Factors Affecting the Preparation and Properties of Electrodeposited

Aug 29, 2006 - Silica Thin Films Functionalized with Amine or Thiol Groups ... Recherche UMR 7564, CNRSsUniVersite´ Henri Poincare´ Nancy I 405, rue d...
0 downloads 0 Views 573KB Size
Download PDF
8366

Langmuir 2006, 22, 8366-8373

Factors Affecting the Preparation and Properties of Electrodeposited Silica Thin Films Functionalized with Amine or Thiol Groups Emilie Sibottier, Ste´phanie Sayen, Fabien Gaboriaud, and Alain Walcarius* Laboratoire de Chimie Physique et Microbiologie pour l’EnVironnement Unite´ Mixte de Recherche UMR 7564, CNRSsUniVersite´ Henri Poincare´ Nancy I 405, rue de VandoeuVre, F-54600 Villers-les-Nancy, France ReceiVed April 11, 2006. In Final Form: July 18, 2006 Well-adherent sol-gel-derived silica films functionalized with amine or thiol groups have been electrogenerated on gold electrodes and both the deposition process and the film properties have been studied by various physicochemical techniques. Electrodeposition was achieved by combining the formation of a self-assembled “nanoglue” on the electrode surface, the sol-gel process, and the electrochemical manipulation of pH to catalyze polycondensation of the precursors. Gold electrodes pretreated with mercaptopropyltrimethoxysilane (MPTMS) were immersed in sol solutions containing the selected precursors (tetraethoxysilane, TEOS, in mixture with (3-aminopropyl)triethoxysilane, APTES, or MPTMS) where they underwent a cathodic electrolysis to generate the hydroxyl ions that are necessary to catalyze the formation of the organosilica films on the electrode surface. Special attention was given to analyze the effects of deposition time and applied potential and to compare APTES and MPTMS films. Characterization was made using quartz crystal microbalance, scanning electron microscopy, cyclic voltammetry, and atomic force microscopy (including in situ monitoring). The electrodeposition process was found to occur at two growing rates: a first slow stage giving rise to rather homogeneous, yet rough, films with thickness in the sub-µm range (increasing continuously when increasing the deposition time), which was followed by a faster gelification step resulting in much thicker (>1 µm) and rougher macroporous deposits. These two successive situations were observed independently on the applied potential except that more cathodic values led to narrower sub-µm ranges (as expected from the larger amounts of the electrogenerated hydroxyl catalyst). Thiol-functionalized silica films were deposited more rapidly than the amine ones and, for both of them, permeability to redox probe was found to decrease when increasing the film thickness because of higher resistance to mass transport.

1. Introduction Implication of silica-based organic-inorganic hybrid materials in electrochemistry is an increasingly growing field of interest.1,2 This relies mostly on their attractive properties likely to be exploited in many applications (chemical and biological sensing, electrocatalysis, long-range charge transfer, solid-state electrochemical devices, and spectroelectrochemistry)1-10 but also on the usefulness of electrochemical techniques to characterize important features of the materials (permeation and recognition properties, redox activity, and mass transport rates).1-3,10 The strategies used to confine the usually nonconductive silicabased materials on electrode surfaces were mostly directed to either the preparation of bulk composite carbon electrodes (powdered materials dispersed in carbon paste11-17 or sol-gel* To whom correspondence should be addressed. Fax: (+33) 3 83 27 54 44. E-mail: [email protected]. (1) Walcarius, A. Chem. Mater. 2001, 13, 3351. (2) Walcarius, A.; Mandler, D.; Cox, J.; Collinson, M. M.; Lev, O. J. Mater. Chem. 2005, 15, 3663. (3) Lev, O.; Wu, Z.; Bharathi, S.; Glezer, V.; Modestov, A.; Gun, J.; Rabinovich, L.; Sampath, S. Chem. Mater. 1997, 9, 2354. (4) Collinson, M. M. Mikrochim. Acta 1998, 129, 149. (5) Walcarius, A. Electroanalysis 1998, 10, 1217. (6) Wang, J.Anal. Chim. Acta 1999, 399, 21. (7) Rabinovich, L.; Lev, O.Electroanalysis 2001, 13, 265. (8) Walcarius, A. Electroanalysis 2001, 13, 701. (9) Collinson, M. M. Trends Anal. Chem. 2002, 21, 30. (10) Walcarius, A. C. R. Chim. 2005, 8, 693. (11) Ribeiro, E. R.; Gushikem, Y. Electroanalysis 1999, 11, 1280. (12) Pessoa, C. A.; Gushikem, Y.; Kubota, L. T. Electrochim. Acta 2001, 46, 2499. (13) Walcarius, A.; Etienne, M.; Sayen, S.; Lebeau, B. Electroanalysis 2003, 15, 414. (14) Wang, Q.; Lu, G.; Yang, B. Langmuir 2004, 20, 1342. (15) Ganesan, V.; Walcarius, A. Langmuir 2004, 20, 3632. (16) Walcarius, A.; Delacote, C.; Sayen, S. Electrochim. Acta 2004, 49, 3775.

