Journal of Electroanalytical Chemistry 664 (2012) 94–99

Contents lists available at SciVerse ScienceDirect

Journal of Electroanalytical Chemistry

journal homepage: www.elsevier .com/locate / je lechem

Self-assembled monolayers formed by 5,10,15,20-tetra(4-pyridyl)porphyrin andcobalt 5,10,15,20-tetra(4-pyridyl)-21H,23H-porphine on iodine-passivated Au(111)as observed using electrochemical scanning tunneling microscopy and cyclicvoltammetry

Scott N. Thorgaard, Philippe Bühlmann ⇑Department of Chemistry, University of Minnesota, 207 Pleasant St. SE, Minneapolis, MN 55455, USA

a r t i c l e i n f o

Article history:Received 17 May 2011Received in revised form 18 October 2011Accepted 21 October 2011Available online 3 November 2011

Keywords:Self assembled monolayersPorphyrinsElectrochemical scanning tunnelingmicroscopyCyclic voltammetryAdsorptionAu(111)

1572-6657/$ - see front matter � 2011 Elsevier B.V. Adoi:10.1016/j.jelechem.2011.10.019

⇑ Corresponding author. Fax: +1 612 626 7541.E-mail address: buhlmann@umn.edu (P. Bühlmann

a b s t r a c t

Self-assembled monolayers (SAMs) of 5,10,15,20-tetra(4-pyridyl)porphyrin (TPyP) and cobalt 5,10,15,20-tetra(4-pyridyl)-21H,23H-porphine (CoTPyP) were formed by in situ equilibrium adsorption ontoAu(111) electrodes passivated by an interposed adlayer of iodine. Adsorption of the porphyrins ontoAu(111) was confirmed by cyclic voltammetry, and each of the SAMs was imaged using electrochemicalscanning tunneling microscopy (EC-STM). The TPyP and CoTPyP SAMs were found to contain highlyordered domains on I–Au(111) at positive potentials where a disordered SAM would be expected for bareAu(111). Cyclic voltammetry was also used to observe in situ the assembly of binary SAMs containingTPyP and CoTPyP as they formed.

� 2011 Elsevier B.V. All rights reserved.

1. Introduction

The formation of self-assembled monolayers (SAMs) under elec-trochemical conditions is of interest to fields such as chemicalsensing, molecular electronics, and electrocatalysis, where SAM-modified electrodes offer surfaces with chemically tailored proper-ties [1–4]. For many applications, preference is given to SAMs thatexhibit high long-range 2-D order and, thereby, extremely well de-fined surface characteristics. Therefore, there is great value in athorough understanding of SAM-ordering processes and self-assembly techniques that promote the formation of ordered SAMs.

Porphyrin derivatives such as 5,10,15,20-tetra(4-pyridyl)por-phyrin (TPyP, Fig. 1) are particularly attractive for the formationof SAMs since they form planar metal complexes that permit theformation of SAMs exposing different metal centers towards thesolution [5,6]. Moreover, the solubility of porphyrin derivativescan be tuned by the choice of the substituents in meso positionand on the pyrrole rings. Indeed, both hydrophilic and hydrophobicporphyrins are commercially available. Porphyrin SAMs have beeninvestigated extensively using scanning tunneling microscopy(STM) in vacuum and inert liquids [7,8], and the formation of

ll rights reserved.

).

porphyrin SAMs on metal surfaces in electrolyte solutions has alsoreceived attention from multiple groups. In particular, Borguetet al. have extensively investigated porphyrin SAMs on bareAu(111) electrodes [9–12]. Importantly, they found that the elec-trode potential can be used to control ordering in TPyP monolayerson bare Au(111). They showed that in 0.1 M HClO4 and at poten-tials above 0.54 V vs. Ag/AgCl, TPyP SAMs form in a disorderedarrangement due to strong interaction between the conjugatedp-system of TPyP and the positively charged electrode surface[9]. In contrast, with the electrode potential set between �0.16and +0.24 V vs. Ag/AgCl, optimal self-assembly of TPyP into highlyordered domains proceeds.

