- Email: [email protected]
Comparative studies on adsorption behavior of thionine on gold nanoparticles with different sizes
Journal of Colloid and Interface Science 327 (2008) 243–250
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Comparative studies on adsorption behavior of thionine on gold nanoparticles with different sizes Yuanhua Ding a,b , Zhuqing Chen b , Ju Xie b , Rong Guo b,∗ a b
School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093, PR China School of Chemistry and Chemical Engineering, Yangzhou University, Yangzhou 225002, PR China
a r t i c l e
i n f o
a b s t r a c t
Article history: Received 18 May 2008 Accepted 29 July 2008 Available online 9 August 2008
The adsorption behavior of thionine on gold nanoparticles of two different mean diameters, 18 and 5 nm, was compared by using UV–vis spectroscopy, fluorescence spectroscopy, Fourier transform infrared (FTIR) spectroscopy, transmission electron microscopy (TEM), dynamic light scattering (DLS), and quantum chemical calculations. It is found that the addition of small particles makes the monomer peak of thionine finally disappear, and the corresponding dimer peak is significantly increased. Small gold nanoparticles make the equilibrium between the monomer and H-type dimer forms of thionine move largely toward the dimer forms. Due to the stronger binding between thionine and small gold nanoparticles, the fluorescence quenching of thionine by small particles is enhanced compared to large particles, and the quenching is both static and dynamic. TEM images indicate that the addition of thionine results in a heavy clustering for small particles, and the resulting thionine–gold nanoclusters of about 45 nm were obtained. Quantum chemical calculations, which were based on the density functional theory (DFT) at the B3LYP level, and infrared spectroscopic studies show that the nitrogen atoms of the NH2 moieties of thionine bind to the gold nanoparticle surfaces. For 18 and 5 nm particles, the surface-to-volume atomic ratios are about 0.0597 and 0.2148, respectively. The higher surface-to-volume atomic ratio and the higher surface free energy result in stronger binding of thionine on small particle surfaces, which can be used to modulate the arrangement of dye molecules on particle surfaces, and thus control the properties of organic–inorganic nanocomposite materials. © 2008 Elsevier Inc. All rights reserved.
Keywords: Gold nanoparticles Thionine Size effect Adsorption behavior Quantum chemical calculation Surface structure parameter
1. Introduction In recent years, significant efforts have been undertaken to investigate the photophysical and photochemical behaviors of multicomponent nanostructured assemblies consisting of photoactive dyes, metals, and semiconductors [1–15]. Such organic–inorganic hybrid materials have various possible applications in developing efficient light energy conversion systems, optoelectronic devices, and sensors [14–18]. Considerable research activities in the field are the design of functionalized metal nanoparticles with optoelectronic properties [12]. Tailoring the optoelectronic properties of metal nanoparticles by organizing organic molecules containing functional groups on metal nanoparticles can yield functionalized organic–inorganic nanocomposite materials [12,19]. It is found that all these properties and applications are related to the surface characteristics of metal nanoparticles [20]. Therefore, it is important to investigate the adsorption behaviors of functional molecules, especially photoactive molecules, on metal nanoparti-
*
Corresponding author. Fax: +86 514 87311374. E-mail address: [email protected] (R. Guo).
0021-9797/$ – see front matter doi:10.1016/j.jcis.2008.07.057
©
2008 Elsevier Inc. All rights reserved.
cles based on the factors influencing surface conditions [20]. There have been several literature reports on dye–metal nanoparticle interactions [12–14,21–29], which are centered on their optical properties for chemical and biological applications[12–14,21–28] and the nanoparticle assembly mediated by dyes [29], but there has little work about the effect of the particle size on the adsorption behavior of dye on metal nanoparticles. Thionine, an important thiazine dye, is usually chosen as the sensitizer as it can yields both singlet and triplet excited states in detectable amounts [30]. Recently, the dye thionine has been used to study photosensitization of large-bandgap semiconductor ZnO [30], and also used to study the electron transfer with DNA [31,32]. However, little research has been reported about the adsorption characteristics of the dye on metal nanoparticles with different sizes. On the other hand, gold nanoparticles have been attracting a great deal of interest as they show unique physicochemical properties, especially the special nonlinear optical properties and surface-enhanced Raman scattering effect, and have potential applications in biology, catalysis, photocatalysis, and microelectronics [15,33–35]. In view of these considerations, thionine and gold nanoparticles have been chosen to investigate the effect of particle
244
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
Chart 1. Molecular structure of thionine.
size on the adsorption characteristics of dyes on metal nanoparticle surfaces. In the present work, the adsorption behavior of thionine on gold nanoparticles with mean diameter of 18 and 5 nm was compared by means of UV–vis spectroscopy, fluorescence spectroscopy, transmission electron microscopy (TEM), dynamic light scattering (DLS), and the effect was analyzed by quantum chemical calculations, Fourier transform infrared (FT-IR) spectroscopy, and the determination of surface structure parameters of gold particles. The size dependence of the aggregation behavior of thionine on gold nanoparticle surfaces, the fluorescence quenching of thionine by gold nanoparticles with different sizes, the formation of gold nanoclusters with different sizes, and surface structure parameters based on the interaction sites on particle surfaces were obtained. The size effect of nanogold on the adsorption behavior of dye will provide important information for the design of functionalized organic–inorganic nanocomposite materials. 2. Materials and methods 2.1. Materials HAuCl4 ·3H2 O, trisodium citrate, and sodium borohydride were purchased from Aldrich. Thionine (Chart 1) was from Sigma and used after purification with high-performance liquid chromatography (HPLC). Triply distilled water was used throughout the work. 2.2. Synthesis of gold nanoparticles Citrate-capped gold nanoparticles of 15–20 nm and 3–6 nm average diameter were prepared by the reduction of HAuCl4 with trisodium citrate or sodium borohydride [36,37]. An 18.5 mL volume of water and 0.5 mL of 1.0 × 10−2 mol/L trisodium citrate were mixed well, and 0.5 mL of chloroauric acid (1.0 × 10−3 mol/L) was then added to the system. The resulting mixture was stirred and then refluxed until the color changed to wine red. In this case, the trisodium citrate itself acted as the reducing agent. From the TEM measurements, the size of the prepared nanoparticles was mainly in the range of 15–20 nm. For preparation of gold nanoparticles of 3–6 nm size, the same concentration of aqueous trisodium citrate and chloroauric acid mentioned in the previous procedure was taken, and the mixture was stirred and then cooled in an ice bath. After 30 min, 0.5 mL of sodium borohydride (0.1 mol/L) was slowly added to the above mixture and stirred under the ice bath condition until the color turned orange. The concentration of the two kinds of gold sols was 2.56 × 10−5 mol/L. 2.3. Absorption and fluorescence measurements UV–vis absorption spectra and emission spectra were recorded with a Shimadzu UV-2550 spectrophotometer and an Edinburgh FLS920 steady state fluorescence spectrometer, respectively, by varying the concentration of gold nanoparticles and keeping thionine concentration constant in aqueous solution. 2.4. Transmission electron microscopy and dynamic light scattering measurements The TEM images of gold nanoparticles before and after the addition of thionine were measured with a Philips Tecnai-12 trans-
mission electron microscope. The TEM samples were prepared by dropping the gold sol onto a copper grid covered with carbon film. The size distribution of gold nanoparticles before and after the addition of thionine was measured by dynamic light scattering. The light scattering experiment was performed on an ALV-5022F laser light scattering spectrometer at a scattering angle of 90◦ , which was equipped with an ALV-6010/160 multiple Tau digital real time correlator and a He–Ne laser (25 mW, 632.8 nm). The final thionine concentration in the gold sol was 1.20 × 10−5 mol/L. 2.5. Quantum chemical calculations The theoretical calculations were carried out using the Gaussian 03 program for Windows [38]. The optimization of the thionine molecular structure and the calculations of the model for the Au-thionine molecule complex were performed using density functional theory method at the B3LYP/Lanl2MB level [39–42]. In the B3LYP functional, the exchange function is of the Becke’s three parameter type, including gradient correction, and the correlation correction involved the gradient-corrected functional of Lee, Yang, and Parr [43]. The split valence type basis sets Lanl2MB were used in the method. 2.6. Infrared spectral measurement The interaction sites of thionine on gold nanoparticle surfaces obtained by quantum chemical calculations were confirmed by FT-IR spectroscopy. The FT-IR spectra of free thionine and the adsorbed species on gold nanoparticle surfaces were measured with a Bruker Tensor 27 FT-infrared spectrometer. The sample of thionine-coated gold nanoparticles was synthesized as follows. A 20 mL volume of citrate-capped gold nanoparticle solution was mixed with 5 mL of saturated thionine solution in 2-propanol and stirred effectively for 24 h. The resulting solution was centrifuged at 12,000 rpm to obtain the precipitate of thionine-covered gold nanoparticles. After it was washed several times with 2-propanol and cold water to remove any unadsorbed thionine and unreacted citrate, the precipitate can be used for infrared experiments. All the experiments were carried out at 25 ± 0.1 ◦ C, unless otherwise mentioned. 3. Results and discussions 3.1. Absorption characteristics Fig. 1 presents the UV–vis absorption spectra of thionine in presence of gold nanoparticles with different sizes at a given thionine concentration (1.20 × 10−5 mol/L). The dot curves in Fig. 1 are the absorption spectra of the prepared gold sols, in which the 520 and 507 nm absorption bands are characteristic of the surface plasmon bands of gold nanoparticles of 15–20 nm and 3–6 nm respectively. As shown by the curve a in Fig. 1, the spectrum of thionine in water exhibits two characteristic absorption bands at 598 and 565 nm. The 598 nm band is a characteristic absorption feature of monomeric form, and the 565 nm shoulder can be attributed to the H-type dimer aggregate [44]. When the 15– 20 nm gold particles of 2.56 × 10−6 mol/L are added, as shown by the curve b in Fig. 1A, the intensities of the 598 and 565 nm bands increase; the peak position of the 598 nm band is almost unchanged, whereas that of the 565 nm band is slightly blueshifted. Moreover, the dimer peak also becomes clear slightly. At the same gold concentration of 2.56 × 10−6 mol/L, the addition of 3–6 nm gold nanoparticles result in the significant decrease in the intensity of the monomer peak at 598 nm, whereas the dimer peak is slightly increased with the peak position a little blue shift (curve d in Fig. 1B). This indicates that the equilibrium between
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
245
(A) Fig. 2. Emission spectra of thionine in aqueous solution containing different concentrations of 15–20 nm gold nanoparticles at 25 ◦ C, thionine concentration: 1.20 × 10−5 mol/L; gold nanoparticle concentration (mol/L): a—0, b—2.56 × 10−6 , c— 5.12 × 10−6 , d—1.02 × 10−5 , e—1.28 × 10−5 , f—1.54 × 10−5 , g—2.05 × 10−5 , h— 2.30 × 10−5 . The inset is the quenching curve for thionine by 15–20 nm gold nanoparticles in aqueous solution. λex = 550 nm.
