Scanning electrochemical microscopy for the direct patterning of a gold surface with organic moities derived from iodonium salt

Electrochemistry Communications 10 (2008) 1230–1234

Contents lists available at ScienceDirect

Electrochemistry Communications journal homepage: www.elsevier.com/locate/elecom

Scanning electrochemical microscopy for the direct patterning of a gold surface with organic moities derived from iodonium salt Tarik Matrab, Catherine Combellas, Frédéric Kanoufi * Laboratoire Environnement Chimie Analytique, ESPCI-Paris Tech, CNRS UMR 7121, 10 rue Vauquelin, 75005 Paris, France

a r t i c l e

i n f o

Article history: Received 19 May 2008 Received in revised form 4 June 2008 Accepted 5 June 2008 Available online 11 June 2008 Keywords: Scanning electrochemical microscopy Electrografting Direct mode Surface pattering Iodonium

a b s t r a c t The scanning electrochemical microscope (SECM) is used in the direct mode to draw patterns of a thin passivating organic layer on a gold electrode surface and to image them. The patterning is ensured by the local electrografting of the organic moieties obtained by reduction of an aryliodonium salt, as evidenced by XPS and SECM line scans. The resolution of the writing process is controlled by the charge injected. Ó 2008 Elsevier B.V. All rights reserved.

1. Introduction The modification of surfaces with specific chemical functionalities results in applications in chemical sensing [1,2], microelectronics [3], corrosion inhibition [4], electrocatalysis and biosensing [5]. For the past decade, the electrochemical reduction of aryldiazonium salts onto conductive surfaces has been widely investigated [6]. This technique allows the covalent attachment on various surfaces of electrogenerated aryl radicals bearing different functionalities. However, the spontaneous grafting of the aryl moieties onto the surface [7] interferes and restricts the patterning of surfaces via diazoniums to the use of micro-contact printing [8] or to local scratching from an electrografted surface [9]. Besides, electrografting of glassy carbon can also be performed by other radical generating species such as diaryliodonium ions [10]. As a result of a lower reduction potential than their diazonium analogues, the iodoniums are less prone to spontaneous grafting of surfaces and are then better candidates for local surface electrografting. The SECM is a convenient tool for local surface derivatization due to its ability to generate micrometric sources of a broad range of reactive chemical species [11]. Moreover, its slow writing speed can be circumvented by using ‘‘stamp” electrodes [12]. In the direct mode, it was used to pattern various conducting surfaces with organic moieties [13–20]. The local introduction of specific chemical or biochemical activity can be obtained either from the local tip-in* Corresponding author. Tel.: +33 1 40 79 45 26. E-mail address: frederic.kanoufi@espci.fr (F. Kanoufi). 1388-2481/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.elecom.2008.06.006

duced desorption of alkanethiols from a thiol-covered gold electrode [13–15] or from the local polymerization of surface with bioactive polymers [16–20], even though the latter strategy requires important efforts to synthesize specific electroactive monomers. Here, we propose to selectively electrograft a conductive surface with organic moieties using the SECM in the direct mode with a commercially available iodonium species, bis-(4-tert-butylphenyl)iodonium p-toluenesulfonate (1). We examine the electrografting of 1 at a gold surface by cyclic voltammetry (CV) and SECM and then use the SECM to micropattern Au surface with aryl radicals derived from 1. 2. Experimental 2.1. Chemicals and electrodes Chemicals were from Aldrich or SDS. Electrografting was conducted either on 1 mm diameter disk gold electrodes or on cm2 area gold plates surfaces (1000 Å gold coated silicon wafers, Aldrich). 2.2. Electrochemical procedures The CV investigation was performed on a 1 mm diameter gold electrode in a glass electrochemical cell containing 10 mL of solution (ACN or water), 0.1 M of supporting electrolyte (NBu4BF4 in ACN, KCl in water). The SECM investigation was performed on gold plated surfaces with a Pt tip (radius, a = 12.5 lm, ratio of glass to

1231

T. Matrab et al. / Electrochemistry Communications 10 (2008) 1230–1234

5.0

3.0

-1 -2

log k

4.0

b'

a'

