Arylthiolate-Protected Silver Quantum Dots - ACS Publications

Oct 28, 2006 - Matthew R. Branham, Alicia D. Douglas, Allan J. Mills,† Joseph B. Tracy,. Peter S. White, and Royce W. Murray*. Kenan Laboratories of...
0 downloads 0 Views 179KB Size
Download PDF
11376

Langmuir 2006, 22, 11376-11383

Arylthiolate-Protected Silver Quantum Dots Matthew R. Branham, Alicia D. Douglas, Allan J. Mills,† Joseph B. Tracy, Peter S. White, and Royce W. Murray* Kenan Laboratories of Chemistry, UniVersity of North Carolina, Chapel Hill, North Carolina 27599-3290 ReceiVed August 7, 2006. In Final Form: September 19, 2006 This paper describes a new, organic-soluble 4-tert-butylbenzyl mercaptan (BBT) monolayer-protected silver cluster (AgBBT MPC) as the first example of a dissolved silver nanoparticle that exhibits quantized one-electron double layer charging (QDL) voltammetry. Polydisperse AgBBT MPCs made by two different synthetic protocols, but with similar average core diameters (2.1 nm), exhibit sharply differing electrochemistry and optical absorbance spectra. A twophase procedure (organic/aqueous, termed Prep A-AgBBT) produced MPCs exhibiting a 475 nm surface plasmon absorbance and QDL voltammetry. Neither property was seen for MPCs made by a single-phase procedure, termed Prep B-AgBBT. The difference is thought to reflect poor passivation to oxide formation in the latter Prep B procedure, which is supported by X-ray photoelectron spectroscopy results. Thermogravimetry, mass spectra, and electrochemistry results suggest an aVerage stoichiometric formula of Ag140BBT53, but transmission electron microscopy shows that the products are also polydisperse and include polycrystalline aggregates. Dry, cast films of both Ag MPC preparations on interdigitated array electrodes exhibit low electron hopping conductivity, compared to Au MPCs.

Introduction Interest in applications of nanometer-sized clusters of metal atoms is broadly based and includes biolabeling,1 luminescent tagging,2 catalysis,3 sensing,4 and release of chemical reagents in biological systems.5 Basic aspects of metal nanoparticles include the 1-3 nm diameter regime, where bulk properties give way to molecular behavior. Synthetic work by Brust et al.6 led to thiolate ligand-stabilized clusters with Au cores in that size range and catalyzed attention to the quantum dot properties of these monolayer-protected clusters (MPCs), notably quantized single-electron double layer (QDL) charging7 and the appearance of a molecule-like band gap.8 Core size-dependent properties are also influenced by the thiolate ligand chemistry.9 The present report concerns Ag MPCs, for which the substantial literature includes two-phase6 and micellar procedures,10 and protecting monolayers ranging from surfactants such as oleate,11 stearate,12 trioctylphosphine oxide,13 and poly(acrylic acid)14 to * Corresponding author. E-mail: [email protected]. † Current address: Department of Chemistry and Centre for Nanoscale Science, University of Liverpool, Liverpool, L69 7ZD, UK. (1) (a) Aubin, M. E.; Morales, D. G.; Hamad-Schifferli, K. Nano Lett. 2005, 5, 519. (b) Grancharov, S. G.; Zeng, H.; Sun, S. H.; Wang, S. X.; O’Brien, S.; Murray, C. B.; Kirtley, J. R.; Held, G. A. J. Phys. Chem. B 2005, 109, 13030. (2) (a) Li, Z. F.; Ruckenstein, E. Nano Lett. 2004, 4, 1463. (b) Gao, X. H.; Nie, S. M. Anal. Chem. 2004, 76, 2406. (3) Eklund, S. E.; Cliffel, D. E. Langmuir 2004, 20, 6012. (4) (a) Malinsky, M. D.; Kelly, K. L.; Schatz, G. C.; Van Duyne, R. P. J. Am. Chem. Soc. 2001, 123, 1471. (b) McFarland, A. D.; Van Duyne, R. P. Nano Lett. 2003, 3, 1057. (c) Riboh, J. C.; Haes, A. J.; McFarland, A. D.; Yonzon, C. R.; Van Duyne, R. P. J. Phys. Chem. B 2003, 107, 1772. (5) Rothrock, A. R.; Donkers, R. L.; Schoenfisch, M. H. J. Am. Chem. Soc. 2005, 127, 9362. (6) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. J. Chem. Soc. Chem. Commun. 1994, 801. (7) Hicks, J. F.; Miles, D. T.; Murray, R. W. J. Am. Chem. Soc. 2002, 124, 13322. (8) (a) Donkers, R. L.; Lee, D.; Murray, R. W. Langmuir 2004, 20, 1945. (b) Jimenez, V. L.; Georganopoulou, D. G.; White, R. J.; Harper, A. S.; Mills, A. J.; Lee, D. I.; Murray, R. W. Langmuir 2004, 20, 6864. (9) (a) Guo, R.; Murray, R. W. J. Am. Chem. Soc. 2005, 127, 12140. (b) Song, Y.; Jimenez, V.; McKinney, C.; Donkers, R.; Murray, R. W. Anal. Chem. 2003, 75, 5088. (10) (a) Taleb, A.; Petit, C.; Pileni, M. P. Chem. Mater. 1997, 9, 950. (b) Pileni, M. P. Langmuir 1997, 13, 3266. (c) Zhang, Z. Q.; Patel, R. C.; Kothari, R.; Johnson, C. P.; Friberg, S. E.; Aikens, P. A. J. Phys. Chem. B 2000, 104, 1176. (11) Wang, W.; Efrima, S.; Regev, O. Langmuir 1998, 14, 602.

