Search

The robust, irreversible adsorption of [omega]-ferrocene hexanethiolate-protected gold nanoparticles (composition ca. {[Au.sub.225][(SC6Fc).sub.43})] on electrodes provides an opportunity to investigate their submonolayer and monolayer films in nanoparticle-free solutions. Observations of nanoparticle adsorption on unmodified electrodes are extended here to Au electrodes having more explicitly controlled surfaces, namely self-assembled monolayers (SAMs) of alkanethioiates with [OMEGA]-sulfonate, carboxylate, and methyl termini, and in different [Bu.sub.4][N.sup.+][X.sup.-] electrolyte [X.sup.-] = [C.sub.7][H.sub.7]S[O.sub.3], CL[O.sub.4.sup.-], C[F.sub.3]S[O.sub.3.sup.-], P[F.sub.6.sup.-], N[O.sub.3.sup.-]) solutions in C[H.sub.2][Cl.sub.2]. The nanoparticle surface coverage ([[GAMMS].sub.NP]) and the stability of the adsorbed nanoparticle film to repeated ferrocene/ferroeenium redox cycling decrease in the order of sulfonate > carboxylate > methyl terminated SAM, with increasing hydrophobicity of X- and with increasing alkyl chain length. The results are consistent with the proposal that the strong surface adsorption is jointly associated with the polyfunctional character of the nanopmlicles, analogous to entropically driven adsorptions of polymeric ions on charged surfaces, and with lateral, ion-bridged nanoparticle-nanoparficle interactions.

The molecular ion of the nanoparticle [Au.sub.25][([SCH.sub.2][CH.sub.2]Ph).sub.18 ([A.sub.25][(SR).sub.18]) is observed at 7394 Da in fast atom bombardment (FAB, Xe atoms) ionization mass spectrometry using a 3-nitrobenzyl alcohol matrix. A distinctive pattern of positive fragment ions is evident in the mass interval 5225-7394 Da, where peaks are seen for successive mass losses equivalent to [R.sub.2]S entities. Because the [Au.sub.25][([SCH.sub.2][CH.sub.2]Ph).sub.18] nanoparticle structure is crystallographically known to consist of a centered [Au.sub.13] icosahedral core surrounded by six [Au.sub.2][(SR).sub.3] semirings, the [R.sub.2]S loses are proposed to represent serial rearrangements and decompositions of the semiring structures. Mass losses equivalent to [R.sub.2][S.sub.2] and [R.sub.2] entities also appear at the lower end of this mass interval. The most intense spectral peak, at m/z = 5246 Da, is assigned to the fragment [Au.sub.25][S.sub.10], from which all of the [CH.sub.2][CH.sub.2]Ph organic units have been cleaved but from which no gold atoms have been lost. A different pattern of fragmentation is observed at lower masses, producing ions corresponding to serial losses of one gold atom and varied numbers of sulfur atoms, which continues down to a [Au.sub.9][S.sub.2] fragment. FAB mass spectra of the Au nanoparticle are much easier to interpret than laser desorption/ionization spectra, but they show more extensive fragmentation than do electrospray and low laser pulse intensity MALDI spectra. The loss of [R.sub.2]S fragmentation in FAB is distinctive and unlike that seen in the other ionization modes. The FAB spectrum for the nanoparticle [Au.sub.25][(S[([CH.sub.2]).sub.9][CH.sub.3]).sub.18] is also reported; its fragmentation parallels that for [Au.sub.25][([SCH.sub.2]-[CH.sub.2]Ph).sub.18], implying that this nanoparticle has the same surprising stellated (staples) structure.

