- Email: [email protected]
Study on the adhesion mechanism of electrodeposited nickel clusters on carbon substrates
Applied Surface Science 255 (2009) 4309–4315
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Study on the adhesion mechanism of electrodeposited nickel clusters on carbon substrates M.F. De Riccardis a,*, D. Carbone a, V. Martina a, M. Re a, B. Bozzini b, L. D’Urzo b a b
ENEA Brindisi Research Centre, SS.7 Appia Km 712, 72100 Brindisi, Italy Dipartimento di Ingegneria dell’Innovazione, Universita` del Salento, via Monteroni, I-73100 Lecce, Italy
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 July 2008 Received in revised form 21 October 2008 Accepted 21 October 2008 Available online 21 November 2008
Growth and adhesion mechanisms of Ni clusters electrodeposited on three different carbon-based substrates have been studied. Glassy carbon, carbon paper and PAN-based fibres have been used as working electrodes and Ni clusters have been electrodeposited from a NiCl26H2O. Ni reduction on carbon substrates has been studied by cyclic voltammetry, chronocoulometry and in situ SERS, whereas the morphological and structural characterization of the interface between Ni clusters and carbon-based substrates has been performed by High Resolution TEM. From our results we can conclude that the precipitation of Ni hydroxides and basic salts in the unbuffered catholyte promotes the adhesion of Ni clusters on the carbon-based substrates considered in this study. This feature of the investigated Ni clusters electrodeposition suggests that it may be a suitable fabrication route for applications in catalytic processes, such as metal-particle catalysed growth of carbon nanotubes. ß 2008 Elsevier B.V. All rights reserved.
Keywords: Ni clusters Electrodeposition SERS TEM
1. Introduction Electrodeposition is a versatile and inexpensive technique that allows the synthesis of metals in several morphologies and structures, including continuous metallic films and nanoparticles, also on 3D substrates. Electrodeposition of metallic clusters has a host of cutting edge prospective applications, such as the catalytic growth of carbon nanotubes (CNT) in CVD processes [1–3]. In order to obtain a reliable catalytic surface for CNT, and in general for the growth of carbon nanostructures (CNS), such as nanofibres or nanowalls, a key property is the adherence of the clusters to the substrate. Outstanding adherence is mandatory in catalytic reactions, in order to avoid clusters coalescence during heatassisted processes. Nevertheless, in the growth of metal catalysed carbon nanostructures, strong interactions between support and metal catalyst, together with minimization of surface mobility of catalyst particles, are highly desirable process highlights. Well adhered metallic coatings can be produced by electrodeposition once the plating parameters, the substrate treatments and the electrolytic bath composition are optimized [4–6]. In previous papers the authors demonstrated that it is possible to obtain well adhered and uniformly distributed Ni clusters, on PAN-based fibres and carbon paper [1,2]. The good adhesion of electrodeposited Ni clusters was demonstrated by performing several experiments,
* Corresponding author. E-mail address: [email protected] (M.F. De Riccardis). 0169-4332/$ – see front matter ß 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2008.10.127
consisting in immersion, magnetic stirring, centrifugation and ultrasonication into different liquids (such as deionized water, ethanol, acetone, n-butanol). Nevertheless, nano-scratch tests were performed on Ni coated supports by means of Atomic Force Microscope (AFM) equipped with a diamond tip; the metallic clusters were moulted and not removed by the diamond tip. The experimental details are reported in [2]. It is well known that Carbon and Nickel not form stable carbides [7]; so the reason of this evidenced strong bond between electrodeposited Ni and carbon support is still undefined. The aim of this paper is to provide a further piece of information to understand the grounds of adhesion between Ni clusters and carbon substrates; for this reason with insight we analysed the interface between Ni electrodeposit and carbon support. Our attention has been directed to Ni and C because their jointed use is highly attractive. In fact remarkable catalytic activity of Ni and amphoteric properties of C make them able to be widely used in metal catalysed reactions as catalyst and support material respectively. Nevertheless, the nature of interfacial bond between metal/carbon or metal/polymer, has been extensively exploited in different technologies, e.g. microelectronics, composites, adhesive bonding and packaging. In particular, a lot of research effort has been devoted to the understanding of the interfacial interactions between carbon-based materials, especially HOPG, and metal deposits [8–11]. In the specific issue of metal-catalysed carbon nanostructure growth, the catalytic activity of metal particles can be more or less dramatically affected by both physical and chemical interactions
4310
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
between substrate and catalyst [11]. For this reason, we are interested in characterising the nature of the bond between Ni clusters and substrate [1–3]. To this aim, we studied the Ni electrodeposition process by combining some classical electrochemical methods – such as cyclic voltammetry and chronocoulometry – with in situ SERS. This technique is typically carried out with precious-metal electrodes, exhibiting a remarkable degree of surface-enhancement. Nevertheless, it has been proved relatively recently (see e.g. [12] for an updated review) that the SERS effect – though at a significantly lower level with respect to precious metals – can be achieved also with Fe- and Pt-group metals. The use of SERS in electrodeposition studies in such transition metals has received very limited attention. Ni electrodeposition has been used in one case to achieve Ni electrode roughening prior to SERS work [13]. Electrodeposited Co, Ni, Pt and Pd nano-wire arrays have been shown to exhibit surface enhancement [14]. Co electrodeposition in the presence of coumarin has also been investigated [15]. In addition to electrochemical analyses on Ni electrodeposition process, structural investigations at the interface between carbon supports (carbon paper, carbon-based fibres and glassy carbon) and the electrodeposited Ni clusters were performed by means of transmission electron microscopy (TEM).
Raman spectra were recorded using a LabRam microprobe confocal system. A 50 long-working-distance objective was used and the excitation line at 632.8 nm was provided by a 12-mW He– Ne laser. The slit and pinhole were set at 200 and 400 mm, respectively, corresponding to a scattering volume of about 3 pL; Raman spectra were acquired with a 600 grid/mm spectrometer. The recorded Raman intensities are directly proportional to the discharge current of the CCD detector. In situ electrochemical measurements were performed in a glass cell with a vertical glassy carbon (GC) electrode (1 3 mm) embedded in a Teflon holder. A limited degree of SERS activity can be attained with GC [16] and Ni [17–19] that can be enhanced during the electrodeposition process, owing to the formation of active Ni clusters. The counter electrode was a Pt wire loop of total area ca. 2 cm2, concentric and coplanar with the gold working electrode. An external Ag/AgCl reference electrode was used and all the potentials were reported on this scale. The applied potential was controlled by an AMEL 2049 potentiostat.