Modification of the electrode surface with a layer of specificallyadsorbed anions offers a complimentary route to controlling SAMordering [13–18]. In pioneering work by Kunitake et al., orderedSAMs of 5,10,15,20-tetrakis(N-methylpyridinium-4-yl)-21H,23H-porphine were formed on Au(111) electrodes modified with a pas-sivating layer of iodine [13,14]. The iodine adlayer weakens theinteraction between the porphyrins and the electrode, allowingthem greater lateral mobility on the surface. This results in the for-mation of highly ordered SAMs in wider and more positive poten-tial windows than what is possible with potential control alone.Besides Au(111), similar results using iodine modification havebeen reported for Ag(111) [15] and Cu(111) [18] electrodes. We

Fig. 1. Structures of 5,10,15,20-tetra(4-pyridyl)porphyrin (TPyP) and cobalt5,10,15,20-tetra(4-pyridyl)-21H,23H-porphine (CoTPyP).

S.N. Thorgaard, P. Bühlmann / Journal of Electroanalytical Chemistry 664 (2012) 94–99 95

have found surface modification with bromine to be a useful alter-native to iodine due to the higher potentials of surface bromineoxidation as well as solution-phase bromide oxidation [19,20].We have been able to observe highly ordered TPyP SAMs at poten-tials between 0.0 and +1.3 V vs. Ag/AgCl on Au(111) electrodes in0.1 M HClO4 solutions containing 150 lM KBr [21].

Here, we apply the iodine passivation technique to 5,10,15,20-tetra(4-pyridyl)porphyrin (TPyP) and cobalt 5,10,15,20-tetra(4-pyridyl)-21H,23H-porphine (CoTPyP) SAMs on Au(111). CoTPyPdiffers from TPyP only by the presence of the cobalt (II) ion inthe porphyrin center (Fig. 1), and we sought to investigate what ef-fect, if any, metallation of the porphyrin would have on SAM for-mation on iodine-passivated Au(111) surfaces (referred to in thefollowing as I–Au(111) surfaces).

2. Experimental

2.1. EC-STM and CV measurements

A Au(111) single crystal electrode was purchased from icryst(Jülich, Germany) and used for all experiments described. The elec-trode consisted of a Au bead formed on the end of a pure Au wireand subsequently cut perpendicular to the (111) face, as describedby Clavilier et al. [22], to yield a Au(111) disk of 2 mm diameter.Before each experiment, the electrode was cleaned by heating tored heat for 5 min in a pure hydrogen flame, followed by coolingin air for 10 s and lastly immersion in pure water. A droplet ofwater was left on the electrode surface during transfer to the elec-trochemical cell to limit surface contamination from the air.

Cyclic voltammetry was performed using a CH InstrumentsElectrochemical Analyzer (CH Instruments, Austin, TX, USA). Thecyclic voltammetry cell was used with a Ag/AgCl reference (Bioan-alytical Systems, West Lafayette, IN, USA; 3 M KCl filling solution,glass frit separation to the sample) and a Pt wire auxiliary elec-trode. Contact between the Au(111) working electrode and thesample solution was made using the hanging meniscus methodto isolate the (111) face [21,23,24]. The Supporting Informationcontains a drawing that illustrates the hanging meniscus contact.Unavoidable wetting of the bead sides during the hanging menis-cus CV may explain small irregularities in the observed peakshapes for all of the CVs shown in this work. However, the overallpeak positions were highly reproducible over several experimentpreparations. Sample solutions were degassed by bubbling highpurity Ar through the stirred solutions for 15 min prior to eachexperiment.

The electrochemical STM consisted of a PicoSPM scan head witha PicoStat auxiliary potentiostat from Molecular Imaging (now Agi-lent Technologies, Santa Clara, CA) and an RHK SPM-100 controller(RHK Technology, Troy, MI, USA). The EC-STM cell, constructed in

house, [17] consisted of a quartz glass cup (1 mL volume) fittedwith a Teflon bushing designed to hold the Au working electrodeand was sealed with a Viton O-ring (Dupont, Newark, DE, USA).The O-ring was stored in pure water for several days before exper-iments to permit leaching of any water-soluble contaminants. Thequartz glass and Teflon pieces of the cell were flushed with piranhasolution (1:3 mixture of 30% hydrogen peroxide and concentratedsulfuric acid) and rinsed with copious amounts of pure water priorto each experiment. Caution: piranha solution is highly oxidizing andshould never be stored in closed containers. For EC-STM experiments,a Ag wire quasi-reference electrode and Pt wire auxiliary electrodewere used. The potential of the Ag wire was verified against the Ag/AgCl reference in solutions identical to those used for EC-STM, andall potentials listed in this work are quoted with respect to the Ag/AgCl electrode.

