- Email: [email protected]
Effect of organic vapors on Au, Ag, and Au–Ag alloy nanoparticle films with adsorbed 2,6-dimethylphenyl isocyanide
Journal of Colloid and Interface Science 411 (2013) 194–197
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Effect of organic vapors on Au, Ag, and Au–Ag alloy nanoparticle films with adsorbed 2,6-dimethylphenyl isocyanide Kwan Kim a,⇑, Kyung Lock Kim a, Kuan Soo Shin b,⇑ a b
Department of Chemistry, Seoul National University, Seoul 151-742, Republic of Korea Department of Chemistry, Soongsil University, Seoul 156-743, Republic of Korea
a r t i c l e
i n f o
Article history: Received 10 July 2013 Accepted 15 August 2013 Available online 28 August 2013 Keywords: Gold–silver alloy Nanoparticles Surface-enhanced Raman scattering Volatile organic compounds Surface potential
a b s t r a c t The physicochemical properties of metallic substrates are affected by the environment in different ways. It is generally difficult to determine these effects because the molecules in the environment interact weakly with metallic substrates. In this work, we demonstrate that even the effect of volatile organic compounds (VOCs) can be identified by utilizing the surface-enhanced Raman scattering of isocyanide molecules. The NC stretching band of 2,6-dimethylphenyl isocyanide (2,6-DMPI) adsorbed on Au, for instance, is blueshifted by 6 cm 1 under an acetone flow and is redshifted by 20 cm 1 under an ammonia flow. The same band of 2,6-DMPI adsorbed on Ag and Au0.5Ag0.5 alloy films is, however, redshifted equally by 8 and 13 cm 1 under acetone and ammonia flows, respectively. This indicates that although the surface plasmons of Au0.5Ag0.5 alloy nanoparticles are clearly distinct from those of Ag (as well as Au) nanoparticles, both Au0.5Ag0.5 and Ag nanoparticles show a similar response to VOCs. These observations led us to conclude that the outermost parts of Au–Ag alloy nanoparticles are enriched with Ag atoms and that only the surfaces of metal nanoparticles, and not the bulk material, are affected by VOCs. Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction Over the past several decades, the properties of metal nanoparticles have been intensively investigated along with their potential use in microelectronics, chemical sensors, data storage, and a host of other applications [1,2]. This high research interest can be attributed to the fact that metal nanoparticles have unique optical, electronic, and magnetic properties that are not found in either isolated atoms or bulk solids [1,2]. Owing to their high surface-to-volume ratios, metal nanoparticles are also expected to function as effective catalysts [1]. Another noteworthy point is that the surface chemical properties of metal nanoparticles are different from their interior properties because of the inevitable presence of dangling bonds at surface sites [1–4]. In alloy nanoparticles, the surface chemical composition may be different from that of their interior owing to the intrinsically different surface tension characteristics of the constituent elements [5]. Such a compositional difference, though small, should be considered seriously, especially when fabricating nanoparticle-based microelectronic devices and sensors because it can affect their performance [6]. Owing to their high surface-to-volume ratio, metal nanoparticles are affected by environment more than bulk metals. Surface plasmons are usually affected little by the vapor of volatile organic ⇑ Corresponding authors. Fax: +82 2 8891568 (K. Kim), fax: +82 2 8144076 (K.S. Shin). E-mail addresses: [email protected] (K. Kim), [email protected] (K.S. Shin). 0021-9797/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2013.08.027
compounds (VOCs) [7]; however, VOC vapors are capable of significantly changing the surface charge and thus the surface potential of metal nanoparticles [8]. Recently, we demonstrated by means of surface-enhanced Raman scattering (SERS) of isocyanides that VOC vapors such as those of acetone and ammonia change the surface potential of Au and Ag nanoparticles [9–11]. On the basis of the blueshift and redshift of the NC stretching band of isocyanides, the surface potential of Au nanoparticles was estimated to change by as much as +0.16 and 0.56 V, respectively, upon contact with acetone and ammonia vapors [10]. It is not evident, however, whether organic vapors affect only the surfaces of metal nanoparticles or their interior parts too. If the vapors affect the interior parts, the NC stretching band of isocyanides on Au–Ag alloy nanoparticles would depend not only on the kind of organic vapors but also on the composition of alloy nanoparticles. In this work, we measured the SERS spectra of 2,6-dimethylphenyl isocyanide (2,6-DMPI) adsorbed on Au, Ag, and Au0.5Ag0.5 alloy nanoparticles in the absence and presence of acetone and ammonia vapors. Moreover, a series of potential-dependent SERS spectra were obtained using indium tin oxide (ITO) electrodes modified consecutively with Au, Ag, or Au0.5Ag0.5 nanoparticles and then 2,6-DMPI. The response of Au0.5Ag0.5 alloy nanoparticles to organic vapors was comparable to that of pure Ag nanoparticles but evidently differed from that of pure Au nanoparticles. This led us to conclude that the outermost parts of Au–Ag alloy nanoparticles were enriched with Ag atoms and that organic vapors affected only the surfaces of metal nanoparticles.
