Gold Nanoparticles

Gold Nanoparticles

(Parte 1 de 4)

13840 DOI: 10.1021/la9019475 Langmuir 2009, 25(24), 13840–13851Published on Web 07/02/2009 ©2009 American Chemical Society

Gold Nanoparticles: Past, Present, and Future†

Rajesh Sardar,‡ Alison M. Funston,§ Paul Mulvaney,*,§ and Royce W. Murray*,‡

‡Kenan Laboratories of Chemistry, University of North Carolina, Chapel Hill, North Carolina 27599, and §School of Chemistry and Bio21 Institute, University of Melbourne, Parkville, Victoria, 3010, Australia

Received June 1, 2009. Revised Manuscript Received June 12, 2009

This perspective reviews recent developments in the synthesis, electrochemistry, and optical properties of gold nanoparticles, with emphasis on papers initiating the developments and with an eye to their consequences. Key aspects of Au nanoparticle synthesis have included the two-phase synthesis of thiolated nanoparticles, the sequestration and reduction of Au salts within dendrimers, the controlled growth of larger particles of well-defined shapes via the seeded approach, and the assembling of a variety of nanoparticle networks and nanostructures. The electrochemistry of thiolated Au nanoparticles is systemized as regions of bulk-continuum voltammetry, voltammetry reflective of quantized double-layer charging, and molecule-like voltammetry reflective of molecular energy gaps. These features are principally determined by the nanoparticle core. Interesting multielectron Au nanoparticle voltammetry is observed when the thiolate ligand shell has been decorated with redox groupings. Another development is that Au nanoparticles were discovered to exhibit unanticipated properties as heterogeneous catalysts, starting with the low-temperature oxidation of CO. Substantial progress has also been made in understanding the surface plasmon spectroscopy of Au nanoparticles and nanorods. The need to investigate the optical properties of metal particles of a single, well-defined shape and size has motivated the development of a number of new techniques, leading to the study of electron transfer and redox catalysis on single nanoparticles.


Gold nanoparticle chemistry and physics has emerged as a broad new subdiscipline in the domain of colloids and surfaces. The unusual optical properties of small gold particles, their sizedependentelectrochemistry,andtheirhighchemicalstabilityhave made them the model system of choice for exploring a wide range of phenomena including self-assembly, biolabeling, catalysis, electron-transfer theories, phase transfer, DNA melting and assays, and crystal growth. This review focuses on gold nanoparticles, specifically those on the small end of the nanoparticle dimension scale. These nanoparticles (NPs) when stabilized or protected by a shell of thiolate ligands display good stability toward aggregation and other modes of decay, which enables attemptsatisolatingdifferentNPsizesandtheexplorationofhow NP properties depend on size (including quantization effects). NPs with fewer than 300 Au atoms can display distinct optical and electronic properties compared to the bulk metal. The thiolated NP stability further enables treating the ligand shell as a chemical platform that can be manipulated to exhibit desired reactivities, polyfunctionalization, and optical properties. The consequence for the past couple of decades has been a very active field of basic nanoscience research and applications of these NPs. An important aspect of Au NPs has been the breadth of their impact; applications range from photonic device fabrications, to sensing of organic and biomolecules, to charge storage systems. Thesubjectistoolargeforacomprehensivetreatmentinthespace available, and we have accordingly made selections of areas to highlight, with apologies to those omitted. Several complementary reviews on nanoparticles have appeared.1

A noteworthy recent observation has been that the ligation chemistry of verysmall thiolate-protected Au nanoparticles isnot simply a head-on thiolate bonding to a consolidated Au core of atoms;a “standard model” assumption1 of Au NP structure.

X-ray crystallographic determinations2a-c of Au102 and Au25 nanoparticle structures show instead that the ligands are bonded inbridgingcoordinationandmanyoftheAuatomsarenotpartof the core but exist with the ligands in semiring or “staple” structures (Figure 1). The extent to which this structural motif extends to the thiolate interfaces of larger Au NPs remains to be seen. The success of DFT calculations2d in reproducing experimental semiring structural facts has put such calculations more firmly into play in understanding this topic.

The above structural discoveries followed detailed investigations of small thiolate-protected Au nanoparticles that had delineated spectroscopic and kinetic properties. The results of those investigations are now open to deeper interpretations in light of the Figure 1 structure. For example, the common procedureofreplacingoneAusurfaceligandwithanother(ligand or place exchange)1 must in view of Figure 1 involve a more complex, multiple bond-breaking pathway.

