J. Phys. Chem. B 2007, 111, 6839-6844
6839
Ruthenium(II) Trisbipyridine Functionalized Gold Nanorods. Morphological Changes and Excited-State Interactions† Meghan Jebb,‡ P. K. Sudeep,‡ P. Pramod,§ K. George Thomas,*,§ and Prashant V. Kamat*,‡ Radiation Laboratory, Department of Chemistry and Biochemistry and Department of Chemical and Biomolecular Engineering, UniVersity of Notre Dame, Notre Dame, Indiana 46556-0579, and Photosciences and Photonics, Regional Research Laboratory (CSIR), TriVandrum 695 019, India ReceiVed: January 26, 2007; In Final Form: March 19, 2007
Gold nanorods synthesized using cetyltrimethylammonium bromide and tetraoctylammonium bromide as stabilizers are functionalized with a thiol derivative of ruthenium(II) trisbipyridyl complex [(Ru(bpy)32+-C5SH] in dodecanethiol using a place-exchange reaction. The changes in the plasmon absorption bands and transmission electron micrographs indicate significant changes in the gold rod morphology during the placeexchange reaction. The (Ru(bpy)32+-C5-SH in its excited state undergoes quick deactivation when bound to gold nanorods. More than 60% of the emission was quenched when [(Ru(bpy)32+-C5-SH] was bound to gold nanorods. Emission decay analysis indicates that the energy transfer rate constant is greater than 1010 s-1.
Introduction Tailoring nanomaterials for optoelectronic, sensor, and light energy conversion applications has emerged as one of the exciting areas of research in this decade.1-8 Modification of the surface of metal nanoparticles with fluorophores is important for designing biological sensors and optoelectronic devices.6,9-15 The binding of chromophores to the metal surface results in quenching of the excited state. Energy transfer and electron transfer are the main deactivation channels for the excited molecules on the metal surfaces.16-18 Most of the studies are concentrated on the spherical gold nanoparticles functionalized with chromophores (see, for example, refs 19-28). There have been several efforts to functionalize gold nanoparticles with the thiol derivatives of ruthenium trisbipyridine (Ru(bpy)32+).25,28-35 Both energy transfer and electron transfer mechanisms have been proposed in these studies to explain the quenching of the excited state (see Scheme 1). Efforts have also been made to utilize these hybrid systems for harvesting light energy.35 In a recent study,36 we reported photophysical properties of Ru(bpy)32+ chromophores on the surface of gold nanoparticles. The excited-state deactivation was modulated by varying the density of chromophores. When in close proximity, the chromophores on the periphery of the gold core undergo electron transfer reaction. The electron transfer products formed during the interaction between excited- and ground-state Ru(bpy)32+ sustained for several nanoseconds when the chromophore was bound to the gold surface. Higher solubility of gold nanorods in aqueous media and the tunability of their longitudinal plasmon absorption band from the visible to the infrared region makes them useful in the design of biological sensors.37-39 In addition, the large number of binding sites makes nanorod structure an attractive candidate for designing light-harvesting systems by manipulating the surface concentration of chromophores. The lower solubility of †
Part of the special issue “Norman Sutin Festschrift”. * Corresponding authors. E-mail:
[email protected] (P.V.K.);
[email protected] (K.G.T.). ‡ University of Notre Dame. § Regional Research Laboratory.