derived ceramic-carbon composites7,18-21) or their deposition as thin films on solid electrode substrates.22-36 This latter approach has often exploited the versatility of the sol-gel process that enables straightforward film formation, mainly by spin-coating or dip-coating, with thickness varying between 100 nm to a few microns. Though simple to apply and compatible with production of various functionalized materials (with ion-exchange27-29 or ligand30-33 properties or redox activity34-36), spin-coating and dip-coating can be applied to basically flat surfaces only and the films are deposited unselectively on both conducting and insulating parts of the substrate. (17) Walcarius, A.; Ganesan, V. Langmuir 2006, 22, 469. (18) Tsionsky, M.; Lev, O. Anal. Chem. 1995, 67, 7, 2409. (19) Sampath, S.; Lev, O. Anal. Chem. 1996, 68, 8, 2015. (20) Oskam, G.; Searson, P. C. J. Phys. Chem. B 1998, 102, 2464. (21) Hua, L.; Tan, S. N. Anal. Chem. 2000, 72, 4821. (22) Collinson, M. M.; Rausch, C. G.; Voigt, A. Langmuir 1997, 13, 3, 7245. (23) Makote, R.; Collinson, M. M. Anal. Chim. Acta 1999, 394, 195. (24) Pandey, P. C.; Upadhyay, S.; Pathak, H. C.; Pandey, C. M. D. Electroanalysis 1999, 11, 950. (25) Wang, B.; Cheng, L.; Dong, S. J. Electroanal. Chem. 2001, 516, 17. (26) Wang, Q.; Lu, G.; Yang, B. Langmuir 2004, 20, 1342. (27) Hsueh, C.; Collinson, M. M. J. Electroanal. Chem. 1997, 420, 243. (28) Petit-Dominguez, M. D.; Shen, H.; Heineman, W. R.; Seliskar, C. J. Anal. Chem. 1997, 69, 9, 703. (29) Guo, Y.; Guadalupe, A. R.; Resto, O.; Fonseca, L. F.; Weisz, S. Z. Chem. Mater. 1999, 11, 1, 135. (30) Wei, H.; Collinson, M. M. Anal. Chim. Acta 1999, 397, 113. (31) Guo, Y.; Guadalupe, A. R. J. Pharm. Biomed. Anal. 1999, 19, 175. (32) Yantasee, W.; Lin, Y.; Li, X.; Fryxell, G. E.; Zemanian, T. S.; Viswanathan, V. V. Analyst 2003, 128, 899. (33) Etienne, M.; Walcarius, A.Electrochem. Commun. 2005, 7, 1449. (34) Audebert, P.; Calas, P.; Cerveau, G.; Corriu, R. J. P.; Costa, N. J. Electroanal. Chem. 1994, 372, 275. (35) Audebert, P.; Cerveau, G.; Corriu, R. J. P.; Costa, N. J. Electroanal. Chem. 1996, 413, 89. (36) Wang, J.; Collinson, M. M. J. Electroanal. Chem. 1998, 455, 127.

10.1021/la060984r CCC: $33.50 © 2006 American Chemical Society Published on Web 08/29/2006