2. Experimental
2.3. TEM characterization
2.1. Ni electrodeposition
Transmission Electron Microscopy (TEM) was carried out by TECNAI G2 F30 operating at an acceleration voltage of 300 kV and with point resolution of 0.205 nm. For TEM observations of electrodeposited Ni clusters, the samples were prepared with three methods, according to the characteristics and the nature of different substrates. With carbon paper, a powder sample surface was obtained by scraping it off the cathode surface and dispersing it in ethylic alcohol; the dispersion was cast on a carbon coated grid. In the case of glassy carbon substrate, two pieces of the same specimen were sandwiched and mechanically thinned down to about 20 mm of thickness, according to the standard preparation method of TEM crosssections. The final thinning was performed by ion beam milling (PIPS Gatan) with Ar ions at 5 kV. For PAN Carbon fibres, a Pt thin layer was deposited by electron beam in order to protect the electrodeposited Ni, during the next Ion beam deposition of a thicker Pt layer necessary for TEM lamella preparation. Then a longitudinal cross-section was prepared with the in situ Lift Out technique with the FEI Strata 400 FIB/SEM system. For all the samples, conventional Bright Field TEM images were obtained from many areas, in order to highlight global morphological features of the electrodeposited Ni clusters, such as size, distribution and shape, while High Resolution TEM images yielded insightful structural information regarding the microstructure of the clusters and the nature of the interface. Nevertheless, High Angle Annular Dark Field (HAADF) STEM images were taken from cross-sections of Ni on glassy carbon and PAN fibres; this kind of analysis has highlighted chemical differences on large areas, thanks to compositional and thickness contrast mechanism. For chemical analysis, Energy Dispersive Spectrometry (EDS) was also carried out with the electron beam focused both on Ni clusters and on their interface with the substrate.
For Ni electrodeposition, a solution of 0.5 M NiCl26H2O with deionized water was prepared and pH was adjusted to 3.0 by HCl addition. Ni clusters were electrodeposited on different carbon substrates, such as PAN carbon-based fibres (obtained by pyrolysis of Polyacrylonitrile), carbon paper, commercially named Papyex, and glassy carbon. Two different set-up of electrolytic cell were used, according to the nature of the substrates: (i) a cylindrical cell in the case of fibre bundles and (ii) a prismatic cell with carbon paper and glassy carbon. The cylindrical cell was equipped with a Pt spiral and a cylindrical assembly of PAN-based fibres as counter-electrode, in order to achieve a high degree of electrodeposit uniformity along both the fibre length and diameter. Uniform deposits on planar substrates were obtained with a plane-parallel electrode configuration in the prismatic cell, featuring a vertical cathodic specimen holder and Pt foil as anode. Voltammetric and potentiostatic measurements were carried out by using a PAR potentiostat model 273A. The potentials were measured against a Ag/AgCl reference electrode and reported on this scale. Different cleaning treatments were applied to the carbon substrates before Ni electrodeposition. PAN fibres, cut from a woven fabric, were treated at 650 8C for 1 h in air to remove the binder. Glassy carbon substrates were treated by boiling the coupons in a solution of HCl and HNO3 (3:1) for 10 min and subsequently rinsing them in deionized water, to remove any organic residue. Carbon paper substrates were washed with acetone and ethylic alcohol, and rinsed in deionized water. The exposed area was 21 cm2 for PAN-based fibres and 0.95 cm2 for carbon paper and glassy carbon electrodes. Experimental details on the electrodeposition process are described elsewhere [2]. After Ni electrodeposition, morphological characterizations of Ni coated carbon substrates surface were performed by using XL40 LaB6 (Philips) SEM. A monomodal narrow diameter distribution was found for the Ni clusters, around 50 nm, under different conditions for each substrate; in the case of glassy carbon at 0.85 V and 20 mC/cm2 total charge, in the case of carbon paper at 0.8 V and 100 mC/cm2
total charge, and in the case of PAN-based fibres at 0.75 V and 7.5 mC/cm2 total charge. 2.2. SERS characterization
3. Results and discussion 3.1. Electrochemistry The more appropriate term for the electrodeposition technique should be electrocrystallization, usually used for electrode pro-
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
cesses which involve formation of a solid state as a result of the reduction of ions in solution, as in the case of metal deposition [4,5]. Electrocrystallization consists of several steps: (i) diffusion of ions in solution towards the electrode surface, (ii) formation of adsorbed atoms, (iii) clustering of ad-atoms to form critical nuclei, (iv) incorporation of adsorbed ions at lattice site and (v) development of crystallographic and morphological characteristics of the deposit. The last step is the most interesting, in particular when the electrodeposited material is used as carbon nanostructures (CNS) catalyst; in fact it was verified that CNS characteristics depend in part on microstructural characteristics of their catalysts (see e.g. [20,21]). Several studies of electrochemical reduction and deposition of Ni based on Bockris’ seminal paper [22] have well-demonstrated that the nickel monohydroxide ion is the intermediate species when a charge transfer mechanism occurs in aqueous unbuffered solutions of simple nickel ions [23,25]. Indeed, the diffusion of ions in solution towards the electrode surface involves Ni(OH)+ species (reaction (1)). In succession, an one-electron reduction reaction takes place forming the adsorbed intermediate, which is believed to be the active intermediate during the metal formation (2): Ni2þ þ H2 O ¼ NiðOHÞþ þ Hþ
(1)
NiðOHÞþ þ e ! NiðOHÞads
(2)
Ni(OH)ads, is reduced to nickel metal via two parallel reactions; the first reaction uses the adsorbed nickel monohydroxide as catalyst, reducing nickel ions to nickel metal (3), and the second reaction allows direct metal reduction from Ni(OH)ads involving one-electron transfer reaction (4): NiðOHÞads þ Ni2þ þ 2e ! Ni þ NiðOHÞads
(3)
NiðOHÞads þ e ! Ni þ OH
(4) 2+
Cui and Lee [23,24] demonstrated that at low Ni concentration in unbuffered aqueous solution, Ni deposition is inhibited by presence of oxygen. Their voltammetric studies demonstrated that at cathode potentials, oxygen reduction causes the increase in the interfacial pH and the precipitation of Ni(OH)2 on the electrode. The hydroxide precipitation takes place at less negative potential than that of Ni reduction, so that the adsorbed intermediate Ni(OH)ads formation is lessened and subsequently nickel nucleation is inhibited. Same authors studied hydroxide formation also in absence of oxygen at low Ni2+ concentration in unbuffered aqueous solution [23,24]. In this case, at high cathode potentials, quite similar to Ni reduction, water discharge causes hydrogen evolution; so pH increases and Ni(OH)2 precipitates on the substrate. When the scan is switched to anodic potentials, the intensity of nickel oxidation peak is much smaller than nickel deposition, irrespective of the presence or absence of oxygen. This phenomenon confirms the decreased Ni formation due to the Ni(OH)2 presence. At high electrolyte concentration, both in presence and in absence of oxygen, Ni deposition is not inhibited by Ni(OH)2. In fact, the adsorbtion of intermediate complex Ni(OH)ads is higher than at low Ni2+ concentration and Ni reduction is favoured [23]. Fig. 1 shows a cyclic voltammetric curve acquired for an aqueous solution of 0.5 M NiCl26H2O at pH 3 on glassy carbon at 20 mV s1. The scan started toward cathodic path and an abrupt increase of cathodic current starting at 0.85 V was observed. This increase means that a reduction process and a concurrent deposition phenomenon were taking place.