K. Kim et al. / Journal of Colloid and Interface Science 411 (2013) 194–197
2. Experimental 2.1. Materials Silver nitrate (AgNO3, 99.9999%), hydrogen chloroaurate (HAuCl4, 99.99%), sodium citrate (Na3C6H5O7, 99.0%), 2,6-DMPI (C9H9N, 96%), sodium perchlorate (NaClO4, 98%), 3-aminopropyltrimethoxysilane (3-APS, 99%), and Pt wire (99.99%) were purchased from Aldrich and used as received. Ammonia solution (NH3, 28.0– 30%) and acetone (C3H6O, 99.5%) were purchased from Samchun Chemical. Chemicals unless otherwise specified were of analytical reagent grade, and highly pure water with resistivity greater than 18.0 MX cm (Millipore Milli-Q System) was used throughout. 2.2. Preparation of Au, Ag, and Au–Ag alloy sols Pure Au and Au0.5Ag0.5 alloy sols were prepared according to the procedure of Link et al. [12]. An aqueous solution of HAuCl4 (95 mL) containing 5.2 mg of Au or a mixture of HAuCl4 and AgNO3 solution (95 mL) containing 2.6 and 1.4 mg of Au and Ag, respectively, was brought to boil; 5 mL of 1% sodium citrate was then added under stirring, and boiling was continued for 30 min. Pure Ag sol was prepared according to the procedure of Lee and Meisel [13]. Initially, 45 mg of AgNO3 was dissolved in 250 mL of water, and the solution was brought to the boil; 1% sodium citrate (5 mL) was then added under stirring, and boiling was continued for 30 min. Transmission electron microscopy (TEM) analyses revealed that all the nanoparticles were spherical, and their average diameters were comparable to one another at 35 nm. 2.3. Preparation of 2,6-DMPI-adsorbed Au, Ag, and Au–Ag films
195
0.2 mW with an average spot diameter of 1 lm. The integration time was 30 s. For potential-dependent SERS measurements, a pure Pt wire was used as a counter electrode, and the potential of the electrochemical cell was controlled using a CH Instruments model 660A potentiostat. All potentials are reported with respect to the saturated Ag/AgCl electrode. 3. Results and discussion Fig. 1(a) shows the UV–vis absorption spectra of pure Au, pure Ag, and Au0.5Ag0.5 alloy sols. Plasmon bands are observed at 530, 460, and 410 nm for the Au, Au0.5Ag0.5, and Ag sols, respectively. The inset of Fig. 1(a) shows a plot for the plasmon maximum versus the Au mole fraction, and a linear relationship between them can be observed [12,15]. For reference, the UV–vis spectrum of a 1:1 mechanical mixture of the Au and Ag sols is shown in Fig. 1(b). Two distinct surface plasmon bands are clearly evident in this spectrum. This indicates that Au–Ag alloy nanoparticles are formed when Au and Ag ions are reduced simultaneously by sodium citrate. Fig. 1(c) shows the UV–vis spectra of Au, Au0.5Ag0.5, and Ag films formed on 3-APS-modified mica substrates. Upon the formation of films, each surface plasmon maximum redshifted by 30 nm. The absorbance of the films at 632.8-nm (the laser wavelength used for Raman spectral measurements) was in the following order: Ag Au > Au0.5Ag0.5 films. Fig. 2(b) and (c) show the SERS spectra of 2,6-DMPI adsorbed on pure Au and pure Ag films, respectively. For reference, the normal Raman (NR) spectrum of 2,6-DMPI in its neat solid state is shown in Fig. 2(a). The NR spectrum shows two high-intensity peaks at 2123 and 640 cm 1 due to the NC and C–NC stretching plus ring breathing vibrations, respectively. Ring-associated bands have comparatively less intensity than the NC-associated bands, as
Au, Ag, and Au0.5Ag0.5 colloidal particles were centrifuged and washed thoroughly with water. The cleaned particles were redispersed in water by sonication (to a particle concentration of 1.8 10 11 M). Au, Ag, and Au0.5Ag0.5 films were prepared by dropping 15 lL of the corresponding sol onto mica (1 mm 30 mm) or ITO (10 mm 30 mm) substrates that had been modified with 3APS according to a literature procedure [14]. After washing with water, the substrates were immersed in a 1 mM ethanolic solution of 2,6-DMPI for 30 min. VOC samples were prepared by first flushing a volumetric flask (1 L) with N2 and then capping it with a rubber septum at atmospheric pressure. Using a syringe, a measured quantity of organics, i.e., acetone or ammonia, was introduced into the volumetric flask through the rubber septum and allowed to vaporize completely. About 50 mL of the resulting mixed gas was extracted using a syringe and was subsequently set to flow through a glass capillary tube (inner diameter: 1.2 mm; length: 75 mm) containing 2,6-DMPI-adsorbed Au, Ag, and Au0.5Ag0.5 films. 2.4. Instrumentation UV–visible (UV–vis) spectra of the samples were obtained using a SCINCO S-4100 spectrometer. TEM images were obtained using a JEM-200CX transmission electron microscope. The flow of organic vapors through the glass capillary tube was controlled using a Sage Instruments model 341 syringe pump. Raman spectra were obtained using a Renishaw Raman System model 2000 spectrometer. The 632.8-nm line from a 17-mW He/Ne laser (Spectra Physics model 127) was used as the excitation source. Raman scattering was detected at 180° geometry using a Peltier-cooled ( 70 °C) CCD camera (400 600 pixels). The Raman band of a silicon wafer at 520 cm 1 was used to calibrate the spectrometer, and the accuracy of the spectral measurements was estimated to be better than 1 cm 1. The typical laser power at the sampling position was
Fig. 1. (a) UV–vis absorption spectra of Au, Ag, and Au0.5Ag0.5 alloy sols. The inset shows the positions of surface plasmon band maxima plotted against the mole fractions of Au. (b) UV–vis spectra of a 1:1 mixture of Au and Ag sols. (c) UV–vis spectra of Au, Ag, and Au0.5Ag0.5 alloy films assembled on mica substrates.
