum is the Raman (λexc = 785 nm) of QA in solid state, while the Ba is the emission spectrum (Raman + Fluorescence with λexc = 532 nm) of 2,9-DMQA at 10−8 M in DMSO. Bottom: Scheme of the deduced interaction and orientation of 2,9-DMQA on the surface as a function of the concentration: (A) at [2,9-DMQA]>3 × 10−6 M Edgeon and (B) [2,9-DMQA] < 3 × 10−6 M parallel.
to the aggregation of 2,9-DMQA on the metal surface in comparison to QA, in agreement with the UV−visible spectra studied above. Quantitative Analysis. The high sensitive of the SERS technique allows the study of the absorption mechanism of these molecules on the metallic surface in terms of the adsorption isotherms of the molecules, as well as estimating the limit of detection (LOD) of the corresponding analyte. The large fluorescence emission of QA SERRS spectra avoids to a large extent to develop an accurate study of the adsorption of this molecule. However, for 2,9-DMQA, this analysis is possible due to the lower fluorescence emission. Thus, we have centered our attention on this molecule. In Figure 5, the total intensity of the SERRS spectra, considered as the total area measured in the 1800−297 cm−1 spectral region (baseline corrected) of the SERRS for three different batches, are plotted, together with the corresponding error, as a function of the concentration. Since the SERS intensity exclusively depends on the number of molecules adsorbed on a plasmonic surface, due to the shortrange character of the SERS effect, the latter plot can be considered as an actual adsorption isotherm of 2,9-DMQA. The adsorption SERRS spectra recorded at different concentrations reveals that no intensity saturation is taking place, in contrast to the Langmuir-type adsorption observed for other adsorbates previously studied.31 The obtained adsorption
aromatic rings, respectively. These two vibrations have a different orientation in relation to the molecular plane, the δa(CH3) being predominantly perpendicular, while the δip(CH) is parallel. Thus, the I1239/I1484 ratio can be employed to follow changes in the orientation occurring when this molecule is adsorbed under different surface coverages, considering the SERS selection rules.30 From the SERRS spectra, we have deduced that the above ratio undergoes an increase as the concentration is raised. This is attributed to a reorientation of 2,9-DMQA from a predominantly parallel orientation at low concentrations to a more perpendicular one at high coverage (Figure 4, bottom scheme). This reorientation is not observed in the case of QA, and this could be attributed to the presence of methyl groups in the structure of 2,9-DMQA. The differences observed in the adsorption mechanism between QA and 2,9-DMQA are also manifested on the fluorescence emission. As in the case of QA, the fluorescence bands of 2,9-DMQA (Figure 4B) in presence of metal nanoparticles are red-shift with respect to the fluorescence of 2,9-DMQA without silver (Figure 4Ba), suggesting that an intermolecular interaction between 2,9-DMQA molecules occurs. As can be seen, the fluorescence emission of 2,9DMQA when adsorbed on the metal is lower than QA and the intensity of the bands are not enhanced below a concentration of 10−4 M. This fact can be again attributed to a lower tendency 759
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
visible spectra reveals a strong aggregation tendency for both molecules and a bathochromic shift effect in the case of the 2,9DMQA in comparison to QA. This differential behavior is attributed to the presence of two methyl groups in 2,9-DMQA. The analysis of SERRS spectra at different concentrations demonstrated a different behavior for these quinacridones regarding their absorption mechanism on silver. While QA is adsorbed perpendicularly to the surface, regardless of the concentration, the adsorption of 2,9-DMQA depends on the surface coverage resulting from different concentrations. At concentrations lower than 3 × 10−6 M, the molecule adopts a predominant parallel orientation, whereas at higher concentrations, the molecule is adsorbed perpendicular to the metal surface. The SEF data indicate a large tendency toward intermolecular interactions of QA with respect to the 2,9-DMQA. Fluorescence bands are shifted toward higher wavelength as a consequence of the molecular aggregation occurring on the metal surface. In addition, the relative intensity of these bands changes due to the surface effect. The lower fluorescence emission observed in the emission spectra of 2,9-DMQA is attributed to the lower aggregation tendency of this molecule because of the presence of the methyl groups, which permitted a quantitative analysis of the adsorption on silver nanoparticles based on the adsorption isotherms. From this study, a BET adsorption model was deduced for 2,9-DMQA, indicating the key role of intermolecular interactions in the adsorption of this molecule on the surface. This is the first time that an adsorption by BET model is observed by SERS. Finally, a limit of detection of 54.4 ppb was calculated for 2,9-DMQA, thus demonstrating the higher sensitivity of the SERS technique in the analysis of quinacridone pigments.