(B) Fig. 1. UV–vis absorption spectra of thionine in aqueous solution in presence of (A) 15–20 nm and (B) 3–6 nm gold nanoparticles at 25 ◦ C, thionine concentration: 1.20 × 10−5 mol/L; gold nanoparticle concentration (mol/L): a—0; b, d—2.56 × 10−6 ; c, e—1.54 × 10−5 . The dot curves are the absorption spectra of the gold sols with different sizes, respectively.
the monomer and dimer forms of thionine moves significantly toward the dimer forms for small particles, but the transition is relatively weak for large particles. When the gold concentration is increased to 1.54 × 10−5 mol/L, it is found that the intensities of the 598 and 565 nm bands increase continuously, and the dimer peak becomes relatively clear for large particles (curve c in Fig. 1A). However, for small particles the intensity of the dimer peak is significantly increased, and the corresponding monomer peak almost disappears (curve e in Fig. 1B). This confirms that the small particles facilitate the transition from monomer to H-dimer forms of thionine. The absorption spectra changes in presence of gold particles with different sizes in Fig. 1 show that the particle size has significant effect on the adsorption behavior of dye. It has been known that gold nanoparticles, prepared by the reduction of HAuCl4 with trisodium citrate or sodium borohydride, have a negative surface charge due to a weakly bound citrate coating [45]. Thus, the cationic dye has opposite charges with respect to the surface charges of gold particles. When gold nanoparticles were added to the system, thionine molecules are adsorbed onto the gold nanoparticle surfaces due to the electrostatic attractive force between the cationic dye molecules and particle surfaces. This often leads to close packing of dye molecules on a charged particle sur-
face [12], and thus makes the local concentration of thionine on particle surfaces increase, correspondingly, its effective absorption cross-sectional area increases. As a result, both the absorption intensities of the 598 and 565 nm bands increase with the addition of gold particles, which is similar to the sensitization of photoactive molecules in micelles [46,47]. However, the close packing of dye molecules on the particle surface also induce the dipolar interactions between dye molecules, and thus results in the transition from monomer to H-type dimer forms of thionine [14]. This is the first report on the transition from monomer to H-type dimer forms of dyes based on metal nanoparticles. Furthermore, gold–dye assemblies coalesce to form larger clusters due to surface charge neutralization and hydrophobic interactions of the adsorbed dyes. This also facilitates the face-to-face arrangement of two thionine molecules, resulting in the transition from the monomer to Hdimer forms of thionine. Strong electronic coupling between the molecules in the formed dye aggregates causes the blue shift of the 565 nm band, which is characteristic of H-type dimer [44]. As the gold concentration is same for the two sets of particles, the smaller is the particle, the higher is the concentration of the particles, and the surface area is correspondingly higher. In the presence of small particles, dye molecules can induce intercluster interactions within the small particles, and such aggregates, which bring adsorbed dye molecules closer, facilitate dimer formation in the assembly of the dye molecules [48]. On the contrary, if the large-sized particles are used (above 20 nm), the aggregation amongst dye molecules is very weak, and thus, the gold particles with mean diameter of 18 and 5 nm were chosen in the present work. From the above analysis, it comes to a conclusion that the addition of small gold particles stimulates the transition from the monomer to dimer forms of thionine on the surface of gold particles. 3.2. Fluorescence quenching of thionine by gold nanoparticles with different sizes In Fig. 2 are shown the fluorescence spectra of thionine in aqueous solution in the presence of different concentrations of 15–20 nm gold nanoparticles. The fluorescence of thionine in water exhibits a maximum at around 619 nm, when it is excited at 550 nm (curve a in Fig. 2). With increasing concentration of gold
246
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
a quadratic relationship between relative intensity and quencher concentration is predicted [49]: I0
Fig. 3. Emission spectra of thionine in aqueous solution containing different concentrations of 3–6 nm gold nanoparticles at 25 ◦ C, thionine concentration: 1.20 × 10−5 mol/L; gold nanoparticle concentration (mol/L): a—0, b—2.56 × 10−6 , c— 5.12 × 10−6 , d—7.68 × 10−6 , e—1.02 × 10−5 , f—1.28 × 10−5 , g—1.54 × 10−5 , h— 1.79 × 10−5 , i—2.30 × 10−5 . The inset is the quenching curve for thionine by 3–6 nm gold nanoparticles in aqueous solution. λex = 550 nm.
nanoparticles, it can be seen from curves b–h in Fig. 2 that the fluorescence of thionine is gradually quenched by gold nanoparticles (2.56 × 10−6 –2.30 × 10−5 mol/L), and the emission peak is slightly blue-shifted. When the 3–6 nm particles are added to the same thionine solution, it is found that with the increasing concentration of gold particles (2.56 × 10−6 –2.30 × 10−5 mol/L), the fluorescence of thionine is significantly quenched by gold nanoparticles, and the quenching efficiency is almost 100% at the gold concentration of 2.30 × 10−5 mol/L (curve i in Fig. 3). Furthermore, the blue shift of the emission peak is increased with small particles added. This indicates that the fluorescence quenching of thionine by small particles is stronger than that by large ones. In addition, the blue shift of the emission peak and the absence of any new emission band in the emission spectra of thionine after the addition of gold particles suggest that the dye aggregates on gold nanoparticle surfaces are nonfluorescent, which is characteristic of H-type aggregates. The fluorescence quenching of thionine by gold nanoparticles with different sizes can be both static and dynamic. The former involves the formation of a nonfluorescent complex between the fluorophore and quencher in their ground states, whereas the latter requires diffusive encounters between the fluorophore and quencher during the lifetime of the excited state of the fluorophore. Assuming that the deactivation of the excited thionine occurs by collision with the quencher in fluid solution (pure dynamic quenching), the relative reduction in emission intensity is equal to the relative reduction in fluorescence lifetime, and the relationship between relative intensity and quencher concentration is linear [49]: I0 I
=
τ0 = 1 + K SV [ Q ], τ
(1)
where I 0 and I are the fluorescence intensities of thionine in absence and presence of quencher Q (gold nanoparticles), respectively; τ0 and τ are the fluorescence lifetimes of thionine in absence and presence of quencher, K SV is the Stern–Volmer quenching constant (L/mol), and [ Q ] is the concentration of the quencher (mol/L). However, when static quenching mechanisms occur through the complexing of the ground state fluorophore and quencher molecules, the relative reduction in emission intensity is greater than the relative reduction in fluorescence lifetime. In this case
= 1 + ( K SV + K eq )[ Q ] + K SV K eq [ Q ]2 , (2) I where K eq is the equilibrium constant for the binding of the fluorophore and quencher. This modified form of the Stern–Volmer equation accounts for an upward curvature of the Stern–Volmer plots. The quenching curve of thionine by large gold nanoparticles is nonlinear, as shown by the inset of Fig. 2, and changes upward with gold particles added, which is better fit by a quadratic function than a linear one. The nonlinearities in the quenching curve indicate that the fluorescence quenching is progressively enhanced with the addition of gold nanoparticles, and also indicate that both static and dynamic quenching occur for the same fluorophore, which is different from the static quenching of chlorophyll a by gold nanoparticles [35]. The upward curvature of the quenching curve in the inset of Fig. 2 also indicates that the binding extent of thionine on particle surfaces increases with the addition of gold particles, which is consistent with the fact that the H-type dimer peak becomes gradually clear with gold nanoparticles added (Fig. 1). Compared to that for large particles, the upward curvature of the quenching curve is more remarkable for small particles (inset of Fig. 3), the quenching value I 0 / I reaches almost 120 at the same gold concentration of 2.30 × 10−5 mol/L, indicating that the static quenching plays a major role in the fluoresce process of thionine by small particles. In addition, the increasing extent of blue shift of the emission peak is larger for small particles than for large particles, which confirms that small particles facilitate the transition from the monomer to dimer forms of thionine on the surface of gold particles, and is consistent with the UV–vis spectral results. For static quenching, a strong binding between thionine and gold nanoparticles provides favorable conditions for the heterogeneous deactivation processes at thionine–gold nanoparticles interface, which is responsible for the upward curvature of the Stern–Volmer plots of the insets of Figs. 2 and 3. 3.3. TEM images of gold sol Fig. 4 shows the TEM images and the corresponding size distribution of large gold nanoparticles before and after the addition of thionine. As can be seen in Fig. 4A, the prepared gold nanoparticles are almost spherical shape, and separated from each other. The particle size is mainly in the range of 15–20 nm in addition to some tiny particles. DLS data indicate that the particle size is centered on about 13 nm, and the calculated average diameter is 18.0 nm. It is noted that the difference of particle sizes between TEM and DLS comes from the different sizing method used by the manufacturers. TEM gives the number-average diameter, whereas DLS give a z-average diameter and also includes an aqueous hydration layer around the particle. After the addition of thionine, as shown in Fig. 4B, the gold particles also exhibit similar particle diameter, but some of gold nanoparticles form close-packing nanoclusters, which is induced by the adsorption of thionine on gold nanoparticle surfaces. DLS data in Fig. 4B indicate that there appear two size distribution peaks, which correspond to about 13 and 84 nm, respectively. The former peak with a smaller hydrodynamic diameter of about 13 nm can be attributed to single thionine-bound particles, and the latter one can be attributed to thionine-bound gold nanoclusters. In the case of gold nanoparticles of average 5 nm, the clustering morphological feature is significantly increased (Fig. 5B), and DLS data also show a new peak corresponding to a hydrodynamic diameter of about 45 nm (Fig. 5B), which suggests the formation of a small-sized gold nanocluster. Therefore, the formation of gold–dye composite particles and gold– dye nanoclusters can be controlled by gold particles of different
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
247
(A)
(B) Fig. 4. TEM images and corresponding particle size distribution of (A) the gold sol (15–20 nm) and (B) the gold sol with 1.20 × 10−5 mol/L thionine. D h denotes the hydrodynamic diameter.
