Langmuir 2004, 20, 4703-4707
4703
Substituent Effects on the Exchange Dynamics of Ligands on 1.6 nm Diameter Gold Nanoparticles Robert L. Donkers,† Yang Song, and Royce W. Murray* Kenan Laboratories of Chemistry, University of North Carolina, Chapel Hill, North Carolina 27599-3290 Received January 28, 2004. In Final Form: March 17, 2004 The kinetics of exchange of phenylethanethiolate ligands (PhC2S) of monolayer-protected clusters (MPCs, average formula Au140(PhC2S)53) by para-substituted arylthiols (p-X-ArSH) are described. 1H NMR measurements of thiol concentrations show that the exchange reaction is initially rapid and gradually slows almost to a standstill. The most labile ligands, exchanging at the shortest reaction times, are thought to be those at defect sites (edges, vertexes) on the nanoparticle core surface. The pseudo-first-order rate constants derived from the first 10% of the exchange reaction profile vary linearly with in-coming arylthiol concentration, meaning that the labile ligands exchange in a second-order process, which is consistent with ligand exchange being an associative process. A linear Hammett relationship with slope F ) 0.44 demonstrates a substituent effect in the ligand place exchange reaction, in which the bimolecular rate constants increase for ligands with electron-withdrawing substituents (1.4 × 10-2 and 3.8 × 10-3 M-1 s-1 for X ) NO2 and 4-OH, respectively). This is interpreted as the more polar Au-S bonds at the defect sites favoring bonding with more electron deficient sulfur moieties. At longer reaction times, where ligands exchange on nondefect (terrace) as well as defect sites, the extent of ligand exchange is higher for thiols with more electrondonating substituents. The difference between short-time kinetics and longer-time pseudoequilibria is rationalized based on differences in Au-S bonding at defect vs nondefect MPC core sites. The study adds substance to the mechanisms of exchange of protecting ligands on nanoparticles. The scope and limitations of 1H NMR spectroscopy for determining rate data are also discussed.
Introduction A key aspect of monolayer-protected metal clusters (MPCs) is how their chemical properties can be manipulated by choice of the chemistry of the surrounding protective monolayer. For Au nanoparticles, where thiolates are the most commonly used ligands, one can vary the thiol employed in the initial MPC synthesis, or one can replace the ligands of an MPC product of a well-studied synthesis with different thiolates.1-9 The latter involves reacting the original MPC with a solution of the thiols of the new, or “in-coming”, thiolate ligands. Using such ligand exchange reactions to prepare mixed monolayer MPCs is abundant in Au MPC chemistry.8-20 * To whom correspondence may be addressed. Email: rwm@ email.unc.edu. † Present address: Steacie Institute for Molecular Sciences, National Research Council of Canada, ON, Canada, K1A 0R6. (1) Ingram, R. S.; Hostetler, M. J.; Murray, R. W. J. Am. Chem. Soc. 1997, 119, 9175. (2) Templeton, A. C.; Hostetler, M. J.; Kraft, C. T.; Murray, R. W. J. Am. Chem. Soc. 1998, 120, 1906. (3) Templeton, A. C.; Hostetler, M. J.; Warmoth, E. K.; Hartshorn, C. M.; Krishnamurthy, V. M.; Forbes, M. D. E.; Murray, R. W. J. Am. Chem. Soc. 1998, 120, 4845. (4) Ingram, R. S.; Murray, R. W. Langmuir 1998, 14, 4115. (5) Templeton, A. C.; Wuelfing, W. P.; Murray, R. W. Acc. Chem. Res. 2000, 33, 27. (6) Templeton, A. C.; Zamborini, F. P.; Wuelfing, W. P.; Murray, R. W. Langmuir 2000, 16, 6682. (7) Hicks, J. F.; Templeton, A. C.; Chen, S.; Sheran, K. M.; Jasti, R.; Murray, R. W.; Debord, J.; Schaaff, T. G.; Whetten, R. L. Anal. Chem. 1999, 71, 3703. (8) Song, Y.; Murray, R. W. J. Am. Chem. Soc. 2002, 124, 7096. (9) Hostetler, M. J.; Templeton, A. C.; Murray, R. W. Langmuir 1999, 15, 3782. (10) Kell, A. J.; Stringle, D. L. B.; Workentin, M. S. Org. Lett. 2000, 3381. (11) Han, L.; Daniel, D. R.; Maye, M. M.; Zhong, C.-J. Anal. Chem. 2001, 73, 4441. (12) Kell, A. J.; Workentin, M. S. Langmuir 2001, 17, 7355. (13) Wuelfing, W. P.; Zamborini, F. P.; Templeton, A. C.; Wen, W.; Yoon, H.; Murray, R. W. Chem. Mater. 2001, 13, 87.