W tips were prepared by mechanically cutting W wire(0.25 mm diameter, 99.95% purity, Alfa Aesar, Ward Hill, MA,USA). To minimize the electrochemical contribution to the tip cur-rent during EC-STM imaging, tips were coated with an electropho-retic paint (Clearclad HSR, Clearclad, Harvey, IL, USA) using aprocedure described previously [25]. STM images were recordedin constant-current mode using tunneling setpoints between 0.5and 2.0 nA, and are shown without any filtering. The potential ofthe tip was kept between 0.00 and +0.20 V for all experiments inorder to prevent the formation of tungsten oxide.

2.2. Reagents

All solutions were prepared using water deionized and charcoaltreated in a Milli-Q Plus reagent grade purification system (Milli-pore, Bedford, MA, USA). The 5,10,15,20-tetra(4-pyridyl)porphyrin(TPyP) was purchased from Sigma–Aldrich (St. Louis, MO, USA) andused without further purification. Analytical reagent grade KBr, KI,and CoCl2 were purchased from Mallinckrodt (St. Louis, MO, USA),and HClO4 and H2SO4 from Fisher Scientific (Pittsburg, PA, USA).

Cobalt 5,10,15,20-tetra(4-pyridyl)-21H,23H-porphine (CoTPyP)was prepared by metallating TPyP using a procedure from the lit-erature [26]. Briefly, tetrapyridylporphyrin was refluxed for 3 hwith an excess of cobalt chloride in a 1:1 acetic acid and dimethylformamide solution. The acetic acid was dried for 12 h over molec-ular sieves before use. The resulting solids were filtered andwashed with pure water and diethyl ether. Mass spectra of thecompound showed fragments with an m/z number and isotopedistribution appropriate for CoTPyP. 1H NMR spectra recorded inchloroform-d with 5% dimethylsulfoxide-d6 and 5% trifluoroaceticacid showed three signals with chemical shifts different from thoseof the free base TPyP, and no other signals but those of the solventwere observed. No change in the 1H NMR spectrum was detectedfor the CoTPyP sample after several weeks of storage, indicatingthat CoTPyP is stable in the acidic medium.

3. Results and discussion

3.1. Cyclic voltammetry

Cyclic voltammetry was used to verify adsorption of TPyP andCoTPyP to Au(111). Fig. 2a shows hanging meniscus cyclic voltam-mograms (CVs) taken using the flame annealed Au(111) electrodein 0.1 M HClO4 solution containing 150 lM KI, before and after theaddition of 30 lM TPyP to the electrochemical cell. In aqueoussolution and at potentials above the potential of zero charge(PZC), iodide adsorbs onto Au(111) electrodes to form a monolayerof iodine, referred to in the following as the iodine adlayer [27–29].Before the introduction of TPyP, the CV features observed below�0.1 V correspond to changing iodine surface coverage as well as

Fig. 2. CVs taken before (dotted line) and after (solid line) the addition of (a) 30 lMTPyP or (b) 30 lM CoTPyP to a cell containing 0.1 M HClO4 and 150 lM KI. The scanrate for all CVs was 100 mV/s.

Fig. 3. CVs recorded in cells containing 0.1 M HClO4, 150 lM KI, and 30 lM CoTPyPalong with 8 lM TPyP (solid line), 30 lM TpyP (dotted line), or 30 lM CoTPyP(dashed line). The scan rate for all CVs was 100 mV/s.

96 S.N. Thorgaard, P. Bühlmann / Journal of Electroanalytical Chemistry 664 (2012) 94–99

lifting of the (p

3 � 22) reconstruction of Au(111), with a denseadlayer of iodine present on the surface at more positive potentials.