An unwalled electrochemical cell formed by a sessile droplet of 1-butyl-3-methylimidazolium hexafluorophosphate resting on electroactive surfaces--self-assembled alkanethiolate monolayers, films of Au nanoparticles and of ferrocene cast from evaporated solution droplets--is used to investigate the voltammetry and surface energy of these electrode/ionic liquid interfaces. Auxiliary and quasi-reference 25-[micro]m-diameter Pt wire electrodes are inserted into the top of the sessile droplet. The cyclic voltammetry of a film of hexanethiolate-protected [Au.sub.140] nanoparticles shows that they undergo quantized double layer charging at the ionic liquid interface, accompanied by insertion of hexafluorophosphate anions into the nanoparticle film. The contact angle of the ionic liquid droplet resting on the nanoparticle film is observed to decrease when the gold electrode potential is controlled to oxidize the interface to the [Au.sub.140.sup.1+] state. The increase in surface charge and the alteration in chemical composition of the oxidized [Au.sub.140] film surface combine to cause a decrease in surface tension at the interface and the ionic liquid to become more wetting of it. Analogous behavior is seen with ferrocene films. In both cases, the changes in contact angles are not reversed upon returning potentials to reducing values. Reductive charging of the [Au.sub.140] surface does not occur; the imidazolium cation does not insert. Films of [Au.sub.38] nanoparticles do not show voltammetric currents for oxidation of the surface to [Au.sub.38.sup.1+] and, correspondingly, no changes in contact angle.

The well-defined one-electron steps in the voltammetry of solutions of the nanoparticles Au38(SC2Ph)24 and Au140(SC6)53 enable preparation of solutions containing for example Au38(SC2Ph)24 and Au 38(SC2Ph)24 [[.sup.(CIO4)].sup.-] nanoparticles in known relative proportions. Differences in pre-exponential terms in the activation equations for the two nanoparticles are a smaller contributor to the rate constant difference and could be partly ascribed to differences in tunneling distances.

We demonstrate the efficacy of ion-pair chromatography for separations of samples of charged, polydisperse, water-soluble gold nanoparticles protected by monolayers of N-acetyl-L-cysteine and of tiopronin ligands. These nanoparticle mixtures have 1-2-nm-diameter Au core sizes as estimated from UV-visible spectra of the separated components. This size range encompasses the transition from bulk metal to molecular properties. The nanoparticle mixtures were resolved, the smallest nano-particles eluting first, on an octadecylsilyl (C18) column using isocratic elution with a methanol/water mobile phase containing tetrabutylammonium fluoride ([Bu.sub.4][N.sup.+][F.sup.-]) and phosphate buffer. The column retention increases with [Bu.sub.4][N.sup.+][F.sup.-] concentration, lowered pH, and decreasing methanol volume fraction. The retention mechanism is dominated by ion-pairing in either the mobile phase or at the stationary/mobile-phase interface. Size exclusion effects, used in many previous nanoparticle separations, are insignificant.

The way in which the efficiency of molecular quenching by Au nanoparticles depends on nanoparticle size is reported for dynamic and static quenching of fluorescence. The results for both the sets of monolayer-protected Au clusters (MPCs) have revealed a clear increase in quenching efficiency at increased MPC core size.

The Brust synthesis of thiolate-protected gold clusters has been modified to produce identifiable proportions of a hexanethiolate-protected [Au.sub.225] core nanoparticle that display quantized double layer charging voltammetry consistent with a [Au.sub.225] core dimension. Transmission electron microscopy (TEM) and thermogravimetric results indicate an average nanoparficle formula of [Au.sub.225][(S[[(C[H.sub.2]).sub.5]-C[H.sub.3])].sub.75]. A simulated pulse voltammogram that accounts for the TEM nanoparticle dispersity matches reasonably well with that of the polydisperse synthetic sample containing the [Au.sub.225] component. In confirmation of the size determination, an HPLC analysis using ratiometric absorbance and electrochemical detectors gives a core radius of 1.0 nm for the [Au.sub.225] nanoparticle.

A molecule-like substituent effect on redox formal potentials in the nanoparticle series Au(sub 38)(SPhX)(sub 24) is discovered. The ligand monolayer of the nanoparticles is shown to play a strong role in determining the electronic energies in the nanoparticle core.

Sequential injections of single electrons (or holes) into the cores of [Au.sub.140] hexanethiolate monolayer-protected clusters (MPCs) occur at measurably different electrochemical potentials owing to the extremely small (subattofarad) values of the single MPC capacitance ([C.sub.MPC]) of the nanoparticle. The potential increment for each sequential injection is [DELTA]V = e/[C.sub.MPC]. The dependence of [DELTA]V on the concentration of supporting electrolyte (from 1 to 100 raM), measured using square wave voltammetry, is shown to be caused, primarily, by changes in the diffuse double layer component ([C.sub.DIFFUSE]) of [C.sub.MPC]. The dependence of [C.sub.DIFFUSE] on [r.sub.core], the radius of the nanoparticle, is considered. Additionally, significant changes in the magnitude of the compact double layer component ([C.sub.COMPACT], equivalent to the Stern layer) of [C.sub.MPC] were induced by adding hydrophobic solvent components such as hexane or dodecane or by introducing hydrophobic electrolyte ions (tetrabutyl-, tetrahexyl-, and tetraoctylammonium, perchlorate, and tetraphenylborate). These changes are interpreted as specific solvation and ion penetration of the hexanethiolate monolayer. For brevity we will refer to these phenomena as solvation/penetration phenomena.