196
K. Kim et al. / Journal of Colloid and Interface Science 411 (2013) 194–197
Fig. 2. (a) NR spectrum of 2,6-DMPI in the neat state, and its SERS spectra on (b) pure Au, (c) pure Ag, and (d) Au0.5Ag0.5 alloy nanoparticle films assembled on mica substrates. (e) Comparison of SERS intensity at 1590 cm 1 and UV–vis absorbance at 632.8-nm for Ag, Au0.5Ag0.5 alloy, and Au nanoparticle films: SERS intensities are normalized to that of a silicon wafer at 520 cm 1, and the error bars represent the average and standard deviation of 10 measurements.
apparent from the peaks at 3046, 1593, and 995 cm 1 in Fig. 2(a) that can be assigned to the CH stretching, ring CC stretching (m8a), and in-plane ring breathing (m18a) plus symmetric CH3 rocking modes of 2,6-DMPI, respectively [15,16]. The spectral peaks in Fig. 2(b) and (c) can be correlated with those in Fig. 2(a), but substantial spectral differences exist not only between the NR and SERS spectra of 2,6-DMPI but also between the two SERS spectra of Ag and Au films. The most noteworthy feature is that the NC stretching peak shifted from 2123 cm 1 (Fig. 2(a)) to 2182 cm 1 (Fig. 2(b)) and then to 2173 cm 1 (Fig. 2(c)) under a N2 gas flow. The NC stretching peak blueshifted by as much as 59 cm 1 and 50 cm 1 upon the adsorption of 2,6-DMPI on Au and Ag, respectively. The substantial blueshift of the NC stretching mode can be attributed to the antibonding character of the carbon lone-pair electrons in the isocyanide group. The donation of these electrons to Au and Ag must have increased the strength of the NC bond, resulting in the blueshift of the NC stretching band [15,16]. Fig. 2(d) shows the SERS spectrum of 2,6-DMPI adsorbed on the Au0.5Ag0.5 alloy film. The SERS spectral pattern in Fig. 2(d) closely resembles that in Fig. 2(c). Not only the peak positions but also the bandwidths and relative intensities are same within the limits of experimental uncertainty. These resemblances can be attributed to the enrichment of Ag atoms in the outermost layer of the alloy nanoparticles [15]. The absolute peak intensity in Fig. 2(d) is nonetheless much smaller than that in Fig. 2(c). This can be understood by referring to the electromagnetic enhancement mechanism [17]. As shown in Fig. 2(e), the absorbance of the Ag film is 1.37 times that of the Au0.5Ag0.5 alloy film at 632.8-nm. The 2,6-DMPI molecules adsorbed on the Ag film would then be exposed to stronger electromagnetic radiation than those on the Au0.5Ag0.5 alloy film. The SERS spectral patterns in Fig. 2(c) and (d) are similar since they are determined solely by the type of metals comprising the outermost surface. The difference between the spectral patterns as well as the peak intensities in Fig. 2(b) and those in Fig. 2(c) and (d) can be similarly understood. Subsequently, we measured Raman spectra under a flow of acetone or ammonia over 2,6-DMPI adsorbed on pure Au, pure Ag, or Au0.5Ag0.5 alloy film. The peaks due to acetone or ammonia were not identifiable [9–11]. A negligible change was observed for the ring modes of 2,6-DMPI, but the NC stretching band was subject to change. Fig. 3(a) and (b) show the NC stretching regions observed under the flow of acetone and ammonia, respectively. When acetone was flowed over the pure Au film, the NC stretching band
Fig. 3. The NC stretching region of 2,6-DMPI measured under (a) acetone and (b) ammonia flows over 2,6-DMPI-adsorbed Au, Au0.5Ag0.5, or Ag nanoparticle films assembled on a mica substrate placed in a glass capillary tube.
blueshifted by 6 cm 1, i.e., from 2182 to 2188 cm 1. This shift was associated with an s-type charge transfer from Au to the unoccupied p-orbital of the oxygen atom of acetone [18]. Because of the transfer, the surface potential of Au nanoparticles moved in the positive direction, resulting in the blueshift of the NC stretching band of 2,6-DMPI. Interestingly, the same band on the pure Ag or Au0.5Ag0.5 alloy film redshifted equally by 8 cm 1, i.e., from 2173 to 2165 cm 1. This observation may suggest that the donation of the lone-pair electrons by the carbonyl oxygen electrons to Ag is far more important than any type of back-donation from Ag to acetone. It is well known that the valence dp orbital in Ag is extremely stable as compared to the same orbital in Au, and hence, the p back-donation bonding in Ag is not efficient [19,20]. On the other hand, when ammonia was flown over the pure Ag or Au0.5Ag0.5 alloy film, the NC stretching band redshifted equally
197
K. Kim et al. / Journal of Colloid and Interface Science 411 (2013) 194–197
tial change of 0.22 V (from +0.04 to (from +0.04 to 0.32 V), respectively.
0.18 V) and
0.36 V
4. Summary and conclusion
Fig. 4. Positions of the NC stretching peak of 2,6-DMPI on ITO electrodes with adsorbed Au, Au0.5Ag0.5, or Ag nanoparticles in 0.1 M NaClO4 aqueous solution, drawn versus the applied potential. The error bars indicate a standard deviation of three independent measurements.