-3

a

-4 -5

2.0

I (µA)

metallic radius 5) moved by a home-made SECM. The approach curves were obtained from 4 mM solutions of a redox mediator (potassium ferrocyanide, FeðCNÞ4 6 , in water, ferrocene, Fc, in ACN under argon) by biasing the tip on the redox mediator oxidation plateau (0.50 vs Ag/AgCl for both probes), the gold substrate was unbiased. A potentiostat/galvanostat (CH660A, IJ Cambria) was used in a three or two-electrode configuration with a 1 mm diameter platinum wire counter-electrode. For SECM or CV experiments, respectively a 1 mm diameter Ag/AgCl or a SCE was the reference. For SECM lithography (with 1) and line scanning (with FeðCNÞ4 6 ), the microelectrode was positioned from approach curves at a distance of 8–10 lm from the substrate and performed by scanning the tip at 5 lm/s above the substrate. The gold surfaces were cleaned by successive oxidation–reduction until reproducible voltammograms were obtained. The electrografting of entire gold surfaces was obtained from chronoamperometric reduction of a 2 mM ACN solution of 1 by biasing them at 1.1 V vs SCE for 300 s. The surfaces were then ultrasonicated in acetone for 10 min.

-6 -0.2 -0.1

c 0

0.1

0.2 0.3

E- E 0 / V

1.0

b 0.0 -1.0 -2.0 -3.0 -0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

E (V / SCE) 2.3. XPS characterization A Thermo VG Scientific ESCALAB 250 system fitted with a micro-focused, monochromatic Al Ka X-ray source (1486.6 eV) and a magnetic lens was used. An X-ray beam of 650 lm was used at a power of 10 mA  15 kV. The spectra were acquired in the constant analyzer energy mode (40 eV pass energy for the narrow regions) and calibrated by setting the C–C/C–H C1s peak at 285 eV. 3. Results and discussion 3.1. CV and SECM examination of surfaces On a 1 mm gold disk electrode, the reduction of 1 shows a broad irreversible wave (peak potential: Ep = 0.9 V vs SCE) which has vanished after the third scan. This is typical of the blocking of the electrode by the aryl moiety that grafts the electrode (Scheme 1a). Entire surfaces of Au electrodes were electrografted by chronoamperometric reduction of 1 at 1.1 V vs SCE. The extent of grafting of the aryl moiety derived from the reduction of 1 has been investigated by CV and SECM. Fig. 1 presents the CV of FeðCNÞ4 oxidation on a 1 mm gold electrode, which 6 has been submitted to either cleaning only (Fig. 1a) or electrochemical reduction of 1 (Fig. 1b) or immersion into the same solution (Fig. 1c). The reduction of 1 induces the grafting of a layer that blocks the electron transfer from the Au electrode to FeðCNÞ4 [21]. A Tafel analysis of the CV allows comparison of 6 0 the apparent standard electron transfer rate constants, kapp . For a quasireversible process, the variation of the electron transfer rate, k, with the potential, E, is given from the current, i(E), by [22–25]

Fig. 1. Cyclic voltammetry of a gold disk (1 mm diameter) in H2O + 0.1 M KCl + 4 mM FeðCNÞ4 6 . (a) Bare, (b) electrografted by 2 mM of 1, and (c) immersed 10 min into a 2 mM solution of 1. Scan rate = 0.2 V/s. Inset: Tafel analysis of the CV (Eq. (1)) for (a0 ) bare and (b’) electrografted electrodes.