Following the introduction of TPyP, the CV changed signifi-cantly. The heights of the large peaks near 0.0 V vs. Ag/AgCl showa linear dependence on the scan rate, consistent with an adsorbedspecies. Adsorbed TPyP is known to undergo reversible two-elec-tron oxidation on Au(111) in dilute HClO4 [10,11]. Using our EC-STM images of the porphyrin monolayers (see below) to roughlyestimate the surface coverage, we calculate that the two-electronoxidation of a TPyP monolayer should result in a charge transferof 1.26 � 10�7 C/mm2. Allowing for some variability due to defects,surface roughness, and the electrolyte meniscus contact, thisagrees well with the observed peak areas for the peaks near0.0 V, and we attribute them to the two-electron oxidation of ad-sorbed TPyP. The large features below �0.1 V ascribed to the iodinelayer also changed shape upon introduction of TPyP. While smallvariations in the electrolyte meniscus contact prevent a quantita-tive comparison of the peak areas, the shift of the peak potentialsmay reflect competing adsorption of TPyP and I� onto Au(111),as has been shown for coumarine and phthalazine adsorption on

Au(100) in sulfate solutions [30]. For this system, as with otherSAMs on halogen-passivated Au, interactions between the halogenand Au(111) are expected to be far stronger than those betweenthe porphyrin and Au(111), making displacement of the halogenlayer by the porphyrin very unlikely [13,21].

An identical experiment was also performed using CoTPyP.Fig. 2b shows CVs taken before and after the addition of 30 lMCoTPyP to a cell containing 0.1 M HClO4 and 150 lM KI. As forTPyP, the CV taken after the introduction of CoTPyP shows manyadditional or altered features that differ from the CV in the porphy-rin-free cell. The many small peaks between �0.14 and +0.10 V vs.Ag/AgCl in the CV for CoTPyP all show a linear dependence of theirheight on the scan rate, and we attribute them to adsorbed CoTPyP.However, the observed peak areas are too small to correspond tothe expected one-electron oxidation of the metal center. We spec-ulate that these smaller peaks may be due to reorganization of theCoTPyP SAM, possibly complicated again by a small degree of wet-ting of the non-(111) surfaces on the sides of the Au bead elec-trode. As no well-resolved peaks attributable to the metal centeroxidation were found in CVs spanning the entire solvent window,it is possible that the metal center oxidation may be convolutedwithin another feature, such as the iodide adsorption in the regionbelow �0.14 V.

Following voltammetry of the single component monolayers,we prepared binary CoTPyP/TPyP SAMs. First, a cell was preparedcontaining 0.1 M HClO4, 150 lM KI, and 30 lM CoTPyP. A cyclicvoltammogram taken in this cell using a hanging meniscusAu(111) electrode was identical to that shown following CoTPyPaddition in Fig. 2b. To create a mixed SAM, several aliquots of TPyPstock were then added to the cell until a CV was observed thatexhibited features of both the TPyP and the CoTPyP SAMs, therebydefining concentrations that would produce a binary SAM. Such aCV was recorded, for example, with a TPyP concentration of8 lM in the cell and is shown as the solid curve in Fig. 3. For com-parison, the single-component TPyP and CoTPyP CVs are alsoshown. The peak at 0.0 V vs. Ag/AgCl showed a linear dependenceof its height on the scan rate and strongly resembles a similar fea-ture in the CV for a pure TPyP SAM. From this we conclude that wegenerated a SAM containing both TPyP and CoTPyP. Importantly,further additions of TPyP after the mixed SAM was observed

Fig. 4. Cyclic voltammogram recorded for a Au(111) working electrode in asolution containing 0.1 M HClO4, 150 lM KI, and 30 lM CoTPyP. Immediately (<5 s)before starting the CV, an aliquot of TPyP stock was added to the cell to produce afinal TPyP concentration between 5 and 10 lM. Thirty successive cycles wererecorded between �0.35 and 0.2 V using a scan rate of 100 mV/s. Arrows markpeaks observed to increase or decrease in size (as indicated by the arrow direction)in successive cycles. (See the Supporting Information for a larger version of thisfigure.)

Fig. 5. EC-STM image of a Au(111) surface in 0.1 M HClO4 with 150 lM KI. TheAu(111) terraces show a Moiré pattern due to the iodine adlayer. The samplepotential was +0.35 V vs. Ag/AgCl.

S.N. Thorgaard, P. Bühlmann / Journal of Electroanalytical Chemistry 664 (2012) 94–99 97

eventually resulted in CVs identical to the one in Fig. 2a, indicatingcomplete replacement of CoTPyP with a uniform TPyP SAM.