smaller ligands such as trimethyl(mercaptoundecyl)ammonium,15 mercaptosuccinic16 and mercaptoacetic17 acids, and alkanethiols.18 The resulting MPCs are polydisperse with core sizes ranging from 2 to 40 nm in diameter. Ag MPCs can be used as substrates for surface-enhanced Raman responses17,19 and can be formed into conductive printed films.20 Large silver MPCs synthesized have included 50-100 nm wide triangles fashioned by4a a nanolithographic technique and then thiolated and dislodged into solution.21 Large Ag nanoparticles can also be formed in biological matrixes,22 and large (∼8 nm) but quite monodisperse Ag nanoparticles have been synthesized in a continuous flow microreactor.23 Metal quantum dot phenomena (single-electron charging) have been observed almost exclusively with Au MPCs,7-9 largely owing to their good stabilityseven at very small sizesswhich facilitates the typical necessary steps to isolate size fractions from polydisperse synthetic products. There is increasing activity in preparing small MPCs from other metals such as Pt,3,24 Pd,15,24a,25 Cu,26 and Ag,16,18,27 but producing stable MPCs of these other metals that are monodisperse enough to observe (12) Aoki, K.; Chen, J. Y.; Yang, N. J.; Nagasawa, H. Langmuir 2003, 19, 9904. (13) Saponjic, Z. V.; Csencsits, R.; Rajh, T.; Dimitrijevic, N. M. Chem. Mater. 2003, 15, 4521. (14) Ung, T.; Giersig, M.; Dunstan, D.; Mulvaney, P. Langmuir 1997, 13, 1773. (15) Cliffel, D. E.; Zamborini, F. P.; Gross, S. M.; Murray, R. W. Langmuir 2000, 16, 9699. (16) Chen, S. H.; Kimura, K. Chem. Lett. 1999, 1169. (17) Li, X. L.; Zhang, J. H.; Xu, W. Q.; Jia, H. Y.; Wang, X.; Yang, B.; Zhao, B.; Li, B. F.; Ozaki, Y. Langmuir 2003, 19, 4285. (18) Shon, Y. S.; Cutler, E. Langmuir 2004, 20, 6626. (19) (a) Lu, Y.; Liu, G. L.; Lee, L. P. Nano Lett. 2005, 5, 5. (b) Emery, S. R.; Haskins, W. E.; Nie, S. M. J. Am. Chem. Soc. 1998, 120, 8009. (20) Li, Y. N.; Wu, Y. L.; Ong, B. S. J. Am. Chem. Soc. 2005, 127, 3266. (21) Haes, A. J.; Zhao, J.; Zou, S. L.; Own, C. S.; Marks, L. D.; Schatz, G. C.; Van Duyne, R. P. J. Phys. Chem. B 2005, 109, 11158. (22) Gardea-Torresdey, J. L.; Gomez, E.; Peralta-Videa, J. R.; Parsons, J. G.; Troiani, H.; Jose-Yacaman, M. Langmuir 2003, 19, 1357. (23) Xue, Z. L.; Terepka, A. D.; Hong, Y. Nano Lett. 2004, 4, 2227. (24) (a) Tan, Y. W.; Dai, X. H.; Li, Y. F.; Zhu, D. B. J. Mater. Chem. 2003, 13, 1069. (b) Kim, K. S.; Demberelnyamba, D.; Lee, H. Langmuir 2004, 20, 556. (c) Yang, J.; Deivaraj, T. C.; Too, H. P.; Lee, J. Y. J. Phys. Chem. B 2004, 108, 2181. (25) Zamborini, F. P.; Gross, S. M.; Murray, R. W. Langmuir 2001, 17, 481. (26) Chen, S. W.; Sommers, J. M. J. Phys. Chem. B 2001, 105, 8816.

10.1021/la062329p CCC: $33.50 © 2006 American Chemical Society Published on Web 10/28/2006

Arylthiolate-Protected Ag Quantum Dots

quantum dot properties has been less successful. QDL charging of other metal MPCs has been seen for dissolved Pd25 and Cu26 MPCs, in the latter case decaying when the Cu surface became oxidized. QDL of Ag MPCs has been reported28 in only one instance, for surfactant-protected, surface-attached, 3.3 nm Ag nanoparticles. The present report is the first to demonstrate a quantized one-electron capacitor effect for dissolved, diffusing Ag MPCs; it also exposes the negative effects of Ag oxide formation on the QDL property. This laboratory has experimented extensively with Au MPCs protected by alkyl- and arylthiolate monolayers, including MPCs in which the metal core is thought to contain about 140 Au atoms (i.e., Au140). Judging from its QDL properties,7 this MPC does not exhibit a molecular HOMO-LUMO energy gap. Another purpose of the present work has been to identify Ag MPCs that are analogues to those of Au in order to compare their physical and chemical properties and reactivities. Two different synthetic protocols were used to prepare Ag MPCs. The preparation from which QDL properties successfully result is a two-phase method,6 termed Prep A-AgBBT MPCs. The second protocol, producing Prep B-AgBBT MPCs and derived from a procedure reported by Murthy et al.,27b is performed in an ethanolic phase. No QDL was seen for the Prep B-AgBBT MPCs, apparently because of significant silver core surface oxidation. In both cases, although numerous different thiols were tested, small size and reasonable stability were best afforded by the approximately cone-shaped ligand 4-tert-butylbenzyl mercaptan (BBT). The size and composition of the MPCs was assessed by UV-vis spectroscopy, transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), thermogravimetric analysis (TGA), and laser desorption-ionization timeof-flight mass spectrometry (LDI-TOF-MS). High-pressure liquid chromatography (HPLC) with UV-vis detection was used in the further study of MPC optical spectra. While both preparations give polydisperse Ag core sizes, the core size that leads to the QDL voltammetry and that is indicated by the TGA results, corresponds to the aVerage formula, Ag140BBT53. Experimental Section Chemicals. Silver nitrate (99.9999%), tetraoctylammonium bromide (Oct4NBr), tetrabutylammonium perchlorate (Bu4NClO4), sodium borohydride, BBT, ethanol, methanol, dichloromethane, and n-heptane obtained from commercial sources were used as received. AgBBT MPC Synthesis. Prep A-AgBBT MPCs were synthesized using a modified version29 of the Brust6 protocol. A mixture of solutions of 1.7 g of AgNO3 in 100 mL of H2O and 5.5 g of Oct4NBr in 100 mL of toluene was magnetically vigorously stirred for 30 min, forming a cloudy cream-colored suspension of (presumably) AgBr particles. A 5.6 mL portion of the thiol BBT was added, and the solution was stirred for another 30 min in an ice bath, after which a solution of 3.8 g of NaBH4 in 100 mL of cold H2O (degassed with Ar for 30 min in an ice bath) was added quickly with vigorous stirring and then stirred under Ar for 1.5 h on an ice bath. The clear aqueous phase was separated and discarded and the toluene phase was dried by rotary evaporation at 25 °C. The resulting black solid was covered with methanol and, after standing overnight in darkness, was collected on a medium porosity fritted glass filter and washed copiously with methanol. The procedure typically yields approximately 1.8 g of AgBBT MPCs in the form of a fine black (27) (a) Huang, T.; Murray, R. W. J. Phys. Chem. B 2003, 107, 7434. (b) Murthy, S.; Bigioni, T. P.; Wang, Z. L.; Khoury, J. T.; Whetten, R. L. Mater. Lett. 1997, 30, 321. (28) Cheng, W. L.; Dong, S. J.; Wang, E. K. Electrochem. Commun. 2002, 4, 412. (29) Hostetler, M. J.; Wingate, J. E.; Zhong, C. J.; Harris, J. E.; Vachet, R. W.; Clark, M. R.; Londono, J. D.; Green, S. J.; Stokes, J. J.; Wignall, G. D.; Glish, G. L.; Porter, M. D.; Evans, N. D.; Murray, R. W. Langmuir 1998, 14, 17.

