Ferrocenated Au Nanoparticle Monolayer Adsorption on Self

Jul 14, 2009 - Martin Greenwald stands in a windowless control room at Massachusetts Institute of Technology, holding. ... 1155 Sixteenth Street N.W...
0 downloads 0 Views 1MB Size
Download PDF
Anal. Chem. 2009, 81, 6960–6965

Ferrocenated Au Nanoparticle Monolayer Adsorption on Self-Assembled Monolayer-Coated Electrodes Rajesh Sardar, Christopher A. Beasley, and Royce W. Murray* Kenan Laboratories of Chemistry, University of North Carolina, Chapel Hill, North Carolina, 27599 The robust, irreversible adsorption of ω-ferrocene hexanethiolate-protected gold nanoparticles (composition ca. {Au225(SC6Fc)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 alkanethiolates with ω-sulfonate, carboxylate, and methyl termini, and in different Bu4N+Xelectrolyte (X- ) C7H7SO3-, ClO4-, CF3SO3-, PF6-, NO3-) solutions in CH2Cl2. The nanoparticle surface coverage (ΓNP) and the stability of the adsorbed nanoparticle film to repeated ferrocene/ferrocenium 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 nanoparticles, analogous to entropically driven adsorptions of polymeric ions on charged surfaces, and with lateral, ion-bridged nanoparticle-nanoparticle interactions. Monolayer ligand-protected Au nanoparticles (NPs) have received substantial research attention given their potential as building blocks for nanoscale assembly creation,1-8 for organic or biomolecular detection,9,10 for device fabrication,11-13 and for * To whom correspondence should be addressed. E-mail: [email protected]. (1) Wang, G.; Murray, R. W. Nano Lett. 2004, 4, 95–101. (2) Warner, M. G.; Hutchison, J. E. Nat. Mater. 2003, 2, 272–277. (3) Wyrwa, D.; Beyer, N.; Schmid, G. Nano Lett. 2002, 2, 419–421. (4) 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. (5) Sardar, R.; Shumaker-Parry, J. S. Nano Lett. 2008, 8, 731–736. (6) Sung, K.-M.; Mosley, D. W.; Peelle, B. R.; Zhang, S.; Jacobson, J. M. J. Am. Chem. Soc. 2004, 126, 5064–5065. (7) Dai, Q.; Worden, J. G.; Trullinger, J.; Huo, Q. J. Am. Chem. Soc. 2005, 127, 8008–8009. (8) Sardar, R.; Heap, T. B.; Shumaker-Parry, J. S. J. Am. Chem. Soc. 2007, 129, 5356–5357. (9) Daniel, M. C.; Astruc, D. Chem. Rev. 2004, 104, 293–346, and references therein. (10) Phillips, R. L.; Miranda, O. R.; You, C,-C.; Rotello, V. M.; Bunz, U, H. F. Angew. Chem., Int. Ed. 2008, 47, 2590–2594. (11) Thomas, K. G.; Kamat, P. V. Acc. Chem. Res. 2003, 36, 888–898. (12) Marinakos, S. M.; Brousseau, L. C., III.; Jones, A.; Feldheim, D. L. Chem. Mater. 1998, 10, 1214–1219.