The kinetics of exchanges of phenylethanethiolate ligands (PhC2S-) of the monolayer-protected clusters (MPCS) Au(sub 38) (SC2Ph)(sub 24) and Au(sub 140) (SC2 Ph)(sub 53) with p-substituted arylthiols (p-x-PhSH), where X = NO2, Br, CH3, OH3 and OH are compared. The rates of later stages of exchange differ for Au(sub 38) and Au(sub 140) cores.

Experimental determinations and comparison of thermal and optical energy barriers for electron transfers in a specific redox polyether melt, namely the metal complex Ru(bpy350)2(CN)2 where bpy350 is 4,4'-dimethyl-2,2'-dipridyl alkyl-bonded to 350 MW polyethylene glycol monomethyl ether is described. The energy barrier results revealed that the observed thermal electron transfer barrier is only an apparent one, being instead controlled by ion atmosphere relaxation processes, whereas the optical energy barrier corresponds to that of the actual electron-transfer process itself.

Redox chemistry in semisolid molten salts ionic liquids of DNA is described in which counterions of the phosphates are redox-active metal complexes with bipyridine ligands labeled with MW 350 poly-(ethylene glycol) (PEG) 'tails. Other redox metal complexes were added to the M(bpy(sub 350) DNA melt: (a) the PEG-tailed metal bipyridine complexes Fe(bpy(sub 350))(sub 3)(ClO(sub 4))(sub 2) and Ru(bpy(sub 350))(sub 3)(ClO(sub 4))(sub 2) and (b) the nontailed complexes Os(bpy)(sub 3)Cl(sub 2) (bpy =2,2'-bipyridine) and Os(bpy)(sub 2)dppzCl(sub 2).

Electrons are transported within polymeric films of alkanethiolate monolayer-protected Au clusters (MPCs) by electron hopping (self-exchange) between the metal cores. The surrounding monolayers, the molecular linkers that generate the network polymer film, or both, presumably serve as tunneling bridges in the electron transfers. This paper introduces a steady-state electrochemical method for measuring electron hopping rates in solvent-wetted and swollen, ionically conductive MPC films. The films are network polymer films of nanoparticles, coated on a rotated disk electrode that is contacted by a solution of a redox species (decamethylferrocene, Cp*Fe). Controlling the electrode potential such that the film mediates oxidation of the redox probe can force control of the overall current onto the rate of electron hopping within the film, which is characterized as the apparent electron diffusion coefficient [D.sub.E]. [D.sub.E] is translated into an apparent electron hopping rate [k.sub.ET] by a cubic lattice model. The experiment is applied to MPC network polymer films linked by [alpha], [omega]-alkanedithiolates and by metal ion-carboxylate connections. We evaluate the dependencies of apparent hopping rate on Cp*Fe concentration, film thickness, electrode potential relative to the Cp*Fe formal potential, film-swelling solvent, and temperature. The apparent hopping rates are in the [10.sup.4]-[10.sup.5] [s.sup.-1] range, which is slower than those for the same kind of MPC films, but in a dry (nonswollen) state measured by electronic conductivities.

Fast-scan cyclic voltammetric (FSCV) detection of the separation of small (< 2-nm core diameter) organothiolate monolayer-protected gold clusters (MPCs) by high-performance liquid chromatography (HPLC) is demonstrated in this paper. Cyclic voltammograms taken at 1-s intervals (scan rate 150 V/s) display distinctive current-potential patterns for different MPCs, aiding in their identification. Previous amperometric detection of MPC separations required multiple HPLC runs at different electrode potentials to construct hydrodynamic current-potential voltammograms. Theory for ratiometric response of photodiode array absorbance and constant potential amperometry detectors was further tested on the fast-scan voltammetric detector and gave reasonable relative core size results for a mixture of [Au.sub.140] and [Au.sub.38] MPCs. In the first analysis of mixed MPC monolayer composition that is not simply an overall average result, preliminary HPLC experiments on MPCs with mixed hexanethiolate and ferrocenyl-octanethiolate monolayers resolved peaks that are shown to correspond to different average numbers of ferrocene-labeled ligands per MPC.