Electrodeposited Silica Thin Films

An elegant alternative was recently proposed by Shacham et al.,37,38 reporting that electrodeposition may help in circumventing the above drawbacks. The basic idea is to manipulate the “twostep” sol-gel preparation procedure39 by an electrochemical control of the pH at the electrode/solution interface.40 Starting from a sol solution where hydrolysis is optimal (i.e., pH 3) and condensation very slow, it is possible to accelerate polycondensation rates by applying a negative potential likely to increase pH at the electrode/solution interface and to generate a silica film on the conductive surface. Such local pH-driven sol-gel film formation resembles the electrolytic deposition of metal hydroxides upon reducing water to increase pH locally thus hydrolyzing the metal ions in the vicinity of the electrode surface.41,42 The film thickness is affected by the applied potential, the electrodeposition time, and the nature of the electrode.37 Moreover, it was clearly demonstrated that film deposition really occurred via an electrogenerated-base mechanism37 and not via electrophoretic deposition, which was otherwise applied to deposit sol-gel films but under high electric fields.42-45 Another related approach is the electrochemical tuning of the solubility of trimethoxysilyl-group-modified monomers to derivatize electrode surfaces with organosilica networks.46 After the pioneering work on the electrodeposition of methyltrimethoxysilane on indium-tin-oxide and gold electrodes,37 the method was extended to the tetramethoxysilane precursor to form nonfunctionalized silica films on various conducting substrates upon the electrochemical generation of OH- species arising from water and/or oxygen reduction.47 On the other hand, electrochemically driven pH decrease was also applied to prepare silica films from colloidal solutions under strictly aqueous conditions.48 When not too thick or too hydrophobic, the electrogenerated silica films were permeable to neutral and positively charged redox probes,47,48 otherwise they could act as protective layers against corrosion.49,50 Zirconia51 and titania52 thin films were also electrogenerated on indiumtin-oxide electrodes. Our group has been interested in extending the above electrodeposition approaches to the preparation of functionalized silica films on electrode surfaces, with possible use in electroanalysis, and to increase the electrode-film adhesion via the use of a “molecular glue”.53,54 In this respect, the formation of a self-assembled monolayer (SAM) of mercaptopropyltrimethoxysilane (MPTMS) on gold was proven to act efficiently as a “molecular glue” between the electrode surface and solgel-derived silica films.55 By combining the SAM technology, (37) Shacham, R.; Avnir, D.; Mandler, D.AdV. Mater. 1999, 11, 384. (38) Shacham, R.; Avnir, D.; Mandler, D. PCT Int. Appl. 2005, WO 2005100642. (39) Brinker, C. J.; Scherer, G. W. Sol-Gel Science; Academic Press: San Diego, 1990. (40) Khun, A. T.; Chan, C. Y. J. Appl. Electrochem. 1983, 13, 189. (41) Joseph, J.; Gomathi, H.; Rao, G. P. Electrochim. Acta 1991, 36, 1537. (42) Zhitomirsky, I. AdV. Colloid Interface Sci. 2002, 97, 279. (43) Castro, Y.; Duran, A.; Moreno, R.; Ferrari, B. AdV. Mater. 2002, 14, 505. (44) Castro, Y.; Ferrari, B.; Moreno, R.; Duran, A. J. Sol-Gel Sci. Technol. 2003, 26, 735. (45) Castro, Y.; Ferrari, B.; Moreno, R.; Duran, A. Surf. Coat. Technol. 2004, 182, 199. (46) Leventis, N.; Chen, M. Chem. Mater. 1997, 9, 2621. (47) Deepa, P. N.; Kanungo, M.; Claycomb, G.; Sherwood, P. M. A.; Collinson, M. M. Anal. Chem. 2003, 75, 5399. (48) Collinson, M. M.; Moore, N.; Deepa, P. N.; Kanungo, M. Langmuir 2003, 19, 7669. (49) Sheffer, M.; Groysman, A.; Mandler, D. Corros. Sci. 2003, 45, 2893. (50) Sheffer, M.; Groysman, A.; Starosvetsky, D.; Savchenko, N.; Mandler, D. Corros. Sci. 2004, 46, 2975. (51) Shacham, R.; Mandler, D.; Avnir, D. Chem.-Eur. J. 2004, 10, 1936. (52) Shacham, R.; Mandler, D.; Avnir, D. J. Sol-Gel Sci. Technol. 2004, 31, 329. (53) Sayen, S.; Walcarius, A. Electrochem. Commun. 2003, 5, 341. (54) Walcarius, A.; Sibottier, E. Electroanalysis 2005, 17, 1716.

Langmuir, Vol. 22, No. 20, 2006 8367

the sol-gel process, and the electrochemically induced deposition method, we have been able to prepare porous silica films functionalized with thiol or amine ligands covalently attached to the silica network.53,54 This involved three successive steps: (1) the formation of a partial SAM of MPTMS on gold, (2) the transfer of the pretreated electrode in a silica sol containing TEOS in mixture with an appropriate amount of MPTMS or (aminopropyl)triethoxysilane (APTES), and (3) the application of a cathodic potential to increase pH at the electrode surface and to induce the co-condensation of TEOS and MPTMS or APTES precursors. The MPTMS/TEOS or APTES/TEOS ratios were found to dramatically affect the film electrode performance (sensitivity, mass transfer rates, permselectivity/recognition properties), the optimum being 10% organosilane in the starting sol.53,54 Thiol- and amine-functionalized thin films were promising for the voltammetric detection of Hg(II)53 and Cu(II),54 respectively, after open-circuit preconcentration. Molecularly imprinted sol-gel films made of APTES electrodeposited on gold nanoparticles immobilized on gold electrodes were also described and applied to cytidine sensing.56 In this paper, we have carefully studied the electrogeneration process, as well as the morphology of the deposits and the film properties, corresponding to the formation of such thiol- and amine-functionalized silica-based materials on MPTMS-treated gold substrates. The main goal was to show the effect of several experimental parameters (electrodeposition potential, preparation time, nature of the organosilane precursor) on the organosilica film characteristics (morphology, thickness, permeability). This has been made by using quartz crystal microbalance (QCM), scanning electron microscopy (SEM), cyclic voltammetry (CV), and atomic force microscopy (AFM). In addition, in situ AFM measurements, performed in the course of film electrodeposition, have permitted us to confirm the effects pointed out by the above (mostly ex situ) techniques. 2. Experimental Section Chemicals. (Organo)silane reagents were used as received: tetraethoxysilane (TEOS, 99%, from Sigma-Aldrich), (3-aminopropyl)triethoxysilane (APTES, 99%, from Sigma-Aldrich), and mercaptopropyltrimethoxysilane (MPTMS, 95%, from Lancaster). Ethanol (95-96%, Merk) and HCl (36%, Prolabo) were used to prepared the sol containing the hydrolyzed precursors, and an electrolyte (0.1 M NaNO3) was added to this sol for film electrodeposition. All aqueous solutions were prepared with highpurity water (18 MΩ cm-1) obtained from a Millipore milliQ water purification system. Solutions containing typically 5 mM Ru(NH3)6Cl3 (in 0.1 M NaNO3) were used in CV experiments. Other chemicals were of analytical grade. Procedure Applied for the Electro-Assisted Generation of the Organo-Silica Films on Gold Electrodes. Three types of gold electrodes have been used: (1) homemade epoxy-sealed gold microdisks (1 mm in diameter, polished on 0.1- or 0.05-µm alumina powder), (2) disposable gold electrodes prepared from recordable gold-type compact disks (Au-CDtrodes54), and (3) gold-coated quartz crystal resonators. The electrodeposition method involved several successive steps. The first one is the formation of a selfassembled monolayer of MPTMS on gold, which was achieved by dropping an aliquot of a MPTMS solution (20 mM in ethanol) on the electrode surface for 10 min. After rinsing in a 50:50 water/ ethanol mixture, the electrode was transferred into the starting sol containing TEOS/APTES or TEOS/MPTMS precursors, which was hydrolyzed for 2.5 h at pH 3 prior to electrodeposition. A typical sol mixture was made of 13.6 mmol of “TEOS + APTES” or “TEOS (55) Wang, J.; Pamidi, P. V. A.; Zanette, D. R. J. Am. Chem. Soc. 1998, 120, 5852. (56) Zhang, Z.; Nie, L.; Yao, S. Talanta 2006, 69, 435.