kðEÞ 

pffiffiffiffi iðEÞ D Ilim  IðEÞð1 þ expðnF=RTðE  E0 ÞÞÞ

ð1Þ

where Ilim is the diffusion limiting current of the sigmoidal convoluted current, I(E), and D is the probe diffusion coefficient (D = 7  106 cm2/s). The extrapolation of log k(E) to E = E0 (Fig. 1) 0 0 leads to kapp . For the oxidation of FeðCNÞ4 6 , kapp decreases from 3 5.5  10 cm/s at an uncovered Au electrode to 2.5  104 cm/s once electrografted. When the gold electrode is immersed into the iodonium solution without potential application, the oxidation of FeðCNÞ4 6 is unchanged, meaning that the surface is uncovered. This is in stark contrast with diazonium salts that spontaneously graft various surfaces, because the radical generation is more endothermic for iodoniums as they are more difficult to reduce than diazoniums [10]. Owing to the dissociative nature of the reduction of these onium salts, this behaviour is a consequence of an I–aryl bond stronger than the N–aryl one [26]. On large Au substrates, the CV analysis was not possible owing to ohmic drop limitations. These surfaces were investigated by SECM in the feedback mode. The SECM approach curves (AC) of the grafted surfaces (Fig. 2) confirm the trends observed on smaller electrodes by CV: the grafting reduces the regeneration of the redox probe at the substrate. A more quantitative analysis could be provided according to Mirkin model for slow heterogeneous kinetics [27,28] at covered substrates, which gives the normalized tip current, IT (IT = i/iT,1 with iT,1 the current measured at infinite dis-

ultramicroelectrode e

Ar. Ar 2 I +

Au

Au I

(1)

Ar Ar

gold gold

vv

Scheme 1. (a) Reductive electrografting of diaryliodonium ions 1 onto gold. (b) SECM local electrografting: the microanode tip held at constant distance from a gold macrocathode.

1232

T. Matrab et al. / Electrochemistry Communications 10 (2008) 1230–1234

ates a polyaryl multilayer (ca. 4 molecular layers) much thinner than that obtained with other aryl moieties [31–34]: the bulky tert-butyl group limits the layer growth [35]. The grafted assembly is more complex than an aryl monolayer, but kapp,SECM at the modified surface is too large to account for electron tunnelling through 2 nm. It is delicate to discriminate between slow permeation of FeðCNÞ3 6 into the multilayer or its reduction at pinholes. In acetonitrile, the AC of the grafted surface with Fc (h) is identical to that of bare Au (e) confirming [29,36] that the hydrophobic layers are permeable to organic solutions [37]. Before and after immersion into an acetonitrile solution of 1, the AC with FeðCNÞ4 6 are unchanged (D,s): the substrate is not spontaneously grafted. An AC was also drawn with the tip held above the center of a locally grafted stripe (see Section 3.2). It exhibits a partial insulating character (), weaker than that observed over a totally grafted surface (), but that can be differentiated from bare ones.

3

2.5

IT

2

1.5

1 0.5

0 0

0.5

1

1.5

2

2.5

3

d/a Fig. 2. Approach curves of gold surfaces for the oxidation of 4 mM of a redox mediator, Fc or FeðCNÞ4 at a Pt tip (radius, a = 12.5 lm). e, h: Fc in acetoni6 trile + 0.1 M NBu4BF4. D, s, , : FeðCNÞ4 in water + 0.1 M KCl. d: electrode– 6 substrate distance. e, D: bare. h, , : electrografted by a 2 mM solution of 1. s: immersed 10 min into a 2 mM solution of 1. : electrografted along a stripe, the tip approaches the center of the stripe. Bold lines: fit by (2) with kapp,SECM = 0.017, 0.009, 0.0028 and 0.0013 cm/s from top to bottom.

tance from the substrate), as a function of (i) the normalized tipsubstrate separation (L = d/a)

IT ¼



 0:78377 0:68 þ 0:3315 expð1:0672=LÞ þ Lð1 þ 1=KÞ 1 þ ð11 þ 7:3KÞ=ðKð110 - 40LÞÞ

c ins  ð1  Iins T =IT Þ þ I T

ð2Þ

(ii) the normalized tip current for conductive and insulating substrates, respectively, IcT and Iins T :