Langmuir, Vol. 22, No. 26, 2006 11377 powder, which was determined by 1H NMR in CD2Cl2 to be free of excess thiol by the absence of sharp resonances thereof. The procedure for Prep B-AgBBT MPCs is given in the Supporting Information. TEM. TEM images (510 000×) were obtained with a side-entry Phillips CM12 electron microscope operating at 100 kV and analyzed using Image J 1.34s software (available publicly online). Highresolution TEM (HR-TEM) images were obtained at 700 000× magnification with a Hitachi HF-2000 electron microscope operating at 200 kV. Samples were prepared by casting and evaporating a droplet of MPC solution (2-3 mg/mL CH2Cl2) onto Formvar-coated Cu grids (400 mesh, Electron Microscopy Sciences). HPLC. HPLC was done with a Waters 600 controller pump, a Rheodyne 7725 injection valve (50 µL sample loop), and a Waters 996 PDA photodiode array detector, using two stainless steel columns (BDS Hypersil C8 stationary phase 250 × 4.6 mm i.d. and BDS Hypersil phenyl 150 × 4.6 mm i.d., both with 5 µm particles and a 120 Å pore size, Keystone Scientific Operations) connected in series (C8 column first). The mobile phase was 10 mM Bu4NClO4/ CH2Cl2 with a flow rate of 0.7 mL/min. The injected MPC solution (∼0.1 mM in CH2Cl2) was prefiltered with a 0.45 µm Nalgene syringe filter with a PTFE membrane. Electrochemistry. Cyclic and square-wave pulse voltammetries were performed on 0.1-0.3 mM solutions (degassed and blanketed with Ar) of AgBBT MPCs in 0.1 M Bu4NClO4/CH2Cl2 using a Bioanalytical Systems 100B electrochemical analyzer. The singlecompartment electrochemical cell contained 1.6-mm diam Pt working, Pt flag counter, and Ag wire quasi-reference electrodes. The working electrode was precleaned by polishing with 0.05 µm alumina powder followed by potential-cycling in 0.1 M H2SO4 for .∼5 min. Ag stripping analysis was performed, when needed after potential-cycling in the AgBBT/Bu4NClO4/CH2Cl2 solution, by rinsing the working electrode with CH2Cl2 and immersing it in aqueous 0.1 M NaNO3. Microelectrode voltammetry was done with Pt microdisks of radii 4.8, 26.2, and 52.4 µm, on a 0.21 mM solution of Prep A-AgBBT and a 0.65 mM solution of Prep B-AgBBT MPCs, in 0.1 M Bu4NClO4/CH2Cl2. Solutions were serially diluted to obtain voltammetry at other concentrations. The microdisk dimensions were verified by voltammetry of a ferrocene solution in 0.1 M Bu4NClO4/CH3CN electrolyte. Electronic Conductivity of AgBBT Film. A film of AgBBT MPCs was dropcast onto an interdigitated array electrode (IDA; Abtech Scientific, Inc., having 50 fingers, each 3000 µm long × 20 µm wide × 0.1 µm high, separated by 20 µm gaps). Droplets of a 5 mg/mL CH2Cl2 solution of MPCs were serially cast and evaporated until the IDA fingers were thoroughly coated and no longer visible under the film, which was dried in a vacuum for 24 h. Linear sweep current-potential (i-E) curves were obtained at room temperature with a home-built potentiostat, and electronic conductivities (σEL, Ω-1 cm-1) were calculated by multiplication of the slopes of the i-E curves by the IDA cell constant (0.04 cm-1) given by the manufacturer. Other Measurements. UV-Vis spectra in n-heptane were obtained with a Shimadzu UV-1601 UV-visible spectrophotometer. Mass spectra were obtained, in positive-ion mode, with a Micromass TOFSPEC equipped with a 337-nm nitrogen laser (Laser Science, Inc., Newton, MA). Thermograms were obtained with a PerkinElmer Pyris 1 thermogravimetric analyzer, from ∼10 mg MPCs in an Al pan, by ramping the temperature from 30 to 600 °C at 15 °C/min. XPS spectra were obtained with a Kratos Analytical Axis Ultra instrument with a hemispherical analyzer and Al KR source. The MPCs were evaporated from CH2Cl2 onto a clean glass slide coated with an evaporated gold film (all peaks referenced to Au 4f7/2 at 83.8 eV).30 1H NMR spectra were obtained using a Bruker 400 MHz Avance spectrometer and CD2Cl2 as the solvent. Powder X-ray (30) Wagner, C. D., Riggs, W. M., Davis, L. E., Moulder, J. F., Muilenberg, G. E., Eds.; Handbook of X-ray Photoelectron Spectroscopy; Perkin-Elmer Corp.: Wellesley, MA, 1979.