by 13 cm 1, i.e., from 2173 to 2160 cm 1. The same band on the pure Au film also redshifted by 20 cm 1, i.e., from 2182 to 2162 cm 1. As the Au nanoparticles became more negatively charged owing to the donation of the lone-pair electrons of ammonia nitrogen to Au, the r donation from the C atom of 2,6-DMPI to Au decreased. Because these carbon lone-pair electrons have an antibonding character, the decreased donation of these electrons to Au would decrease the strength of the NC bond, and hence, the NC stretching peak moved to a lower frequency. In addition, the electron donation capability of ammonia to Ag appeared to be lower than that to Au [10]. Fig. 3(a) shows that the effect of acetone on the Au0.5Ag0.5 alloy film is the same as that on the Ag film, but is clearly distinct from that on the Au film. Fig. 3(b) shows that the effect of ammonia on the Au0.5Ag0.5 alloy film is also the same as that on the Ag film. The effect of ammonia on the Au film clearly differs from that on the Au0.5Ag0.5 alloy film. The present observation is clearly indicative of the exclusive enrichment of Ag atoms in the outermost surface of Au0.5Ag0.5 alloy nanoparticles and suggests further that the organic vapors affect only the surfaces of metal nanoparticles. In this work, organic vapors changed the surface charge and thus the surface potential of metal nanoparticles. The interior part of metal nanoparticles was not affected at all by organic vapors. Otherwise, the peak shift of the NC stretching band due to the Au0.5Ag0.5 alloy film should have been somewhere between those observed for the Ag and Au films. The potential-dependent SERS spectra of 2,6-DMPI on an Au0.5Ag0.5 alloy electrode were also comparable to those assembled on a pure Ag electrode. As shown in Fig. 4, the NC stretching frequency varied linearly against the applied potential, and the curve derived for the Au0.5Ag0.5 alloy electrode overlapped the curve for the pure Ag electrode. However, the slope of the variation in the NC stretching frequency versus the applied potential was nonetheless the same for a pure Au electrode, i.e., 36 cm 1 V 1. Thus, the blueshift and redshift due to acetone and ammonia for pure Au nanoparticles correspond to the potential change of +0.16 V (from +0.16 to +0.32 V) and 0.56 V (from +0.16 to 0.40 V), respectively [10]. On the other hand, the redshifts due to acetone and ammonia for the pure Ag and Au0.5Ag0.5 alloy nanoparticles correspond equally to the poten-
We measured a series of SERS spectra of 2,6-DMPI adsorbed on Au, Ag, and Au0.5Ag0.5 alloy nanoparticle films under acetone and ammonia flows to determine whether VOCs affect only the surfaces of metal nanoparticles or their interior parts too. First, Au0.5Ag0.5 alloy nanoparticles were formed by simultaneously reducing Au and Ag ions with sodium citrate. The surface plasmons of Au0.5Ag0.5 alloy nanoparticles were clearly distinct from those of Ag and Au nanoparticles. Second, the SERS spectral pattern of 2,6-DMPI on the Au0.5Ag0.5 alloy was comparable to that on pure Ag but differed clearly from that on pure Au, suggesting that the outermost part of the Au0.5Ag0.5 nanoparticles was composed of Ag atoms. Third, the NC stretching band of 2,6-DMPI on Au was blueshifted under an acetone flow but was redshifted under an ammonia flow. The same band of 2,6-DMPI on Au0.5Ag0.5 alloy was, however, redshifted under an acetone flow. The amount of redshift was equal to that observed in the case of 2,6-DMPI on pure Ag. Under the ammonia flow, the bands for the pure Ag and Au0.5Ag0.5 alloy films were equally redshifted. This clearly indicates that VOCs affect only the surfaces of metal nanoparticles. By referring to the potentialdependent SERS data, the surface potential of pure Au nanoparticles appeared to increase by +0.16 V under the acetone flow, while it decreased by 0.56 V under the ammonia flow. The surface potential of pure Ag and Au0.5Ag0.5 alloy nanoparticles was in turn decreased equally by 0.22 V under the acetone flow and by 0.36 V under the ammonia flow. Acknowledgments This work was supported by the National Research Foundation of Korea (NRF) grants funded by the Korea government (MSIP) (Nos. 2007-0056095, 2011-0006737, and 2012R1A2A2A01008004). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]
[15] [16] [17] [18] [19] [20]
A. Henglein, Chem. Rev. 89 (1989) 1861. A.P. Alivisatos, J. Phys. Chem. 100 (1996) 13226. L.M. Liz-Marzán, Langmuir 22 (2006) 32. S.K. Ghosh, T. Pal, Chem. Rev. 107 (2007) 4797. F. Tao, M.E. Grass, Y. Zhang, D.R. Butcher, J.R. Renzas, Z. Liu, J.Y. Chung, B.S. Mun, M. Salmeron, G.A. Somorjai, Science 322 (2008) 932. R.V. Benjaminsen, H. Sun, J.R. Henriksen, N.M. Christensen, K. Almdal, T.L. Andresen, ACS Nano 5 (2011) 5864. E.B. Feresenbet, E. Dalcanale, C. Dulcey, D.K. Shenoy, Actuators, B 97 (2004) 211. A. Ulman, Chem. Rev. 96 (1996) 1533. D. Shin, K. Kim, K.S. Shin, Chem. Phys. Chem. 11 (2010) 83. K. Kim, K.L. Kim, D. Shin, J.W. Lee, K.S. Shin, Chem. Commun. 46 (2010) 3753. K. Kim, J.W. Lee, K.S. Shin, Spec. Chim. Acta Part A 100 (2013) 15. S. Link, Z.L. Wang, M.A. El-Sayed, J. Phys. Chem. B 103 (1999) 3529. P.C. Lee, D. Meisel, J. Phys. Chem. 86 (1982) 3391. R.G. Freeman, K.C. Grabar, K.J. Allison, R.M. Bright, J.A. Davis, A.P. Guthrie, M.B. Hommer, M.A. Jackson, P.C. Smith, D.G. Walter, M.J. Natan, Science 267 (1995) 1629. K. Kim, K.L. Kim, J.-Y. Choi, H.B. Lee, K.S. Shin, J. Phys. Chem. C 114 (2010) 3448. Vibrational frequencies of 2,6-DMPI were calculated at the B3LYP/6-31+G(2d, p) level using Gaussian 09 suite program. M. Moskovits, Phys. Chem. Chem. Phys. 15 (2013) 5301. G.S. Shafai, S. Shetty, S. Krishnamurty, V. Shah, D.G. Kanhere, J. Chem. Phys. 126 (2007) 014704. S.A. Wasileski, M.T.M. Koper, M.J. Weaver, J. Am. Chem. Soc. 124 (2002) 2796. B. Liang, L. Andrews, J. Phys. Chem. A 104 (2000) 9156.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Effect of organic vapors on Au, Ag, and Au–Ag alloy nanoparticle films with adsorbed 2,6-dimethylphenyl isocyanide Kwan Kim a,⇑, Kyung Lock Kim a, Kuan Soo Shin b,⇑ a b