8368 Langmuir, Vol. 22, No. 20, 2006 + MPTMS” precursors in 90:10 ratios, 20 mL of ethanol, 20 mL of aqueous solution of 0.1 M NaNO3, and selected amounts of HCl to reach pH 3. A cathodic potential in the range of -1.0 to -1.4 V was then applied for a given period of time to generate the necessary hydroxyl ions targeted to catalyze polycondensation of the precursors on the electrode surface and film deposition. In a final step, the functionalized silica film was rinsed with distilled water and dried overnight at 70 °C. Instrumentation Used to Characterize the Electrodeposition Process and the Film Properties. Electrodeposition was always performed under potentiostatic conditions, most often in stirred medium (150 rpm) to get homogeneous deposits,37,53 using a µ-Autolab potentiostat monitored by the GPES (General Purpose Electrochemical System) software (Eco Chemie). Quartz crystal microbalance (QCM) was used to monitor the formation of the functionalized films. Electrochemical EQCM experiments were performed using a quartz crystal analyzer (QCA917, SEIKO, EG & G) and electrodeposition was made on gold-coated quartz crystal resonators serving as the working electrodes. The reference and counter electrodes were a Ag/AgCl (Metrohm) and a platinum wire, respectively. Various cathodic potentials from 0.0 to -1.0 V up to -1.4 V were applied to the gold substrate in the TEOS/MPTMS or TEOS/APTES sol solutions, using a Palmsens potentiostat. Quartz crystal frequency variations (difference with respect to the fundamental frequency calibrated at 9 MHz) were measured as a function of time, as sampled by an external computer. The permeability of the film was evaluated by cyclic voltammetry (CV) using Ru(NH3)63+ as the redox probe. CV experiments were carried out with the µ-Autolab potentiostat and GPES software. Measurements were performed at room temperature in a conventional three electrode cell, including the film-modified gold working electrode, an Ag/ AgCl reference (Metrohm), and a Pt wire auxiliary electrode. The morphology of the electrodeposited functionalized silica films was first examined by scanning electron microscopy (SEM) using a Philips XL 30 apparatus. Then, both ex situ and in situ atomic force microscopy (AFM) measurements were performed to get further insights in both the film morphology and the electrodeposition process, respectively. Data were acquired at room temperature using a commercial microscope (Thermomicroscope Explorer Ecu+, Veeco Instruments S.A.S.). V-shaped silicon nitride tips (Ref MLCTEXMT-BF, Veeco Instruments) with a spring constant of 0.1 N m-1 (manufacturer specifications) were used for all ex situ and in situ measurements. The images were collected in contact mode with a scan size ranging from 10 to 50 µm.

3. Results and Discussions The general strategy for film deposition was based on pretreating gold with MPTMS to form a self-assembled partial monolayer, transferring the electrode to a suitable precursor solution, and carrying out the electro-assisted generation of the functionalized silica films under potentiostatic conditions.54 It should be mentioned that the MPTMS self-assembled layer, originating from chemisorption due to strong interaction between thiol and gold,57 was permeable to redox probes in solution54 while ensuring good adhesion properties for sol-gel films (both dip- or spin-coated55,58 and electrodeposited53,54) to the underlying gold substrate. Note that permeability was ensured by the non perfectly self-assembled MPTMS layer (which would have exhibited blocking effects if densely packed onto the electrode surface53,58,59) as a result of short contact time for MPTMS on gold and because of possible reductive desorption of thiols from the gold surface when applying a cathodic potential.59,60 This last reaction is reversible (cathodically desorbed MPTMS can (57) Thompson, W. R.; Cai, M.; Ho, M.; Pemberton, J. E. Langmuir 1997, 13, 2291. (58) Thompson, W. R.; Pemberton, J. E. Chem. Mater. 1995, 7, 130. (59) Che, G.; Cabrera, C. R. J. Electroanal. Chem. 1996, 417, 155. (60) Widrig, C. A.; Chung, C.; Porter, M. D. J. Electroanal. Chem. 1991, 310, 335.