11378 Langmuir, Vol. 22, No. 26, 2006

Branham et al.

Figure 1. UV-vis absorbance spectra: (A) Prep A-AgBBT and (B) Prep B-AgBBT normalized at 300 nm for comparison. diffraction data were obtained using a Rigaku Multiflex diffractometer with a Ni-filtered Cu KR source operating at 40 kV and 40 mA.

Results and Discussion Synthesis. There are important differences between the usual6,29 Brust two-phase synthesis of Au MPCs, in which the phase transfer agent Oct4NBr acts to transfer AuCl4- into the organic phase, and the present one for Prep A-AgBBT MPCs, where the action of Oct4NBr instead causes formation of a AgBr colloid. That colloid evidently yields, upon BBT thiol addition, a Agthiolate colloid that is subsequently reduced by borohydride. These details have not been systematically investigated, but provided the reduction step is conducted under Ar, the procedure leads to AgBBT MPCs that are well-protected by thiolate ligand and lack any (appreciable) oxide component of the Ag surface. In contrast, the single-phase27b Prep B-AgBBT MPC synthesis (see Supporting Information) yields MPC cores of a similar size but with distinctly different properties and a significant oxide content of the nanoparticle surface. The behavior and properties of Ag MPCs obtained by the single-phase method depend strongly on the thiol employed. 4-Methoxy-R-toluenethiol and 4-methoxybenzenethiol did not yield useful MPCs. Hexanethiol and phenylethanethiol produced polydisperse (3-5 nm) MPCs that were unstable in chlorinated solvents (aggregating and decomposing within a few minutes) and lacked expected solubility in toluene. The roughly coneshaped, bulky p-substituted BBT ligand produced access to rather small average Ag core diameters and good nanoparticle stability. The detailed reason(s) for this preferential stability is/are unclear. However, the use of the BBT ligand, and excluding oxygen during synthesis in the two-phase method, were clearly important in leading to the small core size nanoparticles described below. Optical Absorbance. Surface plasmon resonance (SPR) absorbance31spectra of metal MPCs are sensitive to changes in core size, especially in the metal-to-molecule size range. For silver nanoparticles, the SPR band intensity and wavelength depend10a,32 on size, being intense and sharp for larger nanoparticles (>2 nm) with λmax at ∼420 nm and, at decreased (1-2 nm) core diameter, becoming less intense, broadened, and shifted to lower energy. UV-vis spectra of both AgBBT preparations are shown in Figure 1. The spectrum of Prep A-AgBBT MPCs (curve A) (31) Kreibig, U.; Vollmer, M. Optical Properties of Metal Clusters; SpringerVerlag: New York, 1995. (32) (a) Mulvaney, P. Langmuir 1996, 12, 788. (b) Taleb, A.; Petit, C.; Pileni, M. P. J. Phys. Chem. B 1998, 102, 2214. (c) Wilcoxon, J. P.; Martin, J. E.; Provencio, P. J. Chem. Phys. 2001, 115, 998.

Figure 2. TEM core size histograms for Prep A-AgBBT (top) and Prep B-AgBBT (bottom) MPCs. Inset: HR-TEM image of Prep A-AgBBT core (see Supporting Information for further examples).

exhibits a broad SPR band at ∼475 nm, which, by being significantly red-shifted from 420 nm, implies32c a small core diameter (∼2 nm). Strikingly, the spectrum of Prep B-AgBBT MPCs (Curve B) shows no SPR band at all. We believe this result reflects an amount of oxide33 coating on the MPC core surface (confirmed below by XPS) sufficient to disrupt the collective oscillation of the conduction-band electrons. Mulvaney32a described analogous damping of SPR by chemisorbed iodide. The small band at ∼420 nm in curve B may be due to a small population of larger, less oxidized MPCs. TEM and Powder X-ray Diffraction (XRD). Figure 2 presents histograms of Ag MPC core diameters determined by TEM; example images are found in the Supporting Information. Samples of Prep A-AgBBT and Prep B-AgBBT MPCs give aVerage core diameters of 2.1 ( 0.7 nm and 2.1 ( 0.9 nm, respectively, with Prep B-AgBBT being somewhat more polydisperse. The marginal resolution of the Ag core edges, unavoidable with clusters this small, also contributes to uncertainty in the apparent average core diameter. Larger core sizes are more easily visualized and can, as a consequence, become overcounted. Also, some of the MPCs in both TEM images have cores that are substantially nonspherical, and the differences in contrast within the same core indicate that they are composed of multiple grains. These nonspherical cores may be the result of aggregation, in which two or a few smaller cores have become fused together. HR-TEM images, such as the inset of Figure 2 for a somewhat larger Prep A-AgBBT MPC core, show that many of the larger MPC cores are polycrystalline (see Supporting Information for additional images), both for spherical and nonspherical cores. The TGA and voltammetric QDL results presented below suggest an average 1.6 nm MPC core size, smaller than the TEM average, which may be biased upward by the polycrystalline aggregates. The polycrystallinity of Ag nanoparticles has been seen previously by electron24a and X-ray diffraction (XRD).27b,34 Figure (33) (a) Kapoor, S. Langmuir 1998, 14, 1021. (b) Yin, Y. D.; Li, Z. Y.; Zhong, Z. Y.; Gates, B.; Xia, Y. N.; Venkateswaran, S. J. Mater. Chem. 2002, 12, 522.

Arylthiolate-Protected Ag Quantum Dots

Figure 3. X-ray diffractograms of Prep A-AgBBT MPCs (top curve) and Ag metal (bottom curve). Application of the Scherrer equation to a rough deconvolution of the Ag(111) peak in the Prep A diffractogram gives an approximate range of core diameters of 1.52.9 nm.