Department of Chemistry, Seoul National University, Seoul 151-742, Republic of Korea Department of Chemistry, Soongsil University, Seoul 156-743, Republic of Korea
a r t i c l e
i n f o
Article history: Received 10 July 2013 Accepted 15 August 2013 Available online 28 August 2013 Keywords: Gold–silver alloy Nanoparticles Surface-enhanced Raman scattering Volatile organic compounds Surface potential
a b s t r a c t The physicochemical properties of metallic substrates are affected by the environment in different ways. It is generally difficult to determine these effects because the molecules in the environment interact weakly with metallic substrates. In this work, we demonstrate that even the effect of volatile organic compounds (VOCs) can be identified by utilizing the surface-enhanced Raman scattering of isocyanide molecules. The NC stretching band of 2,6-dimethylphenyl isocyanide (2,6-DMPI) adsorbed on Au, for instance, is blueshifted by 6 cm 1 under an acetone flow and is redshifted by 20 cm 1 under an ammonia flow. The same band of 2,6-DMPI adsorbed on Ag and Au0.5Ag0.5 alloy films is, however, redshifted equally by 8 and 13 cm 1 under acetone and ammonia flows, respectively. This indicates that although the surface plasmons of Au0.5Ag0.5 alloy nanoparticles are clearly distinct from those of Ag (as well as Au) nanoparticles, both Au0.5Ag0.5 and Ag nanoparticles show a similar response to VOCs. These observations led us to conclude that the outermost parts of Au–Ag alloy nanoparticles are enriched with Ag atoms and that only the surfaces of metal nanoparticles, and not the bulk material, are affected by VOCs. Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction Over the past several decades, the properties of metal nanoparticles have been intensively investigated along with their potential use in microelectronics, chemical sensors, data storage, and a host of other applications [1,2]. This high research interest can be attributed to the fact that metal nanoparticles have unique optical, electronic, and magnetic properties that are not found in either isolated atoms or bulk solids [1,2]. Owing to their high surface-to-volume ratios, metal nanoparticles are also expected to function as effective catalysts [1]. Another noteworthy point is that the surface chemical properties of metal nanoparticles are different from their interior properties because of the inevitable presence of dangling bonds at surface sites [1–4]. In alloy nanoparticles, the surface chemical composition may be different from that of their interior owing to the intrinsically different surface tension characteristics of the constituent elements [5]. Such a compositional difference, though small, should be considered seriously, especially when fabricating nanoparticle-based microelectronic devices and sensors because it can affect their performance [6]. Owing to their high surface-to-volume ratio, metal nanoparticles are affected by environment more than bulk metals. Surface plasmons are usually affected little by the vapor of volatile organic ⇑ Corresponding authors. Fax: +82 2 8891568 (K. Kim), fax: +82 2 8144076 (K.S. Shin). E-mail addresses: [email protected] (K. Kim), [email protected] (K.S. Shin). 0021-9797/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2013.08.027
compounds (VOCs) [7]; however, VOC vapors are capable of significantly changing the surface charge and thus the surface potential of metal nanoparticles [8]. Recently, we demonstrated by means of surface-enhanced Raman scattering (SERS) of isocyanides that VOC vapors such as those of acetone and ammonia change the surface potential of Au and Ag nanoparticles [9–11]. On the basis of the blueshift and redshift of the NC stretching band of isocyanides, the surface potential of Au nanoparticles was estimated to change by as much as +0.16 and 0.56 V, respectively, upon contact with acetone and ammonia vapors [10]. It is not evident, however, whether organic vapors affect only the surfaces of metal nanoparticles or their interior parts too. If the vapors affect the interior parts, the NC stretching band of isocyanides on Au–Ag alloy nanoparticles would depend not only on the kind of organic vapors but also on the composition of alloy nanoparticles. In this work, we measured the SERS spectra of 2,6-dimethylphenyl isocyanide (2,6-DMPI) adsorbed on Au, Ag, and Au0.5Ag0.5 alloy nanoparticles in the absence and presence of acetone and ammonia vapors. Moreover, a series of potential-dependent SERS spectra were obtained using indium tin oxide (ITO) electrodes modified consecutively with Au, Ag, or Au0.5Ag0.5 nanoparticles and then 2,6-DMPI. The response of Au0.5Ag0.5 alloy nanoparticles to organic vapors was comparable to that of pure Ag nanoparticles but evidently differed from that of pure Au nanoparticles. This led us to conclude that the outermost parts of Au–Ag alloy nanoparticles were enriched with Ag atoms and that organic vapors affected only the surfaces of metal nanoparticles.