An understanding of molecular enclosures around nanoparticles is important in a general sense, since both metal and semiconductor nanoparticles are commonly studied or processed while bearing some kind of protecting or capping monolayer. An in-depth understanding of the mechanism and dynamics of ligand place exchange is required to fully appreciate the chemical properties and reactivity of the MPC’s ligand shell. For thiolate-capped Au MPCs, this has proven to be a multifaceted topic, because (a) the nanoparticle surface offers a diversity of ligand binding sitessvertexes, edges, and terraces and pseudoterracesswhich have differing electron densities21 and steric accessibility, and a corresponding diversity of ligand exchange kinetics and thermodynamics at the different sites. In keeping with well-known generalities of surface science, that high reactivity is associated with defect-rich surfaces, we have assumed8,9 that the initial, rapid phase of Au MPC ligand exchange reactions involves the vertex and possibly edge binding sites. Second, (b) the cores of Au MPCs have proven to be somewhat “plastic” in that their sizes can be altered by, for example, etching22 and annealing22-24 conditions. The transfers of Au moieties (14) Ionita, P.; Caragheorgheopol, A. J. Am. Chem. Soc. 2002, 124, 9048. (15) Lin, S.-Y.; Liu, S.-W.; Lin, C.-M.; Chen, C. Anal. Chem. 2002, 74, 330. (16) Wuelfing, W. P.; Murray, R. W. J. Phys. Chem. B 2002, 106, 3139. (17) Zamborini, F. P.; Leopold, M. C.; Hicks, J. F.; Kulesza, P. J.; Malik, M. A.; Murray, R. W. J. Am. Chem. Soc. 2002, 124, 8958. (18) Lee, D.; Donkers, R. L.; Desimone, J. M.; Murray, R. W. J. Am. Chem. Soc. 2003, 125, 1182. (19) Li, D.; Li, J. Surf. Sci. 2003, 522, 105. (20) Pengo, P.; Broxterman, Q. B.; Kaptein, B.; Pasquato, L.; Scrimin, P. Langmuir 2003, 19, 2521. (21) Ha¨kkinen, H.; Barnett, R. N.; Landman, U. Phys. Rev. Lett. 1999, 82, 3264. (22) Schaaff, T. G.; Shafigullin, M. N.; Khoury, J. T.; Vezmar, I.; Whetten, R. L.; Cullen, W. G.; First, P. N.; Gutie´rrez-Wing, C.; Ascensio, J.; Jose-Yacama´n, M. J. J. Phys. Chem. B 1997, 101, 7889.