3 (upper) shows the powder X-ray diffractogram of a sample of Prep A-AgBBT MPCs; the lower diffractogram corresponds to pure Ag metal. The (111), (200), (220), and (311) planes of the Ag face-centered cubic lattice in the MPCs are manifested as broadened peaks as compared to the narrow peaks of the Ag metal diffractogram. The broadening is attributable to the small size of the MPCs and to possible scattering by the BBT monolayer surrounding the Ag core. The poorly defined peaks at the lowest angles may arise from monolayer superlattice diffractions.35 Importantly, the XRD diffractogram in Figure 3 bears little resemblance to those of AgO and Ag2O (see Supporting Information), implying a silver MPC core that is either unoxidized on its surface or bears too little oxide to be detectable by XRD. This conclusion is supported by a lattice spacing analysis of HR-TEM images (see Supporting Information). Application of the Scherrer equation36 to a rough deconvolution of the (111) peak in Figure 3 yields an approximate range of core diameters of 1.5 to 2.9 nm, accounting for uncertainty contributed by peak asymmetry. This range is in rough agreement with the TEM histograms. TGA. The thermal decomposition of Au MPCs has been demonstrated29 to occur by dissociation of the thiolate ligands as volatile disulfides, leaving behind the core metal as the final mass. This is an effective way to determine the organic ligand mass fraction of the MPC, which, upon assigning an average MPC core mass, yields an average chemical formula for the MPC. Figure 4 shows a thermogram representative of both Prep A-AgBBT and Prep B-AgBBT, which gives an average ligand monolayer mass loss of ∼38%. If one assumes, by analogy to results for Au,29 closed-shell truncated octahedral cores of Agn(SR)m with n ) 140, 225, and 314, and m ) 53, 75, and 91,29 the anticipated organic mass losses should be 38, 36, and 32%. The experimental organic mass is closest to a Ag140(BBT)53 formulation, but, given the breadth of the TEM histogram and the detected polycrystalline aggregates, admits to the possibility of an appreciable Ag225 MPC population and smaller MPC sizes than Ag140. We shall term Ag140(BBT)53 the aVerage core composition, but with these caveats in mind. Mass Spectrometry. High-energy laser desorption-ionization mass spectrometry, although giving low mass resolution and (34) (a) Yamamoto, M.; Nakamoto, M. J. Mater. Chem. 2003, 13, 2064. (b) He, S. T.; Yao, J. N.; Xie, S. S.; Gao, H. J.; Pang, S. J. J. Phys. D: Appl. Phys. 2001, 34, 3425. (35) Ang, T. P.; Chin, W. S. J. Phys. Chem. B 2005, 109, 22228. (36) West, A. R. Solid State Chemistry and Its Applications; Wiley: New York, 1984.

Langmuir, Vol. 22, No. 26, 2006 11379

Figure 4. Typical thermogravimetric weight loss curve for AgBBT MPCs.

Figure 5. Positive-ion LDI-TOF-MS mass spectrum of Prep B-AgBBT MPCs. Following ref 37, z ) 1, and desorption of the thiolate ligands accompanies ionization and transfer to the gas phase.

provoking substantial ligand dissociation, has been important in estimations of MPC core masses.8b,37 This experiment was applied to Prep B-AgBBT MPCs (but was unfortunately not available for Prep A-AgBBT MPCs). Figure 5 shows the positive ion mass spectrum of Prep B-AgBBT, where two main peaks are seen, at 15.8 kDa and a smaller one at 28.2 kDa. (The spectral fine structure is just noise without any regular spacing as might be caused by different amounts of residual sulfur ligands. Previous work37b-e has shown that the organic monolayer is generally dissociated from the Au MPCs in the laser desorption-ionization experiment, and that z ) 1.) A mass of 15.1 kDa is expected for a 140-silver atom core, consistent with the larger m/z peak. The broadness of the spectrum in Figure 5 would also accommodate an underlying unresolved 24 kDa peak for Ag MPCs having 225atom cores. The origin of the 28 kDa maximum is unclear; the implied 260 Ag atom count is not a closed-shell structure, at least for a truncated octahedral core geometry.29 Considering the TEM indications of aggregation, perhaps it is an aggregate of Ag140 and Ag116 cores (mass 27.6 kDa). (37) (a) Balasubramanian, R.; Guo, R.; Mills, A. J.; Murray, R. W. J. Am. Chem. Soc. 2005, 127, 8126. (b) Schaaff, T. G. Anal. Chem. 2004, 76, 6187. (c) Schaaff, T. G.; Knight, G.; Shafigullin, M. N.; Borkman, R. F.; Whetten, R. L. J. Phys. Chem. B 1998, 102, 10643. (d) Alvarez, M. M.; Khoury, J. T.; Schaaff, T. G.; Shafigullin, M.; Vezmar, I.; Whetten, R. L. Chem. Phys. Lett. 1997, 266, 91. (e) Schaaff, T. G.; Shafigullin, M. N.; Khoury, J. T.; Vezmar, I.; Whetten, R. L. J. Phys. Chem. B 2001, 105, 8785.

11380 Langmuir, Vol. 22, No. 26, 2006

Branham et al.

0.79 ( 0.17 aF. Assuming that the Ag MPC producing the Figure 6 (top) voltammetry lacks molecularity, and applying the concentric sphere capacitor model,40a,b

CCLU ) 4π0(r/d)(r+d)

Figure 6. Square-wave voltammetry of 0.3 mM Prep A-AgBBT (top) and Prep B-AgBBT (bottom) MPCs, in 0.1 M Bu4NClO4/ CH2Cl2. The upper and lower traces in the top figure correspond to negative- and positive-going scans of potential.