K. Kim et al. / Journal of Colloid and Interface Science 411 (2013) 194–197
2. Experimental 2.1. Materials Silver nitrate (AgNO3, 99.9999%), hydrogen chloroaurate (HAuCl4, 99.99%), sodium citrate (Na3C6H5O7, 99.0%), 2,6-DMPI (C9H9N, 96%), sodium perchlorate (NaClO4, 98%), 3-aminopropyltrimethoxysilane (3-APS, 99%), and Pt wire (99.99%) were purchased from Aldrich and used as received. Ammonia solution (NH3, 28.0– 30%) and acetone (C3H6O, 99.5%) were purchased from Samchun Chemical. Chemicals unless otherwise specified were of analytical reagent grade, and highly pure water with resistivity greater than 18.0 MX cm (Millipore Milli-Q System) was used throughout. 2.2. Preparation of Au, Ag, and Au–Ag alloy sols Pure Au and Au0.5Ag0.5 alloy sols were prepared according to the procedure of Link et al. [12]. An aqueous solution of HAuCl4 (95 mL) containing 5.2 mg of Au or a mixture of HAuCl4 and AgNO3 solution (95 mL) containing 2.6 and 1.4 mg of Au and Ag, respectively, was brought to boil; 5 mL of 1% sodium citrate was then added under stirring, and boiling was continued for 30 min. Pure Ag sol was prepared according to the procedure of Lee and Meisel [13]. Initially, 45 mg of AgNO3 was dissolved in 250 mL of water, and the solution was brought to the boil; 1% sodium citrate (5 mL) was then added under stirring, and boiling was continued for 30 min. Transmission electron microscopy (TEM) analyses revealed that all the nanoparticles were spherical, and their average diameters were comparable to one another at 35 nm. 2.3. Preparation of 2,6-DMPI-adsorbed Au, Ag, and Au–Ag films
195
0.2 mW with an average spot diameter of 1 lm. The integration time was 30 s. For potential-dependent SERS measurements, a pure Pt wire was used as a counter electrode, and the potential of the electrochemical cell was controlled using a CH Instruments model 660A potentiostat. All potentials are reported with respect to the saturated Ag/AgCl electrode. 3. Results and discussion Fig. 1(a) shows the UV–vis absorption spectra of pure Au, pure Ag, and Au0.5Ag0.5 alloy sols. Plasmon bands are observed at 530, 460, and 410 nm for the Au, Au0.5Ag0.5, and Ag sols, respectively. The inset of Fig. 1(a) shows a plot for the plasmon maximum versus the Au mole fraction, and a linear relationship between them can be observed [12,15]. For reference, the UV–vis spectrum of a 1:1 mechanical mixture of the Au and Ag sols is shown in Fig. 1(b). Two distinct surface plasmon bands are clearly evident in this spectrum. This indicates that Au–Ag alloy nanoparticles are formed when Au and Ag ions are reduced simultaneously by sodium citrate. Fig. 1(c) shows the UV–vis spectra of Au, Au0.5Ag0.5, and Ag films formed on 3-APS-modified mica substrates. Upon the formation of films, each surface plasmon maximum redshifted by 30 nm. The absorbance of the films at 632.8-nm (the laser wavelength used for Raman spectral measurements) was in the following order: Ag Au > Au0.5Ag0.5 films. Fig. 2(b) and (c) show the SERS spectra of 2,6-DMPI adsorbed on pure Au and pure Ag films, respectively. For reference, the normal Raman (NR) spectrum of 2,6-DMPI in its neat solid state is shown in Fig. 2(a). The NR spectrum shows two high-intensity peaks at 2123 and 640 cm 1 due to the NC and C–NC stretching plus ring breathing vibrations, respectively. Ring-associated bands have comparatively less intensity than the NC-associated bands, as
Au, Ag, and Au0.5Ag0.5 colloidal particles were centrifuged and washed thoroughly with water. The cleaned particles were redispersed in water by sonication (to a particle concentration of 1.8 10 11 M). Au, Ag, and Au0.5Ag0.5 films were prepared by dropping 15 lL of the corresponding sol onto mica (1 mm 30 mm) or ITO (10 mm 30 mm) substrates that had been modified with 3APS according to a literature procedure [14]. After washing with water, the substrates were immersed in a 1 mM ethanolic solution of 2,6-DMPI for 30 min. VOC samples were prepared by first flushing a volumetric flask (1 L) with N2 and then capping it with a rubber septum at atmospheric pressure. Using a syringe, a measured quantity of organics, i.e., acetone or ammonia, was introduced into the volumetric flask through the rubber septum and allowed to vaporize completely. About 50 mL of the resulting mixed gas was extracted using a syringe and was subsequently set to flow through a glass capillary tube (inner diameter: 1.2 mm; length: 75 mm) containing 2,6-DMPI-adsorbed Au, Ag, and Au0.5Ag0.5 films. 2.4. Instrumentation UV–visible (UV–vis) spectra of the samples were obtained using a SCINCO S-4100 spectrometer. TEM images were obtained using a JEM-200CX transmission electron microscope. The flow of organic vapors through the glass capillary tube was controlled using a Sage Instruments model 341 syringe pump. Raman spectra were obtained using a Renishaw Raman System model 2000 spectrometer. The 632.8-nm line from a 17-mW He/Ne laser (Spectra Physics model 127) was used as the excitation source. Raman scattering was detected at 180° geometry using a Peltier-cooled ( 70 °C) CCD camera (400 600 pixels). The Raman band of a silicon wafer at 520 cm 1 was used to calibrate the spectrometer, and the accuracy of the spectral measurements was estimated to be better than 1 cm 1. The typical laser power at the sampling position was
Fig. 1. (a) UV–vis absorption spectra of Au, Ag, and Au0.5Ag0.5 alloy sols. The inset shows the positions of surface plasmon band maxima plotted against the mole fractions of Au. (b) UV–vis spectra of a 1:1 mixture of Au and Ag sols. (c) UV–vis spectra of Au, Ag, and Au0.5Ag0.5 alloy films assembled on mica substrates.