4311
Fig. 1. Cyclic voltammetric curve acquired from a solution for an aqueous solution of 0.5 M NiCl26H2O at pH 3 on carbon paper at 20 mV s1. The potentials are referred to Ag/AgCl.
When the scan was reversed, a crossover of the cathodic and anodic currents was recorded. This feature is characteristic of a mechanism of nucleation and three-dimensional growth [25]. From the cyclic voltammetry the chosen potential suitable for Ni deposition was of 0.85 V. Fig. 2 shows the current–time response curve in potentiostatic deposition. During deposition, in the primary stage the current got to the maximum value mainly due to nucleation and growth of nickel. In the second stage, starting in few seconds, the nickel reduction continued to occur but the current decreased because of the passivation phenomenon due to hydrogen evolution. After an adequately long time, the deposition current approached the steady state value when the surface was fully covered by Ni deposit [25,26]. 3.2. SERS characterization Ni electrodeposition was studied by in situ SERS experiments. Raman spectra acquisitions have been carried out in the spectroelectrochemical cell, containing the Ni plating solution. A glassy carbon electrode was used because of its high chemical stability; it
Fig. 2. Cronoamperometric curve acquired applying 0.7 V as potential (vs. Ag/ AgCl) to 0.5 M NiCl26H2O solution on carbon paper.
4312
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
Fig. 3. Time-dependent in situ surface Raman spectra recorded during Ni electrodeposition at 0.85 V vs. Ag/AgCl.
was polarized cathodically at different potentials and timedependent surface Raman spectra were recorded during Ni electrodeposition. For potentials in the electrodeposition range, but less negative than 0.85 V, essentially featureless spectra were recorded in the whole accessible wavenumber range (150–4200 cm1). In particular no noticeable spectral features could be detected either in the d(HOH) or in the n(OH) wavenumber ranges (about 1600 and about 3500 cm1, respectively). At 0.85 V and more cathode potentials, a composite band in the n(OH) range, extending from about 3200 cm1 to about 3600 cm1 showed up after a few minutes of electrodeposition and grew in intensity, attaining an asymptotic shape after about 8 min of electrodeposition (Fig. 3). Raman [27] and FT-IR [27,28] literature on corrosion has assigned a similar kind of band to n(OH) (OH stretching) of Fe(OH)2 or related Fe2+-based polymeric species. No similar experimental work on Ni is known to the authors. Of course, vibrational features corresponding to n(OH) modes of adsorbed water are found in the same
Fig. 5. SEM images of Ni clusters electrodeposited on carbon paper (a), glassy carbon (b) and PAN fibres (c).
Fig. 4. In situ surface Raman spectra recorded in the Ni electrodeposition bath, under the following conditions: (A) during Ni electrodeposition at 0.85 V vs. Ag/ AgCl; (B) same as previous, but in the catholyte about 10 mm above the electrode; (C) and (D) applying +0.80 V vs. Ag/AgCl after Ni electrodeposition obtained by using 0.85 V vs. Ag/AgCl for 10 min. Spectra (C) and (D) have been shifted downwards by 1200 a.u., for clarity of presentation; spectrum (D) has also been expanded by a factor 2.
wavenumber range [29] and it is not possible to distinguish among hydroxides, oxi-hydroxides, basic salts, hydration water and adsorbed water just on the basis of the band position. This point has been further proved on the basis of DTF computations, not reported for brevity, of n(OH) bands corresponding to water and the mono- and polymeric Ni(II) hydroxides listed in [30]. The composite nature of the band is ascribed to different types of hydroxides – or to the corresponding hydration-water modes – coherently with the complex colloidal nature of these species (see, e.g. [30]). In the present case, we reckon that the build-up of the spectral feature shown in Figs. 3 and 4A,B can be related to some form of sparingly soluble or colloidal Ni(II)-containing species, precipitating at the cathode, owing to local alkalinization, in the absence of an interfacial buffer, such as H3BO3 (see, e.g. [31]);
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
4313
Fig. 6. BF TEM images in cross-section of Ni clusters electrodeposited on carbon paper (a), glassy carbon (b) and PAN fibres (c). Their shape resulted quite globular.
deconvolution analysis and detailed spectra assignment is beyond the scope of this paper and would not affect our conclusions. In Fig. 4 Raman spectra acquired during the application of cathodic potentials are reported. Spectrum A of Fig. 4 is referred to a confocal volume centred at the electrode surface, whereas spectrum B in the same Figure was measured by centring the confocal volume within the catholyte, about 10 mm above the electrode. The signal, associated with n(OH) band, is more intense within the catholyte where the analytical volume contains a higher amount of Ni(II)-related precipitates, forming according to reactions (1) and (2). Again in Fig. 4, the portions C and D of the spectrum recorded during the stripping at +0.8 V on Ni layer previously electrodeposited at 0.85 V for 10 min, do not show any well-defined n(OH) band. Instead of the n(OH) band, spectrum D exhibits other modes at 350 (strong) and 1170 cm1 (faint). Nakamoto [32] and Iwasita [28] interpreted similar features as
n(M–O–M) and d(MOOH), respectively. By analogy, we can associate these features to Ni(II)-containing precipitates. We reckon that these pieces of evidence are enough to prove that some forms of Ni(II) basic precipitate accumulate at the cathode during Ni electrodeposition under the specified conditions. This statement is coherent with the integral electrochemical data discussed above. 3.3. SEM and TEM characterization Ni electrodeposition was performed on different carbon substrates: glassy carbon, carbon paper, and PAN fibres. For the three different kinds of substrates, SEM images and conventional Bright Field (BF) images showed that the Ni electrodeposit consists of several clusters (Figs. 5 and 6). In all the three samples these clusters are polycrystals with a grain size of few nm.
4314
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
Fig. 7. HRTEM image of a small area of the interface of PAN fibres (a), glassy carbon (b) and carbon paper (c).