Figure 5. Plot of SERRS intensity of 2,9-DMQA as a function of the dye concentration. Inset: BET function.
isotherm curve fits a BET model,32 and it is similar to other examples found in the literature,33−35 which suggests the importance of the intermolecular interactions in this molecule. The parameters related to the adsorption can be deduced from the BET adsorption model according to the following expression: A=
A m KC Cm + (K − 2)C − (k − 1/Cm)C 2
([1])
where A is the total intensity of each SERRS spectrum; Am is the total intensity of SERRS spectra if a monolayer is formed on the metal surface; C is the 2,9-DMQA concentration, Cm is the concentration of saturation of a monolayer; and, finally, K is the adsorption constant which is defined as K = eE1−EL/RT, with EL being the adsorption enthalpy of a monolayer, E1 the adsorption enthalpy of consecutive layers, R the gas constant, and T the temperature. As far as we know, this is the first time that a BET model adsorption behavior is deduced from SERS data. This is probably due to the fact that quinacridones have a large tendency toward aggregation due to the favored intermolecular interactions. The adsorption parameters Am, K, and Cm deduced from the analysis of the adsorption isotherm are summarized in Table 3.
■
Corresponding Author
*E-mail: [email protected] (S.S.-C.); edpuertonevado@ gmail.com (E.d.P.). Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work has been financially supported by MINECO (Projects FIS2010-15405, and CONSOLIDER CSD20070058/TCP) and the Comunidad de Madrid (MICROSERES II Project S2009/TIC1476). E.d.P. acknowledges CSIC and FSE 2007-2013 for a JAE_CSIC predoctoral grant.
Table 3. BET Parameters Deduced from the Analysis of the Adsorption Isotherm Am
K
1.2 × 106 ± 4 × 105
650
Cm 1.1 × 10−4 ± 1 × 10
−5
R
LOD (ng/ mL)
0.98
54.4
AUTHOR INFORMATION
■
REFERENCES
(1) Jaffe, E. E. High Performance Pigments, Second ed.; Faulkner, E. B., Schwartz, R. J., Eds.; WILEY-VCH Verlag GmbH & Co. KGaA: Weinheim, 2009; pp 298−329. (2) Meier, H.; Albrecht, W. Photoelectric Properties of 2,9-DimethylQuinacridone. Z. Phys.Chem. 1986, 148, 171−183. (3) Chen, T. L.; Chen, J. J. A.; Catane, L.; Ma, B. Fully Solution Processed P-I-N Organic Solar Cells with an Industrial Pigment Quinacridone. Org. Electron. 2011, 12, 1126−1131. (4) Jabbour, G. E.; Kawabe, Y.; Shaheen, S. E.; Wang, J. F.; Morrell, M. M.; Kippelen, B.; Peyghambarian, N. Highly Efficient and Bright Organic Electroluminescent Devices with an Aluminum Cathode. Appl. Phys. Lett. 1997, 71, 1763. (5) Blochwitz, J.; Pfeiffer, M.; Hofmann, H.; Leo, K. Non-Polymeric Oleds with a Doped Amorphous Hole Transport Layer and Operating
The limit of detection (LOD) was calculated by considering the concentration at which a SERRS intensity (A) is three times higher than the average noise of the 1800−297 cm−1 spectral region of the Raman spectrum of the blank. The calculated LOD was 54.4 ppb, which points out the high sensitivity of the SERS technique in the detection of this kind of compounds.