Cyclic voltammetry also permitted the in situ observation of thereplacement of CoTPyP by TPyP during SAM formation by record-ing a large number of CV scans in continuous succession immedi-ately after a single injection of TPyP into the cell containingCoTPyP. This data is shown in Fig. 4. The first trace of the 30-cycleCV is nearly identical to that shown in Fig. 2b for a pure CoTPyPSAM. In subsequent traces, peaks attributed to CoTPyP, namelythe small features near �0.1 V, were found to shrink, while a peaknear 0 V appears and increases in size. We attribute these changesto gradual replacement of CoTPyP by TPyP on the surface. Theshrinking peaks at �0.07 and �0.12 V, attributed to CoTPyP, arewell resolved and occur in a potential region devoid of peaks inthe pure TPyP CV, and therefore it appears possible to use theirintegrated areas to make a quantitative determination of ratio ofCoTPyP to TPyP on the surface. However, because these peakscould not be assigned to any single redox or capacitive process, itis questionable whether the area of the sharp peaks and the surfacecoverage of CoTPyP are directly proportional to one another. In acontrol experiment, CVs in the absence of TPyP showed no chang-ing features over 30 cycles.

3.2. EC-STM observations

EC-STM imaging of the TPyP and CoTPyP SAMs was performedin 0.1 M HClO4 containing 150 lM KI. Except where indicated be-low, all images were recorded with the sample potential set at+0.35 V vs. Ag/AgCl. This potential is below the onset of solutionphase iodide oxidation, and was chosen to avoid any complicationsdue to the formation of polyiodide species [19]. When the Au(111)terraces were first imaged before the introduction of the porphy-rins, a Moiré pattern showing threefold symmetry was observed(see Fig. 5). This pattern is due to lattice size mismatch betweenadsorbed iodine and the underlying Au(111) substrate, and isindicative of a highly ordered iodine adlayer on the surface [19,31].

After well-defined terraces were observed, either TPyP or CoT-PyP stock in 0.1 M H2SO4 was added drop-wise to the EC-STM cellto produce a final porphyrin concentration of �30 lM. For bothTPyP and CoTPyP, highly ordered domains were observed to formon the surface 20–40 min after introduction. Fig. 6 shows represen-tative EC-STM images of TPyP SAMs formed on I–Au(111). Similarto earlier observations of TPyP on Br–Au(111) [21], the moleculesformed domains with rows that meet at 120� angles, indicatingalignment with the close-packed substrate surface. Possibly dueto effects of the tip geometry on resolution, images showing theindividual pyridyl groups of TPyP were not obtained in our exper-iments; each TPyP molecule in the SAM instead appeared as aroughly diamond-shaped object. However, the typical distance be-tween nearest-neighbor molecules in the TPyP SAM was 1.6 nm,which is in good agreement with existing reports of TPyP SAMs[9,21], and this was sufficient to estimate the surface coverage aswas needed to interpret the CV data described above. Fig. 6d showsin white a possible orientation of the porphyrins in the monolayer,which we propose based on the observed dimensions of the unitcell and earlier reports on porphyrin SAMs [9,13,14,21]. Singlemolecule defects in the SAM could also be observed, such as shownin the lower right-hand corner of Fig. 6d, where the dark spot in theimage indicates one missing TPyP molecule.

In EC-STM imaging of both TPyP and CoTPyP SAMs, monolayerdomains were typically not separated by sharp domain boundariesbut by broad, dark, irregular bands across the Au(111) terraces.These disordered regions likely contain fewer porphyrin moleculesthan the highly ordered domains. The dark bands were never ob-served on terraces in the absence of TPyP or CoTPyP (i.e., with io-dine alone on the surface). Therefore, their sudden appearanceafter the introduction of TPyP or CoTPyP served as an indicator ofporphyrin adsorption even in images where single molecules werenot resolved. This can be seen in Fig. 6b, where a white, dashed linemarks a darkened band cutting across a single Au(111) terrace, afeature that we attribute to a disordered boundary region betweendomains of TPyP.

Using cyclic voltammetry with a cell containing 0.1 M HClO4

and 150 lM KI, we found the onset of solution phase iodide oxida-tion to occur at +0.5 V vs. Ag/AgCl (data not shown). We performed

Fig. 6. Representative EC-STM images of a TPyP SAM on I–Au(1 11). The sample potential was +0.35 V vs. Ag/AgCl (images a and b) or +0.51 V vs. Ag/AgCl (images c and d).