sizes, which further supports the explanations for the above spectra changes. It is also noted that the nanoparticles in the gold nanoclusters are almost individually isolated, as can be observed from the TEM images of Figs. 4B and 5B. This is in contrast to the simple saltinduced aggregation of nanoparticles in which the nanoparticles are usually fused together due to the collapse of electrical double layers. In addition, the formation of the H-type dimer of dye thionine on the particle surfaces may contribute to bridging the neighboring nanoparticles. 3.4. Quantum chemical structure of thionine and model for thionine–Au complex To understand the significant changes of the adsorption behavior in presence of gold nanoparticles with different sizes, the binding of thionine on gold particles was first analyzed in terms of the thionine molecular structure. The nitrogen atoms of the NH2 moieties of thionine are usually electron-rich, which are likely to bind to the particles. The nitrogen and sulfur atoms of the central heterocyclic ring of thionine are less-electron rich due to delocalization of electrons to electron deficient phenyl ring, which can be supported by the following results of quantum chemical calculations (Table 1). It has been reported that the electron deficiency of atoms make them unlikely to bind directly to the nanoparticle surfaces [37]. Therefore, it can be inferred that the nitrogen atoms of the NH2 moieties of thionine strongly bind to gold nanoparticle surfaces. The binding of the amino groups to
the surfaces of gold nanoparticles has been used in fabricating the nanogold self-assembly through an amine-terminated monolayer of cystamine [50], or in achieving the organization of pyrene chromophores around gold nanoparticles through the surface binding of 1-methylaminopyrene (Py-CH2 NH2 , molecular probes) to gold nanoparticles [51]. To confirm the interaction sites of the thionine on gold nanoparticle surfaces, quantum chemical calculations were performed to obtain the optimized molecular structure of thionine and the model for thionine–Au complex (Fig. 6). The calculating results indicate that thionine molecules show planar structure, and at both of the NH2 ends of the molecule, the bond angles of H– N–H and C–N–H, e.g., H19 –N17 –H18 and C1 –N17 –H19 , are 117.70◦ and 121.00◦ , respectively, and the bond distances of N–H and C–N bonds, e.g. N17 –H19 and C1 –N17 , are 1.0408 and 1.3788 Å, respectively (Fig. 6A). From the natural charge of the relevant atoms of the molecule (Table 1), it can be known that the nitrogen atoms of the –NH2 moieties of the molecule, N17 and N21 , are most electron-rich, and easily offer their lone electron pairs to the electron deficient gold nanoparticles, which provides possibility of the binding of the –NH2 on the particle surfaces. Assuming that gold atoms and the nitrogen atoms of the –NH2 moieties of thionine form complex, the obtained data (Table 2) indicate that, at both of the NH2 ends of the molecule, the bond angle of H–N–H decreases from 117.70◦ to 107.84◦ , and that of C–N–H decreases from 121.00◦ to 111.01◦ ; whereas the bond distance of N–H bond increases from 1.0408 to 1.0544 Å, and that of C–N bond increases significantly from 1.3788 to 1.4602 Å. However, the bond angles of C–S–C and
248
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
(A)
(B) Fig. 5. TEM images and corresponding particle size distribution of (A) the gold sol (3–6 nm) and (B) the gold sol with 1.20 × 10−5 mol/L thionine. D h denotes the hydrodynamic diameter. Table 1 Natural charge of relevant atoms of the optimized thionine molecule at the DFT level of theory Symbol Natural charge
N17 −0.3215
N21 −0.3215
S26 −0.1396
N20 −0.1091
(A)
(B) Fig. 6. Optimized molecular structure of (A) thionine and (B) the modeled thionine– Au complex obtained from the DFT level of theory.
C–N–C, and the bond distances of C–S and C–N bonds at the central heterocyclic ring hardly change. On the other hand, assuming that gold atoms and the nitrogen or sulfur atom of the central heterocyclic ring of the thionine form complex, the bond angles of C–N–C and C–S–C, and the bond distances of C–N and C–S bonds of the ring are almost unchanged (data not shown), suggesting that the nitrogen and sulfur atoms of the central heterocyclic ring do not bind to the gold particle surfaces. The remarkable decrease of the bond angles of H–N–H and C–N–H, and the increase of the bond distances of N–H and C–N bonds at both of the NH2 ends of the molecule (Table 2) indicate that the –NH2 binds to the gold particle surfaces with the lone electron pair of the nitrogen atom pointing to the particle surface. This makes the N–H bonds deviate from the particle surface, and turn to the direction of C–N bond, and thus, weaken the N–H bonds. It is noted that the calculations of the interactions between a thionine molecule and a gold particle is very complex, as a particle contains many gold atoms (about 3859–180,072 gold atoms per particle). Considering that the binding of thionine on gold particles occurs at the gold atoms on the particle surfaces, two gold atoms, which were from a particle or from two particles respectively, were used to study the interactions of thionine with particles by the DFT method. This can approximately give specific interaction sites of thionine on particle surfaces. Such a calculation method has been reported in studying the adsorption behavior of biologically significant 2-aminobenzothiazole molecule on colloid silver particles by surface-enhanced Raman spectroscopy [52].
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
Table 2 Relevant structural parameters of thionine and the changes upon binding to Au at the DFT level of theory Thionine
Thionine–Au
Bond lengths (Å) 1.0544 1.4602 1.9071 1.4041
N17 –H19 C1 –N17 C2 –S26 C3 –N20
1.0408 1.3788 1.9055 1.3925
H19 –N17 –H18 C1 –N17 –H19 C2 –S26 –C7 C3 –N20 –C6
Interatomic angles (deg) 117.70 107.84 121.00 111.01 100.25 99.62 120.99 119.57
Diff.
−0.0136 −0.0814 −0.0016 −0.0116 9.86 9.99 0.63 1.42
249
Table 3 Surface structure parameters of gold nanoparticles with different sizes Particle size (nm)
N Au
α
N sur
18 5
180,072 3859
0.0597 0.2148
10,750 829
N Au —aggregation number of a gold particle; α —surface-to-volume atomic ratio of a gold particle; N Suf —the number of gold atoms on the surface of a gold particle.
number per gold particle (N Au ), the number of gold atoms per particle, was calculated using the formula [48]:
N Au = 59 nm−3
π ( D M ), 6
(3)
where D M is the mean diameter of the particles. And the surfaceto-volume atomic ratio (α ) of metal nanoparticles is defined by [48]:
α=
Fig. 7. FT-IR spectra of (a) free thionine and (b) the adsorbed thionine on gold nanoparticles.