Sibottier et al.

Figure 1. Cyclic voltammograms recorded in a solution containing 5 × 10-3 M Ru(NH3)6Cl3 (in 0.1 M NaNO3), using a bare gold electrode (a) or thiol-functionalized silica film modified gold electrodes (b-e) prepared by electrogeneration at -1.2 V for 10 s (b), 20 s (c), 30 s (d), and 60 s (e), from a sol solution containing 90% TEOS and 10% MPTMS as the precursors. Scan rate: 20 mV s-1.

be anodically re-deposited59) so that the “nanoglue” character of the MPTMS SAM was simply maintained by applying an anodic potential (i.e., 0.0 V) at the end of the electrodeposition process. Hydrodynamic conditions were usually applied (i.e., rotating disk electrode spinning at 150 rpm) because it gave rise to more homogeneous deposits in comparison to those obtained without stirring.37,53 3.1. Effect of Electrodeposition Time. Figure 1 shows typical cyclic voltammograms recorded in a solution containing ruthenium(III) hexaammine using gold electrodes coated with thiolfunctionalized silica films electrodeposited at -1.2 V for increasing periods of time. Clearly, as the electrolysis time is growing, as the amount of electrodeposited organosilica is expected to increase, as peak currents were found to continuously decrease and CV curves to shift progressively from well-defined peaks to wave-shaped signals of lower intensity. This behavior is explained by the presence of thiol-functionalized silica layers becoming less permeable to external redox probes the longer the electrodeposition time was (restricted motion of Ru(NH3)63+ across the porous medium), and the sigmoidal form of the curves recorded at longer electrolysis times suggests the presence of pores inducing diffusion change from linear to radial (as reported for ultramicroelectrode ensembles61). Film formation was very fast (a few seconds) in comparison to pure silica (TMOS based) or methylated sol-gel films,37,47 whereas CV curves changing from peaked to wave forms were also observed with other kinds of sol-gel films.37,55,62 SEM examination of the thiol-functionalized silica films electrodeposited at -1.2 V for increasing periods of time (Figure 2) reveals that rather homogeneous thin layers were obtained for short electrodeposition duration while rougher and much thicker deposits were observed after longer times. Thinner films can be evidenced via the polishing scratches (from 0.1-µm alumina polishing) on the underlying gold substrate (Figure 2A), which remained clearly visible after film formation for 10 s (Figure 2B) and even noticeable on the SEM picture of the 20 s electrode(61) Gao, Z.; Siow, K. S. Electrochim. Acta 1997, 42, 315. (62) Collinson, M. M.; Wang, H.; Makote, R.; Khramov, A. J. Electroanal. Chem. 2002, 519, 65.

Electrodeposited Silica Thin Films

Langmuir, Vol. 22, No. 20, 2006 8369

Figure 2. Scanning electron micrographs of a bare gold electrode (A) or thiol-functionalized silica film modified gold electrodes (B-E) prepared by electrogeneration at -1.2 V for 10 s (B), 20 s (C), 30 s (D), and 60 s (E), from the same TEOS/MPTMS sol solution as in Figure 1.

posited film (Figure 2C). In this last case, one can see that the film, though homogeneously and uniformly deposited on the electrode surface, is constituted of closely packed small particles. This agrees well with the base-catalyzed polycondensation process involving a sol of particulate nature.39 Increasing the deposition time by only 10 s resulted in a much thicker deposit consisting of aggregates of particles that are linked together in a threedimensional random arrangement (Figure 2D), as also shown the 60 s electrodeposited film (Figure 2E). The macroporosity of these films can be due to the formation of hydrogen bubbles as a consequence of electrogeneration of the base catalyst (from proton and/or water reduction). The existence of such submicrometric pores can contribute to explain the radial diffusion regime observed by CV as the wave-shaped voltammetric signals were only observed (see curves d and e in Figure 1) for the thicker films macroporous deposits (Figure 2D,E). It seems from the above results that electrodeposition time can be adjusted either to low values to prepare rather uniform, yet rough, thin deposits, or to higher values to get much thicker macroporous films, with an abrupt evolution between these two situations. This is better evidenced by in situ monitoring of the film growing by quartz crystal microbalance (QCM). As shown in Figure 3A, the frequency-time plots recorded at the gold electrode following a potential step from 0 V to a suitable cathodic value (in the -1.0 to -1.4 V range) reveal the existence of two distinct regions. The first region was characterized by slowfrequency variations due to slow condensation of the sol into thin layers on the electrode surface, which was followed by an abrupt frequency drop indicating gelification of much higher quantities of the (organo)silica precursors. These observations agree well with the SEM pictures in Figure 2 indicating a sudden thin-to-thick film transformation for electrodeposition times differing by only 10 s (compare parts C and D in Figure 2). 3.2. Effect of Electrodeposition Potential. The time at which the above abrupt variation occurred (tdrop) was strongly dependent on the applied potential (see, e.g., Figure 3A for thiol-

Figure 3. Frequency-time plots obtained by quartz crystal microbalance at gold electrodes in 90:10 TEOS/MPTMS (A) or TEOS/ APTES (B) sol solutions following potential steps from 0.0 V to selected cathodic values (in the -1.0 to -1.4 V range).

functionalized films). As the potential was made more negative, tdrop was found to decrease and this trend was always observed when repeating the experiments, even if variability was rather high (standard deviations for tdrop ranging from 10 to 40%, n )

8370 Langmuir, Vol. 22, No. 20, 2006

Sibottier et al.