The interface between Ni clusters and carbon substrates was analysed with particular attention. A typical HRTEM image of a small area of the interface of PAN fibres is reported in Fig. 7a, where some reticular fringes in a Ni cluster, well oriented with respect to the electron beam, and some Moire` fringes, due to the overlapping of more grains, are evidenced. Similar features are visible in Fig. 7b for Ni electrodeposited on glassy carbon and in Fig. 7c for carbon paper. Due to the presence of these features, only for glassy carbon and PAN fibres substrates it was possible to prepare suitable samples for more appropriate investigations at interface. In HRTEM images the structure of the interface was not particularly evident, despite the favourable observation geometry. Same results about the interface structure were obtained also in HRTEM images taken after varying the sample tilt, in order to exclude any detrimental effect of a not well perpendicular interface to the electron beam.
For chemical investigation, EDS spectra were collected focusing the electron beam both in the clusters and in some points at interface. Fig. 8a and b are HAADF STEM images of Ni clusters on PAN fibres and glassy carbon respectively, indicating the points for EDS spectra acquisition. By comparison of the ratio between the integrated intensity of the Ok peak and that of Nik peak (after background subtraction) for the clusters and for their interface, oxygen content at interface results systematically higher than at Table 1 Ratios of the intensities of Ok and Nik lines recorded in inner points of Ni clusters and at the interface with carbon supports.
PAN fibres Glassy carbon
Inner points
Interface points
0.025 0.007 0.02 0.01
0.06 0.02 0.4 0.1
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
4315
the first stage of Ni reduction from the unbuffered nickel chloride solution. As the strict bond Ni–C is not favourite, we deduce that the adhesion of the Ni clusters to substrate is due to the presence of these Ni(II)-containing precipitates, that makes the electrodeposited Ni clusters so obtained suitable for application in catalytic growth processes. Acknowledgements The authors want to thank Mr. Martino Palmisano for TEM preparation, Dr. Massimo Catalano and Mr. Maurizio Russo of IMM of Lecce for their technical support in TEM sample preparation with PIPS and D. Wall ad. F. Tatti, Application Specialists for Dual Beam systems FEI, for FIB preparation. The authors are indebted to an anonymous referee for his/her constructive criticism and suggestions concerning the discussion of the Raman spectra and the interpretation of the n(OH) band. References
Fig. 8. HAADF STEM images of Ni clusters electrodeposited on PAN fibres (a) and on glassy carbon (b), indicating the points where EDS spectra were acquired.
inner positions (Table 1). This finding is in agreement with the results of the other analytical studies discussed above. 4. Conclusions Ni clusters were electrodeposited on different carbon-based substrates. The Ni reduction was studied by combining some classical electrochemical methods with in situ SERS. The interface between Ni clusters and substrates was structurally analysed by means of transmission electron microscopy and relative techniques (BF, HAADF STEM, HRTEM, EDS). All findings are concordant to say that the bond between Ni clusters and the carbon substrate is based on some form of Ni(II)-containing precipitates forming in
[1] T. Dikonimos Makris, R. Giorgi, N. Lisi, L. Pilloni, E. Salernitano, M.F. De Riccardis, D. Carbone, Fullerenes Nanotubes Carbon Nanostruct. 13 (Suppl. 1) (2005) 383. [2] M.F. De Riccardis, D. Carbone, Appl. Surf. Sci. 252 (2006) 5403. [3] M.F. De Riccardis, D. Carbone, Th. Dikonimos Makris, R. Giorgi, N. Lisi, E. Salernitano, Carbon 44 (2005) 671. [4] C.M. Reddy, R.S. Gaston, C.M. Weikart, H.K. Yasuda, Prog. Org. Coat. 33 (1998) 225. [5] R. Greef, R. Peat, L.M. Peter, D. Pletcher, J. Robinson, Instrumental Methods in Electrochemistry by Southampton Electrochemistry Group, John Wiley & Sons, New York, 1985. [6] J.O.’M. Bockris, Modern Electrochemistry, vols. 1–2, Plenum Press, New York, 1998. [7] T.B. Massalski, Binary Alloy Phase Diagrams, ASM International, 1990. [8] P. Marcus, C. Hinnen, Surf. Sci. 392 (1997) 134. [9] D.Q. Yang, E. Sacher, J. Phys. Chem. B 109 (2005) 19329. [10] P. Bebin, R.E. Prud’Homme, J. Polym. Sci. B: Polym. Phys. 40 (2002) 82. [11] R.L. Vander Wal, T.M. Ticich, V.E. Curtis, Carbon 39 (2001) 2277. [12] B. Ren, G.-K. Liu, X.-B. Lian, Zh.-L. Yang, Z.-Q. Tian, Anal. Bioanal. Chem. 388 (2007) 29. [13] Q.-J. Huang, X.-Q. Li, J.-L. Yao, B. Ren, W.-B. Cai, J.-S. Gao, B.-W. Mao, Z.-Q. Tian, Surf. Sci. 427–428 (1999) 162. [14] J.L. Yao, J. Tang, D.Y. Wu, D.M. Sun, K.H. Xue, B. Ren, B.W. Mao, Z.Q. Tian, Surf. Sci. 514 (2002) 108. [15] B. Bozzini, L. D’Urzo. J. Electrochem. Soc., submitted for publication. [16] M. Veres, M. Fu¨le, S. To´th, M. Koo´s, I. Po´csik, Diamond Relat. Mater. 13 (2004) 1412. [17] P.G. Cao, R.-N. Gu, B. Ren, Z.-Q. Tian, Chem. Phys. Lett. 366 (2002) 440. [18] Q.J. Huang, J.L. Yao, B.-W. Mao, R.A. Gu, Z.Q. Tian, Chem. Phys. Lett. 271 (1997) 101. [19] Q.-J. Huang, X.-F. Lin, Z.-L. Yang, J.-W. Hu, Z.-Q. Tian, J. Electroanal. Chem. 563 (2004) 121. [20] F. Ding, K. Bolton, Nanotechnology 17 (2006) 543. [21] Y. Wang, Z. Luo, B. Li, P.S. Ho, Z. Yao, L. Shi, E.N. Bryan, R.J. Nemanich, J. Appl. Phys. 101 (2007) 124310. [22] J.O.’M. Bockris, D. Drazic, A.R. Despic, Electrochim. Acta 4 (1961) 325. [23] C.Q. Cui, Jim Y. Lee, J. Electrochem. Soc. 141 (1994) 2030. [24] C.Q. Cui, Jim Y. Lee, Electrochim. Acta 40 (1995) 1653. [25] E. Gomez, C. Muller, W.G. Proud, E. Valles, J. Appl. Electrochem. 22 (1992) 872. [26] R. Orinakova, A. Turonova, D. Kladekova, M. Galova, R.M. Smith, J. Appl. Electrochem. 36 (2006) 957. [27] B. Bozzini, C. Mele, V. Romanello, Mater. Corros. 58 (2007) 362. [28] T. Iwasita, F.C. Nart, in: H. Gerischer, C.W. Tobias (Eds.), Advances in Electrochemical Science and Engineering, vol. 4, VCH, Weinheim (D), 1995. [29] M. Ito, M. Nakamura, Faraday Discuss. 121 (2002) 71. [30] C.F. Baes, R.E. Mesmer, The Hydrolysis of Cations, J. Wiley & Sons, New York, 1976 . [31] P.L. Cavallotti, B. Bozzini, L. Nobili, G. Zangari, Electrochim. Acta 39 (1994) 1123. [32] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, Part B, J. Wiley & Sons Inc., New York, 1997.