To verify the interaction sites obtained by quantum chemical calculations, the FT-IR spectra of free thionine and the adsorbed thionine were examined (Fig. 7). The absorption bands at 1600 and 1500 cm−1 , as shown by the curve a in Fig. 7, are assigned to the skeletal vibration of the phenyl ring of thionine [53], which are also shown by the adsorbed thionine (curve b in Fig. 7), confirming the presence of thionine on the gold nanoparticle surfaces. The absorption bands at 3310 and 3150 are ascribed to the N–H stretching vibration of the amino moieties (curve a in Fig. 7) [53], which are not observed in the IR spectrum of the adsorbed thionine (curve b in Fig. 7), indicating that the nitrogen atoms of the NH2 moieties of thionine bind strongly to gold nanoparticle surfaces. However, as the absorption band of C–S bond in the range of 710–570 nm is weak and broad [53], it is meaningless to observe the difference between the IR spectrum of free thionine and that of the adsorbed thionine. On the curve b in Fig. 7, the absorption band at 3427 cm−1 is due to traces of water in the nanoparticle sample, and the absorption bands at 2930 and 2856 cm−1 are due to traces of citrate impurity in the sample. From the above analysis, it can be seen that the IR results are consistent with those obtained by quantum chemical calculations. Therefore, the interaction sites obtained by quantum chemical calculations and FT-IR spectroscopy confirm the binding mode of thionine on gold particles, and can be used for further surface structure analysis of gold particles of different sizes. 3.5. Surface structure analysis of gold particles The changes of adsorption behavior of thionine on gold nanoparticles with different sizes can be further analyzed in terms of surface structure parameters of gold particles. The aggregation
3d R
,
(4)
where d and R denote the atomic diameter and particle size, respectively. Assuming d = 3.58 Å for atomic gold, the obtained surface parameters are shown in Table 3. From Table 3, it can be seen that an 18 nm gold particle with 10,750 gold atoms on the surface can accommodate large number of dye molecules through the formation of multilayers of dye molecules which contain planar ring capable of forming dimers. The number of thionine molecules corresponding to an 18 nm gold particle is larger than that corresponding to a 5 nm particle as the number of large particles decrease in solution. Thus, higher thionine concentration per large gold nanoparticle exists in solution, and is expected to induce extensive dimerization. On the other hand, as the gold concentration is same for the two sets of particles, the smaller of the particle, the larger is the concentration of the particles, and the total surface area is correspondingly larger, although the number of surface gold atoms of a small particle is relatively small. The data in Table 3 clearly indicate that a small particle have a larger surface-to-volume atomic ratio, 0.2148, than a large one. As a result, the observed dye aggregation in the presence of small particles can be interpreted by considering that dye molecules can induce intercluster interactions within the small particles, and such aggregates, which bring adsorbed dye molecules closer, facilitate dimer formation in the assembly of the dye molecules. Furthermore, the molar surface area increases with decreasing particle size, then results in an increase in molar free energy of the small particles compared to large ones, and thus, the aggregation of particles is also the driving force for the strong dimerization of thionine molecules. TEM images indicate that the addition of thionine in the small particle system induces stronger clustering than in large ones, which confirm the above analysis. It should be noted that the preparation procedure of the two kinds of gold nanoparticles was different, and thus the surface coverage of citrate ions as well as the surface charge of the gold nanoparticles were also different. These factors also contribute to the difference in the adsorption behavior of thionine on gold nanoparticles with different sizes. However, as the binding of citrate ions on gold nanoparticles is relatively weak [45], the effects of these factors on the adsorption behavior of thionine are little. 4. Summary The size effect of gold particle on the adsorption behavior has been obtained by using various experimental techniques and quantum chemical calculations. The small particles stimulate the equilibrium between the monomer and H-dimer forms of thionine to move significantly toward the dimer forms. Due to stronger binding of thionine to small particles, the quenching of the dye by
250
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
small particles is more remarkable than that by large particles. The binding of thionine to gold particles results in the formation of gold nanoclusters with different sizes, and DLS indicates that the size of nanoclusters is 45 and 84 nm for the 5 and 18 nm particles, respectively. Quantum chemical calculation and IR results indicate that the nitrogen atoms of the NH2 moieties of thionine strongly bind to gold nanoparticle surfaces, and the number of surface gold atoms are 3859 and 180,072 for a 5 or 18 nm gold particle, respectively. The higher surface-to-volume atomic ratio and the higher surface free energy of small particles result in the significant adsorption behavior of thionine on the particles. The above results will provide important guidance for controlling the properties of organic–inorganic nanocompo site materials. Acknowledgment This work was supported by the National Natural Science Foundation of China (20633010). References [1] A. Henglein, J. Phys. Chem. 97 (1993) 5457. [2] P. Mulvaney, M. Giersig, A. Henglein, J. Phys. Chem. 97 (1993) 7061. [3] U. Kreibig, M. Vollmer, Optical Properties of Metal Clusters, Springer, Berlin, 1995. [4] L.M. Liz-Marzán, M. Giersig, P. Mulvaney, Langmuir 12 (1996) 4329. [5] R.G. Freeman, M.B. Hommer, K.C. Grabar, M.A. Jackson, M.J. Natan, J. Phys. Chem. 100 (1996) 718. [6] A. Martino, S.A. Yamanaka, J.S. Kawola, D.A. Loy, Chem. Mater. 9 (1997) 423. [7] G.L. Hornyak, C.J. Patrissi, C.R. Martin, J. Phys. Chem. B 101 (1997) 1548. [8] A. Henglein, Chem. Mater. 10 (1998) 444. [9] P. Xu, H. Yanagi, Chem. Mater. 11 (1999) 2626. [10] H. Behar-Levy, D. Avnir, Chem. Mater. 14 (2002) 1736. [11] V. Subramanian, E.E. Wolf, P.V. Kamat, J. Phys. Chem. B 107 (2003) 7479. [12] K. George Thomas, P.V. Kamat, Acc. Chem. Res. 36 (2003) 888. [13] A.R. Clapp, I.L. Medintz, B.R. Fisher, G.P. Anderson, H. Mattoussi, J. Am. Chem. Soc. 127 (2005) 1242. [14] N. Chandrasekharan, P.V. Kamat, J. Hu, G. Jonesll, J. Phys. Chem. B 104 (2000) 11103. [15] M.-C. Daniel, D. Astruc, Chem. Rev. 104 (2004) 293. [16] J. Hu, J. Zhang, F. Liu, K. Kittredge, J.K. Whitesell, M.A. Fox, J. Am. Chem. Soc. 123 (2001) 1464. [17] B.I. Ipe, K. George Thomas, S. Barazzouk, S. Hotchandani, P.V. Kamat, J. Phys. Chem. B 106 (2002) 18. [18] P.K. Sudeep, B.I. Ipe, K. George Thomas, M.V. George, S. Barazzouk, S. Hotchandani, P.V. Kamat, Nano Lett. 2 (2002) 29. [19] K. George Thomas, J. Zajicek, P.V. Kamat, Langmuir 18 (2002) 3722. [20] H.M. Eckenrode, S.-H. Jen, J. Han, A.-G. Yeh, H.-L. Dai, J. Phys. Chem. B 109 (2005) 4646. [21] N. Kometuni, M. Tsubonishi, T. Fujita, K. Asami, Y. Yonezawa, Langmuir 17 (2001) 578.
[22] S. Franzen, J.C.W. Folmer, W.R. Glomm, R. O’Neal, J. Phys. Chem. A 106 (2002) 6533. [23] J. Zhang, J.R. Lakowicz, J. Phys. Chem. B 119 (2005) 8701. [24] P.C. Lee, D. Meisel, J. Phys. Chem. 86 (1982) 3391. [25] F. Cannone, M. Collini, L. D’Alfonso, G. Baldini, G. Chirico, G. Tallarida, P. Pallavicini, Nano Lett. 7 (2007) 1070. [26] F. Cannone, G. Chirico, A.R. Bizzarri, S. Cannistraro, J. Phys. Chem. B 110 (2006) 16491. [27] J. Seelig, K. Leslie, A. Renn, S. Kuhn, V. Jacobsen, M. van de Corput, C. Wyman, V. Sandoghdar, Nano Lett. 7 (2007) 685. [28] T.L. Jennings, M.P. Singh, G.F. Strouse, J. Am. Chem. Soc. 128 (2006) 5462. [29] I.-I.S. Lim, F. Goroleski, D. Mott, N. Kariuki, W. Ip, J. Luo, C.J. Zhong, J. Phys. Chem. B 110 (2006) 6673. [30] B. Patrick, P.V. Kamat, J. Phys. Chem. 96 (1992) 1423. [31] G.D. Reid, D.J. Whittaker, M.A. Day, D.A. Turton, V. Kayser, J.M. Kelly, G.S. Beddard, J. Am. Chem. Soc. 124 (2002) 5518. [32] C. Dohno, E.D.A. Stemp, J.K. Barton, J. Am. Chem. Soc. 125 (2003) 9586. [33] V. Subramanian, E.E. Wolf, P.V. Kamat, J. Am. Chem. Soc. 126 (2004) 4943. [34] R. Shenhar, V.M. Rotello, Acc. Chem. Res. 36 (2003) 549. [35] S. Barazzouk, P.V. Kamat, S. Hotchandani, J. Phys. Chem. B 109 (2005) 716. [36] J. Turkevitch, P.C. Stevenson, J. Hillier, Discuss. Faraday Soc. 11 (1951) 55. [37] R.T. Tom, V. Suryanarayanan, P.G. Reddy, S. Baskaran, T. Pradeep, Langmuir 20 (2004) 1909. [38] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery Jr., R.E. Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.V. Ortiz, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe, P.M.W. Gill, B.G. Johnson, W. Chen, M.W. Wong, J.L. Andres, M. Head-Gordon, E.S. Replogle, J.A. Pople, Gaussian 98, revision A.7, Gaussian, Inc., Pittsburgh, PA, 1998. [39] A.D. Becke, Phys. Rev. A 38 (1988) 3098. [40] A.D. Becke, J. Chem. Phys. 98 (1993) 5648. [41] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785. [42] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 270. [43] P.J. Stevens, F.J. Devlin, C.F. Chabalowski, M.J. Frisch, J. Phys. Chem. 98 (1994) 11623. [44] S. Das, P.V. Kamat, J. Phys. Chem. B 103 (1999) 209. [45] C. Lee, S. Kim, C. Yoon, M. Gong, B.K. Choi, K. Kim, S.W. Joo, J. Colloid Interface Sci. 271 (2004) 41. [46] C.A. Bunton, L. Sepulveda, J. Phys. Chem. 83 (1979) 680. [47] M.A. Awan, S.S. Shah, Colloids Surf. A 122 (1997) 97. [48] S.K. Ghosh, A. Pal, S. Nath, S. Kundu, S. Panigrahi, T. Pal, Chem. Phys. Lett. 412 (2005) 5. [49] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Plenum Press, New York, 1983. [50] C.R. Raj, T. Okajima, T. Ohsaka, J. Electroanal. Chem. 543 (2003) 127. [51] K. George Thomas, P.V. Kamat, J. Am. Chem. Soc. 122 (2000) 2655. [52] J. Sarkar, J. Chowdhury, M. Ghosh, R. De, G.B. Talapatra, J. Phys. Chem. B 109 (2005) 12861. [53] E. Pretsch, P. Bühlmann, C. Affolter, Structure Determination of Organic Compounds: Tables of Spectral Data, third ed., Springer-Verlag, Berlin, 2000.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Comparative studies on adsorption behavior of thionine on gold nanoparticles with different sizes Yuanhua Ding a,b , Zhuqing Chen b , Ju Xie b , Rong Guo b,∗ a b
School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093, PR China School of Chemistry and Chemical Engineering, Yangzhou University, Yangzhou 225002, PR China
a r t i c l e
i n f o
a b s t r a c t
Article history: Received 18 May 2008 Accepted 29 July 2008 Available online 9 August 2008
The adsorption behavior of thionine on gold nanoparticles of two different mean diameters, 18 and 5 nm, was compared by using UV–vis spectroscopy, fluorescence spectroscopy, Fourier transform infrared (FTIR) spectroscopy, transmission electron microscopy (TEM), dynamic light scattering (DLS), and quantum chemical calculations. It is found that the addition of small particles makes the monomer peak of thionine finally disappear, and the corresponding dimer peak is significantly increased. Small gold nanoparticles make the equilibrium between the monomer and H-type dimer forms of thionine move largely toward the dimer forms. Due to the stronger binding between thionine and small gold nanoparticles, the fluorescence quenching of thionine by small particles is enhanced compared to large particles, and the quenching is both static and dynamic. TEM images indicate that the addition of thionine results in a heavy clustering for small particles, and the resulting thionine–gold nanoclusters of about 45 nm were obtained. Quantum chemical calculations, which were based on the density functional theory (DFT) at the B3LYP level, and infrared spectroscopic studies show that the nitrogen atoms of the NH2 moieties of thionine bind to the gold nanoparticle surfaces. For 18 and 5 nm particles, the surface-to-volume atomic ratios are about 0.0597 and 0.2148, respectively. The higher surface-to-volume atomic ratio and the higher surface free energy result in stronger binding of thionine on small particle surfaces, which can be used to modulate the arrangement of dye molecules on particle surfaces, and thus control the properties of organic–inorganic nanocomposite materials. © 2008 Elsevier Inc. All rights reserved.