■
CONCLUSIONS Surface-Enhanced Raman Spectroscopy has proven to be a suitable technique for the detection of QA and 2,9-DMQA, despite their low solubility. The analysis of their corresponding 760
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
Langmuir
Article
Voltages Down to 3.2 V to Achieve 100 Cd/M2. Synth. Met. 2002, 127, 169−173. (6) Li, C.; Xiao, B.; Li, X.; Zhao, Y.; Hou, J.; Liu, S. Quinacridone Sub-Monolayer as Efficient Emitter in Oleds. J. Lumin. 2007, 122− 123, 623−625. (7) del Puerto, E.; Domingo, C.; Sanchez-Cortes, S.; García-Ramos, J. V.; Aroca, R. F. Plasmon Enhanced Spectroscopy of N,N0Dialkylquinacridones Used as Codopants in Oleds. J. Phys. Chem. C 2011, 115, 16838. (8) Senju, T.; Sakai, M.; Mizuguchi, J. Cohesion of G-Quinacridone and 2,9-Dimethylquinacridone in the Solid State. Dyes Pigments 2007, 75, 449−453. (9) Mizuguchi, J.; Senju, T. Solution and Solid-State Spectra of Quinacridone Derivatives as Viewed from the Intermolecular Hydrogen Bond. J. Phys. Chem.B 2006, 110, 19154−19161. (10) del Puerto, E.; Sanchez-Cortes, S.; Garcia-Ramos, J. V.; Domingo, C. Solution SERS of an Insoluble Synthetic Organic PigmentQuinacridone QuinoneEmploying Calixarenes as Dispersive Agent. Chem. Commun. 2011, 47, 1854. (11) del Puerto, E.; Cuesta, A.; Sanchez-Cortes, S.; J.V., G.-R.; Domingo, C. Electrochemical Sers Study on a Copper Electrode of the Insoluble Organic Pigment Quinacridone Quinone Using Ionic Liquids (BMIMCl and TBAN) as Dispersing Agents. Analyst 2013, 138, 4670−4676. (12) Lincke, G. On Quinacridones and Their Supramolecular Mesomerism within the Crystal Lattice. Dyes Pigments 2002, 52, 169−181. (13) Keller, U.; Miillen, K.; De Feyter, S.; De Schryver, F. C. Hydrogen-Bonding and Phase-Forming Behavior of a Soluble Quinacridone. Adv. Mater. 1996, 8, 6. (14) Binant, C.; Lautie, A. Identification of Quinacridone Pigment Mixtures Using Resonance Raman Microspectrometry. Appl. Spectrosc. 1989, 43 (5), 851−855. (15) Binant, C.; Guineau, B.; Lautie, A. The Application of Electronic and Vibrational Spectroscopy Techniques to the Identification of Quinacridone Pigments in Vehicle Paint Systems. J. Soc. Dyers Colour. 1998, 106 (5−6), 187−191. (16) Lincke, G. A Review of Thirty Years of Research on Quinacridones. X-Ray Crystallography and Crystal Engineering. Dyes Pigments 2000, 44, 101−122. (17) Ann Menke, C.; Rivenc, R.; Learner, T. The Use of Direct Temperature-Resolved Mass Spectrometry (Dtms) in the Detection of Organic Pigments Found in Acrylic Paints Used by Sam Francis. Int. J. Mass Spectrom. 2009, 284 (1−3), 2−11. (18) Kirbya, D. P.; Khandekara, N.; Sutherlandb, K.; Priceb, B. A. Applications of Laser Desorption Mass Spectrometry for the Study of Synthetic Organic Pigments in Works of Art. Int. J. Mass Spectrom. 2009, 284 (1−3), 115−122. (19) Jurasekova, Z.; del Puerto, E.; Bruno, G.; Garcia-Ramos, J. V.; Sanchez-Cortes, S.; Domingo, C. Extractionless Non-Hydrolysis Surface-Enhanced Raman Spectroscopic Detection of Historical Mordant Dyes on Textile Fibers. J. Raman Spectrosc. 