6960

Analytical Chemistry, Vol. 81, No. 16, August 15, 2009

electronic sensor applications.14,15 There are also fundamental aspects of very small, ligand-protected Au NPs ( carboxylate > methyl, with increasing hydrophobicity of X- and with increasing alkanethiolate chain length. The results are consistent with the hypothesis that the strong surface adsorption is entropically associated with the polyfunctional character of the nanoparticles, analogous to polymeric polyions on charged surfaces,36 and also with lateral nanoparticle-nanoparticle ion-bridged interactions. EXPERIMENTAL SECTION Chemicals. Sodium 3-mercapto-1-propanesulfonate, 8-mercaptooctanoic acid (MOA, 95%), 11-mercaptoundecanoic acid (MUA, 95%), 16-mercaptohexadecanoic acid (MHA, 90%) t-octylammonium bromide (Oct4NBr, >98%), sodium borohydride (NaBH4, >98%), t-butylammonium perchlorate (Bu4NClO4, >99%), tbutylammonium p-toluenesulfonate (Bu4NC7H7SO3, puress), t-butylammonium hexafluorophosphate (Bu4NPF6, puress), t-butylammonium nitrate (Bu4NNO3), and t-butylammonium trifluoromethanesulfonate (Bu4NCF3SO3) were from Aldrich. Toluene (reagent grade), acetonitrile (Optima), methylene chloride (HPLC grade), tetrahydrofuran (HPLC grade), and ethanol (HPLC grade) from Fisher were used as received. HAuCl4 · xH2O (from 99.999% pure gold) was synthesized using a literature procedure37 and stored in a freezer at -20 °C. Water was purified using a Barnstead NANOpure system (18 MΩ). Ferrocene hexanethiol (HSC6Fc) was synthesized by refluxing (35) Stiles, R. L.; Balasubramanian, B.; Feldberg, S. W.; Murray, R. W. J. Am. Chem. Soc. 2008, 130, 1856–1865. (36) Bertrand, P.; Jonas, A.; Laschewsky, A.; Legras, R. Macromol. Rapid Commun. 2000, 21, 319–348, and references therein. (37) Brauer, G. Handbook of Preparative Inorganic Chemistry; Academic Press: New York, 1965.

a mixture of (1.11 g, 3.17 mmol) ω-bromohexane ferrocene38 and thiourea (0.600 g, 7.88 mmol) in ethanol (50 mL) overnight. The reaction mixture was neutralized with NaOH (aq), refluxed for a further 3 h, acidified with HCl to pH ∼2, diluted with water, and extracted with CH2Cl2, thoroughly washing the organic extract phase water. The material obtained after rotary evaporation of the CH2Cl2 product solution was chromatographed on silica gel with ethyl acetate/hexanes. 1H NMR (400 MHz, CD2Cl2) of the thiol gave the appropriate NMR peaks: (δ) 4.0 (m, 9 H), 2.49 (q, J ) 7.2 Hz, 2 H), 2.30 (t, J ) 7.6 Hz, 2 H), 1.56 (m, 2 H), 1.46 (m, 2 H), and 1.32 (m, 5 H) ppm with no dithiol peaks present and no significant line broadening, indicating that the ferrocene groups were in the reduced state. MPC Synthesis. Au225(SC6Fc)43 was synthesized as previously reported.39 Briefly, 3.19 g of HAuCl4 · xH2O dissolved in 100 mL of deionized water was mixed with 5.20 g of Oct4NBr dissolved in 200 mL of toluene and stirred for 30 min at room temperature, giving a clear aqueous phase and an orange-brown toluene phase. The separated organic phase was mixed and stirred for 20 min with added ω-ferrocenyl hexanethiol (2:1 ligand-to-Au mole ratio). The now colorless Au(I)-thiolate polymer solution was cooled to 0 °C, and 3.8 g of NaBH4, dissolved in 10 mL of water, also cooled to 0 °C, was added with vigorous stirring; the solution immediately turned black and some gas evolution occurred. The rapid stirring was continued for 1 h at 0 °C, after which the dark organic phase was collected and the solvent was removed on a rotary evaporator at room temperature. The black solid suspension was stirred in 400 mL of acetonitrile for 6 h, and the solid product was collected and washed with acetonitrile on a fine glass frit. The average Au core size and number of ferrocene ligands were previously determined39 on the basis of transmission electron microscopy, UV-vis absorbance spectroscopy, coulometry, and cyclic and differential pulse voltammetries. Self-Assembled Monolayer (SAM) Formation. Disk Au working electrodes (1.6 mm diameter) were polished with 0.25 µm diamond paste (Buehler), sonicated in water for 5 min, and cleaned electrochemically by potential cycling in a 0.10 M H2SO4 solution. The electrodes were thoroughly rinsed with NANOpure water, CH2Cl2, and dried under a stream of N2. Selfassembled monolayers (SAMs) were prepared by immersing the electrodes in 1.0 mM methanol solutions of the desired thiol for 24 h. The electrodes were rinsed with methanol and dried under N2 flow. In the case of carboxylic acid-terminated thiols, the electrodes were additionally rinsed with 1 mM aqueous NaOH, followed by NANOpure water and methanol, dried under N2 flow, and used fresh. Electrochemistry. Adsorption voltammetry of typical 0.05 mM AuFc NPs in degassed CH2Cl2 solutions containing varying concentrations of electrolyte was done on a SAM-coated working Au disk, in cells containing a Pt wire counter and Ag/ AgCl/3 M KCl (aq) electrodes, using a Model 760C CH (CH Instruments) electrochemical analyzer. Except as otherwise noted, AuFc NP adsorption was induced by a single cyclical potential scan of the SAM-coated electrode potential from -0.1 (38) Yu, C. J.; Wang, H.; Wan, Y. J.; Yowanto, H.; Kim, J. C.; Donilon, L. H.; Tao, C. L.; Strong, M.; Chong, Y. C. J. Org. Chem. 2001, 66, 2937–2942. (39) Wolfe, R. L.; Balasubramanian, R.; Tracy, J. B.; Murray, R. W. Langmuir 2007, 23, 2247–2254.