Synthesis of Thiolate-Protected Au Nanoparticles. The synthesis of small, monodisperse nanoparticles is a major challenge in nanotechnology research. Smaller particles experience increased driving forces to aggregate to diminish surface energy, so a protective coating, or “capping”, is necessary during synthesis to keep them in a finely dispersed state. Of the two distinct

†Part of the “Langmuir 25th Year: Nanoparticles synthesis, properties, and assemblies” special issue. *Corresponding authors., rwm@email.

(2) (a) Jadzinsky, P. D.; Calero, G.; Ackerson, C. J.; Bushnell, D. A.; Kornberg,

Sardar et al. Perspective

strategies for making metal nanoparticles, (1) the top-down approach and (2) the bottom-up technique, the latter is by far the more common and effective. Metal ions are reduced by a reducingreagent;they nucleate to form nanoparticles while in the presence of protective ligands. Either the stoichiometry of the available metalion supply or the passivation of furthergrowthby theprotectiveligandscandeterminetheeventualnanoparticlesize.

A seminal early nanoparticle synthesis3a waso fA u55(Ph3P)Cl, whereby the complex [Au(Ph3P)Cl] was reduced with diborane gas. This nanoparticle proved to be difficult to isolate in a pure state and was somewhat unstable. A veryinfluential3bsubsequent synthesis protocol is outlined in Scheme 1, where following phase transfer of a gold salt into an organic medium;using a phasetransfer agent such as tetraoctylammonium bromide;and the addition of an organic thiol, an excess of a strong reducing agent;such as sodium borohydride;is added, with the rapid production of thiolate-protected Au nanoparticles. The parameters of the “Brust synthesis” have been thoroughly explored.3c A major advantage is rough control over the average size of the synthesized particles by the choice of the reduction condition and the gold salt-to-thiol molar ratio used, but even then, in general postsynthesis size fractionation is necessary.

SynthesisofAuNanoparticleswithinDendrimers.Various dendrimer molecules have also been used to stabilize metal nanoparticles (including Au). Pioneering work by Crooks and co-workers4a substantially defined the nanoparticle stability inside the dendritic network and its further utilization (Figure 2). The advantagesof the dendrimer approach include the following: (1) The completeness of reduction of the metal ions sequestered inside the dendrimer yields some stoichiometric control over the nanoparticle size. The dendrimer arms are very effective at preventing nanoparticle aggregation. (2) The loose steric aspects of dendrimers also allow the encapsulated nanoparticles to participate in various catalytic reactions. (3) The terminal groups on the dendrimer periphery can be modified for purposes of nanoparticle solubility in different media, among other goals.

The encapsulation of metal ions inside dendrimers involves, variously,electrostaticandcomplexationinteractions;thedendrimer interior includes coordinating groups such as -OH, NH2,o r -COOH. Stable monodisperse Au NPs can be synthesized using quaternary ammonium-terminated fourth- and sixth-generation dendrimers without further purification steps. Water-soluble Au8 nanodots have also been prepared within hydroxy-terminated second- and fourth-generation (G2-OH and G4-OH) dendrimers,4b and they fluoresce at short wavelengths that change with the size of the Au nanoparticle. In contrast, thiolated Au nanoparticles do not emit at shorter wavelengths but instead emit at relatively size-independent near-infrared wavelengths.

Making Networks and Films of Thiolate-Protected Au

Nanoparticles. The formation of mono- and multilayer films of small-ligand-stabilized metal nanoparticles is a significant research topic given the variety of possible applications (separations, chemical sensors) and their significance in fundamental science (such as the ligand dependence of electron hopping conductivity within the films). The formation of gold colloid monolayerswasreportedinLangmuirin1993usingelectrophoretic deposition,5 and the sophistication of the approaches has grown rapidly. Techniques1c employed to form multilayers include (1) ligand exchanges combined with dithiol linkers, (2) electrostatic or coordinative interactions with groupings present on the nanoparticle surface, (3) covalent linking using amide coupling, and (4) the LB method. A particularly straightforward method is coordinative coupling between carboxylate or sulfonate groups on the nanoparticle ligand shell with metal ions (such as Cu2þ or Zn2þ). Multilayer films can be prepared in simple two-step dip and rinse cycles as illustrated in Figure 3.6a Polyelectrolytes such as cationic poly(allylamine hydrochloride) can also be used; in this case, the nanoparticle is itself effectively a polyelectrolyte imitating the layer-by-layer (LBL) technique.6b-6d By choice of the organic polyelectrolyte charge complement, this procedure is effective at building multilayers whether the nanoparticle ligand shell is cationic or anionic.