Figure 4. Variation of film thickness as a function of the electrodeposition time for amine-functionalized silica films electrogenerated at -1.2 V, from a sol solution containing 90% TEOS and 10% APTES as the precursors.

Figure 5. Influence of the aging time of the sol (90:10 TEOS/ APTES) on the transition time from slow to fast deposition of aminefunctionalized silica films on gold electrodes prepared by electrogeneration at -1.15 V, as measured from quartz crystal microbalance experiments (as in Figure 3).

3). Such behavior is fully explained by the electrogeneration of increasing amounts of OH- species by time unit when applying more cathodic potentials, as confirmed by measurement of higher current densities (see Figure A in the Supporting Information). Interestingly, integrating the current-time curves from the beginning (t ) 0) to the time at which the abrupt variation occurred (tdrop) gave charge values of the same order of magnitude independently on the applied potential (i.e., 3.3 mC for -1.0 V; 3.6 mC for -1.1 V; 3.7 mC for -1.2 V; 4.9 mC for -1.4 V). This suggests that formation of thick macroporous films (Figure 2D,E) requires the generation of a minimum amount of OHspecies, which can be made by applying either a highly negative potential for a short time or lesser cathodic values for longer times. When applying conditions maintaining the OH- amount below the threshold value, thinner films displaying no apparent macroporosity were obtained (Figure 2B,C). An additional comment relative to current-time curves presented in the Supporting Information (Figure A) is the very slow decrease of the electrolysis currents (after the expected decay in the first few seconds) during growing of the films, indicating that they remained highly porous even when becoming thicker and thicker. The characteristic profile of the curves plotting the amount of the electrodeposited material as a function of the electrolysis time and the abrupt variation of tdrop with the applied potential was observed for both thiol- and amine-functionalized silica films (Figure 3), except that film formation was slower when using APTES instead of MPTMS as the organosilane precursor (see section 3.3. for explanation). The two regions discussed above are even better evidenced by measuring the film thickness (by AFM) as a function of the electrodeposition time, as illustrated in Figure 4 for amine-functionalized films. Before tdrop, the film thickness remained in the sub-micrometric range and increased almost linearly with the duration of the electrodeposition. This confirms that gelification occurred immediately after the generation of the base catalysts at the electrode surface, in agreement with what has been reported for electrodeposited nonfunctionalized TMOS-based silica films.47 Much thicker deposits (>1 µm) were observed after tdrop, growing faster with the electrodeposition time and showing larger variations from one film to another prepared in the same conditions. The origin of the two regions is not fully understood at this stage but a possible interpretation might be proposed on the basis of the particular mechanism likely to induce gelation of a

base-catalyzed sol by involving the nucleation of small particles.39 In the present case, gelification is due to the electrogeneration of the base catalysts at the electrode/solution interface, leading to the formation of colloidal (organo)silica particles in the diffusion layer. Because of the presence of a higher OHconcentration close to the electrode surface, the film started to grow by aggregating the small particles in the form of a thin film while other colloidal (organo)silica particles were formed in the whole volume of the diffusion layer where OH- species were present as well. The existence of an alkaline local pH over several hundreds of microns was pointed out by the observation of a purple color at the electrode/solution interface when introducing phenolphthalein in the medium. After a selected period of time (the shorter the time, the more cathodic the applied potential was), these colloidal particles are expected to attain a critical mass (or concentration) facilitating their aggregation and falling down as a whole gel on the electrode surface. This possible explanation is also sustained by the fact that the transition time was found to shorten when increasing the hydrolysis time, especially for the first hours and then tended to stabilize (Figure 5), as a consequence of enhanced particulate state of the sol.39 3.3. Comparison between Amine- and Thiol-Functionalized Films. One knows from the above electrochemical QCM measurements that electrodeposition of both amine- and thiolfunctionalized silica films seems to occur according to the same mechanism, giving rise to homogeneous thin films growing slowly and continuously at small deposition times and to much thicker macroporous deposits later on (Figure 3). The permeability of these films to the Ru(NH3)63+ redox probe, as characterized by cyclic voltammetry using the film electrodes dried beforehand, followed the same trend for amine-functionalized films (data not shown, but similar as in Ref.54) as for the thiol-ones (Figure 1). Also, the SEM examination of the electrodeposited aminefunctionalized films has revealed a morphology similar to that of the thiol-ones (aggregates made of smal particles, see Figure B in the Supporting Information); thus, it is consistent with the particulate nature of the base-catalyzed sol as previously reported for pure silica (nonfunctionalized) TMOS-based electrodeposited films.47 There was however a significant functional group effect on the electrodeposition process applied to get amine- or thiolfunctionalized silica films. This is best illustrated in Figure 3 showing that film formation was much slower when using APTES