Analytical Chemistry, Vol. 81, No. 16, August 15, 2009

6961

to +1.0 to -0.1 V vs Ag/AgCl 3 M KCl (aq), in unstirred 0.05 mM AuFc NP solutions in CH2Cl2. The electrode was rinsed with CH2Cl2 and placed in NP-free fresh electrolyte for inspection of the adsorption by cyclic voltammetry of the ferrocenated NP layer. In a few experiments, the freshly prepared SAM-modified electrode was simply immersed in a 0.05 mM ferrocenated NP in degassed CH2Cl2 for 15 min, rinsed with fresh electrolyte solution, and transferred to fresh electrolyte, as above. The CH2Cl2 solutions from which adsorption occurred, and to which the electrode was transferred, contained various concentrations of Bu4N+X- electrolyte (X) ClO4-, PF6-, NO3-, CF3SO3-, or p-toluene sulfonate). Adsorbed surface coverages, ΓNP (mol NP/cm2), were calculated from the charge, Q, under the ferrocene voltammetric current peaks (corrected for background currents) in the NP-free electrolyte solution, Q ) nAFΓNP

Figure 1. Cyclic voltammetry (single scan) at 0.5 V/s of 0.05 mM AuFc NPs in different 1.0 M electrolytes in CH2Cl2 at 3-mercapto-1propanesulfonate SAMs on Au working electrodes.

(1)

where n is the average number of ferrocenes39 per NP, 43, and A is the Au working electrode area, not corrected for microscopic roughness. The experimentally determined values of ΓNP are compared to estimates of a model AuFc NP monolayer, ΓMONO,NP ≈ 9.0 × 10-12 mol/cm2 (assuming 18.1 nm2/AuFc NP and based on an overall nanoparticle radius of 2.4 nm). RESULTS AND DISCUSSION We recently discussed the irreversible adsorption of AuFc NPs from electrolyte/CH2Cl2 solutions onto a Pt electrode.35 The adsorption coverage (ΓNP) was at monolayer and submonolayer levels and was sensitive to the electrolyte concentration and choice of electrolyte anion. We postulated that the adsorption was due to multiple ion-pair bridges between electrolyte anions specifically adsorbed on the electrode surface and ferrocenium ions on AuFc NPs. From the sharp, pointy current peaks of the ferrocene voltammetry, it was additionally inferred that the adsorbed AuFc NPs interacted (attractively) laterally with each other, again through ion bridging. Here, we investigate the above hypothesis using electrodes coated with sulfonate, carboxylate, and methyl-terminated alkanethiolate SAMs, of varying alkyl chain lengths. The observed adsorption behavior is introduced with 3-mercapto-1-propanesulfonate SAMs, i.e., a negatively charged electrode interface. Figure 1 shows cyclic voltammetry at this electrode in a 0.05 mM AuFc NP solution in 1.0 M Bu4N+X-/CH2Cl2 (X- ) ClO4-, PF6-, C7H7SO3-). The oxidation peak current is smaller than the ensuing reduction peak current, because the oxidative redox process induces the AuFc NP to become strongly adsorbed to the electrode surface. After a single cyclical potential scan, the electrode is removed, rinsed, and transferred to fresh, nanoparticle-free 1.0 M electrolyte in CH2Cl2, giving voltammetry as seen in Figure 2. The oxidation and reduction peaks have a symmetrical shape and have equal peak currents (ip,c ) ip,a) as expected for stable surface-confined ferrocene species. This voltammetry is stable for repeated potential cycling, losing only ∼3% of the surface coverage (ΓNP) after 25 potential scans. (More extensive stability tests were not done.) 6962