10.1021/la0497494 CCC: $27.50 © 2004 American Chemical Society Published on Web 04/30/2004
4704
Langmuir, Vol. 20, No. 11, 2004
between MPC cores surely also involves ligand transfer reactions and indeed may be controlled by them. Third, (c) both metal transfer25 and ligand exchanges8,9 are accelerated by oxidative electronic charging of the Au core, indicating increased reactivity of the core defect sites and/or surface-migratory aptitude of bound ligands. The oxidative charging can be accomplished either electrochemically8 or by admitting dioxygen26 to the solution. The sensitivity of exchanges to oxidation can be rationalized by considering the differences between Au-S bonding at MPC defect (vertex, for example, with Au:S ratio of 1:1) and terrace (with Au:S ratio of 3:1) sites.21,27 The polarity of the partially ionic Au-S bonds becomes enhanced at the defect sites, leading (along with obvious steric advantages) to more facile ligand displacement relative to terrace sites. Oxidative core charging promotes the bond polarity. Fourth, (d) while the kinetics of the earliest and most rapid phase of ligand exchange are8,9 second orders suggestive of an associative exchange mechanismsthe rate order plots are not always ideal, displaying intercepts. Observations8 in the presence of dioxygen have suggested a possible competing, dissociative, exchange pathway, which may involve some kind of Au(I) thiolate species. Given all these features, the ligand exchange dynamics of MPCs is obviously a relatively complex topic. Our previous investigations8,9 of ligand exchange kinetics concerned alkanethiolate-coated Au MPCs (1.6 and 2.2 nm diameter cores and average formulas Au140(ligand)53 and Au314(ligand)91). These reactions follow a 1:1 stoichiometry, liberating one out-going ligand from the MPC monolayer (as a thiol) for each in-coming one, and without participation of disulfides or other oxidized forms of sulfur. The exchange reaction is initially rapids presumably being reaction of ligands at core defect sites (vertexes, edges)sand then slows as ligands on nondefect (terrace) core sites reactseither by mechanisms such as those at defect sitessbut much more sluggishlysor by intramolecular migration of ligands to defect sites. While the fast reaction phase was a second-order process8,9 s first order in both in-coming ligand and MPCsan intercept in the kinetic order plot indicated that the second-order dependence was not ideal. The late reaction phase becomes very slow, and many ligands typically remain unexchanged. The previous investigation involved precipitating and isolating the exchanged MPCs at timed intervals, determining the monolayer composition of the purified exchange products by 1H NMR. In the present study, we select ligands that allow in situ 1H NMR monitoring of the solution populations of in-coming and out-going thiolate ligands, which enhances the reaction sampling frequency and avoids uncertainties in the efficacy of reaction quenching. The exchange reaction is of para-substituted arylthiols (spanning a wide range of Hammett substituent values, σ ) -0.37 to 0.78) with phenylethanethiolatecoated MPCs of average formula Au140(PhC2S)53 (1.6 nm core diameter). The observed substituent effects are the first reported for Au nanoparticle ligand exchange kinetics and open a useful window on electronic effects of the incoming ligand. The fast phase of the exchange reaction is a cleanly defined second-order processsagain consistent with an associative exchange mechanism. The initial (23) Hicks, J. F.; Miles, D. T.; Murray, R. W. J. Am. Chem. Soc. 2002, 124, 13322. (24) Miles, D. T.; Murray, R. W. Anal. Chem. 2003, 75, 1251. (25) Song, Y.; Huang, T.; Murray, R. W. J. Am. Chem. Soc. 2003, 125, 11694. (26) Myung, N.; Bae, Y.; Bard, A. J. Nano Lett. 2003, 3, 747. (27) Ha¨berlen, O. D.; Chung, S.-C.; Stener, M.; Ro¨sch, N. J. Chem. Phys. 1997, 106, 5189.
Donkers et al.