Electrodeposited Silica Thin Films

Langmuir, Vol. 22, No. 20, 2006 8371

Figure 6. 3D AFM images (images size 10 by 10 µm, x/y/z ratios taken as 1/1/10) of a bare gold electrode (A) or thiol-functionalized silica film modified gold electrodes (B-D) prepared by electrogeneration at -1.2 V for 10 s (B), 20 s (C), and 60 s (D), from the same TEOS/ MPTMS sol solution as in Figure 1.

instead of MPTMS as the organosilane precursor (compare parts “A” and “B” of the figure where longer tdrop were observed when preparing amine-functionalized deposits at the same applied potential as for the thiol-ones). This behavior was also noticeable from the permeability characterization by CV for which the drop in the intensity of the voltammetric signals was observed after longer electrodeposition times for APTES-based films as for MPTMS, but this observation must be taken with great care as thiol-functionalized films are more hydrophobic than the amineones and such enhanced hydrophobicity may induce additional restriction to mass transport.53,63 A possible explanation to longer tdrop for amine-functionalized films can be proposed on the basis of the basic character of amine groups (pKa of APTES equal to 9.564), which are present in the deposition medium at a rather high concentration (34 mM). Sol solutions were always prepared at pH 3 (to ensure high hydrolysis rates and negligible polycondensation39), using a strong acid, so that much more protons were necessary to reach pH 3 when using APTES than when using MPTMS (about 35 mM HCl was added in the first case, whereas 1 mM was enough in the second case). Consequently, greater amounts of acidic species (free protons and ammonium groups) needed to be electrolyzed before starting to generate hydroxyl ions in a TEOS/APTES sol than in a TEOS/ MPTMS medium. Also, reduction of a weak acid (i.e., alkylammonium) occurs at more cathodic values than reduction of free protons and the cathodic barrier due to water reduction is located at lower potential values in the presence of a base in the medium (see Figure C in the Supporting Information for an illustration of the case of pH 3 media with or without added butylamine, as studied by CV). Organo-functional group effects may also contribute to influence polycondensation rates.65,66 All of the above considerations may help us to understand the fact that electrodeposition of amine-functionalized silica films was slower than for getting thiol-containing deposits by the same process. (63) Niedziolka, J.; Palys, B.; Nowakowski, R.; Opallo, M. J. Electroanal. Chem. 2005, 578, 239. (64) Coche-Guerente, L.; Desprez, V.; Labbe, P. J. Electroanal. Chem. 1998, 458, 73. (65) Prabakar, S.; Assink, R. A. J. Non-Cryst. Solids 1997, 211, 39. (66) Delak, K. M.; Sahai, N. Chem. Mater. 2005, 17, 3221.

3.4. Ex Situ and in Situ AFM Characterizations. AFM was also used as a powerful tool to characterize the electrodeposited films, by providing further information of the surface roughness and morphology of the deposits. This technique was first applied on dried films and then, in situ, to characterize films growing in the course of the gelification process. Uniformly carved gold surfaces were also used in addition to the flat electrode substrates to point out the interest of the electrodeposition method to prepare thin films of regular thicknesses on very rough surfaces and to show how the roughness of the films evolves when passing from thin to thick deposits. Figure 6 depicts 3D views of AFM images obtained for a freshly polished bare gold electrode (part A), for two thin thiolfunctionalized silica films electrodeposited for small times on the gold surface (parts B and C), as well as a much thicker deposit of the same material electrodeposited for a longer time (part D). Comparing parts A and B of the figure indicates that the small polishing scratches visible on the bare substrate (from 0.05-µm alumina polishing) are rapidly filled with the electrodeposited material, confirming the specificity of the electrodeposition method to coat small-size patterns constituting the underlying conductive support.38 The particulate nature of the electro-assisted base-catalyzed gelification process was also noticeable at the early times of electrodeposition, the thinner deposits being characterized by closely packed nanoparticles (see part B in Figure 6). These particles seem to act as nuclei for further growing of the film, inducing thereby significant increase in the roughness of the film (see part C in Figure 6). These variations were dramatically more pronounced when passing to the thicker films for which the random threedimensional arrangement of bigger aggregates of nanoparticles gave rise to less compact and extremely rough deposits (see part D in Figure 6). The above observations are consistent with mean roughness data for samples of Figure 6. Indeed, the root-meansquared roughness (RMS) of the film electrodeposited for 10 s (RMS ) 8.5 nm, Figure 6B) was of the same order of magnitude as that characterizing the bare gold electrode (RMS ) 6.3 nm, Figure 6A), whereas the deposit obtained after 20 s electrodeposition was hardly higher (RMS ) 12.1 nm, Figure 6C). On the

8372 Langmuir, Vol. 22, No. 20, 2006

Sibottier et al.