Analytical Chemistry, Vol. 81, No. 16, August 15, 2009

Figure 2. Cyclic voltammetry at 0.5 V/s of AuFc NPs adsorbed on 3-mercapto-1-propanesulfonate SAMs on Au by the single cyclical potential scan of Figure 1, after transfer to NP-free 1.0 M electrolyte/ CH2Cl2. Table 1. Electrolyte Effects on Surface Coverage of AuFc NPs Adsorbed on 3-Mercapto-1-propanesulfonate SAMs on a Au Electrode from a 0.05 mM NP Solution/ CH2Cl2a,b,c electrolyte Bu4NC7H7SO3 Bu4NPF6 Bu4NClO4

[electrolyte] (M) ΓNP (mol/cm2) 1.0 1.0 1.0 0.1 0.01

4.1 × 10-12 6.7 × 10-12 8.2 × 10-12 4.3 × 10-12 1.0 × 10-12

ΓNP (mol/cm2)d (ref 35) no film detected 6.3 × 10-12 2.9 × 10-12 1.4 × 10-12 no film detected

a The potential scan rate is 0.5 V/s. b Adsorption induced by a single potential scan between -0.1 and +1.0 V, at 0.5 V/s. c ΓNP was the same for various scan rates from 0.01 to 2.0 V/s. d Adsorption induced by 25 cyclical potential scans, unmodified Pt electrode.

The nanoparticle surface coverages (ΓNP) calculated from the charge under the ferrocene oxidation peaks in Figure 2, and other data taken at a lowered Bu4NClO4 electrolyte concentration, are presented in Table 1. The ΓNP for Bu4NClO4 electrolyte, 8.2 × 10-12 mol/cm2, is approximately a monolayer level (ΓMONO,NP ≈ 9.0 × 10-12 mol/cm2). ΓNP is less, but not greatly so, for the other two electrolytes and at a lowered electrolyte concentration.

Figure 3. (A) Schematic representation of ion-induced adsorption of a ferrocenium cation on a bare Pt electrode surface. Ion-pair bridges are proposed (ref 35) to form between adsorbed electrolyte anions on the electrode surface and NPs. (B) Cartoon representing adsorption of NPs on a negatively charged SAM surface. The adsorption is promoted by lateral ion bridges between neighbor NPs.

Table 1 also shows previously measured35 ΓNP values for similar AuFc NPs, adsorbed from the same electrolyte solutions, but on unmodified Pt electrode surfaces. In every case, the amount of adsorption is elevated on the sulfonate SAM electrode surface. Adsorption is seen at low electrolyte concentration, and for the p-toluenesulfonate electrolyte, where none was detected at the unmodified Pt surfaces. Further, the adsorption on unmodified Pt surfaces was formed relatively slowly; the previous ΓNP values shown in the Table were the result of 25 cyclical potential scans through the ferrocene wave, as opposed to a single scan in the present work. The data in Table 1 show that the earlier proposal of the importance of interactions of nanoparticles with surface ions, illustrated by the cartoon of Figure 3A, was basically correct. In the present work, the electrode surface always has a negative charge, due to the SAM (Figure 3B), and the previous uncontrolled variations in the amount of specifically adsorbed electrolyte anion on the Pt surface matter less. In the present work, the ion-pair association is between the SAM sulfonate and the nanoparticle ferrocenium sites. The primary picture that emerges is that the choice of the surface-bridging anion matters mildly in terms of the amounts of nanoparticle adsorption but that the SAM-enforced presence of an anionic electrode surface ensures strong adsorption of the polyfunctional ferrocenated nanoparticle. How can simple ion-ion association produce such persistent adsorption? The most important point is that multiple ion-ion interactions occur between the surfaces of the polycationic NP and the polyanionic SAM. The situation is analogous to interactions between oppositely charged polyelectrolytes or between polyelectrolytes and oppositely charged surfaces as in the so-called layer-by-layer adsorption experiment introduced by Decher.40 In most cases, the electrostatic interactions between the surface and polyelectrolyte, caused not just by ion-ion attractions but by their multiplicity, result in virtually irreversible adhesion of the (40) Decher, G. Science 1997, 277, 1232–1237.