SCHEME 1: Excited-State Interactions between Thiol Derivatized Ruthenium(II) Trisbipyridine and Gold Nanorod
the gold nanorods in organic solvents poses a major problem to directly functionalize thiol derivatives in organic media. To overcome this problem, we first synthesized the gold nanorods in aqueous surfactant solution adopting a photochemical preparation method,40 and then transferred the nanorods into dodecanethiol by mixing the aqueous solution in the presence of a small amount of acetone. This procedure allowed us to functionalize the gold nanorods with the thiol derivative of Ru(bpy)32+ complex. We report herein the functionalization of gold nanorods with a thiol derivative of ruthenium(II) trisbipyridine to study the excited-state interactions of the hybrid material. Experimental Section Materials and Methods. Cetyltrimethylammonium bromide, tetraoctylammonium bromide, silver nitrate, and hydrogen tetrachloroaurate(III) trihydrate, were purchased from SigmaAldrich and used as received. Dodecanethiol, methylene chloride, and acetonitrile were obtained from Fisher Chemicals and used as received. Acetone was obtained from VWR International
10.1021/jp070701j CCC: $37.00 © 2007 American Chemical Society Published on Web 05/04/2007
6840 J. Phys. Chem. B, Vol. 111, No. 24, 2007 and used as received. The thiol derivative bis(2,2′-bipyridine) (5-(4′-methyl-2,2′-bipyridin-4-yl)pentane-1-thiol)ruthenium(II) [(Ru(bpy)32+-C5-SH] was synthesized using the method described earlier.36 Optical Measurements. The absorption spectra were recorded using a Varian CARY50 Bio UV-vis spectrophotometer. Emission spectra were recorded using an SLM 8000 photon counting spectrofluorimeter. Emission lifetimes were measured using a Horiba Jobin Yvon single photon counting system, and the fluorescence decay measurements were further analyzed using the IBH software library. Laser Flash Photolysis. Nanosecond laser flash photolysis experiments were performed with a 355 nm laser pulse (5 mJ, pulse width 6 ns) from a Quanta Ray Nd:YAG laser system. Kinetic traces at appropriate wavelengths were assembled from the time-resolved data. All measurements were conducted at room temperature. The experiments were performed in a 6 mm quartz cell, and all the solutions were deaerated with high-purity nitrogen. Synthesis of Gold Nanorods. Gold nanorods were prepared by adopting a photochemical method that employs UV irradiation to facilitate slow growth of rods.40 We used tetraoctylammonium bromide as a cosurfactant instead of tetradodecylammonium bromide. The growth solution was prepared by dissolving 440 mg of cetyltrimethylammonium bromide (CTAB) and 4.5 mg of tetraoctylammonium bromide (TOAB) in 15 mL of water and transferring it to a cylindrical quartz tube (length 15 cm and diameter 2 cm). To this solution 1.25 mL of 0.024 M HAuCl4 solution was added along with 325 µL of acetone and 225 µL of cyclohexane. A small amount of AgNO3 (250 µL of 0.01 M) was also introduced. It is reported that the presence of AgNO3 is essential for the growth of the Au nanorod and the length:diameter aspect ratio can be varied by varying the AgNO3 concentration.41 The quartz tube was closed with a rubber stopper through which a glass rod was inserted (15 cm length and 1 cm diameter). The glass rod helps to reduce the effective thickness of the solution and facilitates uniform absorption of the light through a thin solution layer. The photochemical reaction was carried out using 300 nm irradiation in a Rayonet Photochemical Reactor for 18 h. Gold nanorods prepared by the photochemical method were first purified by centrifugation. The residue obtained after 10 min of centrifugation (7000 rpm) was dispersed in 2 mL of 0.7 M CTAB solution and kept undisturbed at 50 °C for 12 h. Upon cooling, excess CTAB crystallized and was separated by filtration. The filtrate contains monodisperse Au nanorods and was used directly in subsequent studies. The formation of a gold nanorod and its aspect ratio was confirmed from transmission electron microscopic analysis. A drop of a dilute solution of Au nanorods was allowed to dry on a carbon coated copper grid and then probed using a JEOL JEM100sx electron microscope. The average length and diameter of rods employed in the present investigation are 50.0 and 20.0 nm, respectively, with an average aspect ratio of 2.5. Surface Functionalization of Au Rods with Ru(bpy)32+C5-SH. The place-exchange method has been widely employed to functionalize gold nanoparticles with thiol derivatives of organic molecules.42,43 A modified place-exchange process was developed to link Ru(bpy)32+-C5-SH to gold nanorods. Two methods were found to be successful in binding the Ru(bpy)32+C5-SH to the gold rods. In the first method, gold nanorods in aqueous solution were added to a solution of Ru(bpy)32+-C5SH dissolved in acetonitrile. The mixture was reacted at room temperature with constant stirring for 24 h, and periodic