Keywords: Gold nanoparticles Thionine Size effect Adsorption behavior Quantum chemical calculation Surface structure parameter
1. Introduction In recent years, significant efforts have been undertaken to investigate the photophysical and photochemical behaviors of multicomponent nanostructured assemblies consisting of photoactive dyes, metals, and semiconductors [1–15]. Such organic–inorganic hybrid materials have various possible applications in developing efficient light energy conversion systems, optoelectronic devices, and sensors [14–18]. Considerable research activities in the field are the design of functionalized metal nanoparticles with optoelectronic properties [12]. Tailoring the optoelectronic properties of metal nanoparticles by organizing organic molecules containing functional groups on metal nanoparticles can yield functionalized organic–inorganic nanocomposite materials [12,19]. It is found that all these properties and applications are related to the surface characteristics of metal nanoparticles [20]. Therefore, it is important to investigate the adsorption behaviors of functional molecules, especially photoactive molecules, on metal nanoparti-
*
Corresponding author. Fax: +86 514 87311374. E-mail address: [email protected] (R. Guo).
0021-9797/$ – see front matter doi:10.1016/j.jcis.2008.07.057
©
2008 Elsevier Inc. All rights reserved.
cles based on the factors influencing surface conditions [20]. There have been several literature reports on dye–metal nanoparticle interactions [12–14,21–29], which are centered on their optical properties for chemical and biological applications[12–14,21–28] and the nanoparticle assembly mediated by dyes [29], but there has little work about the effect of the particle size on the adsorption behavior of dye on metal nanoparticles. Thionine, an important thiazine dye, is usually chosen as the sensitizer as it can yields both singlet and triplet excited states in detectable amounts [30]. Recently, the dye thionine has been used to study photosensitization of large-bandgap semiconductor ZnO [30], and also used to study the electron transfer with DNA [31,32]. However, little research has been reported about the adsorption characteristics of the dye on metal nanoparticles with different sizes. On the other hand, gold nanoparticles have been attracting a great deal of interest as they show unique physicochemical properties, especially the special nonlinear optical properties and surface-enhanced Raman scattering effect, and have potential applications in biology, catalysis, photocatalysis, and microelectronics [15,33–35]. In view of these considerations, thionine and gold nanoparticles have been chosen to investigate the effect of particle
244
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
Chart 1. Molecular structure of thionine.
size on the adsorption characteristics of dyes on metal nanoparticle surfaces. In the present work, the adsorption behavior of thionine on gold nanoparticles with mean diameter of 18 and 5 nm was compared by means of UV–vis spectroscopy, fluorescence spectroscopy, transmission electron microscopy (TEM), dynamic light scattering (DLS), and the effect was analyzed by quantum chemical calculations, Fourier transform infrared (FT-IR) spectroscopy, and the determination of surface structure parameters of gold particles. The size dependence of the aggregation behavior of thionine on gold nanoparticle surfaces, the fluorescence quenching of thionine by gold nanoparticles with different sizes, the formation of gold nanoclusters with different sizes, and surface structure parameters based on the interaction sites on particle surfaces were obtained. The size effect of nanogold on the adsorption behavior of dye will provide important information for the design of functionalized organic–inorganic nanocomposite materials. 2. Materials and methods 2.1. Materials HAuCl4 ·3H2 O, trisodium citrate, and sodium borohydride were purchased from Aldrich. Thionine (Chart 1) was from Sigma and used after purification with high-performance liquid chromatography (HPLC). Triply distilled water was used throughout the work. 2.2. Synthesis of gold nanoparticles Citrate-capped gold nanoparticles of 15–20 nm and 3–6 nm average diameter were prepared by the reduction of HAuCl4 with trisodium citrate or sodium borohydride [36,37]. An 18.5 mL volume of water and 0.5 mL of 1.0 × 10−2 mol/L trisodium citrate were mixed well, and 0.5 mL of chloroauric acid (1.0 × 10−3 mol/L) was then added to the system. The resulting mixture was stirred and then refluxed until the color changed to wine red. In this case, the trisodium citrate itself acted as the reducing agent. From the TEM measurements, the size of the prepared nanoparticles was mainly in the range of 15–20 nm. For preparation of gold nanoparticles of 3–6 nm size, the same concentration of aqueous trisodium citrate and chloroauric acid mentioned in the previous procedure was taken, and the mixture was stirred and then cooled in an ice bath. After 30 min, 0.5 mL of sodium borohydride (0.1 mol/L) was slowly added to the above mixture and stirred under the ice bath condition until the color turned orange. The concentration of the two kinds of gold sols was 2.56 × 10−5 mol/L. 2.3. Absorption and fluorescence measurements UV–vis absorption spectra and emission spectra were recorded with a Shimadzu UV-2550 spectrophotometer and an Edinburgh FLS920 steady state fluorescence spectrometer, respectively, by varying the concentration of gold nanoparticles and keeping thionine concentration constant in aqueous solution. 2.4. Transmission electron microscopy and dynamic light scattering measurements The TEM images of gold nanoparticles before and after the addition of thionine were measured with a Philips Tecnai-12 trans-
mission electron microscope. The TEM samples were prepared by dropping the gold sol onto a copper grid covered with carbon film. The size distribution of gold nanoparticles before and after the addition of thionine was measured by dynamic light scattering. The light scattering experiment was performed on an ALV-5022F laser light scattering spectrometer at a scattering angle of 90◦ , which was equipped with an ALV-6010/160 multiple Tau digital real time correlator and a He–Ne laser (25 mW, 632.8 nm). The final thionine concentration in the gold sol was 1.20 × 10−5 mol/L. 2.5. Quantum chemical calculations The theoretical calculations were carried out using the Gaussian 03 program for Windows [38]. The optimization of the thionine molecular structure and the calculations of the model for the Au-thionine molecule complex were performed using density functional theory method at the B3LYP/Lanl2MB level [39–42]. In the B3LYP functional, the exchange function is of the Becke’s three parameter type, including gradient correction, and the correlation correction involved the gradient-corrected functional of Lee, Yang, and Parr [43]. The split valence type basis sets Lanl2MB were used in the method. 2.6. Infrared spectral measurement The interaction sites of thionine on gold nanoparticle surfaces obtained by quantum chemical calculations were confirmed by FT-IR spectroscopy. The FT-IR spectra of free thionine and the adsorbed species on gold nanoparticle surfaces were measured with a Bruker Tensor 27 FT-infrared spectrometer. The sample of thionine-coated gold nanoparticles was synthesized as follows. A 20 mL volume of citrate-capped gold nanoparticle solution was mixed with 5 mL of saturated thionine solution in 2-propanol and stirred effectively for 24 h. The resulting solution was centrifuged at 12,000 rpm to obtain the precipitate of thionine-covered gold nanoparticles. After it was washed several times with 2-propanol and cold water to remove any unadsorbed thionine and unreacted citrate, the precipitate can be used for infrared experiments. All the experiments were carried out at 25 ± 0.1 ◦ C, unless otherwise mentioned. 3. Results and discussions 3.1. Absorption characteristics Fig. 1 presents the UV–vis absorption spectra of thionine in presence of gold nanoparticles with different sizes at a given thionine concentration (1.20 × 10−5 mol/L). The dot curves in Fig. 1 are the absorption spectra of the prepared gold sols, in which the 520 and 507 nm absorption bands are characteristic of the surface plasmon bands of gold nanoparticles of 15–20 nm and 3–6 nm respectively. As shown by the curve a in Fig. 1, the spectrum of thionine in water exhibits two characteristic absorption bands at 598 and 565 nm. The 598 nm band is a characteristic absorption feature of monomeric form, and the 565 nm shoulder can be attributed to the H-type dimer aggregate [44]. When the 15– 20 nm gold particles of 2.56 × 10−6 mol/L are added, as shown by the curve b in Fig. 1A, the intensities of the 598 and 565 nm bands increase; the peak position of the 598 nm band is almost unchanged, whereas that of the 565 nm band is slightly blueshifted. Moreover, the dimer peak also becomes clear slightly. At the same gold concentration of 2.56 × 10−6 mol/L, the addition of 3–6 nm gold nanoparticles result in the significant decrease in the intensity of the monomer peak at 598 nm, whereas the dimer peak is slightly increased with the peak position a little blue shift (curve d in Fig. 1B). This indicates that the equilibrium between
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
245
(A) Fig. 2. Emission spectra of thionine in aqueous solution containing different concentrations of 15–20 nm gold nanoparticles at 25 ◦ C, thionine concentration: 1.20 × 10−5 mol/L; gold nanoparticle concentration (mol/L): a—0, b—2.56 × 10−6 , c— 5.12 × 10−6 , d—1.02 × 10−5 , e—1.28 × 10−5 , f—1.54 × 10−5 , g—2.05 × 10−5 , h— 2.30 × 10−5 . The inset is the quenching curve for thionine by 15–20 nm gold nanoparticles in aqueous solution. λex = 550 nm.