polyelectrolyte,41,42 The thermodynamic stabilization is presumed to be entropic36 and is well-known for multidentate metal chelates.43,44 A supporting factor in the robust nanoparticle adsorption is the presence of lateral nanoparticle-nanoparticle interactions, in which ion bridges are formed between neighbor ferroceniums. Hydrophobic interactions may also occur (for reduced NPs) between ferrocenes and indeed between ferrocenes on the same nanoparticle. The lateral interactions in the adsorbed NP layer and/or intra-nanoparticle ferrocene-ferrocene interactions are revealed by the narrowness of the ferrocene voltammetric peaks; this reflects a neighbor-attractive surface activity effect.45-47 The full width half maxima (Efwhm) of the current peaks are 35, 37, and 49 mV for Bu4NClO4, Bu4NPF6, and Bu4NC7H7SO3 electrolytes, respectively. These values are much lower than ideally expected (90.6 mV)47 for surfaced-confined species that do not interact laterally with one another, i.e., whose activities are simply proportional to their fractional surface coverages. The lateral interactions depend on the electrolyte anions and could be the source of differences in ΓNP observed among them (Table 1). The p-toluenesulfonate electrolyte, for example, is both the most bulky and hydrophobic anion of those in Table 1; its space requirements in the NP adsorbed layer and its better solvation by the CH2Cl2 solvent could cause the lower observed ΓNP value and higher observed Efwhm value in its solutions. It (41) Harris, J. J.; DeRose, P. M.; Bruening, M. L. J. Am. Chem. Soc. 1999, 121, 1978–1979. (42) Xiao, K. P.; Harris, J. J.; Park, A.; Martin, C. M.; Pradeep, V.; Bruening, M. L. Langmuir 2001, 17, 8236–8241. (43) Busch, D. H. Chem. Rev. 1993, 93, 847–860. (44) Leininger, S.; Olenyuk, B.; Stang, P. J. Chem. Rev. 2000, 100, 853–908, and references therein. (45) Anson, F. C.; Barclay, D. J. Anal. Chem. 1968, 40, 1791–1798. (46) Takada, K.; Diaz, D. J.; Abruna, H. D.; Cuadrado, I.; Casado, C.; Alonso, B.; Moran, M.; Losada, J. J. Am. Chem. Soc. 1997, 119, 10763–10773. (47) Bard, A. J.; Faulkner, L. Electrochemical Techniques: Fundamentals and Applications, 2nd ed.; Wiley: New York, 2001.

Analytical Chemistry, Vol. 81, No. 16, August 15, 2009

6963

Table 2. Surface Coverage of AuFc NPs on Modified ω-Carboxylate Alkanethiolate SAM Surfaces, Using Varying Alkane Chain Lengtha,b electrolyte

thiols used for SAMs preparation

ΓNP (mol/cm2)

Bu4NClO4

c

7.4 × 10-12 4.8 × 10-12 1.5 × 10-12 1.5 × 10-12

MOA MUAd MHAd OT:MUAb,d

a The potential scan rate is 0.5 V/s, 1.0 M electrolyte. b Mixed SAM on a Au electrode surface was prepared by OT and MUA at a ratio of 7:1. c ΓNP was the same for various scan rates from 0.01 to 2.0 V/s. d These ΓNP data were not examined as function of scan rate.

Figure 4. Cyclic voltammetry of AuFc NPs adsorbed on the electrode in NP-free 1.0 M Bu4NClO4/CH2Cl2 at 0.5 V/s. The Au working electrodes were functionalized with carboxylic acid-terminated thiols of varying chain length (as indicated) which were rinsed with 1.0 mM aq. NaOH solution before adsorption was induced by a single potential scan in NP solutions (see Experimental Section).