196
K. Kim et al. / Journal of Colloid and Interface Science 411 (2013) 194–197
Fig. 2. (a) NR spectrum of 2,6-DMPI in the neat state, and its SERS spectra on (b) pure Au, (c) pure Ag, and (d) Au0.5Ag0.5 alloy nanoparticle films assembled on mica substrates. (e) Comparison of SERS intensity at 1590 cm 1 and UV–vis absorbance at 632.8-nm for Ag, Au0.5Ag0.5 alloy, and Au nanoparticle films: SERS intensities are normalized to that of a silicon wafer at 520 cm 1, and the error bars represent the average and standard deviation of 10 measurements.
apparent from the peaks at 3046, 1593, and 995 cm 1 in Fig. 2(a) that can be assigned to the CH stretching, ring CC stretching (m8a), and in-plane ring breathing (m18a) plus symmetric CH3 rocking modes of 2,6-DMPI, respectively [15,16]. The spectral peaks in Fig. 2(b) and (c) can be correlated with those in Fig. 2(a), but substantial spectral differences exist not only between the NR and SERS spectra of 2,6-DMPI but also between the two SERS spectra of Ag and Au films. The most noteworthy feature is that the NC stretching peak shifted from 2123 cm 1 (Fig. 2(a)) to 2182 cm 1 (Fig. 2(b)) and then to 2173 cm 1 (Fig. 2(c)) under a N2 gas flow. The NC stretching peak blueshifted by as much as 59 cm 1 and 50 cm 1 upon the adsorption of 2,6-DMPI on Au and Ag, respectively. The substantial blueshift of the NC stretching mode can be attributed to the antibonding character of the carbon lone-pair electrons in the isocyanide group. The donation of these electrons to Au and Ag must have increased the strength of the NC bond, resulting in the blueshift of the NC stretching band [15,16]. Fig. 2(d) shows the SERS spectrum of 2,6-DMPI adsorbed on the Au0.5Ag0.5 alloy film. The SERS spectral pattern in Fig. 2(d) closely resembles that in Fig. 2(c). Not only the peak positions but also the bandwidths and relative intensities are same within the limits of experimental uncertainty. These resemblances can be attributed to the enrichment of Ag atoms in the outermost layer of the alloy nanoparticles [15]. The absolute peak intensity in Fig. 2(d) is nonetheless much smaller than that in Fig. 2(c). This can be understood by referring to the electromagnetic enhancement mechanism [17]. As shown in Fig. 2(e), the absorbance of the Ag film is 1.37 times that of the Au0.5Ag0.5 alloy film at 632.8-nm. The 2,6-DMPI molecules adsorbed on the Ag film would then be exposed to stronger electromagnetic radiation than those on the Au0.5Ag0.5 alloy film. The SERS spectral patterns in Fig. 2(c) and (d) are similar since they are determined solely by the type of metals comprising the outermost surface. The difference between the spectral patterns as well as the peak intensities in Fig. 2(b) and those in Fig. 2(c) and (d) can be similarly understood. Subsequently, we measured Raman spectra under a flow of acetone or ammonia over 2,6-DMPI adsorbed on pure Au, pure Ag, or Au0.5Ag0.5 alloy film. The peaks due to acetone or ammonia were not identifiable [9–11]. A negligible change was observed for the ring modes of 2,6-DMPI, but the NC stretching band was subject to change. Fig. 3(a) and (b) show the NC stretching regions observed under the flow of acetone and ammonia, respectively. When acetone was flowed over the pure Au film, the NC stretching band
Fig. 3. The NC stretching region of 2,6-DMPI measured under (a) acetone and (b) ammonia flows over 2,6-DMPI-adsorbed Au, Au0.5Ag0.5, or Ag nanoparticle films assembled on a mica substrate placed in a glass capillary tube.
blueshifted by 6 cm 1, i.e., from 2182 to 2188 cm 1. This shift was associated with an s-type charge transfer from Au to the unoccupied p-orbital of the oxygen atom of acetone [18]. Because of the transfer, the surface potential of Au nanoparticles moved in the positive direction, resulting in the blueshift of the NC stretching band of 2,6-DMPI. Interestingly, the same band on the pure Ag or Au0.5Ag0.5 alloy film redshifted equally by 8 cm 1, i.e., from 2173 to 2165 cm 1. This observation may suggest that the donation of the lone-pair electrons by the carbonyl oxygen electrons to Ag is far more important than any type of back-donation from Ag to acetone. It is well known that the valence dp orbital in Ag is extremely stable as compared to the same orbital in Au, and hence, the p back-donation bonding in Ag is not efficient [19,20]. On the other hand, when ammonia was flown over the pure Ag or Au0.5Ag0.5 alloy film, the NC stretching band redshifted equally
197
K. Kim et al. / Journal of Colloid and Interface Science 411 (2013) 194–197
tial change of 0.22 V (from +0.04 to (from +0.04 to 0.32 V), respectively.
0.18 V) and
0.36 V
4. Summary and conclusion
Fig. 4. Positions of the NC stretching peak of 2,6-DMPI on ITO electrodes with adsorbed Au, Au0.5Ag0.5, or Ag nanoparticles in 0.1 M NaClO4 aqueous solution, drawn versus the applied potential. The error bars indicate a standard deviation of three independent measurements.