98 S.N. Thorgaard, P. Bühlmann / Journal of Electroanalytical Chemistry 664 (2012) 94–99

additional EC-STM imaging of the TPyP SAMs with the sample po-tential set between +0.5 and +0.8 V in order to probe the effect ofpolyiodide formation on the porphyrin SAM. However, we sawno significant change in the TPyP SAM structure at these potentials.Figs. 6a and b were recorded with the sample potential set to+0.35 V, while images 6c and d were recorded with the sample po-tential set to +0.51 V.

Very similar structures could be seen for CoTPyP. Representa-tive images of CoTPyP SAMs formed on Au(111) in 0.1 M HClO4

solutions containing 150 lM KI are shown in Fig. 7. Analogous towhat was observed for TPyP in Fig. 6b, the dark, irregular linemarked with a white arrow in Fig. 7b is a boundary where CoTPyP

Fig. 7. Representative EC-STM images of CoTPyP SAMs on I–

domains of differing orientations meet. The other lines in the im-age (marked with black arrows) are single atom steps on the Ausurface. Domain boundaries and atomic step were easily distin-guished from one another by examining image cross-sections.The monolayer orientation was observed to persist over singleatom steps on the Au(111) surface, suggesting that adsorbate–adsorbate interactions carry over between adjacent Au(111) ter-races. As with TPyP above, the individual pyridyl groups of CoTPyPwere not resolved in these images. Most likely because of the lim-ited resolution, the characteristic bright spot [32] of the cobaltatom often observed in cobalt porphyrin monolayers was notresolved from the rest of the molecule, and each CoTPyP molecule

Au(111). The sample potential was +0.35 V vs. Ag/AgCl.

S.N. Thorgaard, P. Bühlmann / Journal of Electroanalytical Chemistry 664 (2012) 94–99 99

instead appeared as a bright, roughly diamond-shaped object. Theorientation of the individual porphyrin molecules in the monolayeris not directly evident from the STM images but based on the ob-served spacing between nearest-neighbor rows of CoTPyP(1.6 nm) and previous reports of porphyrin SAMs [9,13,14,21], wepropose the monolayer structure illustrated in white in Fig. 7a.

For CoTPyP, we also tried stepping the potential of the workingelectrode to increasingly positive values while imaging with EC-STM. No change in the SAM structure was seen at potentials upto +0.85 V vs. Ag/AgCl. However, CoTPyP domains could not be ob-served after steps to +1.05 V vs. Ag/AgCl. This may be the result ofCoTPyP desorption following oxidation of the underlying iodinelayer. The CoTPyP SAM was again observed after the potentialwas stepped back to +0.35 V vs. Ag/AgCl.

4. Conclusion

CoTPyP and TPyP SAMs were formed by in situ equilibriumadsorption on iodine-modified Au(111) electrodes in 0.1 M HClO4

solutions containing 150 lM KI. Adsorption of the porphyrins onthe Au(111) surface was confirmed using cyclic voltammetry. Thistechnique was also used to establish the formation of a two-com-ponent, mixed monolayer of CoTPyP and TPyP. EC-STM observa-tions found that both the TPyP and CoTPyP SAMs containedhighly ordered domains.

This work adds to the very limited number of reports on the useof halide modified Au(111) for the formation of organic SAMs un-der electrochemical conditions [13,14,21]. The observation of TPyPSAMs on I–Au(111) echoes work by Kunitake et al. using a similarmolecule, TMPyP [13,14]. There, TMPyP SAMs were formed on anI–Au(111) electrode in neat 0.1 M HClO4, the iodine layer havingbeen generated ex situ by electrode modification in a separate KIsolution. CoTPyP is here the first metalloporphyrin to be imagedon halide modified Au(111). The similar structures of the single-component TPyP and CoTPyP SAMs suggest that metal atom coor-dination to the porphyrin center does not significantly influencethe porphyrin’s adsorption behavior onto I–Au(111). Indeed, theCV experiments indicated formation of mixed monolayers contain-ing both TPyP and CoTPyP.

Acknowledgements

The authors thank Clearclad (Harvey, IL, USA) for samples of theHSR electropaint that was used to insulate the STM tips.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.jelechem.2011.10.019.