IcT ¼ 0:78377=L þ 0:3315 expð1:0672=LÞ þ 0:68 Iins T

ð3Þ

¼ 1=ð0:15 þ 1:5358=L þ 0:58 expð1:14=LÞ þ 0:0908 exp½ðL  6:3Þ=ð1:017LÞÞ

ð4Þ

and (iii) K = kapp,SECMd/D with kapp,SECM the apparent limiting heterogeneous rate constant. The fit at uncovered substrates (e) is poor as a diffusion limited positive feedback was expected, however, a lower limit of kapp,SECM was estimated from the best fit of the approach curves with (2) at small L. The AC of covered Au with FeðCNÞ4 agrees with the CV 6 analysis and previous studies on surfaces electrografted by diazonium salts [29]. The reduction of 1 on Au forms a compact 2 nm thick (from ellipsometry) organic layer that blocks the ET () to FeðCNÞ3 as kapp,SECM decreases by a factor of 13 upon grafting. 6 0 The values of kapp,SECM (see caption of Fig. 2) are higher than kapp obtained from the CV investigation. On one hand, the CVs provide the electron transfer rate at zero driving force. On the other hand, even though the gold substrate is unbiased during SECM approach curves, its rest potential in a FeðCNÞ4 6 solution is generally much more negative (by more than 200 mV) than the E0 value of this redox probe, its regeneration at the gold substrate is then largely favoured and the SECM provides rates measured at large overpotential. The kinetic limitation observed during SECM analysis of covered substrates could come either from electron tunnelling through the passivating layer or from the probe transport within pinholes or its permeation through the layer. DFT calculations propose that a phenyl radical binds strongly to Au and stands perpendicularly to the surface [30]. The electrografting of 1 gener-

3.2. SECM direct patterning of 1 on Au Layers of an aryl moiety were selectively deposited onto a gold substrate by the SECM direct mode (Scheme 1b). The substrate acts as the working cathode for the reduction of 1, while the Pt tip held at <10 lm from the substrate acts as the counter-microanode (in a three- or two-electrode configuration). In this mode, pattern dimensions can be controlled by the duration of the electrochemical transformation [17,19] and by the substrate potential, current or charge [38]. A two-electrode configuration allows a precise control of the current flowing through the electrodes, their potentials are defined by two-electrode potentiometry. The voltammogram of the reduction of a 20 mM iodonium solution at the gold cathode (Pt tip is both the counter and reference electrode) shows a reduction peak at 2.5 V (vs Pt) with a peak current of ip = 30 nA (Fig. 3a). The iodonium dissociative reduction forbids its regeneration at the Pt microelectrode, the latter, which is inserted in a large glass sheath (RG  5), acts as a non-interacting electrode that hinders iodonium diffusion as an insulator of diameter 125 lm. Thanks to this confinement, the gold substrate behaves as a 25 lm diameter microelectrode, held at 8 lm of an insulating surface, which gives: Ip (L = 8/12.5) = ip/ip,1 = 0.35. Therefore, the ip value (Fig. 3a) corresponds to a limiting current of ip,1 = 86 nA, similar to the 94 nA observed at a 25 lm Pt tip. This demonstrates the confinement of the electric and diffusion fields and the possible electrografting localization. Since in the absence of iodonium, no electrografting occurs, 1 was then reduced at Au during a CV which was paused at 2.6 V for 20 s before the scan is continued. During this pause, the electrode potential is maintained at 2.6 V and the electrode current decreases steadily and slowly as a consequence of the electrode grafting, the amount of charge involved is then controlled and 2.1 lC was injected in the system. Fig. 3b shows the SECM line scan, performed with an aqueous FeðCNÞ4 6 , of the resulting pattern formed on Au. The current decrease reveals the formation of a 110 lm wide passivating layer on the substrate. The electrografting is confined in a much larger region than the 25 lm expected from the CV of Fig. 3a. This could indicate that the grafting is not very efficient at this time (iodoniums graft less efficiently than their diazonium counterparts) and requires longer times of radical generation. The injection of a greater charge (5 lC) generates a larger pattern (210 lm diameter) showing that the control of the charge allows localization of the electrografting in this configuration. In a three-electrode configuration, the CV of a 2 mM iodonium solution at Au leads to the flowing of 250 lC and to a large pattern (diameter 600 lm). The Pt tip cannot sustain such charge and most of the charge leaks through the reference electrode. Owing to the

1233

T. Matrab et al. / Electrochemistry Communications 10 (2008) 1230–1234

1.1

20

b 1

-20

0.9

1

IT

i (nA)

a 0

-40

0.8

-60

0.7

-80 -3

-2.5

-2

-1.5

-1

-0.5

0.6 -300

0

2

-200

-100

E (V/Pt microelectrode)

0

100

200

300

distance (µm) 3.8

1.1

c

d 3.4

[C] / [Au]