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Study on the adhesion mechanism of electrodeposited nickel clusters on carbon substrates M.F. De Riccardis a,*, D. Carbone a, V. Martina a, M. Re a, B. Bozzini b, L. D’Urzo b a b
ENEA Brindisi Research Centre, SS.7 Appia Km 712, 72100 Brindisi, Italy Dipartimento di Ingegneria dell’Innovazione, Universita` del Salento, via Monteroni, I-73100 Lecce, Italy
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 July 2008 Received in revised form 21 October 2008 Accepted 21 October 2008 Available online 21 November 2008
Growth and adhesion mechanisms of Ni clusters electrodeposited on three different carbon-based substrates have been studied. Glassy carbon, carbon paper and PAN-based fibres have been used as working electrodes and Ni clusters have been electrodeposited from a NiCl26H2O. Ni reduction on carbon substrates has been studied by cyclic voltammetry, chronocoulometry and in situ SERS, whereas the morphological and structural characterization of the interface between Ni clusters and carbon-based substrates has been performed by High Resolution TEM. From our results we can conclude that the precipitation of Ni hydroxides and basic salts in the unbuffered catholyte promotes the adhesion of Ni clusters on the carbon-based substrates considered in this study. This feature of the investigated Ni clusters electrodeposition suggests that it may be a suitable fabrication route for applications in catalytic processes, such as metal-particle catalysed growth of carbon nanotubes. ß 2008 Elsevier B.V. All rights reserved.
Keywords: Ni clusters Electrodeposition SERS TEM
1. Introduction Electrodeposition is a versatile and inexpensive technique that allows the synthesis of metals in several morphologies and structures, including continuous metallic films and nanoparticles, also on 3D substrates. Electrodeposition of metallic clusters has a host of cutting edge prospective applications, such as the catalytic growth of carbon nanotubes (CNT) in CVD processes [1–3]. In order to obtain a reliable catalytic surface for CNT, and in general for the growth of carbon nanostructures (CNS), such as nanofibres or nanowalls, a key property is the adherence of the clusters to the substrate. Outstanding adherence is mandatory in catalytic reactions, in order to avoid clusters coalescence during heatassisted processes. Nevertheless, in the growth of metal catalysed carbon nanostructures, strong interactions between support and metal catalyst, together with minimization of surface mobility of catalyst particles, are highly desirable process highlights. Well adhered metallic coatings can be produced by electrodeposition once the plating parameters, the substrate treatments and the electrolytic bath composition are optimized [4–6]. In previous papers the authors demonstrated that it is possible to obtain well adhered and uniformly distributed Ni clusters, on PAN-based fibres and carbon paper [1,2]. The good adhesion of electrodeposited Ni clusters was demonstrated by performing several experiments,
* Corresponding author. E-mail address: [email protected] (M.F. De Riccardis). 0169-4332/$ – see front matter ß 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2008.10.127
consisting in immersion, magnetic stirring, centrifugation and ultrasonication into different liquids (such as deionized water, ethanol, acetone, n-butanol). Nevertheless, nano-scratch tests were performed on Ni coated supports by means of Atomic Force Microscope (AFM) equipped with a diamond tip; the metallic clusters were moulted and not removed by the diamond tip. The experimental details are reported in [2]. It is well known that Carbon and Nickel not form stable carbides [7]; so the reason of this evidenced strong bond between electrodeposited Ni and carbon support is still undefined. The aim of this paper is to provide a further piece of information to understand the grounds of adhesion between Ni clusters and carbon substrates; for this reason with insight we analysed the interface between Ni electrodeposit and carbon support. Our attention has been directed to Ni and C because their jointed use is highly attractive. In fact remarkable catalytic activity of Ni and amphoteric properties of C make them able to be widely used in metal catalysed reactions as catalyst and support material respectively. Nevertheless, the nature of interfacial bond between metal/carbon or metal/polymer, has been extensively exploited in different technologies, e.g. microelectronics, composites, adhesive bonding and packaging. In particular, a lot of research effort has been devoted to the understanding of the interfacial interactions between carbon-based materials, especially HOPG, and metal deposits [8–11]. In the specific issue of metal-catalysed carbon nanostructure growth, the catalytic activity of metal particles can be more or less dramatically affected by both physical and chemical interactions
4310
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
between substrate and catalyst [11]. For this reason, we are interested in characterising the nature of the bond between Ni clusters and substrate [1–3]. To this aim, we studied the Ni electrodeposition process by combining some classical electrochemical methods – such as cyclic voltammetry and chronocoulometry – with in situ SERS. This technique is typically carried out with precious-metal electrodes, exhibiting a remarkable degree of surface-enhancement. Nevertheless, it has been proved relatively recently (see e.g. [12] for an updated review) that the SERS effect – though at a significantly lower level with respect to precious metals – can be achieved also with Fe- and Pt-group metals. The use of SERS in electrodeposition studies in such transition metals has received very limited attention. Ni electrodeposition has been used in one case to achieve Ni electrode roughening prior to SERS work [13]. Electrodeposited Co, Ni, Pt and Pd nano-wire arrays have been shown to exhibit surface enhancement [14]. Co electrodeposition in the presence of coumarin has also been investigated [15]. In addition to electrochemical analyses on Ni electrodeposition process, structural investigations at the interface between carbon supports (carbon paper, carbon-based fibres and glassy carbon) and the electrodeposited Ni clusters were performed by means of transmission electron microscopy (TEM).
Raman spectra were recorded using a LabRam microprobe confocal system. A 50 long-working-distance objective was used and the excitation line at 632.8 nm was provided by a 12-mW He– Ne laser. The slit and pinhole were set at 200 and 400 mm, respectively, corresponding to a scattering volume of about 3 pL; Raman spectra were acquired with a 600 grid/mm spectrometer. The recorded Raman intensities are directly proportional to the discharge current of the CCD detector. In situ electrochemical measurements were performed in a glass cell with a vertical glassy carbon (GC) electrode (1 3 mm) embedded in a Teflon holder. A limited degree of SERS activity can be attained with GC [16] and Ni [17–19] that can be enhanced during the electrodeposition process, owing to the formation of active Ni clusters. The counter electrode was a Pt wire loop of total area ca. 2 cm2, concentric and coplanar with the gold working electrode. An external Ag/AgCl reference electrode was used and all the potentials were reported on this scale. The applied potential was controlled by an AMEL 2049 potentiostat.