should also be evident that if lateral interactions occur at submonolayer coverages, the adsorption would occur as islands. The procedure to induce adsorption (see Experimental Section) involves generating the polycation (ferrocenium) form of the AuFc NP in a single cyclical potential scan. To assess the importance of the polyion ferrocenium charge, “no-scan” experiments were performed in which the sulfonate SAM-coated electrode was simply immersed in the nanoparticle solution (containing 1 M Bu4NClO4), rinsed, and transferred to fresh NP-free electrolyte. The observed voltammetry (see Supporting Information, Figure S-1) is similar to that in Figure 2, but the surface coverage, ΓNP ) 1.4 × 10-12 mol/cm2, is substantially lower. This was unsurprising, since at open circuit, only a tiny portion of the ferrocene on the nanoparticle would be in the oxidized, cationic form. These adsorbed AuFc NPs are also less resistant to repeated redox cycling (between -0.1 and +1.0 V); over 25 cycles the peak currents decreased to a greater extent (57%) showing that this adsorbed layer is less robust. Base-washed ω-carboxylic acid alkanethiolate SAMs provide alternative anionic surfaces for adsorption of the AuFc NPs and for assessing the effect of SAM thickness. SAMs prepared with 8-mercapto-1-octanoic acid (MOA), 11-mercapto-1-undecanoic acid (MUA), and 16-mercapto-1-hexadecanoic acid (MHA) were freshly rinsed with 1.0 mM NaOH to produce a carboxylate ion electrode surface. The alkyl chains thus varied from seven to fifteen methylene units. Adsorption was again induced by a single potential scan in a NP solution, and the electrodes were transferred to NP-free 1 M Bu4NClO4/CH2Cl2 electrolyte solution. A simple expectation of this experiment might be that ΓNP would be independent of the alkanethiolate chain length. The experiments showed, however (Figure 4 and Table 2), that increasing chain length substantially decreases the quantity of adsorbed AuFc NPs. Clearly, the model of Figure 3B is lacking in some respect. One possibility is that the diminished distancedependent electronic coupling associated with thicker SAMs results in incomplete electrolysis of all the ferrocene sites on the adsorbed AuFc NPs. Measured coverages did not change with 6964

Analytical Chemistry, Vol. 81, No. 16, August 15, 2009

Figure 5. Cyclic voltammetry of AuFc NPs adsorbed on basewashed HSC7CO2H SAMs on Au, in different 1.0 M electrolytes in NP-free CH2Cl2, at 0.5 V/s.

potential scan rate for MOA (or for other experiments cited above), but different scan rates were not tested for longer chain thiols, MUA, and MHA. Another possibility is that carboxylate sites on the longer alkanethiolate chains are, due to structural features of the carboxylate/CH2Cl2 interface, less accessible to ferrocenium ion association. Yet another is that the positive electrode potential lowers the basicity of the carboxylate headgroup, especially for the shorter alkanethiolate chain lengths. Rather than speculate further, we leave this interesting issue to future work. The electrolyte employed in the carboxylate SAM experiments matters, as shown by the voltammetry of Figure 5 and data in Table 3. Using a series of Bu4N+X- electrolytes, there is little change in the appearance of the adsorbed AuFc NP voltammetry, but the values of ΓNP vary mildly, but systematically, in the order of X- ) NO3- > ClO4- > CF3SO3- > PF6- > C7H7SO3-.48-50 This is in order of the decreasing hydrophilicity of X- and suggests that anions less well solvated by the nonpolar CH2Cl2 medium are more aggressive at lateral nanoparticle-nanoparticle ion bridging than are well-solvated ones like p-toluenesulfonate. The adsorption was also investigated on a mixed SAM-coated Au electrode surface in order to reduce the number of ion-pair (48) Chen, S.; Pei, R. J. Am. Chem. Soc. 2001, 123, 10607–10615. (49) Ruiz, V.; Colina, A.; Heras, M. A.; Lopez-Palacios, J. J. Phys. Chem. C 2007, 111, 4277–4284. (50) Laaksonen, T.; Virginia, V.; Liljeroth, P.; Quinn, B. M. J. Phys. Chem. C 2008, 112, 15637–15642.

Table 3. Effect of Electrolyte Anion on Surface Coverage of AuFc NPs Adsorbed on a Base-Rinsed 8-Mercapto-1-octanoic Acid SAM on a Au Electrode, from a 0.05 mM AuFc NP Solution/CH2Cl2, in a Single Potential Scana,b electrolyte

[electrolyte] (M)

ΓNP (mol/cm2)a

Bu4NC7H7SO3 Bu4NPF6 Bu4NCF3SO3 Bu4NClO4 Bu4NNO3

1.0 1.0 1.0 1.0 1.0

2.6 × 10-12 5.8 × 10-12 6.0 × 10-12 7.4 × 10-12 7.8 × 10-12

a The potential scan rate is 0.5 V/s. b The electrode was rinsed and transferred to an NP-free 1.0 M electrolyte. Scan between -0.1 and +1.0 V.