Electrochemistry of AgBBT MPCs. The electrochemical properties of Ag MPCs have been relatively under-researched compared to those of Au MPCs. Au MPCs from Au38 to Au2257,8b,38 exhibit well-defined, sequential, one-electron charging peaks, which, for the larger versions follow QDL principles,7,38 and for the smaller MPCs exhibit a voltammetric pattern indicating the coalescence of core molecular orbitals. The sole report28 of Ag MPC QDL is for surfactant-protected, surface-attached, 3.3 nm nanoparticles. Figure 6 presents the square-wave voltammetry of solutions of Prep A-AgBBT and Prep B-AgBBT MPCs. That for Prep A-AgBBT MPCs (top curve) shows a roughly evenly spaced progression of current peaks (indicated by asterisks) lying astride the potential of zero core charge (PZC) of the MPC core (see vertical dashed line). The PZC is judged, as before,39 as lying at the potential of the overall capacitance minimum. The QDL charging of Prep A-AgBBT nanoparticles means that they have double-layer capacitances so small that large shifts in potential occur upon transfers of single electrons as the MPCs diffuse to the electrode and their double layers are charged. The figure indicates the overall double-layer charge state of the cores for several of the current peaks. A plot (Supporting Information) of apparent core charge against potential spacing gives an average spacing of ∆V ) 211 ( 53 mV, which is related40 to the capacitance CCLU for one-electron double-layer charging of the MPC by

CCLU ) e/∆V

(1)

where e is the electron charge (C). This relation gives CCLU ) (38) Wolfe, R. L.; Murray, R. W. Anal. Chem. 2006, 78, 1167. (39) Chen, S. W.; Murray, R. W. J. Phys. Chem. B 1999, 103, 9996. (40) (a) Chen, S. W.; Murray, R. W.; Feldberg, S. W. J. Phys. Chem. B 1998, 102, 9898. (b) Hicks, J. F.; Templeton, A. C.; Chen, S. W.; Sheran, K. M.; Jasti, R.; Murray, R. W.; Debord, J.; Schaaf, T. G.; Whetten, R. L. Anal. Chem. 1999, 71, 3703. (c) Green, S. J.; Stokes, J. J.; Hostetler, M. J.; Pietron, J.; Murray, R. W. J. Phys. Chem. B 1997, 101, 2663.

(2)

where  is the monolayer dielectric constant ( ) 4.7 for BBT thiol),41 0 is the permittivity of free space, r is the MPC core radius (m), and d is the monolayer chain length (9.1 × 10-10 m), the ensuing estimate of r ) 0.84 nm comfortably lies within the Figure 2 histogram and is consistent with the 1.6-1.7 nm diameter expected for a Ag140 core. An intriguing feature of Figure 6 (top) is that the current peaks flanking the PZC are separated by a somewhat larger spacing (∼320 mV) than the 211 mV average. A barely emergent energy gap for the Prep A-AgBBT MPCs would cause such an effect, but so would a diffuse double-layer contribution.42 Samples of much higher purity will be required for a clearer evaluation. A current background continuum underlies the square-wave voltammetry QDL peaks of the Prep A-AgBBT MPCs in Figure 6 (top). The resolution of the QDL current peaks, relative to the background, is modest but similar to that seen in early40b studies of Au MPCs where the monodispersity was low. The underlying background currents are attributed to the electron-transfer charging of a mixture of other MPC core dimensions, analogous to previous experimental results29,40c,43 and supported by simulations40a,44 of observations for polydisperse Au MPCs. From the relative magnitudes of the QDL current peaks and background continuum, we estimate that the MPCs that produce the peaks are roughly 20% of the total MPC population. In contrast, the square-wave voltammetry of solutions of Prep B-AgBBT MPCs (Figure 6, bottom) shows a mostly featureless current continuum and no QDL charging peaks. Although Prep A-AgBBT and Prep B-AgBBT MPCs have similar average diameters and thiolate monolayer content, the absence of QDL peaks for the latter could possibly arise from the Ag140 MPC component simply being present in much lower population than in the Prep A-AgBBT MPC samples. However, when one additionally considers the complete absence of a surface plasmon absorbance for Prep B-AgBBT MPCs (Figure 1), a more profound difference between the two MPC core-monolayer interfaces seems probable. The cores of Prep B-AgBBT MPCs do not act as a discretely double-layer chargeable, metal-like surface nor do they display the collective surface electronic oscillations characteristic of the plasmon resonance. We propose that the presence of a partial oxide layer on the cores of Prep B-AgBBT MPCs is responsible for both behaviors. The presence of a substantial oxygen content is shown by XPS elemental results (see below). Bulk electrolysis45 of these MPCs in degassed Bu4NClO4/CH2Cl2 was not successful in reducing the surface oxygen and regenerating QDL properties such as those of Prep A-AgBBT MPCs. Similarly featureless voltammograms were observed for Ag MPCs prepared (in both synthetic procedures) with other thiol ligands such as phenylethanethiol, hexanethiol, dodecanethiol, and 4-methoxy-R-toluenethiol. These ligands apparently do not serve to prevent oxidation of the silver core. (41) Lide, D. R., Ed. CRC Handbook of Chemistry and Physics; CRC Press: Boca Raton, FL, 2004; Vol. 85. (42) Guo, R.; Georganopoulou, D.; Feldberg, S. W.; Donkers, R.; Murray, R. W. Anal. Chem. 2005, 77, 6516. (43) Green, S. J.; Pietron, J. J.; Stokes, J. J.; Hostetler, M. J.; Vu, H.; Wuelfing, W. P.; Murray, R. W. Langmuir 1998, 14, 5612. (44) Miles, D. T.; Leopold, M. C.; Hicks, J. F.; Murray, R. W. J. Electroanal. Chem. 2003, 554, 87. (45) Song, Y.; Murray, R. W. J. Am. Chem. Soc. 2002, 124, 7096.

Arylthiolate-Protected Ag Quantum Dots

Langmuir, Vol. 22, No. 26, 2006 11381

Figure 7. Cyclic voltammetry of 0.3 mM Prep A-AgBBT (top) and Prep B-AgBBT (bottom) MPCs, in 0.1 M Bu4NClO4/CH2Cl2.