(B) Fig. 1. UV–vis absorption spectra of thionine in aqueous solution in presence of (A) 15–20 nm and (B) 3–6 nm gold nanoparticles at 25 ◦ C, thionine concentration: 1.20 × 10−5 mol/L; gold nanoparticle concentration (mol/L): a—0; b, d—2.56 × 10−6 ; c, e—1.54 × 10−5 . The dot curves are the absorption spectra of the gold sols with different sizes, respectively.
the monomer and dimer forms of thionine moves significantly toward the dimer forms for small particles, but the transition is relatively weak for large particles. When the gold concentration is increased to 1.54 × 10−5 mol/L, it is found that the intensities of the 598 and 565 nm bands increase continuously, and the dimer peak becomes relatively clear for large particles (curve c in Fig. 1A). However, for small particles the intensity of the dimer peak is significantly increased, and the corresponding monomer peak almost disappears (curve e in Fig. 1B). This confirms that the small particles facilitate the transition from monomer to H-dimer forms of thionine. The absorption spectra changes in presence of gold particles with different sizes in Fig. 1 show that the particle size has significant effect on the adsorption behavior of dye. It has been known that gold nanoparticles, prepared by the reduction of HAuCl4 with trisodium citrate or sodium borohydride, have a negative surface charge due to a weakly bound citrate coating [45]. Thus, the cationic dye has opposite charges with respect to the surface charges of gold particles. When gold nanoparticles were added to the system, thionine molecules are adsorbed onto the gold nanoparticle surfaces due to the electrostatic attractive force between the cationic dye molecules and particle surfaces. This often leads to close packing of dye molecules on a charged particle sur-
face [12], and thus makes the local concentration of thionine on particle surfaces increase, correspondingly, its effective absorption cross-sectional area increases. As a result, both the absorption intensities of the 598 and 565 nm bands increase with the addition of gold particles, which is similar to the sensitization of photoactive molecules in micelles [46,47]. However, the close packing of dye molecules on the particle surface also induce the dipolar interactions between dye molecules, and thus results in the transition from monomer to H-type dimer forms of thionine [14]. This is the first report on the transition from monomer to H-type dimer forms of dyes based on metal nanoparticles. Furthermore, gold–dye assemblies coalesce to form larger clusters due to surface charge neutralization and hydrophobic interactions of the adsorbed dyes. This also facilitates the face-to-face arrangement of two thionine molecules, resulting in the transition from the monomer to Hdimer forms of thionine. Strong electronic coupling between the molecules in the formed dye aggregates causes the blue shift of the 565 nm band, which is characteristic of H-type dimer [44]. As the gold concentration is same for the two sets of particles, the smaller is the particle, the higher is the concentration of the particles, and the surface area is correspondingly higher. In the presence of small particles, dye molecules can induce intercluster interactions within the small particles, and such aggregates, which bring adsorbed dye molecules closer, facilitate dimer formation in the assembly of the dye molecules [48]. On the contrary, if the large-sized particles are used (above 20 nm), the aggregation amongst dye molecules is very weak, and thus, the gold particles with mean diameter of 18 and 5 nm were chosen in the present work. From the above analysis, it comes to a conclusion that the addition of small gold particles stimulates the transition from the monomer to dimer forms of thionine on the surface of gold particles. 3.2. Fluorescence quenching of thionine by gold nanoparticles with different sizes In Fig. 2 are shown the fluorescence spectra of thionine in aqueous solution in the presence of different concentrations of 15–20 nm gold nanoparticles. The fluorescence of thionine in water exhibits a maximum at around 619 nm, when it is excited at 550 nm (curve a in Fig. 2). With increasing concentration of gold
246
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
a quadratic relationship between relative intensity and quencher concentration is predicted [49]: I0
Fig. 3. Emission spectra of thionine in aqueous solution containing different concentrations of 3–6 nm gold nanoparticles at 25 ◦ C, thionine concentration: 1.20 × 10−5 mol/L; gold nanoparticle concentration (mol/L): a—0, b—2.56 × 10−6 , c— 5.12 × 10−6 , d—7.68 × 10−6 , e—1.02 × 10−5 , f—1.28 × 10−5 , g—1.54 × 10−5 , h— 1.79 × 10−5 , i—2.30 × 10−5 . The inset is the quenching curve for thionine by 3–6 nm gold nanoparticles in aqueous solution. λex = 550 nm.
nanoparticles, it can be seen from curves b–h in Fig. 2 that the fluorescence of thionine is gradually quenched by gold nanoparticles (2.56 × 10−6 –2.30 × 10−5 mol/L), and the emission peak is slightly blue-shifted. When the 3–6 nm particles are added to the same thionine solution, it is found that with the increasing concentration of gold particles (2.56 × 10−6 –2.30 × 10−5 mol/L), the fluorescence of thionine is significantly quenched by gold nanoparticles, and the quenching efficiency is almost 100% at the gold concentration of 2.30 × 10−5 mol/L (curve i in Fig. 3). Furthermore, the blue shift of the emission peak is increased with small particles added. This indicates that the fluorescence quenching of thionine by small particles is stronger than that by large ones. In addition, the blue shift of the emission peak and the absence of any new emission band in the emission spectra of thionine after the addition of gold particles suggest that the dye aggregates on gold nanoparticle surfaces are nonfluorescent, which is characteristic of H-type aggregates. The fluorescence quenching of thionine by gold nanoparticles with different sizes can be both static and dynamic. The former involves the formation of a nonfluorescent complex between the fluorophore and quencher in their ground states, whereas the latter requires diffusive encounters between the fluorophore and quencher during the lifetime of the excited state of the fluorophore. Assuming that the deactivation of the excited thionine occurs by collision with the quencher in fluid solution (pure dynamic quenching), the relative reduction in emission intensity is equal to the relative reduction in fluorescence lifetime, and the relationship between relative intensity and quencher concentration is linear [49]: I0 I
=
τ0 = 1 + K SV [ Q ], τ
(1)
where I 0 and I are the fluorescence intensities of thionine in absence and presence of quencher Q (gold nanoparticles), respectively; τ0 and τ are the fluorescence lifetimes of thionine in absence and presence of quencher, K SV is the Stern–Volmer quenching constant (L/mol), and [ Q ] is the concentration of the quencher (mol/L). However, when static quenching mechanisms occur through the complexing of the ground state fluorophore and quencher molecules, the relative reduction in emission intensity is greater than the relative reduction in fluorescence lifetime. In this case
= 1 + ( K SV + K eq )[ Q ] + K SV K eq [ Q ]2 , (2) I where K eq is the equilibrium constant for the binding of the fluorophore and quencher. This modified form of the Stern–Volmer equation accounts for an upward curvature of the Stern–Volmer plots. The quenching curve of thionine by large gold nanoparticles is nonlinear, as shown by the inset of Fig. 2, and changes upward with gold particles added, which is better fit by a quadratic function than a linear one. The nonlinearities in the quenching curve indicate that the fluorescence quenching is progressively enhanced with the addition of gold nanoparticles, and also indicate that both static and dynamic quenching occur for the same fluorophore, which is different from the static quenching of chlorophyll a by gold nanoparticles [35]. The upward curvature of the quenching curve in the inset of Fig. 2 also indicates that the binding extent of thionine on particle surfaces increases with the addition of gold particles, which is consistent with the fact that the H-type dimer peak becomes gradually clear with gold nanoparticles added (Fig. 1). Compared to that for large particles, the upward curvature of the quenching curve is more remarkable for small particles (inset of Fig. 3), the quenching value I 0 / I reaches almost 120 at the same gold concentration of 2.30 × 10−5 mol/L, indicating that the static quenching plays a major role in the fluoresce process of thionine by small particles. In addition, the increasing extent of blue shift of the emission peak is larger for small particles than for large particles, which confirms that small particles facilitate the transition from the monomer to dimer forms of thionine on the surface of gold particles, and is consistent with the UV–vis spectral results. For static quenching, a strong binding between thionine and gold nanoparticles provides favorable conditions for the heterogeneous deactivation processes at thionine–gold nanoparticles interface, which is responsible for the upward curvature of the Stern–Volmer plots of the insets of Figs. 2 and 3. 3.3. TEM images of gold sol Fig. 4 shows the TEM images and the corresponding size distribution of large gold nanoparticles before and after the addition of thionine. As can be seen in Fig. 4A, the prepared gold nanoparticles are almost spherical shape, and separated from each other. The particle size is mainly in the range of 15–20 nm in addition to some tiny particles. DLS data indicate that the particle size is centered on about 13 nm, and the calculated average diameter is 18.0 nm. It is noted that the difference of particle sizes between TEM and DLS comes from the different sizing method used by the manufacturers. TEM gives the number-average diameter, whereas DLS give a z-average diameter and also includes an aqueous hydration layer around the particle. After the addition of thionine, as shown in Fig. 4B, the gold particles also exhibit similar particle diameter, but some of gold nanoparticles form close-packing nanoclusters, which is induced by the adsorption of thionine on gold nanoparticle surfaces. DLS data in Fig. 4B indicate that there appear two size distribution peaks, which correspond to about 13 and 84 nm, respectively. The former peak with a smaller hydrodynamic diameter of about 13 nm can be attributed to single thionine-bound particles, and the latter one can be attributed to thionine-bound gold nanoclusters. In the case of gold nanoparticles of average 5 nm, the clustering morphological feature is significantly increased (Fig. 5B), and DLS data also show a new peak corresponding to a hydrodynamic diameter of about 45 nm (Fig. 5B), which suggests the formation of a small-sized gold nanocluster. Therefore, the formation of gold–dye composite particles and gold– dye nanoclusters can be controlled by gold particles of different
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
247
(A)
(B) Fig. 4. TEM images and corresponding particle size distribution of (A) the gold sol (15–20 nm) and (B) the gold sol with 1.20 × 10−5 mol/L thionine. D h denotes the hydrodynamic diameter.