Jebb et al. sonication. It was then washed three to five times with dichloromethane and supernatant was discarded to remove unbound dye molecules. The product was then resuspended in water. The UV-visible absorption of the suspension was measured to ensure the presence of absorption corresponding to the bound Ru(bpy)32+-C5-SH. In the second method the gold nanorods from water were first extracted into dodecanethiol/acetone. A mixture of 2:3:4 water:dodecanethiol:acetone ratio was able to transfer the nanorods into the organic layer. Formation of covalent bonds with the thiol group facilitated phase transfer of gold nanorods. It took approximately 5 h for the nanorods to migrate from water to the dodecanethiol layer. The upper organic layer was extracted and added to sodium sulfate to remove any residual water. The Au nanorod solution was then added to Ru(bpy)32+-C5-SH previously dissolved in a drop of acetone. The mixture was reacted for 24 h at room temperature with constant stirring and periodic sonication. To purify, the reaction mixture was washed with dichloromethane centrifuged at 5000 rpm for 3 min to remove unbound Ru(bpy)32+-C5-SH. This was repeated three to five times, until the supernatant removed after centrifugation did not show any unbound dye. The product could then be resuspended in either water or dichloromethane. (Note that binding with Ru(bpy)32+-C5-SH facilitates the dissolution of functionalized gold nanorods in nonpolar solvents.) The binding of Ru(bpy)32+-C5-SH to the surface of Au nanorods (Scheme 2) was confirmed by recording the UV-visible absorption spectrum. Most of the experiments reported here were carried out using this second synthetic approach. As presented in the Supporting Information, the maximum number of thiol molecules that can be accommodated on a single gold rod is ∼23 800. On the basis of the absorbance values, we expect ∼4000 molecules are exchanged with Ru-C5-SH molecules. Results and Discussion Interaction between Ru(bpy)32+-C5-SH and Au Nanorods. The gold nanorods prepared by the photochemical method exhibit transverse and longitudinal plasmon absorption bands at 542 and 690 nm, respectively. The absorption spectrum of surfactant stabilized gold nanorods in water is shown in Figure 1A (trace a). As shown earlier,44-46 these two absorption bands correspond to the gold nanorods having an aspect ratio (length to diameter) of approximately 2.5. When they are suspended in dodecanethiol, we observe dampening of both longitudinal and transverse plasmon absorption bands. In addition, a small red shift (20-30 nm) in both absorption peaks can also be seen in spectrum b (Figure 1A). The interaction of gold nanorods with dodecanethiol results in dampening and broadening of the plasmon absorption bands. Earlier studies of interaction of gold particles with thiols have established the response of plasmon absorption to surface modifications.47,48 Despite the broadening effect and red shift in the absorption, the retention of both longitudinal and transverse plasmon absorption features confirmed the absence of morphological changes upon their binding to dodecanethiol. (As will be seen in the next section, we have obtained independent confirmation using transmission electron microscopy.) The gold nanorods suspended in dodecanethiol were then subjected to place-exchange reaction with Ru(bpy)32+-C5-SH using the methods described in the Experimental Section. The Ru(bpy)32+-C5-SH functionalized gold nanorods that were separated by centrifugation were purified by washing with dichloromethane. Absorption spectra of the gold nanorods before
Ru(II)trisbipyridine Functionalized Gold Nanorods
J. Phys. Chem. B, Vol. 111, No. 24, 2007 6841
SCHEME 2: Synthetic Steps of Gold Nanorod Functionalizationa
a
CTAB/TOAB surfactant capping is first exchanged with n-dodecanethiol followed by its reaction with Ru(bpy)32+-C5-SH.
Figure 1. (A, left) Absorption spectrum of Au nanorods in (a) water and (b) dodecanethiol. (B, right) Absorption spectra of (a) Au nanorods in dodecanethiol, (b) Ru(bpy)32+-C5-SH in dichloromethane, and (c) Au rod-Ru(II)thiol dye in dichloromethane after purification.