2. Experimental
2.3. TEM characterization
2.1. Ni electrodeposition
Transmission Electron Microscopy (TEM) was carried out by TECNAI G2 F30 operating at an acceleration voltage of 300 kV and with point resolution of 0.205 nm. For TEM observations of electrodeposited Ni clusters, the samples were prepared with three methods, according to the characteristics and the nature of different substrates. With carbon paper, a powder sample surface was obtained by scraping it off the cathode surface and dispersing it in ethylic alcohol; the dispersion was cast on a carbon coated grid. In the case of glassy carbon substrate, two pieces of the same specimen were sandwiched and mechanically thinned down to about 20 mm of thickness, according to the standard preparation method of TEM crosssections. The final thinning was performed by ion beam milling (PIPS Gatan) with Ar ions at 5 kV. For PAN Carbon fibres, a Pt thin layer was deposited by electron beam in order to protect the electrodeposited Ni, during the next Ion beam deposition of a thicker Pt layer necessary for TEM lamella preparation. Then a longitudinal cross-section was prepared with the in situ Lift Out technique with the FEI Strata 400 FIB/SEM system. For all the samples, conventional Bright Field TEM images were obtained from many areas, in order to highlight global morphological features of the electrodeposited Ni clusters, such as size, distribution and shape, while High Resolution TEM images yielded insightful structural information regarding the microstructure of the clusters and the nature of the interface. Nevertheless, High Angle Annular Dark Field (HAADF) STEM images were taken from cross-sections of Ni on glassy carbon and PAN fibres; this kind of analysis has highlighted chemical differences on large areas, thanks to compositional and thickness contrast mechanism. For chemical analysis, Energy Dispersive Spectrometry (EDS) was also carried out with the electron beam focused both on Ni clusters and on their interface with the substrate.
For Ni electrodeposition, a solution of 0.5 M NiCl26H2O with deionized water was prepared and pH was adjusted to 3.0 by HCl addition. Ni clusters were electrodeposited on different carbon substrates, such as PAN carbon-based fibres (obtained by pyrolysis of Polyacrylonitrile), carbon paper, commercially named Papyex, and glassy carbon. Two different set-up of electrolytic cell were used, according to the nature of the substrates: (i) a cylindrical cell in the case of fibre bundles and (ii) a prismatic cell with carbon paper and glassy carbon. The cylindrical cell was equipped with a Pt spiral and a cylindrical assembly of PAN-based fibres as counter-electrode, in order to achieve a high degree of electrodeposit uniformity along both the fibre length and diameter. Uniform deposits on planar substrates were obtained with a plane-parallel electrode configuration in the prismatic cell, featuring a vertical cathodic specimen holder and Pt foil as anode. Voltammetric and potentiostatic measurements were carried out by using a PAR potentiostat model 273A. The potentials were measured against a Ag/AgCl reference electrode and reported on this scale. Different cleaning treatments were applied to the carbon substrates before Ni electrodeposition. PAN fibres, cut from a woven fabric, were treated at 650 8C for 1 h in air to remove the binder. Glassy carbon substrates were treated by boiling the coupons in a solution of HCl and HNO3 (3:1) for 10 min and subsequently rinsing them in deionized water, to remove any organic residue. Carbon paper substrates were washed with acetone and ethylic alcohol, and rinsed in deionized water. The exposed area was 21 cm2 for PAN-based fibres and 0.95 cm2 for carbon paper and glassy carbon electrodes. Experimental details on the electrodeposition process are described elsewhere [2]. After Ni electrodeposition, morphological characterizations of Ni coated carbon substrates surface were performed by using XL40 LaB6 (Philips) SEM. A monomodal narrow diameter distribution was found for the Ni clusters, around 50 nm, under different conditions for each substrate; in the case of glassy carbon at 0.85 V and 20 mC/cm2 total charge, in the case of carbon paper at 0.8 V and 100 mC/cm2
total charge, and in the case of PAN-based fibres at 0.75 V and 7.5 mC/cm2 total charge. 2.2. SERS characterization
3. Results and discussion 3.1. Electrochemistry The more appropriate term for the electrodeposition technique should be electrocrystallization, usually used for electrode pro-
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
cesses which involve formation of a solid state as a result of the reduction of ions in solution, as in the case of metal deposition [4,5]. Electrocrystallization consists of several steps: (i) diffusion of ions in solution towards the electrode surface, (ii) formation of adsorbed atoms, (iii) clustering of ad-atoms to form critical nuclei, (iv) incorporation of adsorbed ions at lattice site and (v) development of crystallographic and morphological characteristics of the deposit. The last step is the most interesting, in particular when the electrodeposited material is used as carbon nanostructures (CNS) catalyst; in fact it was verified that CNS characteristics depend in part on microstructural characteristics of their catalysts (see e.g. [20,21]). Several studies of electrochemical reduction and deposition of Ni based on Bockris’ seminal paper [22] have well-demonstrated that the nickel monohydroxide ion is the intermediate species when a charge transfer mechanism occurs in aqueous unbuffered solutions of simple nickel ions [23,25]. Indeed, the diffusion of ions in solution towards the electrode surface involves Ni(OH)+ species (reaction (1)). In succession, an one-electron reduction reaction takes place forming the adsorbed intermediate, which is believed to be the active intermediate during the metal formation (2): Ni2þ þ H2 O ¼ NiðOHÞþ þ Hþ
(1)
NiðOHÞþ þ e ! NiðOHÞads
(2)
Ni(OH)ads, is reduced to nickel metal via two parallel reactions; the first reaction uses the adsorbed nickel monohydroxide as catalyst, reducing nickel ions to nickel metal (3), and the second reaction allows direct metal reduction from Ni(OH)ads involving one-electron transfer reaction (4): NiðOHÞads þ Ni2þ þ 2e ! Ni þ NiðOHÞads
(3)
NiðOHÞads þ e ! Ni þ OH
(4) 2+
Cui and Lee [23,24] demonstrated that at low Ni concentration in unbuffered aqueous solution, Ni deposition is inhibited by presence of oxygen. Their voltammetric studies demonstrated that at cathode potentials, oxygen reduction causes the increase in the interfacial pH and the precipitation of Ni(OH)2 on the electrode. The hydroxide precipitation takes place at less negative potential than that of Ni reduction, so that the adsorbed intermediate Ni(OH)ads formation is lessened and subsequently nickel nucleation is inhibited. Same authors studied hydroxide formation also in absence of oxygen at low Ni2+ concentration in unbuffered aqueous solution [23,24]. In this case, at high cathode potentials, quite similar to Ni reduction, water discharge causes hydrogen evolution; so pH increases and Ni(OH)2 precipitates on the substrate. When the scan is switched to anodic potentials, the intensity of nickel oxidation peak is much smaller than nickel deposition, irrespective of the presence or absence of oxygen. This phenomenon confirms the decreased Ni formation due to the Ni(OH)2 presence. At high electrolyte concentration, both in presence and in absence of oxygen, Ni deposition is not inhibited by Ni(OH)2. In fact, the adsorbtion of intermediate complex Ni(OH)ads is higher than at low Ni2+ concentration and Ni reduction is favoured [23]. Fig. 1 shows a cyclic voltammetric curve acquired for an aqueous solution of 0.5 M NiCl26H2O at pH 3 on glassy carbon at 20 mV s1. The scan started toward cathodic path and an abrupt increase of cathodic current starting at 0.85 V was observed. This increase means that a reduction process and a concurrent deposition phenomenon were taking place.