sizes, which further supports the explanations for the above spectra changes. It is also noted that the nanoparticles in the gold nanoclusters are almost individually isolated, as can be observed from the TEM images of Figs. 4B and 5B. This is in contrast to the simple saltinduced aggregation of nanoparticles in which the nanoparticles are usually fused together due to the collapse of electrical double layers. In addition, the formation of the H-type dimer of dye thionine on the particle surfaces may contribute to bridging the neighboring nanoparticles. 3.4. Quantum chemical structure of thionine and model for thionine–Au complex To understand the significant changes of the adsorption behavior in presence of gold nanoparticles with different sizes, the binding of thionine on gold particles was first analyzed in terms of the thionine molecular structure. The nitrogen atoms of the NH2 moieties of thionine are usually electron-rich, which are likely to bind to the particles. The nitrogen and sulfur atoms of the central heterocyclic ring of thionine are less-electron rich due to delocalization of electrons to electron deficient phenyl ring, which can be supported by the following results of quantum chemical calculations (Table 1). It has been reported that the electron deficiency of atoms make them unlikely to bind directly to the nanoparticle surfaces [37]. Therefore, it can be inferred that the nitrogen atoms of the NH2 moieties of thionine strongly bind to gold nanoparticle surfaces. The binding of the amino groups to
the surfaces of gold nanoparticles has been used in fabricating the nanogold self-assembly through an amine-terminated monolayer of cystamine [50], or in achieving the organization of pyrene chromophores around gold nanoparticles through the surface binding of 1-methylaminopyrene (Py-CH2 NH2 , molecular probes) to gold nanoparticles [51]. To confirm the interaction sites of the thionine on gold nanoparticle surfaces, quantum chemical calculations were performed to obtain the optimized molecular structure of thionine and the model for thionine–Au complex (Fig. 6). The calculating results indicate that thionine molecules show planar structure, and at both of the NH2 ends of the molecule, the bond angles of H– N–H and C–N–H, e.g., H19 –N17 –H18 and C1 –N17 –H19 , are 117.70◦ and 121.00◦ , respectively, and the bond distances of N–H and C–N bonds, e.g. N17 –H19 and C1 –N17 , are 1.0408 and 1.3788 Å, respectively (Fig. 6A). From the natural charge of the relevant atoms of the molecule (Table 1), it can be known that the nitrogen atoms of the –NH2 moieties of the molecule, N17 and N21 , are most electron-rich, and easily offer their lone electron pairs to the electron deficient gold nanoparticles, which provides possibility of the binding of the –NH2 on the particle surfaces. Assuming that gold atoms and the nitrogen atoms of the –NH2 moieties of thionine form complex, the obtained data (Table 2) indicate that, at both of the NH2 ends of the molecule, the bond angle of H–N–H decreases from 117.70◦ to 107.84◦ , and that of C–N–H decreases from 121.00◦ to 111.01◦ ; whereas the bond distance of N–H bond increases from 1.0408 to 1.0544 Å, and that of C–N bond increases significantly from 1.3788 to 1.4602 Å. However, the bond angles of C–S–C and
248
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
(A)
(B) Fig. 5. TEM images and corresponding particle size distribution of (A) the gold sol (3–6 nm) and (B) the gold sol with 1.20 × 10−5 mol/L thionine. D h denotes the hydrodynamic diameter. Table 1 Natural charge of relevant atoms of the optimized thionine molecule at the DFT level of theory Symbol Natural charge
N17 −0.3215
N21 −0.3215
S26 −0.1396
N20 −0.1091
(A)
(B) Fig. 6. Optimized molecular structure of (A) thionine and (B) the modeled thionine– Au complex obtained from the DFT level of theory.
C–N–C, and the bond distances of C–S and C–N bonds at the central heterocyclic ring hardly change. On the other hand, assuming that gold atoms and the nitrogen or sulfur atom of the central heterocyclic ring of the thionine form complex, the bond angles of C–N–C and C–S–C, and the bond distances of C–N and C–S bonds of the ring are almost unchanged (data not shown), suggesting that the nitrogen and sulfur atoms of the central heterocyclic ring do not bind to the gold particle surfaces. The remarkable decrease of the bond angles of H–N–H and C–N–H, and the increase of the bond distances of N–H and C–N bonds at both of the NH2 ends of the molecule (Table 2) indicate that the –NH2 binds to the gold particle surfaces with the lone electron pair of the nitrogen atom pointing to the particle surface. This makes the N–H bonds deviate from the particle surface, and turn to the direction of C–N bond, and thus, weaken the N–H bonds. It is noted that the calculations of the interactions between a thionine molecule and a gold particle is very complex, as a particle contains many gold atoms (about 3859–180,072 gold atoms per particle). Considering that the binding of thionine on gold particles occurs at the gold atoms on the particle surfaces, two gold atoms, which were from a particle or from two particles respectively, were used to study the interactions of thionine with particles by the DFT method. This can approximately give specific interaction sites of thionine on particle surfaces. Such a calculation method has been reported in studying the adsorption behavior of biologically significant 2-aminobenzothiazole molecule on colloid silver particles by surface-enhanced Raman spectroscopy [52].
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
Table 2 Relevant structural parameters of thionine and the changes upon binding to Au at the DFT level of theory Thionine
Thionine–Au
Bond lengths (Å) 1.0544 1.4602 1.9071 1.4041
N17 –H19 C1 –N17 C2 –S26 C3 –N20
1.0408 1.3788 1.9055 1.3925
H19 –N17 –H18 C1 –N17 –H19 C2 –S26 –C7 C3 –N20 –C6
Interatomic angles (deg) 117.70 107.84 121.00 111.01 100.25 99.62 120.99 119.57
Diff.
−0.0136 −0.0814 −0.0016 −0.0116 9.86 9.99 0.63 1.42
249
Table 3 Surface structure parameters of gold nanoparticles with different sizes Particle size (nm)
N Au
α
N sur
18 5
180,072 3859
0.0597 0.2148
10,750 829
N Au —aggregation number of a gold particle; α —surface-to-volume atomic ratio of a gold particle; N Suf —the number of gold atoms on the surface of a gold particle.
number per gold particle (N Au ), the number of gold atoms per particle, was calculated using the formula [48]:
N Au = 59 nm−3
π ( D M ), 6
(3)
where D M is the mean diameter of the particles. And the surfaceto-volume atomic ratio (α ) of metal nanoparticles is defined by [48]:
α=
Fig. 7. FT-IR spectra of (a) free thionine and (b) the adsorbed thionine on gold nanoparticles.