and after functionalization are shown in Figure 1B. The appearance of the 460 nm band confirms the binding of Ru(bpy)32+-C5-SH to gold nanorods. On the basis of the absorbance values in Figure 4B, we estimate ∼4000 Ru(bpy)32+-C5-SH molecules are exchanged with dodecanethiol. The final coverage has an approximate ratio of 1:6 for Ru(bpy)32+-C5-SH:dodecanethiol (see Supporting Information). The binding of Ru(bpy)32+-C5-SH also affects the plasmon absorption band. The characteristic transverse and longitudinal plasmon absorption bands of rods merge to form a single broad absorption band with a maximum around 590 nm. Earlier studies on the interaction of gold nanoparticles with thiols have shown
that the longitudinal band is sensitive to the edge-to-edge interactions among gold rods.41 For example, these studies indicated a red shift of the longitudinal plasmon absorption of gold nanorods as the rods were assembled linearly through the hydrogen bonding. The change in the plasmon absorption band during the self-exchange reaction is likely to arise from the surface interactions or loss of plasmon characteristics of rods because of morphological changes. We followed the binding of Ru(bpy)32+-C5-SH to the surface gold rods by monitoring the absorption spectral changes. A few selected spectra recorded during this period of place-exchange reaction are shown in Figure 2. It is evident that during the
6842 J. Phys. Chem. B, Vol. 111, No. 24, 2007
Jebb et al.
Figure 2. Absorption changes of gold nanorods in dodecanethiol. The spectra were recorded in the (A) presence and (B) absence of Ru(bpy)32+C5-SH dye at different intervals of time.
Figure 3. (A) TEM image of CTAB/TOAB stabilized Au nanorods in water. (B) TEM image of Au nanorods in dodecanethiol. (C) TEM image of Au nanorods after binding with Ru thiol and purification.
first 3 h there is a decrease in the near-infrared absorption (region where longitudinal plasmon absorption dominates). Over a 29-h period, we observed the disappearance of the longitudinal plasmon absorption band and broadening of the transverse band with a small red shift (Figure 2A). These spectral profiles show that the absorption changes were gradual. This further supports the argument that the absorption changes were evolving as the place-exchange reaction proceeded to completion. As a control experiment, the absorption spectra of the gold rods in dodecanethiol were also recorded during the same period. (Figure 2B). Only a small decrease (∼10%) in the overall absorption is seen. The spectra recorded during the reaction time scale retained the feature of dual absorption peaks corresponding to transverse and longitudinal peaks. This control experiment further ensures that dodecanethiol itself has little or no effect on the morphology of the gold nanorods over a long period of time. On the basis of these spectral measurements, we can conclude that the appearance of a single absorption band in Figure 2A arises from the binding Ru(bpy)32+-C5-SH dye. If indeed the absorption changes seen during the placeexchange reaction with Ru(bpy)32+-C5-SH were due to morphological changes, we should be able to characterize them through transmission electron microscopy (TEM). The TEM images of the gold nanorods in water and dodecanethiol are shown in parts A and B, respectively, of Figure 3. The TEM image of the gold nanorods obtained after the place-exchange reaction with Ru(bpy)32+-C5-SH and purification is shown in Figure 3C. The gold nanorods in dodecanethiol are well separated and exhibit minimal changes in shape and size as a result of thiol interactions. In contrast, the TEM image obtained with the Ru(bpy)32+-C5-SH-bound gold nanorod sample shows aggregated rods with distorted structures. This indicates that the surface modification of the gold nanorods with Ru(bpy)3C5-SH leads to the destruction of their protective capping layer. Such an alteration of the surface further leads to the aggregation effects. As observed in thiol functionalized gold nanoparticles,49 we expect etching of the surface by Ru(bpy)32+-C5-SH during
the place-exchange reaction. The addition of electrons to gold nanorods with aspect ratios ranging from 2 to 4 has also been found to change the morphology as a result of surface charging effects.50 The redox nature of the Ru(bpy)32+ moiety is likely to make the thiol interaction more reactive and induces surface corrosion. The ordered structure of gold nanorods seen in dodecanethiol thus becomes distorted as the surfactant capping layer from the surface is removed. This in turn increases the probability of the aggregation of rods. Excited-State Interactions. Although they were distorted morphologically, we found that the Ru(bpy)32+-C5-SH modified gold nanorods were photochemically active. To assess the excited-state interaction between gold rods and chromophore, we monitored the emission behavior