The small QDL peaks seen for Prep A-AgBBT MPCs in the square-wave voltammetry of Figure 6 (top) are scarcely visible in the (macroelectrode) cyclic voltammetry (CV) of Figure 7 (top), but the two most prominent square-wave voltammetry oxidation peaks, at approximately +0.23 and +0.50 V (ostensibly, these are the Ag1402+/1+ and Ag1403+/2+ formal potentials), are readily seen in the CV. Waves at these potentials are also seen in the CV (Figure 7, lower) of solutions of the Prep B-AgBBT MPCs. For both MPCs, the corresponding reduction waves are absent on the reverse potential scan. An issue was thus whether some non-QDL process was enhancing the currents at the potentials of the Ag1402+/1+ and Ag1403+/2+ QDL double-layer chargings. The wave at +0.5 V in the Figure 7 Prep B-AgBBT CV is a diffusion-controlled reaction of the MPCs, according to (Supporting Information) the linear plot46 of its peak current against [potential scan rate]1/2. Diffusion-controlled MPC voltammetry of larger (5 and 11 nm) Ag clusters has been reported previously.12,14 Aoki et al.12 reported an oxidation wave at +1.1 V (vs Ag/AgXO) and a reduction at -0.6 V for stearate-stabilized 5 nm Ag MPCs in cyclohexane/CH3CN solvent. CH3CN is a good coordinating solvent for Ag+ ions. The oxidation current was more than 9 times larger than the reduction and was assigned to (extensive) oxidation of the Ag atoms in the core to Ag+. The reduction current was assigned to reduction of the silver stearate protecting layer, leaving Ag0 on the working electrode. This process is akin to reductive desorption47 of a self-assembled thiolate monolayer from Au(111). The results in Figure 7, in a solvent (CH2Cl2) much less coordinating (than acetonitrile) for Ag+, and for a thiolateprotected (as opposed to a carboxylate-protected) Ag MPC, do not show the reported12 reduction step, and the oxidation steps appear at much less positive potentials. Understanding the difference between these results and that between the Figure 7 Prep A-AgBBT and Prep B-AgBBT MPC voltammetry is aided by evaluation of the effective value(s) of n, the number of electrons transferred per MPC in the Figure 7 oxidation waves. The substantial background currents made doing this by coulometry unappealing, so we chose a microdisk electrode approach. Microdisk limiting currents follow the relation46

ILIM ) 4nFrDC

(3)

where F is the Faraday, C is the MPC concentration (based on MW ) 24600 g/mol), r is the microelectrode radius (cm), and (46) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications, 2nd ed.; John Wiley & Sons: New York, 2001. (47) Walczak, M. M.; Popenoe, D. D.; Deinhammer, R. S.; Lamp, B. D.; Chung, C. K.; Porter, M. D. Langmuir 1991, 7, 2687.

Figure 8. Typical microelectrode voltammogram of 0.21 mM Prep A-AgBBT MPCs, in 0.1 M Bu4NClO4/CH2Cl2. Microelectrode voltammetry of Prep B-AgBBT MPCs results in similarly shaped voltammograms.

D ) 2.6 × 10-6 cm2/s is the MPC diffusion coefficient (estimated previously48 by Taylor dispersion for similar size Au MPCs). The experiment necessarily entails the assumption of an aVerage Ag140 composition and voltammetric behavior. The microelectrode voltammetry of solutions of Prep A-AgBBT and Prep B-AgBBT both show (Figure 8) a single broadened feature, which is apparently an unresolved overlap of those seen in Figure 7. Voltammetry taken using three microelectrodes (r ) 5, 25, 50 µm) at five different MPC concentrations gives plots of plateau current (at +0.8 V) versus r at each concentration (Supporting Information) that are reasonably linear, with slopes determining the values of n. These determinations give n ) 5.3 ( 1.0 and 2.1 ( 0.2 electrons per MPC for the Prep A-AgBBT and Prep B-AgBBT MPCs, respectively, at +0.8 V. Comparing these numbers to Figure 6, the pattern of QDL peaks indicates that three and possibly four electrons should have been transferred to charge the double layers of the Prep A-AgBBT MPCs to the +0.8 V potential. The microelectrode determination gives a larger n, by one to two electrons. The strong inference is that some electrode reaction in addition to the QDL charging processes must be occurring in the two oxidation waves prominent in Figure 7 (and more prominent than other QDL features in Figure 6). The n ) 2.1 electron result for the Prep B-AgBBT MPCs is consistent with the same reaction occurring for them, noting that no obvious QDL peaks are involved in its voltammetry. The nature of the parasitic reaction that adds to the QDL peak currents is unclear; it may be oxidation of a Ag core site to a AgXO form or loss of one to two atoms from the core as dissolved Ag+. The small number of electrons transferred does show, on the other hand, that, unlike the previous study12 of stearate-stabilized large Ag MPCs, the Figure 8 oxidation waves do not correspond to a massive dissolution of the Ag cores of these nanoparticles. (This is not surprising considering the differences in solvent and ligands.) Finally, we call attention to the small oxidation peak at -0.5 V in the Figure 7 Prep B-AgBBT voltammogram. This is a stripping peak of a Ag0 submonolayer on the working electrode; a small amount of Ag+ becomes reduced from the MPC solution during the preceding negative potential scan. (The fact that it is Ag0 was shown by disconnecting the negatively potentiostated working electrode in the CH2Cl2 MPC solution, transferring it (48) Wuelfing, W. P.; Templeton, A. C.; Hicks, J. F.; Murray, R. W. Anal. Chem. 1999, 71, 4069.

11382 Langmuir, Vol. 22, No. 26, 2006

Branham et al.