Multimer Assemblies of Au Nanoparticles. Multimers of nanoparticles are interesting chemical objects. A seemingly general approach to preparing them, say by ligand exchanges on the nanoparticle producing linker connections, encounters the challenge of controlling the number of such exchanges, with its underlying statistics. A sample containing an average of three exchanged ligands will actually contain a distribution of exchanged ligands, and excessive exchanges will yield nanoparticle aggregates. A solid-phase approach7 can limit the number of linker ligand exchange sites by providing them with large (in comparison) polystyrene beads as illustrated in Figure 4. In the first step, a single acid-terminated thiolate is incorporated onto a 2 nm butanethiolate-protected nanoparticle; that entity is then released from the bead, and the resulting singly functionalized nanoparticle can become linked into dimers by amide coupling with aliphatic diamines. This method produced

Figure 1. Simplified X-ray crystal structure of [Oct4Nþ]

[Au25(SC2Ph)18-] adapted from ref 2e. Phenylethylene R groups (shown only for one semiring) have two different bridging thiolate coordinations. The -SR-Au-SR-Au-SR- semirings in the reduced MPC do not align in a plane, exhibiting a puckering of thesemiringthatishighlightedbybrightercoloring.(Au=yellow, S=violet,C=green,andHnotshown).Reprintedwithpermission from ref 2e. Copyright 2008 American Chemical Society.

(3) (a) Schmid, G.; Pfeil, F.; Boese, R.; Bandermann, F.; Meyer, S.; Calis, G. H.

M.; van der Velden, J. W. A. Chem. Ber. 1981, 114, 3634. (b) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. J. Chem. Soc., Chem. Commun. 1994, 801– 802. (c) 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, 14,1 7–30. (4) (a)Scott, R. W. J.; Wilson, O. M.; Crooks, R. M. J. Phys. Chem. B2005, 109, 692–704. (b) Zheng, J.; Petty, J. T.; Dickson, R. M. J. Am. Chem. Soc. 2003, 125, 7780– 7781.

Perspective Sardar et al.

nanoparticleswhere60-70%weredimers,alongwithsomesingle particles and trimers.

In another nanostructure assembly creation, Stellacci and coworkers8 prepared nanoparticles with mixed monolayers of functional ligands such as nonanethiolate and 4-methylbenzenethiolate. The disparities between these functional ligands caused a form of striped surface ordering, as in Figure 5A,B. Subsequent ligand exchange reactions (with 1-mercaptoundecanoic acid, MUA), according to the “hairy ball theorem”, were directed to the “poles” of the nanoparticles. Coupling of the MUA acid groups with a diamine (1,6-diaminohexane) leads to the formation of 1D nanoparticle chains. The interparticle distance inside the chain could be controlled by using different diamines such as O,O0-bis(2 aminoethyl)octadecaethylene glycol (EGDA).

Gold as a Catalytic Nanoparticle. Although gold nanoparticles have been used for many different purposes, their catalytic properties were for decades considered to be weak or absent. It was an exciting discovery then when Haruta and Hutchings simultaneously and independently9a,9b showed that gold could be very active, in particular, for the heterogeneous low-temperature oxidation of CO. It was found that bare gold nanoparticles were not active but when on a metal oxide support, such as R-

Fe2O3, became excellent catalysts for the oxidation of CO. It was first considered that the high activity resulted from a new type of composite oxide catalyst, but after a detailed electron microscopy study, it was found that the most active catalysts were small gold nanoparticlesapproximately2-4nmindiameter.Thecatalytically active nanoparticles form a reconstructed structure with the

Figure 2. Cartoon depicting the synthesis of various metal nanoparticles using dendrimers as capping ligands. Reprinted with permission from ref 4a. Copyright 2005 American Chemical Society.

Scheme 1. Scheme for Gold Nanoparticle Synthesis by the Brust Two-Phase Approach

(8) DeVries, G. A.; Brunnbauer, M.; Hu, Y.; Jackson, A. M.; Long, B.; Neltner, B. T.; Uzun, O.; Wunsch, B. H.; Stellacci, F. Science 2007, 315, 358–361.

(9) (a) Sanchez, R. M. T.; Ueda, A.; Tanaka, K.; Haruta, M. J. Catal. 1997, 168, 125–127. (b) Hashmi, A.S. K.; Hutchings, G. J.Angew. Chem., Int. Ed. 2006, 45,7896– 7936. (c) Thomas, K. G.; Kamat, P. V. Acc. Chem. Res. 2003, 36, 8–898.