References

[1] A. Ulman, Chem. Rev. 1996 (1996) 1533–1554.[2] J.V. Barth, G. Costantini, K. Kern, Nature 437 (2005) 671–679.[3] J.C. Love, L.A. Estroff, J.K. Kriebel, R.G. Nuzzo, G.M. Whitesides, Chem. Rev. 105

(2005) 1103–1170.[4] H. Uehara, Y. Okawa, Y. Sasaki, K. Uosaki, Chem. Lett. 38 (2009) 148–149.[5] K.M. Smith, Porphyrins and Metalloporphyrins, Elsevier, Amsterdam, 1975.[6] K.M. Kadish, K.M. Smith, R. Guilard, The Porphyrin Handbook, Academic Press,

San Diego, 2000.[7] J. Otsuki, Coord. Chem. Rev. 254 (2010) 2311–2341.[8] T. Oshiro, T. Ito, P. Bühlmann, Y. Umezawa, Anal. Chem. 73 (2001) 878–883.[9] Y. He, T. Ye, E. Borguet, J. Am. Chem. Soc. 124 (2002) 11964–11970.

[10] Y. He, E. Borguet, Angew. Chem. Int. Ed. 46 (2007) 6098–6101.[11] T. Ye, Y. He, E. Borguet, J. Phys. Chem. B 110 (2006) 6141–6147.[12] Q. Yuan, Y. Xing, E. Borguet, J. Am. Chem. Soc. 132 (2010) 5054–5060.[13] M. Kunitake, N. Batina, K. Itaya, Langmuir 11 (1995) 2337–2340.[14] M. Kunitake, U. Akiba, N. Batina, K. Itaya, Langmuir 13 (1997) 1607–1615.[15] K. Ogaki, N. Batina, M. Kunitake, K. Itaya, J. Phys. Chem. 100 (1996) 7185–7190.[16] L.-J. Wan, S. Shundo, J. Inukai, K. Itaya, Langmuir 16 (2000) 2164–2168.[17] S. Yoshimoto, T. Sawaguchi, J. Am. Chem. Soc. 130 (2008) 15944–15949.[18] N.T.M. Hai, B. Gasparovic, K. Wandelt, P. Broekmann, Surf. Sci. 601 (2007)

2597–2602.[19] X. Gao, M.J. Weaver, J. Am. Chem. Soc. 114 (1992) 8544–8551.[20] O.M. Magnussen, B.M. Ocko, J.X. Wang, R.R. Adzic, J. Phys. Chem. 100 (1996)

5500–5508.[21] S.N. Thorgaard, P. Bühlmann, Langmuir 26 (2010) 7133–7137.[22] J. Clavilier, R. Faure, G. Guinet, R. Durand, J. Electroanal. Chem. 107 (1980) 205–

209.[23] D. Dickertmann, F.D. Koppitz, J.W. Schultze, Electrochim. Acta 21 (1976) 967–

971.[24] P.J. Weldford, B.A. Brookes, V. Climent, R.G. Compton, J. Electroanal. Chem. 513

(2001) 8–15.[25] S.N. Thorgaard, P. Bühlmann, Anal. Chem. 79 (2007) 9224–9228.[26] N.M. Barbosa Neto, L. De Boni, C.R. Mendonça, L. Misoguti, S.L. Queiroz, L.R.

Dinelli, A.A. Batista, S.C. Zilio, J. Phys. Chem. B 109 (2005) 17340–17345.[27] O.M. Magnussen, Chem. Rev. 102 (2002) 679–726.[28] J. Lipkowski, Z. Shi, A. Chen, B. Pettinger, C. Bilger, Electrochim. Acta 43 (1998)

2875–2888.[29] A. Chen, Z. Shi, D. Bizzotto, J. Lipkowski, B. Pettinger, C. Bilger, J. Electroanal.

Chem. 467 (1999) 342–353.[30] D.M. Kolb, Prog. Surf. Sci. 51 (1996) 109–173.[31] T. Yamada, N. Batina, K. Itaya, J. Phys. Chem. 99 (1995) 8817–8823.[32] S. Yoshimoto, A. Tada, K. Suto, R. Narita, K. Itaya, Langmuir 19 (2003) 672–677.