Figure 8. In situ AFM monitoring of the growth of an electrogenerated thiol-functionalized silica film on a gold CDtrode: the two lines in red and yellow corresponds to the height profiles measured on the AFM image for the hollow parts of the electrode and the two other lines (pink and blue) depict the top of the electrode. The dashed black lines represent the average evolution profiles. The inset is a 3D image of the full deposit.

Figure 7. (A) SEM picture of the gold CDtrode. (B) Selected height profiles recorded in situ in the course of the electrodeposition of a thiol-functionalized silica film from a sol solution containing 90% TEOS and 10% MPTMS as the precursors. The curves correspond to the bare substrate (a) and to AFM lines recorded after 10 s (b), 40 s (c), 100 s (d), 125 s (e), 170 s (f), and 210 s (g).

contrary, much higher values were calculated for thicker films (i.e., RMS ) 57.0 nm for a thiol-functionalized film electrodeposited for 60 s, Figure 6D). Such changes in film roughness, as well as the morphology of the deposits, were similar for both thiol- and amine-functionalized materials. Functionalized silica films were also electrodeposited on gold surfaces displaying a regular streaked structure made of trenches of 1-µm width separated by mounds of 500-nm width and 100nm height relative to the base (see part A in Figure 7). These electrodes were fabricated from recordable gold-type compact disks and called Au-Cdtrodes.67 From the electrochemical point of view, they behave similarly to the conventional commercially available gold electrodes67-69 and they are likely to act as suitable supports for electrodeposition of silica-based materials.53,54 Such a CDtrode substrate allows us to record in situ AFM images during the electrodeposition process. Part B of Figure 7 depicts (67) Angnes, L.; Richter, E. M.; Augelli, M. A.; Kume, G. H. Anal. Chem. 2000, 72, 5503. (68) Richter, E. M.; Augelli, M. A.; Kume, G. H.; Mioshi, R. N.; Angnes, L. Fresenius J. Anal. Chem. 2000, 366, 444. (69) Richter, E. M.; Augelli, M. A.; Magarotto, S.; Angnes, L. Electroanalysis 2001, 13, 760.

AFM profiles of the AFM image (inset of Figure 8) measured perpendicularly to the band arrays at various electrodeposition times (expressed as the number of AFM lines recorded over a 50-µm distance, each scan lasting 1.25 s, 300 × 300 scan lines). As it can be seen, the thin deposits formed at the beginning of the electrodeposition process follow quite well the regular topography of the gold substrate, demonstrating the interest of this method to produce thin films uniformly deposited on hilly surfaces. As far as the amount of electrodeposited material increased, the “memory shape” arisen from the underlying support progressively disappeared and the film surface roughness increased (as discussed above for thicker deposits). This occurred for films thicker by about two times over the step height of the gold arrays. In the other spatial direction, the AFM lines parallel to the trenches and mounds in the course of the electrochemically induced gelification process are plotted following the Y line number that is related to the electrodeposition time (Figure 8). A real-time evidence that the film begins to grow slowly for smaller electrodeposition times and then much more rapidly when applying the cathodic potential for longer times is clearly pointed out. Such a trend was observed independently by the fact that AFM lines were recorded in the hollow parts of the electrode or on the mounds. Within the variations inherent to the experiment, the average evolutions (dashed lines on Figure 8) resemble very much the QCM profiles, showing good agreement between the amount of gelified material and the thickness of the resulting films. The results also point out that the hilly topography of the electrode surface is maintained for thin deposits while the lines recorded on the base or the top of the electrode arrays tended to merge in a single profile (see the right part of Figure 8).

4. Conclusions Electrogeneration of OH- species at the surface of MPTMSpretreated gold electrodes, which were immersed in sol solutions containing APTES/TEOS or MPTMS/TEOS precursor mixtures, led to the deposition of amine- or thiol-functionalized silica films with properties and characteristics depending on the electrodeposition conditions. Film thickness was found to increase and film porosity to decrease, by increasing the electrodeposition time of shifting the applied potential toward more cathodic values. The whole process was characterized by two successive and distinct

Electrodeposited Silica Thin Films

rates, starting by a slow deposition stage leading to thin porous deposits, which was followed by a much faster film growing in the form of rough macroporous coatings, as demonstrated by in situ QCM and AFM measurements. The main parameter governing the evolution from thin to thick films seemed to be the amount of electrogenerated OH- catalyst rather than the speed at which these species were produced, this latter controlling the time required to get thick deposits. The film morphology was typical of particulate polycondensation, independently on the type of organo-functional groups incorporated in the final material.

Langmuir, Vol. 22, No. 20, 2006 8373

Acknowledgment. The authors thank A. Kohler for SEM experiments. We are also grateful to M. Etienne for fruitful discussions and help in EQCM and AFM experiments. Supporting Information Available: Current time curves recorded at gold electrodes in 90:10 TEOS/MPTMS sol solutions (Figure A). SEM of an amine-functionalized silica film modified gold electrode (Figure B). Cyclic voltammograms recorded in HCL solution with or without butylamine (Figure C). This material is available free of charge via the Internet at http://pubs.acs.org. LA060984R