Sardar et al. Perspective

substrate, and CO adsorption would proceed on the adjacent metal oxide. The reaction is thought to involve carbonate-like intermediates decomposing to CO2 upon desorption from the surface. This catalytic discovery spurred a substantial body of other studies on heterogeneous gold catalysis, including the hydrogenation of alkenes or alkynes, hydrosilylation, oxidation of alcohols, and photocatalysis.9c Au nanoparticles chemically attached to glassy carbon electrodes have also been used for the oxidation of CO and CH3OH resulting in the formation of CO2.1b

Electrochemical Reactivities of Au Nanoparticles

Three categories of electrochemical behavior of solutions of thiolate-protected Au nanoparticles have been delineated. These categoriescrosstheso-calledmetal-to-moleculetransitionregionof nanoparticle size, and their voltammetries are labeled1c bulkcontinuum voltammetry,quantized double-layer charging voltammetry,andmolecule-likevoltammetry.Theobservedvoltammetric currents in all three are controlled by mass transport rates; they differ in the potential dependence of thecurrents(justas is thecase for small redoxmolecules).

Another form of nanoparticle voltammetry involves cases where the protecting ligand shell is itself electroactive. The redox-labeled voltammetric reactions add to those underlying the nanoparticle itself. Another form of electroactivity is that of a stronglyheld, redox-activeconjugatewith the nanoparticleligand shell, which has been used in various forms of DNA sensing in particular.

Bulk-Continuum Voltammetry. Just as ionic space charges;the electrical double layer;exist at all electrified metal/electrolyte solution interfaces, nanoparticles in solutions (colloids, metal sols, regardless of the metal, and semiconductor nanoparticles) have double layers with ionic surface excesses on the solution side that reflect any net electronic charge residing on the metal nanoparticle surface (or its capping ligand shell). In this light, one can say that all metal-like nanoparticles are intrinsically electroactive and act as electron donor/ acceptors to the quantitative extent of their double-layer capacitances. The electron charge storage capacity, per nanoparticle, depends on the nanoparticle size (surface area), nanoparticle double-layer capacitance

CDL, and potential (relative to nanoparticle zero charge). This capacity can be quite substantial; for example, a 10-nm-diameter nanoparticle with CDL=120 aF (equivalent to 40 μF/cm2)c an store ∼750 e/V. This capacity is capable, as a “colloidal micro- electrode”, of driving electrochemical reactions such as proton reductiontoH2.Thequantitativedemonstrationsofthisproperty by Henglein et al.10a and Gratzel et al.,10b starting in 1979, represented the beginning of the modern understanding of the electrochemistry of metal nanoparticles.

In the earliest experiments, nanoparticles were charged by chemical reactions. The transition to electrochemical control by

Figure 3. Schemefor forming multilayerfilms frommixed-monolayer Au nanoparticles. Adapted from ref 6c. Reprinted with permission from ref 1c. Copyright 2008 American Chemical Society.

Figure 5. Syntheticapproachof1Dnanoparticlechainformation. Reprinted with permission from ref 8. Copyright 2007 American Association for the Advancement of Science.

Figure 4. Synthetic outline of gold-nanoparticle dimer synthesis. Reprinted with permission from ref 7. Copyright 2004 The Royal Society of Chemisty.

(10) (a) Henglein, A. J. Phys. Chem. 1979, 83, 2209. (b) Kiwi, J.; Gratzel, M. J. Am. Chem. Soc. 1979, 101, 7214.

Perspective Sardar et al.

Ung et al.1 was made by showing that solutions of Ag nanoparticles capped with poly(acrylic acid) could also be charged at (macroscopic) working electrodes, diffusing to undergo electron transfer at electrode/electrolyte interfaces. Although the observed current-potential curves (at potentials insufficiently positive to oxidize Ag) were featureless, the currents exceeded background values and varied with [time]-1/2 (in potential step experiments) and with [rotation rate]1/2 (in rotated disk electrode experiments);characteristics of current control by diffusing entities according to the well-known Cottrell and Levich equations, respectively,

IðtÞ¼ nFADNP 1=2CNP BULK where DNP and CNP BULK are the Ag nanoparticle diffusion coefficientandconcentration.Thus,eventhoughametalnanoparticle yields no voltammetric wave and exhibits no discrete “formal potential”, its electrochemical charging is observable and demonstrable by the application of mass transport criteria.

(Parte 1 de 4)