Reversed-phase HPLC chromatographic separations of Au monolayer protected clusters (MPCs) in the size range 1-2 nm (diameter) are detected electrochemically (ECLC) on the basis of double layer charging of the metal-like nanoparticles by the detector electrode. Discrete double-layer charging peaks are detected for the nanoparticles with resolution similar to that of the photodiode array detector. The ECLC experiment further seeks not solely to detect the nanoparticle peaks, but to estimate the nanoparticle sizes from the electrochemical responses. Hydrodynamic current--potential curves of nanoparticles eluted in different peaks were obtained by repeating chromatographic separations at different working electrode potentials; the [E.sub.1/2] separations between the voltammetric waves give estimates of cluster capacitance and, thus, size. Other approaches to core size determination are discussed, including determination based upon relative values of optical absorbance and current for each eluted peak. The data show that, at least for alkanethiolate MPCs, the elution order on the HPLC column employed is small clusters first and that there are significant differences in reversed-phase chromatographic retention times for clusters with apparently very similar core sizes over the range of 1-2-nm diameter.

The electron-transfer rates of the CO(super II/I) couple are controlled by the diffusive relaxation of the ionic atmosphere around the reaction pair, and the inactivity of the bound Co complexes can be attributed to the very low mobility of the anionic phosphate groups in the DNA counterion. The study shows that the measured Co(super II/I) charge transport and electron-rate constants reflect the diffusive mobility of the perchlorate counterion more than the intrinsic Co(super II/I).

An end-opened single wall carbon nanotubes are solubilized by aniline and then purified by flash chromatography. Other solubilizing procedures used in the study are explained.

This paper describes low-temperature voltammetry of purified hexanethiolate-coated monolayer-protected [Au.sub.140] clusters (C6 MPCs). Lowered temperatures enhance the resolution of quantized double layer (QDL) charging peaks in differential pulse voltammetry (DPV) observations. As many as 13 resolved peaks are seen in illustrative voltammetry at 263 K in C[H.sub.2][Cl.sub.2] solvent, and the concept of voltammetric peak capacity is introduced. For the one-electron MPC charge steps surrounding the [E.sub.PZC] of the MPC (small numbers of electrons added or removed from the core), the capacitance [C.sub.CLU] of the MPCs (measured from the voltage spacing between charging peaks) increases by ~15% as the solvent temperature is lowered from 273 to 210 K. The experimental [C.sub.CLU] temperature dependency (d[ln([C.sub.CLU])]/dT [approximately equal to] -0.0025, in 0.1M electrolyte) is discussed in light of temperature dependencies of the compact and diffuse double layer capacitances. It is concluded that the observed temperature dependence is probably a mixed diffuse, compact dependence. The regular voltage spacing of MPC charging peaks near the potential of zero charge is generally consistent with electrical double layer properties, but the irregular pattern of charging of the nanoparticles seen at higher charge states suggests intervention of the incipient molecular behavior of [Au.sub.140] cores in the spacing of energies at which further electrons are added or removed.

The sorption of carbon dioxide in the extremely poisonous, semisolid hybrid redox polyether melt, [Co-(phenanthroline)3](MePEG-SO3)2, wherein MePEG-SO3 is a MW 350 polyether made sulfonate anion, significantly enhances charge movement in the molten salt material. Moreover, since the counterion relaxation is restricted by polyether 'solvent' variations, this is a novel mode of solvent dynamics restriction of electron transfer.

Polydisperse samples of Au nanoparticles protected with monolayers of hexanethiolate ligands (C6 MPCs) and with mixed monolayers of hexanethiolate and mercaptoundecanoic acid (C6/MUA MPCs) have been chromatographically separated using C8 120-[Angstrom] columns and acetone/ toluene mobile phase. The spectral details of eluted peaks and of quantized double-layer charging features in the differential pulse voltammetry of collected fractions were used to show that the elution orders of C6 MPC mixtures and of C6/MUA MPC mixtures were different For C6 MPCs, the smallest MPCs were eluted first, whereas the smallest C6/MUA MPCs were eluted last. The reversal of order of elution was rationalized in terms of intermolecular interactions with the stationary phase, dominant for the C6 MPC, being suppressed by the heightened polarity of the monolayer surface of the C6/MUA MPCs, making a size exclusion mechanism dominant. The range of apparent core diameters of the separated nanoparticles was 1.3-2 nm.