by 13 cm 1, i.e., from 2173 to 2160 cm 1. The same band on the pure Au film also redshifted by 20 cm 1, i.e., from 2182 to 2162 cm 1. As the Au nanoparticles became more negatively charged owing to the donation of the lone-pair electrons of ammonia nitrogen to Au, the r donation from the C atom of 2,6-DMPI to Au decreased. Because these carbon lone-pair electrons have an antibonding character, the decreased donation of these electrons to Au would decrease the strength of the NC bond, and hence, the NC stretching peak moved to a lower frequency. In addition, the electron donation capability of ammonia to Ag appeared to be lower than that to Au [10]. Fig. 3(a) shows that the effect of acetone on the Au0.5Ag0.5 alloy film is the same as that on the Ag film, but is clearly distinct from that on the Au film. Fig. 3(b) shows that the effect of ammonia on the Au0.5Ag0.5 alloy film is also the same as that on the Ag film. The effect of ammonia on the Au film clearly differs from that on the Au0.5Ag0.5 alloy film. The present observation is clearly indicative of the exclusive enrichment of Ag atoms in the outermost surface of Au0.5Ag0.5 alloy nanoparticles and suggests further that the organic vapors affect only the surfaces of metal nanoparticles. In this work, organic vapors changed the surface charge and thus the surface potential of metal nanoparticles. The interior part of metal nanoparticles was not affected at all by organic vapors. Otherwise, the peak shift of the NC stretching band due to the Au0.5Ag0.5 alloy film should have been somewhere between those observed for the Ag and Au films. The potential-dependent SERS spectra of 2,6-DMPI on an Au0.5Ag0.5 alloy electrode were also comparable to those assembled on a pure Ag electrode. As shown in Fig. 4, the NC stretching frequency varied linearly against the applied potential, and the curve derived for the Au0.5Ag0.5 alloy electrode overlapped the curve for the pure Ag electrode. However, the slope of the variation in the NC stretching frequency versus the applied potential was nonetheless the same for a pure Au electrode, i.e., 36 cm 1 V 1. Thus, the blueshift and redshift due to acetone and ammonia for pure Au nanoparticles correspond to the potential change of +0.16 V (from +0.16 to +0.32 V) and 0.56 V (from +0.16 to 0.40 V), respectively [10]. On the other hand, the redshifts due to acetone and ammonia for the pure Ag and Au0.5Ag0.5 alloy nanoparticles correspond equally to the poten-
We measured a series of SERS spectra of 2,6-DMPI adsorbed on Au, Ag, and Au0.5Ag0.5 alloy nanoparticle films under acetone and ammonia flows to determine whether VOCs affect only the surfaces of metal nanoparticles or their interior parts too. First, Au0.5Ag0.5 alloy nanoparticles were formed by simultaneously reducing Au and Ag ions with sodium citrate. The surface plasmons of Au0.5Ag0.5 alloy nanoparticles were clearly distinct from those of Ag and Au nanoparticles. Second, the SERS spectral pattern of 2,6-DMPI on the Au0.5Ag0.5 alloy was comparable to that on pure Ag but differed clearly from that on pure Au, suggesting that the outermost part of the Au0.5Ag0.5 nanoparticles was composed of Ag atoms. Third, the NC stretching band of 2,6-DMPI on Au was blueshifted under an acetone flow but was redshifted under an ammonia flow. The same band of 2,6-DMPI on Au0.5Ag0.5 alloy was, however, redshifted under an acetone flow. The amount of redshift was equal to that observed in the case of 2,6-DMPI on pure Ag. Under the ammonia flow, the bands for the pure Ag and Au0.5Ag0.5 alloy films were equally redshifted. This clearly indicates that VOCs affect only the surfaces of metal nanoparticles. By referring to the potentialdependent SERS data, the surface potential of pure Au nanoparticles appeared to increase by +0.16 V under the acetone flow, while it decreased by 0.56 V under the ammonia flow. The surface potential of pure Ag and Au0.5Ag0.5 alloy nanoparticles was in turn decreased equally by 0.22 V under the acetone flow and by 0.36 V under the ammonia flow. Acknowledgments This work was supported by the National Research Foundation of Korea (NRF) grants funded by the Korea government (MSIP) (Nos. 2007-0056095, 2011-0006737, and 2012R1A2A2A01008004). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]
[15] [16] [17] [18] [19] [20]
A. Henglein, Chem. Rev. 89 (1989) 1861. A.P. Alivisatos, J. Phys. Chem. 100 (1996) 13226. L.M. Liz-Marzán, Langmuir 22 (2006) 32. S.K. Ghosh, T. Pal, Chem. Rev. 107 (2007) 4797. F. Tao, M.E. Grass, Y. Zhang, D.R. Butcher, J.R. Renzas, Z. Liu, J.Y. Chung, B.S. Mun, M. Salmeron, G.A. Somorjai, Science 322 (2008) 932. R.V. Benjaminsen, H. Sun, J.R. Henriksen, N.M. Christensen, K. Almdal, T.L. Andresen, ACS Nano 5 (2011) 5864. E.B. Feresenbet, E. Dalcanale, C. Dulcey, D.K. Shenoy, Actuators, B 97 (2004) 211. A. Ulman, Chem. Rev. 96 (1996) 1533. D. Shin, K. Kim, K.S. Shin, Chem. Phys. Chem. 11 (2010) 83. K. Kim, K.L. Kim, D. Shin, J.W. Lee, K.S. Shin, Chem. Commun. 46 (2010) 3753. K. Kim, J.W. Lee, K.S. Shin, Spec. Chim. Acta Part A 100 (2013) 15. S. Link, Z.L. Wang, M.A. El-Sayed, J. Phys. Chem. B 103 (1999) 3529. P.C. Lee, D. Meisel, J. Phys. Chem. 86 (1982) 3391. R.G. Freeman, K.C. Grabar, K.J. Allison, R.M. Bright, J.A. Davis, A.P. Guthrie, M.B. Hommer, M.A. Jackson, P.C. Smith, D.G. Walter, M.J. Natan, Science 267 (1995) 1629. K. Kim, K.L. Kim, J.-Y. Choi, H.B. Lee, K.S. Shin, J. Phys. Chem. C 114 (2010) 3448. Vibrational frequencies of 2,6-DMPI were calculated at the B3LYP/6-31+G(2d, p) level using Gaussian 09 suite program. M. Moskovits, Phys. Chem. Chem. Phys. 15 (2013) 5301. G.S. Shafai, S. Shetty, S. Krishnamurty, V. Shah, D.G. Kanhere, J. Chem. Phys. 126 (2007) 014704. S.A. Wasileski, M.T.M. Koper, M.J. Weaver, J. Am. Chem. Soc. 124 (2002) 2796. B. Liang, L. Andrews, J. Phys. Chem. A 104 (2000) 9156.