associations between SAM carboxylate ions and nanoparticle ferrocenium sites. A ratio of 7:1 octanethiol (OT)/MUA was used to prepare the SAM, calculating that on average ca. only one MUA site would be present per cross-sectional footprint of the AuFc NP. Adsorption was induced by a single cyclical potential scan, as above, and the electrode was transferred to NP-free 1.0 M Bu4NClO4/CH2Cl2. The observed coverage of AuFc NPs, 1.5 × 10-12 mol/cm2, is less than that on the fully carboxylated MUA SAM electrode, ΓNP ) 4.8 × 10-12 mol/cm2, see Table 2. The cyclic voltammetry also shows a narrowed peak shape with Efwhm of 40 mV (see Supporting Information, Figure S-2). The combined low coverage and narrow peak shape suggest that the AuFc NPs may form patchy islands on the electrode surface, perhaps induced by patchy islands of the carboxylic acidterminated units of the SAM.51-53 Finally, one would predict that if the SAM monolayer were entirely neutral (e.g., methylene terminated), adsorption of the AuFc NP might be altogether choked off. This was examined using alkanethiolate SAMs of varied chain length: 1-octanethiol (OT), 1-dodecanethiol (DDT), and 1-hexadecanethiol (HDT). Figure 6 shows voltammetry of AuFc NPs adsorbed on these SAMs using a single potential scan in NP solution in 1.0 M Bu4NClO4/CH2Cl2. The resulting adsorbed coverage seen in NP-free electroyte, while observable, was greatly diminished, and decreased from OT SAMs (ΓNP ) 4.8 × 10-13 mol/cm2) to HDT (ΓNP ) 1.8 × 10-13 mol/cm2). The voltammetric peaks displayed diffusion-like “tails”, and indeed, the adsorbed layers persisted only for a few cyclical potential scans (see Supporting Information, Figures S-3 and S-4). That any adsorption occurred at all may be attributable to lateral nanoparticle-nanoparticle interactions. From these experiments, one draws the conclusion that the nanoparticle-electrode surface polyion interactions are more essential to a robust adsorbed AuFc NP film than are lateral interactions, however prominent the latter’s effects are on voltammetric peak shape. (51) Chen, S.; Li, L.; Boozer, C. L.; Jiang, S. Langmuir 2000, 16, 9287–9293. (52) Yu, J.-J.; Tan, Y. H.; Li, X.; Kuo, P.-K.; Liu, G.-Y. J. Am. Chem. Soc. 2006, 128, 11574–11581. (53) No effects associated with SAM instability were observed during the potential scanning over -0.1 to +1.0 V.

Figure 6. Cyclic voltammetry of AuFc NPs adsorbed on alkanethiolate SAMS, as indicated, in NP-free 1.0 M Bu4NClO4/CH2Cl2 at 0.5 V/s.

CONCLUSIONS The results presented involve the important general conclusions that the cationic NPs are most robustly adsorbed on anionic surfaces, notably those fashioned using SAMs. The adsorption of AuFc NPs on unmodified electrodes depends on the electrolyte through electrolyte anion adsorption on the electrode; that on SAM-modified electrodes depends on the electrolyte mainly by the supporting role of lateral nanoparticle-nanoparticle interactions. There are three different kinds of interactions influencing nanoparticle adsorption: (i) interaction between AuFc NPs and anionic SAM surfaces, (ii) lateral interactions caused by bridging electrolyte ions between charged nanoparticles, and (iii) specific adsorption interactions between electrolyte ions and a naked electrode surface. Important in the robustness of adsorption is the multiplicity of such interactions for each adsorbed nanoparticle, making it an analogy to polyelectrolyte adsorption on charged surfaces. That analogy leads to the prediction that any polyionic nanoparticle should become strongly adherent to a charged electrode surface, which is under exploration with promising early results.54 ACKNOWLEDGMENT This research was supported in part by grants from the National Science Foundation and the Office of Naval Research. SUPPORTING INFORMATION AVAILABLE Graphs for comparison of surface coverage in the presence of a methyl-terminated thiol modified Au electrode and a kinetic plot for the number of scans. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review May 13, 2009. Accepted June 27, 2009. AC9010364 (54) Sardar, R.; Beasley, C. A.; Murray, R. W. UNC-Chapel Hill. Unpublished results, 2009.

Analytical Chemistry, Vol. 81, No. 16, August 15, 2009

6965