Ethidium intercalation has been investigated as a means of inducing binding of Au nanoparticles to DNA. The ethidium sites are attached to the nanoparticles as thiolate ligands, using 3,8-diamino-5-mercaptododecy1-6-phenylphenanthridinium (ethidium thiolate). Each nanoparticle bears only one or two ethidium thiolate ligands. The rest of the thiolate monolayer ligands on the monolayer-protected Au clusters (MPCs) were either N-(2-mercaptopropionyl)glycine (tiopronin/ethidium MPC) or trimethyl(mercaptoundecyl)ammonium (TMA/ethidium MPC). In solution mixtures of DNA and MPCs, the energy-transfer quenching of the ethidium ligands by the metal-like MPC core is partially released by ethidium binding to DNA, as observed by an increase in the intensity of ethidium fluorescence. Binding of the cationic TMA/ethidium MPC to DNA was efficient and rapid. The negatively charged tiopronin/ethidium MPC, in contrast, exhibits slow intercalation kinetics, relative to ethidium cation not attached to an MPC. The slow kinetics were analyzed as two competing binding interactions. The tiopronin/ethidium MPC binding to DNA was imaged by AFM.

Measurements of the rate of electron transfers between electrodes and attached monolayers of gold nanoparticles (monolayer-protected clusters, MPCs) are described. The attached MPC monolayers exhibit well-defined single electron double layer charging voltammetry of the Au cores.

Description about the Co(III/II) and Co(II/I) voltammetry of thin films of a polyether-tailed cobalt complex melt that are contacted by a CO2 phase and which coat an electrochemical cell composed of working micro-, reference and counter electrodes exposed on a miniature platform is provided. The adaption of voltammetric methods to CO2-based media is challenging owing to low dielectric strength.

The report states that the enantionselective chemisorption of (S)-and (R)-propylene oxide was measured on a Pd(III) surface, which was chirally modified using (S)-and (R)-2 butanol. The reflection - absorption infrared spectroscopic (RAIRS) data suggests that absorbed 2-butanol forms 2-butoxide species, when heated to ~150 K and converts to a ketone with a concomitant loss in chirality at 200 K.

The study was conducted to investigate the adsorption of ethyl vinyl ketone (EVK) on the Ge (100)-2x1 and Si(100)-2x1 surfaces using multiple internal reflection infrared spectroscopy and density functional theory quantum chemistry calculations. The results of the study suggest that due to the high reactivity of clean semiconductor surfaces, thermodynamic selectivity and control play important roles in their fucntionlization favouring the use of Ge for selective attachment of multifunctional organics.

DFT(B3LYP/6-31G(super *) and ab initio molecular orbital theory (QCISD/cc-pVDZ), which is used to investigate several possible mechanisms involving free radical intermediates as well as their protonated forms for processes related to the coenzyme B(sub 12)-dependent rearrangement catalyzed by ethanolamine ammonia lyase is presented.

A density functional theory study of the ruthenium-catalyzed olefin metathesis reactions is presented. It is stated that the calculated insertion barrier for the NHC systems was lower than that of the (PCy(sub 3))(sub 2)Cl(sub 2)Ru=CHPh system.

The study was conducted to report the electronic conductivity properties of network polymer films of Au(sub 140) nonoparticles. The film conductivities respond to exposure to organic vapours, decreasing in electronic conductivity and increasing in mass(quartz crystal microgravimetry, QCM).

Research into the rate of electron self-exchange reactions between discretely charged metal-like cores of nanoparticles in multilayer films is presented. It was demonstrated that the diffusion-like electron transport in multilayer films can be used to obtain nanoparticle-nanoparticle electron-transfer dynamics.

New liquid ionic materials based on DNA salts of polyether-tailed cations are reported to have electrochemical properties.

Solid-state, mixed-valent monolayer-protected Au clusters were studied to determine their electronic conductivity.