To verify the interaction sites obtained by quantum chemical calculations, the FT-IR spectra of free thionine and the adsorbed thionine were examined (Fig. 7). The absorption bands at 1600 and 1500 cm−1 , as shown by the curve a in Fig. 7, are assigned to the skeletal vibration of the phenyl ring of thionine [53], which are also shown by the adsorbed thionine (curve b in Fig. 7), confirming the presence of thionine on the gold nanoparticle surfaces. The absorption bands at 3310 and 3150 are ascribed to the N–H stretching vibration of the amino moieties (curve a in Fig. 7) [53], which are not observed in the IR spectrum of the adsorbed thionine (curve b in Fig. 7), indicating that the nitrogen atoms of the NH2 moieties of thionine bind strongly to gold nanoparticle surfaces. However, as the absorption band of C–S bond in the range of 710–570 nm is weak and broad [53], it is meaningless to observe the difference between the IR spectrum of free thionine and that of the adsorbed thionine. On the curve b in Fig. 7, the absorption band at 3427 cm−1 is due to traces of water in the nanoparticle sample, and the absorption bands at 2930 and 2856 cm−1 are due to traces of citrate impurity in the sample. From the above analysis, it can be seen that the IR results are consistent with those obtained by quantum chemical calculations. Therefore, the interaction sites obtained by quantum chemical calculations and FT-IR spectroscopy confirm the binding mode of thionine on gold particles, and can be used for further surface structure analysis of gold particles of different sizes. 3.5. Surface structure analysis of gold particles The changes of adsorption behavior of thionine on gold nanoparticles with different sizes can be further analyzed in terms of surface structure parameters of gold particles. The aggregation
3d R
,
(4)
where d and R denote the atomic diameter and particle size, respectively. Assuming d = 3.58 Å for atomic gold, the obtained surface parameters are shown in Table 3. From Table 3, it can be seen that an 18 nm gold particle with 10,750 gold atoms on the surface can accommodate large number of dye molecules through the formation of multilayers of dye molecules which contain planar ring capable of forming dimers. The number of thionine molecules corresponding to an 18 nm gold particle is larger than that corresponding to a 5 nm particle as the number of large particles decrease in solution. Thus, higher thionine concentration per large gold nanoparticle exists in solution, and is expected to induce extensive dimerization. On the other hand, as the gold concentration is same for the two sets of particles, the smaller of the particle, the larger is the concentration of the particles, and the total surface area is correspondingly larger, although the number of surface gold atoms of a small particle is relatively small. The data in Table 3 clearly indicate that a small particle have a larger surface-to-volume atomic ratio, 0.2148, than a large one. As a result, the observed dye aggregation in the presence of small particles can be interpreted by considering that dye molecules can induce intercluster interactions within the small particles, and such aggregates, which bring adsorbed dye molecules closer, facilitate dimer formation in the assembly of the dye molecules. Furthermore, the molar surface area increases with decreasing particle size, then results in an increase in molar free energy of the small particles compared to large ones, and thus, the aggregation of particles is also the driving force for the strong dimerization of thionine molecules. TEM images indicate that the addition of thionine in the small particle system induces stronger clustering than in large ones, which confirm the above analysis. It should be noted that the preparation procedure of the two kinds of gold nanoparticles was different, and thus the surface coverage of citrate ions as well as the surface charge of the gold nanoparticles were also different. These factors also contribute to the difference in the adsorption behavior of thionine on gold nanoparticles with different sizes. However, as the binding of citrate ions on gold nanoparticles is relatively weak [45], the effects of these factors on the adsorption behavior of thionine are little. 4. Summary The size effect of gold particle on the adsorption behavior has been obtained by using various experimental techniques and quantum chemical calculations. The small particles stimulate the equilibrium between the monomer and H-dimer forms of thionine to move significantly toward the dimer forms. Due to stronger binding of thionine to small particles, the quenching of the dye by
250
Y. Ding et al. / Journal of Colloid and Interface Science 327 (2008) 243–250
small particles is more remarkable than that by large particles. The binding of thionine to gold particles results in the formation of gold nanoclusters with different sizes, and DLS indicates that the size of nanoclusters is 45 and 84 nm for the 5 and 18 nm particles, respectively. Quantum chemical calculation and IR results indicate that the nitrogen atoms of the NH2 moieties of thionine strongly bind to gold nanoparticle surfaces, and the number of surface gold atoms are 3859 and 180,072 for a 5 or 18 nm gold particle, respectively. The higher surface-to-volume atomic ratio and the higher surface free energy of small particles result in the significant adsorption behavior of thionine on the particles. The above results will provide important guidance for controlling the properties of organic–inorganic nanocompo site materials. Acknowledgment This work was supported by the National Natural Science Foundation of China (20633010). References [1] A. Henglein, J. Phys. Chem. 97 (1993) 5457. [2] P. Mulvaney, M. Giersig, A. Henglein, J. Phys. Chem. 97 (1993) 7061. [3] U. Kreibig, M. Vollmer, Optical Properties of Metal Clusters, Springer, Berlin, 1995. [4] L.M. Liz-Marzán, M. Giersig, P. Mulvaney, Langmuir 12 (1996) 4329. [5] R.G. Freeman, M.B. Hommer, K.C. Grabar, M.A. Jackson, M.J. Natan, J. Phys. Chem. 100 (1996) 718. [6] A. Martino, S.A. Yamanaka, J.S. Kawola, D.A. Loy, Chem. Mater. 9 (1997) 423. [7] G.L. Hornyak, C.J. Patrissi, C.R. Martin, J. Phys. Chem. B 101 (1997) 1548. [8] A. Henglein, Chem. Mater. 10 (1998) 444. [9] P. Xu, H. Yanagi, Chem. Mater. 11 (1999) 2626. [10] H. Behar-Levy, D. Avnir, Chem. Mater. 14 (2002) 1736. [11] V. Subramanian, E.E. Wolf, P.V. Kamat, J. Phys. Chem. B 107 (2003) 7479. [12] K. George Thomas, P.V. Kamat, Acc. Chem. Res. 36 (2003) 888. [13] A.R. Clapp, I.L. Medintz, B.R. Fisher, G.P. Anderson, H. Mattoussi, J. Am. Chem. Soc. 127 (2005) 1242. [14] N. Chandrasekharan, P.V. Kamat, J. Hu, G. Jonesll, J. Phys. Chem. B 104 (2000) 11103. [15] M.-C. Daniel, D. Astruc, Chem. Rev. 104 (2004) 293. [16] J. Hu, J. Zhang, F. Liu, K. Kittredge, J.K. Whitesell, M.A. Fox, J. Am. Chem. Soc. 123 (2001) 1464. [17] B.I. Ipe, K. George Thomas, S. Barazzouk, S. Hotchandani, P.V. Kamat, J. Phys. Chem. B 106 (2002) 18. [18] P.K. Sudeep, B.I. Ipe, K. George Thomas, M.V. George, S. Barazzouk, S. Hotchandani, P.V. Kamat, Nano Lett. 2 (2002) 29. [19] K. George Thomas, J. Zajicek, P.V. Kamat, Langmuir 18 (2002) 3722. [20] H.M. Eckenrode, S.-H. Jen, J. Han, A.-G. Yeh, H.-L. Dai, J. Phys. Chem. B 109 (2005) 4646. [21] N. Kometuni, M. Tsubonishi, T. Fujita, K. Asami, Y. Yonezawa, Langmuir 17 (2001) 578.
[22] S. Franzen, J.C.W. Folmer, W.R. Glomm, R. O’Neal, J. Phys. Chem. A 106 (2002) 6533. [23] J. Zhang, J.R. Lakowicz, J. Phys. Chem. B 119 (2005) 8701. [24] P.C. Lee, D. Meisel, J. Phys. Chem. 86 (1982) 3391. [25] F. Cannone, M. Collini, L. D’Alfonso, G. Baldini, G. Chirico, G. Tallarida, P. Pallavicini, Nano Lett. 7 (2007) 1070. [26] F. Cannone, G. Chirico, A.R. Bizzarri, S. Cannistraro, J. Phys. Chem. B 110 (2006) 16491. [27] J. Seelig, K. Leslie, A. Renn, S. Kuhn, V. Jacobsen, M. van de Corput, C. Wyman, V. Sandoghdar, Nano Lett. 7 (2007) 685. [28] T.L. Jennings, M.P. Singh, G.F. Strouse, J. Am. Chem. Soc. 128 (2006) 5462. [29] I.-I.S. Lim, F. Goroleski, D. Mott, N. Kariuki, W. Ip, J. Luo, C.J. Zhong, J. Phys. Chem. B 110 (2006) 6673. [30] B. Patrick, P.V. Kamat, J. Phys. Chem. 96 (1992) 1423. [31] G.D. Reid, D.J. Whittaker, M.A. Day, D.A. Turton, V. Kayser, J.M. Kelly, G.S. Beddard, J. Am. Chem. Soc. 124 (2002) 5518. [32] C. Dohno, E.D.A. Stemp, J.K. Barton, J. Am. Chem. Soc. 125 (2003) 9586. [33] V. Subramanian, E.E. Wolf, P.V. Kamat, J. Am. Chem. Soc. 126 (2004) 4943. [34] R. Shenhar, V.M. Rotello, Acc. Chem. Res. 36 (2003) 549. [35] S. Barazzouk, P.V. Kamat, S. Hotchandani, J. Phys. Chem. B 109 (2005) 716. [36] J. Turkevitch, P.C. Stevenson, J. Hillier, Discuss. Faraday Soc. 11 (1951) 55. [37] R.T. Tom, V. Suryanarayanan, P.G. Reddy, S. Baskaran, T. Pradeep, Langmuir 20 (2004) 1909. [38] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery Jr., R.E. Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.V. Ortiz, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe, P.M.W. Gill, B.G. Johnson, W. Chen, M.W. Wong, J.L. Andres, M. Head-Gordon, E.S. Replogle, J.A. Pople, Gaussian 98, revision A.7, Gaussian, Inc., Pittsburgh, PA, 1998. [39] A.D. Becke, Phys. Rev. A 38 (1988) 3098. [40] A.D. Becke, J. Chem. Phys. 98 (1993) 5648. [41] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785. [42] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 270. [43] P.J. Stevens, F.J. Devlin, C.F. Chabalowski, M.J. Frisch, J. Phys. Chem. 98 (1994) 11623. [44] S. Das, P.V. Kamat, J. Phys. Chem. B 103 (1999) 209. [45] C. Lee, S. Kim, C. Yoon, M. Gong, B.K. Choi, K. Kim, S.W. Joo, J. Colloid Interface Sci. 271 (2004) 41. [46] C.A. Bunton, L. Sepulveda, J. Phys. Chem. 83 (1979) 680. [47] M.A. Awan, S.S. Shah, Colloids Surf. A 122 (1997) 97. [48] S.K. Ghosh, A. Pal, S. Nath, S. Kundu, S. Panigrahi, T. Pal, Chem. Phys. Lett. 412 (2005) 5. [49] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Plenum Press, New York, 1983. [50] C.R. Raj, T. Okajima, T. Ohsaka, J. Electroanal. Chem. 543 (2003) 127. [51] K. George Thomas, P.V. Kamat, J. Am. Chem. Soc. 122 (2000) 2655. [52] J. Sarkar, J. Chowdhury, M. Ghosh, R. De, G.B. Talapatra, J. Phys. Chem. B 109 (2005) 12861. [53] E. Pretsch, P. Bühlmann, C. Affolter, Structure Determination of Organic Compounds: Tables of Spectral Data, third ed., Springer-Verlag, Berlin, 2000.