reaction rates are sensitive to the electronic character of the aryl substituent, becoming accelerated for electronwithdrawing substituents, an observation explained using polar effects with the in-coming ligand and the nanoparticle exchange site. The previously mentioned migratory aptitudes are consistent with dipole effects induced by the substituent of the exchanging ligands. Interestingly, the slower phase of the reaction leads to a pseudoequilibrium product in which ligands with more electron donating substituents are favored in the MPC monolayer. Experimental Section Chemicals. 2-Phenylethanethiol (PhC2SH, 99%), tetra-noctylammonium bromide (Oct4NBr, 98%), deuterium oxide (99.9%) and sodium borohydride (99%), p-X-ArSH ligands (Aldrich, toluene (Fisher), and d2-methylene chloride (Cambridge Isotope Laboratories, Inc.) were all used as received. Hydrogen tetrachloroaurate trihydrate (from 99.999% pure gold) was prepared using a literature procedure28 and stored in a freezer at -20 °C. Low conductivity water was obtained with a Millipore Nanopure water purification system. Synthesis of p-X-ArSD Thiols. In the general procedure for X ) NO2, CH3, and OCH3, 3 mL of deuterium oxide was added to ca. 100 mg of the thiol in 3 mL of methylene chloride, and the two-phase mixture was vigorously stirred for a day. The organic layer was separated and dried over sodium sulfate, filtered, and rotary evaporated. Complete D/H exchange was confirmed by complete loss of the SH signal in the 1H NMR spectrum in dry CD2Cl2. The deuterated thiols were used immediately after being characterized. Synthesis of Au140(PhC2S)53. We have recently described preparation of this MPC.29 Briefly, the reaction is a Brust style MPC preparation30,31 in which 3.1 g of HAuCl4‚3H2O is phasetransferred into toluene using 1.1 molar equiv of tetra-noctylammonium bromide, reacted with a 3-fold molar excess of phenylethanethiol at 0 °C until the red solution turns clear, and lastly adding with rapid stirring 10 equiv of freshly prepared sodium borohydride in 60 mL of ice cold water followed by stirring at 0 °C for 20 h. The aqueous layer was removed and the organic layer washed with 100 mL of water followed by rotary evaporation to a black oil that was dissolved in 300 mL of ethanol overnight, ultimately precipitating the particles leaving a clear, colorless solution. The Au140(PhC2S)53 MPCs were isolated by filtering the remaining precipitate, washing it copiously with acetonitrile to remove tetra-n-octylammonium bromide, disulfide, and Au38(PhC2S)24 impurities. Ligand Exchange Kinetics by 1H NMR Spectroscopy (General Procedure). 1H NMR spectra were collected on a Bruker AC500 spectrometer, of solution mixtures of Au140(PhC2S)53 MPCs (19 mg/2.0 mL) and p-X-ArSH ligands (amounts varied based upon desired reactant ratios) in CD2Cl2 solutions. Sublimed ferrocene (2 mg/2.0 mL) was used as an internal standard. Changes in the relative NMR signals of the in-coming p-X-ArSH ligand and the ferrocene standard were recorded by integrating the NMR peaks (as shown in Figure 1 and in Supporting Information for an example exchange reaction where X ) 4-methoxybenzenethiol). In a typical procedure, a 1H NMR spectrum is acquired of the initial solution of ferrocene and p-XArSH, the MPCs are added and rapidly mixed, and the sample is placed in the preshimmed spectrometer for collection (15 s) of 1H NMR spectra (at regular periods over the 1-9 ppm (vs TMS) region for four transients at set intervals using in-house software). Average spectra of these sets were acquired each 1.5 min for mole ratios 1:1. Spectra were taken in at regularly timed intervals, taking four scans over 15 s every 1-2 min was a typical procedure. Signals for p-X-ArS- thiolates bound to the Au MPC core were suppressed by using a null value for the relaxation delay. The chemical shift range for peak integration was constant throughout each experiment, and reactions were performed for hours until no further change in the concentration of the p-X-ArSH ligand could be observed. A 1:1 place exchange reaction was confirmed based on relative changes in the p-X-ArSH ligand NMR signals and those of liberated PhC2SH ligands. Determination of the Extent of Ligand Exchange at Long Reaction Times. The reaction mixture was set aside for ca. 4 days and then rotary evaporated and washed extensively with ethanol and acetonitrile to remove unattached thiols, which was confirmed by the absence of sharp 1H NMR signals in its spectrum. A small crystal of iodine was added to the cleaned sample, which decomposes2,3 the MPCs to a Au precipitate and a disulfide mixture of the monolayer’s ligands, from which the relative quantities of each kind of ligand could be determined by integration of the respective NMR signals.