4311
Fig. 1. Cyclic voltammetric curve acquired from a solution for an aqueous solution of 0.5 M NiCl26H2O at pH 3 on carbon paper at 20 mV s1. The potentials are referred to Ag/AgCl.
When the scan was reversed, a crossover of the cathodic and anodic currents was recorded. This feature is characteristic of a mechanism of nucleation and three-dimensional growth [25]. From the cyclic voltammetry the chosen potential suitable for Ni deposition was of 0.85 V. Fig. 2 shows the current–time response curve in potentiostatic deposition. During deposition, in the primary stage the current got to the maximum value mainly due to nucleation and growth of nickel. In the second stage, starting in few seconds, the nickel reduction continued to occur but the current decreased because of the passivation phenomenon due to hydrogen evolution. After an adequately long time, the deposition current approached the steady state value when the surface was fully covered by Ni deposit [25,26]. 3.2. SERS characterization Ni electrodeposition was studied by in situ SERS experiments. Raman spectra acquisitions have been carried out in the spectroelectrochemical cell, containing the Ni plating solution. A glassy carbon electrode was used because of its high chemical stability; it
Fig. 2. Cronoamperometric curve acquired applying 0.7 V as potential (vs. Ag/ AgCl) to 0.5 M NiCl26H2O solution on carbon paper.
4312
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
Fig. 3. Time-dependent in situ surface Raman spectra recorded during Ni electrodeposition at 0.85 V vs. Ag/AgCl.
was polarized cathodically at different potentials and timedependent surface Raman spectra were recorded during Ni electrodeposition. For potentials in the electrodeposition range, but less negative than 0.85 V, essentially featureless spectra were recorded in the whole accessible wavenumber range (150–4200 cm1). In particular no noticeable spectral features could be detected either in the d(HOH) or in the n(OH) wavenumber ranges (about 1600 and about 3500 cm1, respectively). At 0.85 V and more cathode potentials, a composite band in the n(OH) range, extending from about 3200 cm1 to about 3600 cm1 showed up after a few minutes of electrodeposition and grew in intensity, attaining an asymptotic shape after about 8 min of electrodeposition (Fig. 3). Raman [27] and FT-IR [27,28] literature on corrosion has assigned a similar kind of band to n(OH) (OH stretching) of Fe(OH)2 or related Fe2+-based polymeric species. No similar experimental work on Ni is known to the authors. Of course, vibrational features corresponding to n(OH) modes of adsorbed water are found in the same
Fig. 5. SEM images of Ni clusters electrodeposited on carbon paper (a), glassy carbon (b) and PAN fibres (c).
Fig. 4. In situ surface Raman spectra recorded in the Ni electrodeposition bath, under the following conditions: (A) during Ni electrodeposition at 0.85 V vs. Ag/ AgCl; (B) same as previous, but in the catholyte about 10 mm above the electrode; (C) and (D) applying +0.80 V vs. Ag/AgCl after Ni electrodeposition obtained by using 0.85 V vs. Ag/AgCl for 10 min. Spectra (C) and (D) have been shifted downwards by 1200 a.u., for clarity of presentation; spectrum (D) has also been expanded by a factor 2.
wavenumber range [29] and it is not possible to distinguish among hydroxides, oxi-hydroxides, basic salts, hydration water and adsorbed water just on the basis of the band position. This point has been further proved on the basis of DTF computations, not reported for brevity, of n(OH) bands corresponding to water and the mono- and polymeric Ni(II) hydroxides listed in [30]. The composite nature of the band is ascribed to different types of hydroxides – or to the corresponding hydration-water modes – coherently with the complex colloidal nature of these species (see, e.g. [30]). In the present case, we reckon that the build-up of the spectral feature shown in Figs. 3 and 4A,B can be related to some form of sparingly soluble or colloidal Ni(II)-containing species, precipitating at the cathode, owing to local alkalinization, in the absence of an interfacial buffer, such as H3BO3 (see, e.g. [31]);
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
4313
Fig. 6. BF TEM images in cross-section of Ni clusters electrodeposited on carbon paper (a), glassy carbon (b) and PAN fibres (c). Their shape resulted quite globular.
deconvolution analysis and detailed spectra assignment is beyond the scope of this paper and would not affect our conclusions. In Fig. 4 Raman spectra acquired during the application of cathodic potentials are reported. Spectrum A of Fig. 4 is referred to a confocal volume centred at the electrode surface, whereas spectrum B in the same Figure was measured by centring the confocal volume within the catholyte, about 10 mm above the electrode. The signal, associated with n(OH) band, is more intense within the catholyte where the analytical volume contains a higher amount of Ni(II)-related precipitates, forming according to reactions (1) and (2). Again in Fig. 4, the portions C and D of the spectrum recorded during the stripping at +0.8 V on Ni layer previously electrodeposited at 0.85 V for 10 min, do not show any well-defined n(OH) band. Instead of the n(OH) band, spectrum D exhibits other modes at 350 (strong) and 1170 cm1 (faint). Nakamoto [32] and Iwasita [28] interpreted similar features as
n(M–O–M) and d(MOOH), respectively. By analogy, we can associate these features to Ni(II)-containing precipitates. We reckon that these pieces of evidence are enough to prove that some forms of Ni(II) basic precipitate accumulate at the cathode during Ni electrodeposition under the specified conditions. This statement is coherent with the integral electrochemical data discussed above. 3.3. SEM and TEM characterization Ni electrodeposition was performed on different carbon substrates: glassy carbon, carbon paper, and PAN fibres. For the three different kinds of substrates, SEM images and conventional Bright Field (BF) images showed that the Ni electrodeposit consists of several clusters (Figs. 5 and 6). In all the three samples these clusters are polycrystals with a grain size of few nm.