Figure 9. XPS spectra of the Ag and O regions for Prep A-AgBBT (left) and Prep B-AgBBT (right) MPCs. Table 1. O/Ag Elemental Ratios in AgBBT MPCs

to an MPC-free aqueous 0.1 M NaNO3 solution, and conducting a stripping potential sweep. No Ag0 stripping is seen unless the electrode has been held at a reducing potential in the MPC solution.) Whether the Ag+ reduced was an impurity in the sample or originated on the core itself is unknown, but, by comparison of the Figure 7 -0.5 V current to that at +0.2 V, its extent is much less than one electron per MPC. Electronic Conductivity. The electronic conductivities (σEL) of dry, cast films of Prep A-AgBBT and Prep B-AgBBT MPCs on IDAs are both rather low, being measured as ∼3.5 × 10-9 Ω-1cm-1 for both MPCs. For comparison, nonmixed valent films49 of Au140(Shexyl)53 MPCs have a much larger σEL (∼3 × 10-5 Ω-1cm-1), while those of Au38(SCH2CH2Ph)24 have a similar σEL (6 × 10-9 Ω-1cm-1). The Au MPC conductivities are interpreted as reflecting the rates of electron hopping between neighboring MPCs, under the influence of an electrical gradient. Electronic conductivity in the dry state depends on both activation barrier energies for the electron transfers between Au140+ and Au140° and the carrier population (such as the relative proportions of Au140+ and Au140° charge states). The barrier energy for electron hopping in mixed valent Au38+/Au38° is much larger than that for Au140 and is the main reason for its lower conductivity. Against this backdrop, possible reasons for the slow electron hopping in Prep A-AgBBT and Prep B-AgBBT MPC films are a low carrier population and/or an intrinsically slow electron-transfer process with a large energy barrier. Barrier energies were not measured in the present study. Also, considering the difference in oxide content, it is surprising that the Prep A-AgBBT and Prep B-AgBBT MPC film conductivities are similar, so there may be additional, undiscovered factors.

XPS. XPS detects the elements present in an MPC29,50 sample and can semiquantitatively determine their relative populations. Figure 9 shows spectra of the Ag and O binding energy regions for Prep A-AgBBT and Prep B-AgBBT. In both spectra, the positions of the larger Ag 3d5/2 peaks are consistent with the Ag0 metal binding energy (367.9 eV).30 The smaller peaks at ∼370 and 376 eV for the Prep B-AgBBT MPCs and the much larger O 1s peak suggests some core metal oxidation manifested as Ag2O, as seen previously50 with other Ag nanoparticles. The higher binding energy Ag3d5/2 peaks are notably absent in the Prep A-AgBBT spectrum, and the O 1s signal is very small, barely above the background. O/Ag atomic ratios from these spectra are presented in Table 1. The oxygen content in the Prep A-AgBBT MPCs amounts, at most, to a few atoms per average Ag140 core. The large O/Ag ratio for Prep B-AgBBT MPCs implies significant core oxidation, confirming the reason for the lack of QDL charging. The XPS result suggests ∼70 oxygen atoms per average Ag140 core, equivalent to a complete monolayer. The HPLC separation below shows that oxygen content is probably not evenly distributed among the MPCs, so some may be very highly oxidized and some less so.

(49) (a) Wuelfing, W. P.; Green, S. J.; Pietron, J. J.; Cliffel, D. E.; Murray, R. W. J. Am. Chem. Soc. 2000, 122, 11465. (b) Choi, J.-P.; Murray, R. W. J. Am. Chem. Soc. 2006, 128, 10496-10502.

(50) (a) Pol, V. G.; Srivastava, D. N.; Palchik, O.; Palchik, V.; Slifkin, M. A.; Weiss, A. M.; Gedanken, A. Langmuir 2002, 18, 3352. (b) Sun, X. P.; Dong, S. J.; Wang, E. K. Macromolecules 2004, 37, 7105.

O/Ag ratio

Prep A-AgBBT

Prep B-AgBBT

theoreticala calculatedb correctedc

0 0.05 0.04

0 0.58 0.51

a Based on Ag BBT . b Calculated as I /I 140 53 o Ag ) (Ao/σo)/(AAg/σAg), where A is the peak area and σ is the elemental cross section. c The correction factor depends on the electron escape depths and attenuation lengths (see Supporting Information).

Arylthiolate-Protected Ag Quantum Dots

Langmuir, Vol. 22, No. 26, 2006 11383

chromatogram of Prep A-AgBBT MPCs; two well-shaped peaks are visible between two smaller peaks and severe tailing. The UV-vis spectra (Figure 10, bottom) of these peaks clarify the presence of two dominant MPC populations in the sample. The peaks a and b spectra are reminiscent of that of Prep B-AgBBT (see Figure 1), whereas a red-shifted SP band appears in the spectra of peaks c and d. This separation shows that Prep A-AgBBT MPC samples contain some oxidized MPCs (notably peak b) and a large population of unoxidized MPCs (notably peak c). Acknowledgment. This research was supported in part by grants from the National Science Foundation and the Office of Naval Research. The authors thank Wallace Ambrose (UNC School of Dentistry) for TEM aid and Benjamin Pierce (UNC Chemistry Department) for assistance with TGA.

Figure 10. HPLC chromatogram (top) and corresponding UV-vis absorbance spectra (bottom) of Prep A-AgBBT MPCs. The spectra are normalized at 300 nm for comparison; the spectra of peaks a and d are noisy due to the low absorbance of their peaks.

Supporting Information Available: Procedure for the synthesis of Prep B-AgBBT MPCs; TEM images of Prep A-AgBBT and Prep B-AgBBT; HR-TEM images and fast Fourier transform lattice spacing analysis of Prep A-AgBBT MPCs; X-ray diffractograms of AgO and Ag2O; Z plot of the Prep A-AgBBT square-wave voltammetric peaks shown in Figure 6; plot of peak current vs the square root of the scan rate for the peak at +500 in Figure 7; plot of microelectrode currents vs radius for different concentrations of Prep B-AgBBT; and an explanation of the XPS elemental ratio correction factor. This material is available free of charge via the Internet at http://pubs.acs.org. LA062329P

HPLC. Chromatographic separation of MPCs is a useful tool51 to analyze sizes and the separation of MPCs by core size and monolayer polarity.37a,38,52 Figure 10 (top) is a representative

(51) Wilcoxon, J. P.; Martin, J. E.; Provencio, P. Langmuir 2000, 16, 9912. (52) Jimenez, V. L.; Leopold, M. C.; Mazzitelli, C.; Jorgenson, J. W.; Murray, R. W. Anal. Chem. 2003, 75, 199.