2010, 41, 1455−1461. (20) Jurasekova, Z.; Garcia-Ramos, J. V.; Domingo, C.; SanchezCortes, S. Surface-Enhanced Raman Scattering of Flavonoids. J. Raman Spectrosc. 2006, 37 (11), 1239−1241. (21) Leona, M.; Decuzzi, P.; Kubic, T. A.; Gates, G.; Lombardi, J. R. Nondestructive Identification of Natural and Synthetic Organic Colorants in Works of Art by Surface Enhanced Raman Scattering. Anal. Chem. 2011, 83, 3990−3993. (22) Wustholz, K. L.; Brosseau, C. L.; Casadio, F.; Van Duyne, R. P. Surface-Enhanced Raman Spectroscopy of Dyes: From Single Molecules to the Artists’ Canvas. Phys. Chem. Chem. Phys. 2009, 11, 7350−7359. (23) Pozzi, F.; Lombardi, J. R.; Bruni, S.; Leona, M. Sample Treatment Considerations in the Analysis of Organic Colorants by Surface-Enhanced Raman Scattering. Anal. Chem. 2012, 84 (8), 3751− 3757.
(24) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. G. Gaussian 09, Gaussian, Inc.: Wallingford CT, 2009. (25) Leopold, N.; Lendl, B. A New Method for Fast Preparation of Highly Surface Enhanced Raman Scattering (SERS) Active Colloids at Room Temperature by Reduction of Silver Nitrate with Hydroxylamine Hydrochloride. J. Phys.Chem.B 2003, 107, 5723−5727. (26) Merrick, J. P.; Moran, D.; Radom, L. An Evaluation of Harmonic Vibrational Frequency Scale Factor. J.Phys. Chem. A 2007, 111, 11683−11700. (27) DeSaja-Gonzalez, J.; al., e. Surface-Enhanced Fluorescence and Serrs Spectra of N-Octadecyl-3,4:9,10-Perylenetetracarboxylic Monoanhydride on Silver Island Films. Spectrochim. Acta A 1997, 53, 173− 181. (28) Gentili, P. L.; Clementi, C.; Romani, A. Ultraviolet−Visible Absorption and Luminescence Properties of Quinacridone−Barium Sulfate Solid Mixtures. Appl. Spectrosc. 2010, 64, 923−929. (29) Moritz, A. G. The N−H Stretching Frequencies and Molecular Configurations of Some Aromatic Amines. Spectrochim. Acta 1960, 16 (10), 1176−1183. (30) Moskovits, M.; Suh, J. S. Surface Selection Rules for SurfaceEnhanced Raman Spectroscopy: Calculations and Application to the Surface-Enhanced Raman Spectrum of Phtalazine on Silver. J. Phys. Chem. 1984, 88, 5526−5530. (31) Izquierdo-Lorenzo, I.; Alda, I.; Sanchez-Cortes, S.; GarciaRamos, J. V. Adsorption and Detection of Sport Doping Drugs on Metallic Plasmonic Nanoparticles of Different Morphology. Langmuir 2012, 28, 8891−8901. (32) Brunauer, S.; Emmett, P. H.; Teller, E. Adsorption of Gases in Multimolecular Layers. J. Am. Chem. Soc. 1938, 60, 309. (33) Mikhalev, Y.; Øye, H. A. Absorption of Metallic Sodium in Carbon Cathode Materials. Carbon 1996, 34 (1), 37−41. (34) Koh, K.; Wong-Foy, A. G.; Matzger, A. J. A Porous Coordination Copolymer with over 5000 M2/G Bet Surface Area. J. Am. Chem. Soc. 2009, 131 (12), 4184−4185. (35) Dormant, L. M.; Adamson, A. W. Application of the Bet Equation to Heterogeneous Surfaces. J. Colloid Interface Sci. 1972, 38 (1), 285−289.
761
dx.doi.org/10.1021/la403625u | Langmuir 2014, 30, 753−761