IT

1.05

1

3

2.6

0.95

0.9 -400

2.2

-200

0

200

1.8 -2000

-1000

distance(µm)

0

1000

2000

distance(μm)

Fig. 3. Local electrografting in (a, b) 20 mM or (c, d) 2 mM solution of 1 in ACN + 0.1 M NBu4BF4 of a gold substrate in a (a, b) two- or (c, d) three-electrode configuration, 25 lm diameter Pt tip (a,b) held or (c, d) moved at 5 lm/s at 8 lm from the substrate. (a) Cyclic voltammetry, v = 0.1 V/s. (b, c) SECM line scan with FeðCNÞ4 of the 6 electrografted patterns. (b) (1) before and (2) after injection of 2.1 lC. (d) XPS line scan analysis of the C1s and Au4f at%. Distances = 0 at the pattern center.

short experiment duration and lifetime [39] of the intermediate aryl radical species formed upon reduction of 1, this radical should not diffuse over such large distance. The expansion of the grafting is rather explained by the fast reaction of the radical with the surface turning it into a passivating domain, displacing the electrografting to vicinal uncovered domains. This patterning was limited to a much smaller spot (180 lm) by conveniently imposing a low constant current (0.5 lA for 10 s) at the substrate. Finally, we drew lines by moving the Pt microanode, at fixed d, above the cathode substrate. The SECM line scan performed perpendicularly to the electrografted pattern reveals (Fig. 3c) a 450 lm wide stripe. This pattern was analyzed by XPS (Fig. 3d) by scanning a 650 lm large spot perpendicularly to the pattern. The C/Au atomic ratio is maximum in the electrografted area. If taken as the width of this maximum signal the electrografted stripe width is approximately 500 lm, in agreement with the SECM estimate. 4. Conclusion We selectively patterned a conducting surface with an organic thin layer by the electrogeneration of highly reactive radical species. Although diazonium ions are the ‘‘radical” generators most often used to electrograft surfaces, we focused on iodonium ions, which are less prone to spontaneous chemical grafting. Local electrografting of a gold surface is obtained from the iodonium reduction by

the SECM in the direct mode: the gold substrate is used as the working cathode while the SECM tip is the counter-anode. This enables to write patterns of very thin organic layers onto a gold surface. The reduction rate and time control the resolution of the writing process. The best results are obtained when using a two-electrode configuration and controlling the charge that flows through the circuit. The local electrografting was imaged by XPS and SECM, as the organic patterns block the electron transfer to aqueous redox probes. We are currently working on enhancing the writing resolution by improving the electric field confinement at the substrate surface. This might be performed either by using smaller microelectrodes held closer to the surface, or by using more reactive radical source or appropriate potential pulses [40]. Acknowledgments Carole Bilem, ITODYS, Université Paris 7 is thanked for helpful assistance in XPS acquisition. The ‘‘Agence Nationale de la Recherche” is gratefully acknowledged for its financial support via the ANR-06-BLAN-0368 project. References [1] B.P. Corgier, C.A. Marquette, L.J. Blum, J. Am. Chem. Soc. 127 (2005) 18328. [2] G.G. Wildgoose, M. Pandurangappa, N.S. Lawrence, L. Jiang, T.G.J. Jones, R.G. Compton, Talanta 60 (2003) 887.