of surface-bound Ru(bpy)32+-C5-SH. The strong emission of the dye at 630 nm (trace a in Figure 4A) is quenched after binding to the surface of the gold nanorods (trace b in Figure 4A). Since gold rods also exhibit absorption at the excitation wavelength (∼50% at 460 nm), care was taken to account for this absorption while recording the emission spectra. The decreased fluorescence yield suggests that a large fraction of the excited dye undergoes deactivation as it interacts with gold nanorods. In our earlier studies we have shown that chromophores bound to gold nanoparticles undergo excited-state deactivation by an energy or electron transfer process.2,5,21,51 The excited Ru(bpy)32+-C5-SH dye emission at 630 nm was further analyzed by comparing the emission decay of the unbound dye (Figure 4B) with that of the bound dye. The dye alone shows monoexponential fluorescence decay with a lifetime of 0.4 µs in methylene chloride. The emission decay of the dyebound gold nanorods exhibited an additional fast deactivation pathway. The emission decay was analyzed using a biexponential kinetic fit. The fast component decays within the laser pulse duration of 100 ps. The longer component had a lifetime similar to that of unbound dye (viz., 0.4 µs). The fast emission decay component of the Ru(bpy)32+-C5-SH bound to Au nanorods parallels the excited-state quenching seen in the emission spectra (Figure 4A). The slower decay component, on the other hand, represents the unquenched component of the bound Ru(bpy)32+C5-SH. As discussed in earlier studies,5,8,19 smaller metal particles cannot quench all the bound chromophores. Furthermore, the excited-state interactions can be modulated by charging of metal particles.52,53 On the basis of the initial fast decay component and emission yield, we estimate that more than 60% of the bound dye Ru(bpy)32+-C5-SH is able to interact with gold rods from its excited state. The obvious question in the present case is whether the interaction between excited Ru(bpy)32+-C5-SH and gold nano-
Ru(II)trisbipyridine Functionalized Gold Nanorods
J. Phys. Chem. B, Vol. 111, No. 24, 2007 6843
Figure 4. (A) Emission of Ru(bpy)32+-C5-SH (a) unbound and (b) bound to gold nanorods in methylene chloride. Excitation was at 460 nm. (B) Fluorescence lifetime profiles of Ru(bpy)32+-C5-SH in the (a) unbound and (b) bound states.
Figure 5. Transient absorption spectra of Ru(bpy)32+-C5-SH in methylene chloride recorded 200 ns after 355 nm laser pulse excitation (a) before and (b) after binding to gold nanorods (inset) Decay of triplet excited state at 380 nm: (a) unbound Ru(bpy)3-C5-SH; (b) Ru(bpy)32+C5-SH bound to gold rods.
rods involves energy transfer or electron transfer. To probe these processes, we recorded transient absorption spectra immediately following 355 nm laser pulse excitation (Figure 5). Ru(bpy)32+C5-SH shows characteristic triplet-triplet (T-T) absorption with a difference absorption maximum at 380 nm and ground-state beaching at 460 nm. (Note that the negative absorption in the 550-700 nm region arises from the prompt emission of the excited Ru(bpy)32+-C5-SH. The magnitude of the emission decreases when bound to gold nanorods.) The decay behavior (lifetime of 0.4 µs) parallels the emission decay, thus confirming the identity of the triplet excited state. When the suspension of gold nanorods functionalized with Ru(bpy)32+-C5-SH was excited with a 355 nm laser pulse, we observed transient absorption spectra similar to that of the triplet excited state. Lower absorption values of spectrum b in Figure 5 confirmed the quenching of the excited state within the laser pulse duration. The unquenched component contributed to the weak absorption. We carefully investigated for the possible formation of electron transfer products. Neither the reduced (Ru(bpy)3+) nor oxidized (Ru(bpy)33+) form of the chromophore can be seen during the transient decay. The failure to observe electron transfer products indicates that electron transfer is not a major pathway in the deactivation of the excited state. On the basis of the absorption and emission studies, we can infer that the excited-state deactivation pathway for the Ru(bpy)32+-C5-SH bound to gold nanorods is dominated by an energy transfer process. Compari-
son of the emission lifetimes in Figure 4B suggests that the rate constant for energy transfer is greater than 1010 s-1. According to Dulkeith et al.,19 the quenching of the surfacebound chromophore is caused not only by an increased nonradiative rate but also by a drastic decrease in the dye’s radiative rate. The size-dependent fluorescence decay analysis indicated that the radiative rate exhibits a pronounced minimum for gold particles of around 8 nm diameter. The deactivation of the excited chromophore can occur via energy transfer or electron transfer processes. When the gold particles are sufficiently small (