The study describes experiments exploring the role of the redox-inactive counterion in hybrid redox polyether molten salts. As these materials are undiluted, they intrinsically contain 'solvent-separated ion pairs', where the 'solvent' is the polyether tail attached to the redox moiety.

This paper describes microelectrode voltammetry measurements of serf-diffusion coefficients and of heterogeneous Co(II/III) electron-transfer rate constants (k(super o)) in undiluted molten salts of three cobalt tris(bipyridine) perchlorate complexes in which the bipyridine ligands are 'tailed' with poly(propylene oxide) and poly(ethylene oxide) oligomers. The self-diffusion coefficients are measured with potential step chronoamperometry and range from 10(super-12) to 10(super -17) cm(super 2)/s, while the quasi-reversible reaction rate constants are measured using cyclic voltammetry and small potential steps and range from 10(super -7) to 10(super -12) cm/s. The k(super o) measurements are unusual in that when rate constants become smaller, the reaction remains quasi-reversible, because of concurrently decreasing self-diffusion rates. The measurements are, furthermore, accomplished in the face of uncompensated resistances that range from mega- to gigaohms, which is made possible by the combination of microelectrode properties and small diffusivities. The melt in which serf-diffusion and k(super o) values are smallest is at a temperature below its nominal glassing transition and in the regime of molecule-scale diffusion profiles.

Issues concerning the effect of core charge and solvent refractive index on the surface plasmon absorbance of alkanethiolate monolayer-protected gold clusters are presented.

Results show that activation of homogeneous electron transfers in molten salts encounter large activation barriers. Data further indicate that physical diffusion rates and viscosity is dependent on the number of ethylene oxide segments in the polyether tails.

A study is presented of the solid state voltammetry of the two reduction steps of an anthrequinone molten salt and its disulfonated analogue with focus on the charge transport rates.

Issues concerning ensembles of nanoparticles that are chemically anchored at a macroscopic gold electrode/electorate solution interface which undergo an electrochemical quantized charge are discussed.

We describe the use of the Taylor dispersion method to determine the diffusion coefficients of monolayer protected clusters (MPCs). The MPCs have nanometer Au cores ([approximately] 1- 5 nm diameter) coated with a passivating, self-assembled monolayer of thiolate ligands ([approximately]40-600). The results are found to agree with electrochemically measured diffusion coefficients, validating both approaches for measuring nanoparticle diffusivities. Taylor diffusion coefficients are used to calculate MPC hydrodynamic diameters under the sticking and the slip boundary conditions of the Stokes-Einstein equation, which are compared to the overall MPC diameter-estimated as the core diameter (measured by TEM) plus the length (2L) of two fully extended ligands. Better agreement between calculated and experimental hydrodynamic diameters of alkanethiolate-coated MPCs in organic solvents are obtained using slip boundary conditions, implying free draining of solvent through the outer portion of the extended chains. The hard-sphere (sticking) version of the Stokes-Einstein relation predicts unrealistically small hydrodynamic diameters. Similar but less clear-cut results are obtained for tiopronin-coated MPC ions in aqueous media.

This report describes how the electrochemical double-layer capacitances of nanometer-sized alkanethiolate monolayer-protected Au clusters (MPCs) dissolved in electrolyte solution depend on the alkanethiolate chain length (C4 to C16). The double-layer capacitances of individual MPCs ([C.sub.CLU]) are sufficiently small (sub-attoFarad, aF) that their metal core potentials change by > 0.1 V increments for single electron transfers at the electrode/solution interface. Thus, the current peaks observed are termed 'quantized double layer charging peaks', and their spacing on the potential axis varies with [C.sub.CLU]. Differential pulse voltammetric measurements of [C.sub.CLU] in solutions of core-size-fractionated (i.e., monodisperse) MPCs are compared to a simple theoretical model, which considers the capacitance as governed by the thickness of a dielectric material (the monolayer, whose chain length is varied) between concentric spheres of conductors (the Au core and the electrolyte solution). The experimental results fit the simple model remarkably well. The prominent differential pulse voltammetric charging peaks additionally establish this method, along with high-resolution transmission electron microscopy and laser ionization-desorption mass spectrometry, as a tool for evaluating the degree of monodispersity of MPC preparations. We additionally report on a new tactic for the preparation of monodisperse MPCs with hexanethiolate monolayers.