1234

T. Matrab et al. / Electrochemistry Communications 10 (2008) 1230–1234

[3] R.L. McCreery, J. Dieringer, A.O. Solak, B. Snyder, A.M. Novak, W. McGovern, S. Duvall, J. Am. Chem. Soc. 126 (2004) 6200. [4] P. Viel, C. Bureau, G. Deniau, G. Zalczer, G. Lecayon, J. Electroanal. Chem. 470 (1999) 14. [5] B.D. Bath, H.B. Martin, R.M. Wightman, M.R. Anderson, Langmuir 17 (2001) 7032. [6] J. Pinson, F. Podvorica, Chem. Soc. Rev. 34 (2005) 429. and references therein. [7] C. Combellas, F. Kanoufi, J. Pinson, F.I. Podvorica, Chem. Mater. 17 (2005) 3968. [8] D.J. Garrett, J. Lehr, G.M. Miskelly, A.J. Downard, J. Am. Chem. Soc. 129 (2007) 15456. [9] F. Anariba, S.H. DuVall, R.L. McCreery, Anal. Chem. 75 (2003) 3837. [10] K.H. Vase, A.H. Holm, K. Norman, S.U. Pedersen, K. Daasbjerg, Langmuir 23 (2007) 3786. [11] G. Wittstock, M. Burchardt, S.E. Pust, Y. Shen, C. Zhao, Angew. Chem. Int. Ed. 46 (2007) 1584. [12] C. Combellas, A. Fuchs, F. Kanoufi, Anal. Chem. 76 (2004) 3612. [13] G. Wittstock, W. Schuhmann, Anal. Chem. 69 (1997) 5059. [14] T. Wilhelm, G. Wittstock, Electrochim. Acta 47 (2001) 275. [15] T.H. Treutler, G. Wittstock, Electrochim. Acta 48 (2003) 2923. [16] W. Schuhmann, C. Kranz, H. Wohlschlager, J. Strohmeier, Biosens. Bioelectron. 12 (1997) 1157. [17] S.A.G. Evans, J.M. Elliott, L.M. Andrews, P.N. Bartlett, P.J. Doyle, G. Denuault, Anal. Chem. 74 (2002) 1322. [18] S. Szunerits, N. Knorr, R. Calemczuk, T. Livache, Langmuir 20 (2004) 9236. [19] E. Fortin, Y. Defontaine, P. Mailley, T. Livache, S. Szunerits, Electroanalysis 17 (2005) 495.

[20] S.A.G. Evans, K. Brakha, M. Billon, P. Mailley, Guy Denuault, Electrochem. Commun. 7 (2005) 135. [21] A. Laforgue, T. Addou, D. Bélanger, Langmuir 21 (2005) 6855. [22] J.C. Imbeaux, J.-M. Savéant, J. Electroanal. Chem. 44 (1973) 169. [23] J.-M. Savéant, D. Tessier, J. Electroanal. Chem. 65 (1975) 57. [24] A.J. Bard, L.R. Faulkner, Electrochemical Methods, Fundamentals and Applications, second ed., Wiley, 2001. p. 250. [25] C. Cannes, F. Kanoufi, A.J. Bard, J. Electroanal. Chem. 547 (2003) 83. [26] C. Costentin, M. Robert, J.-M. Savéant, Chem. Phys. 324 (2006) 30. [27] A.J. Bard, M.V. Mirkin, P.R. Unwin, D.O. Wipf, J. Phys. Chem. 96 (1992) 1861. [28] C. Wei, A.J. Bard, M.V. Mirkin, J. Phys. Chem. 99 (1995) 16033. [29] D. Zigah, M. Pellissier, F. Barrière, A.J. Downard, P. Hapiot, Electrochem. Commun. 9 (2007) 2387. [30] D. Jiang, B.G. Sumpter, S. Dai, J. Am. Chem. Soc. 128 (2006) 6030. [31] S. Ranganathan, I. Steidel, F. Anariba, R.L. McCreery, Nano Lett. 1 (2001) 491. [32] J.K. Kariuki, M.T. McDermott, Langmuir 17 (2001) 5947. [33] F. Anariba, S.H. DuVall, R.L. McCreery, Anal. Chem. 75 (2003) 4680. [34] P.A. Brooksby, A.J. Downard, J. Phys. Chem. B 109 (2005) 8791. [35] C. Combellas, F. Kanoufi, J. Pinson, F.I. Podvorica, J. Am. Chem. Soc., (2008), doi:10.1021/ja8018912. [36] A. Adenier, C. Combellas, F. Kanoufi, J. Pinson, F.I. Podvorica, Chem. Mater. 18 (2006) 2021. [37] C. Cannes, F. Kanoufi, A.J. Bard, Langmuir 18 (2002) 8134. [38] R.C. Tenent, D.O. Wipf, J. Electrochem. Soc. 150 (2003) E131–E139. [39] R.G. Kryger, J.P. Lorand, R. Neal, N.R. Herron, J. Am. Chem. Soc. 99 (1977) 7589. [40] J.A. Kenney, G.S. Hwang, Nanotechnology 16 (2005) S309.