Results and Discussion Ligand Exchange Kinetics. The kinetics of exchanges of para-substituted arylthiolate ligands (p-X-ArSH, where X ) NO2, CN, Br, CH3, OCH3, OH) for phenylethanethiolate ligands on Au140(PhC2S)53 MPCs were followed using 1H NMR spectra of the reaction mixture as exemplified in Figure 1. A more detailed example is given in Supporting Information, typical experimental details are given in the Experimental Section, and Table 1 gives a summary of reagent concentrations. The course of the reaction was determined from the decrease in concentration, relative to a ferrocene internal standard, of the thiol of the in-coming p-X-ArS- ligand. The 1H NMR peak intensities of the X substituents CH3 and OCH3 were used in those cases, and peak intensities for the aryl protons were used for the other thiols. For reasons explained below, once exchanged onto the MPC, resonances for protons on the p-X-ArS- ligand become essentially invisible to the NMR spectrometer, as are those for the original MPC-
X substituent
KP-E (M-1 s-1)a
in-coming/MPC ligand ratiob
no. of exchanged ligands after 4 daysc
NO2 CN Br CH3 OCH3 OH
1.4 × 10-2 8.5 × 10-3 6.4 × 10-3 4.3 × 10-3 4.0 × 10-3 3.8 × 10-3
1.4 1.2 1.2 1.4 1.7 1.7
21 23 24 26 28 28
a Exchanging ligand concentration varied at a constant [MPC] of 2.8 × 10-4 M. In a plot of kobs against in-coming thiol concentration (i.e., Figure 4), the intercepts on the kobs axis are all -CH3 > -NO2, that is, the apparent equilibrium reached at (33) McClelland, R. A.; Kanagasabapathy, V. M.; Danait, N. S.; Steenken, S. J. Am. Chem. Soc. 1992, 114, 1816.
Figure 6. Plot showing the number of p-X-ArSH ligands (out of 53) incorporated into the MPC monolayer at long reaction times as a function of exchanging ligand excess relative to the total PhC2SH monolayer ligand concentration. X ) (O) NO2, (4) CH3, and (3) OCH3.
long reaction times favors ligands with more electrondonating substituents. This is at first glance a counterintuitive result; the second-order rate constants increase as -X is a more electron withdrawing substituent, but the apparent equilibrium position favors electron-donating substituents. However, one must consider that the actual exchanging sites on the Au core surface are different during the short-time second-order reaction phase and during the long-time phase approaching the apparent equilibrium. The former are proposed to be the defect sites (vertexes and edges) and the latter the more terrace-like sites where the Au-S bonding is less polar. The kinetic results may be interpreted as the more polar defect-site Au-S bonds being richer in S character than their nondefect counterparts, thus being stabilized by thiolates connected to electron-withdrawing groups. Bonding to terrace-like regions of the cluster core where the Au-S bonding ratio is higher is favored by an electron-rich thiolate ligand for similar reasons. This is consistent with recent observations by the Ulman laboratory, who observed the highest rate of self-assembled monolayer formation on gold surfaces with p-X-biphenylthiols containing electron-donating substituents.34 This trend was attributed to increased Au-S interaction from electronrich thiol moieties. Thus the higher extent of long-time exchange on terrace-like sites may arise from a higher Au-S bond thermodynamic stability provided by electrondonating substituents on the p-X-ArSH ligands. Acknowledgment. This research was supported by the University of North Carolina and grants from the National Science Foundation and the Office of Naval Research. R.L.D. acknowledges the National Sciences and Engineering Research Council of Canada for a Postdoctoral Fellowship. Supporting Information Available: NMR spectra illustrating ligand exchange at different reaction times are shown. This material is available free of charge via the Internet at http://pubs.acs.org. LA0497494 (34) Liao, S.; Schnidman, Y.; Ulman, A. J. Am. Chem. Soc. 2000, 122, 3688.