4314
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
Fig. 7. HRTEM image of a small area of the interface of PAN fibres (a), glassy carbon (b) and carbon paper (c).
The interface between Ni clusters and carbon substrates was analysed with particular attention. A typical HRTEM image of a small area of the interface of PAN fibres is reported in Fig. 7a, where some reticular fringes in a Ni cluster, well oriented with respect to the electron beam, and some Moire` fringes, due to the overlapping of more grains, are evidenced. Similar features are visible in Fig. 7b for Ni electrodeposited on glassy carbon and in Fig. 7c for carbon paper. Due to the presence of these features, only for glassy carbon and PAN fibres substrates it was possible to prepare suitable samples for more appropriate investigations at interface. In HRTEM images the structure of the interface was not particularly evident, despite the favourable observation geometry. Same results about the interface structure were obtained also in HRTEM images taken after varying the sample tilt, in order to exclude any detrimental effect of a not well perpendicular interface to the electron beam.
For chemical investigation, EDS spectra were collected focusing the electron beam both in the clusters and in some points at interface. Fig. 8a and b are HAADF STEM images of Ni clusters on PAN fibres and glassy carbon respectively, indicating the points for EDS spectra acquisition. By comparison of the ratio between the integrated intensity of the Ok peak and that of Nik peak (after background subtraction) for the clusters and for their interface, oxygen content at interface results systematically higher than at Table 1 Ratios of the intensities of Ok and Nik lines recorded in inner points of Ni clusters and at the interface with carbon supports.
PAN fibres Glassy carbon
Inner points
Interface points
0.025 0.007 0.02 0.01
0.06 0.02 0.4 0.1
M.F. De Riccardis et al. / Applied Surface Science 255 (2009) 4309–4315
4315
the first stage of Ni reduction from the unbuffered nickel chloride solution. As the strict bond Ni–C is not favourite, we deduce that the adhesion of the Ni clusters to substrate is due to the presence of these Ni(II)-containing precipitates, that makes the electrodeposited Ni clusters so obtained suitable for application in catalytic growth processes. Acknowledgements The authors want to thank Mr. Martino Palmisano for TEM preparation, Dr. Massimo Catalano and Mr. Maurizio Russo of IMM of Lecce for their technical support in TEM sample preparation with PIPS and D. Wall ad. F. Tatti, Application Specialists for Dual Beam systems FEI, for FIB preparation. The authors are indebted to an anonymous referee for his/her constructive criticism and suggestions concerning the discussion of the Raman spectra and the interpretation of the n(OH) band. References
Fig. 8. HAADF STEM images of Ni clusters electrodeposited on PAN fibres (a) and on glassy carbon (b), indicating the points where EDS spectra were acquired.
inner positions (Table 1). This finding is in agreement with the results of the other analytical studies discussed above. 4. Conclusions Ni clusters were electrodeposited on different carbon-based substrates. The Ni reduction was studied by combining some classical electrochemical methods with in situ SERS. The interface between Ni clusters and substrates was structurally analysed by means of transmission electron microscopy and relative techniques (BF, HAADF STEM, HRTEM, EDS). All findings are concordant to say that the bond between Ni clusters and the carbon substrate is based on some form of Ni(II)-containing precipitates forming in
[1] T. Dikonimos Makris, R. Giorgi, N. Lisi, L. Pilloni, E. Salernitano, M.F. De Riccardis, D. Carbone, Fullerenes Nanotubes Carbon Nanostruct. 13 (Suppl. 1) (2005) 383. [2] M.F. De Riccardis, D. Carbone, Appl. Surf. Sci. 252 (2006) 5403. [3] M.F. De Riccardis, D. Carbone, Th. Dikonimos Makris, R. Giorgi, N. Lisi, E. Salernitano, Carbon 44 (2005) 671. [4] C.M. Reddy, R.S. Gaston, C.M. Weikart, H.K. Yasuda, Prog. Org. Coat. 33 (1998) 225. [5] R. Greef, R. Peat, L.M. Peter, D. Pletcher, J. Robinson, Instrumental Methods in Electrochemistry by Southampton Electrochemistry Group, John Wiley & Sons, New York, 1985. [6] J.O.’M. Bockris, Modern Electrochemistry, vols. 1–2, Plenum Press, New York, 1998. [7] T.B. Massalski, Binary Alloy Phase Diagrams, ASM International, 1990. [8] P. Marcus, C. Hinnen, Surf. Sci. 392 (1997) 134. [9] D.Q. Yang, E. Sacher, J. Phys. Chem. B 109 (2005) 19329. [10] P. Bebin, R.E. Prud’Homme, J. Polym. Sci. B: Polym. Phys. 40 (2002) 82. [11] R.L. Vander Wal, T.M. Ticich, V.E. Curtis, Carbon 39 (2001) 2277. [12] B. Ren, G.-K. Liu, X.-B. Lian, Zh.-L. Yang, Z.-Q. Tian, Anal. Bioanal. Chem. 388 (2007) 29. [13] Q.-J. Huang, X.-Q. Li, J.-L. Yao, B. Ren, W.-B. Cai, J.-S. Gao, B.-W. Mao, Z.-Q. Tian, Surf. Sci. 427–428 (1999) 162. [14] J.L. Yao, J. Tang, D.Y. Wu, D.M. Sun, K.H. Xue, B. Ren, B.W. Mao, Z.Q. Tian, Surf. Sci. 514 (2002) 108. [15] B. Bozzini, L. D’Urzo. J. Electrochem. Soc., submitted for publication. [16] M. Veres, M. Fu¨le, S. To´th, M. Koo´s, I. Po´csik, Diamond Relat. Mater. 13 (2004) 1412. [17] P.G. Cao, R.-N. Gu, B. Ren, Z.-Q. Tian, Chem. Phys. Lett. 366 (2002) 440. [18] Q.J. Huang, J.L. Yao, B.-W. Mao, R.A. Gu, Z.Q. Tian, Chem. Phys. Lett. 271 (1997) 101. [19] Q.-J. Huang, X.-F. Lin, Z.-L. Yang, J.-W. Hu, Z.-Q. Tian, J. Electroanal. Chem. 563 (2004) 121. [20] F. Ding, K. Bolton, Nanotechnology 17 (2006) 543. [21] Y. Wang, Z. Luo, B. Li, P.S. Ho, Z. Yao, L. Shi, E.N. Bryan, R.J. Nemanich, J. Appl. Phys. 101 (2007) 124310. [22] J.O.’M. Bockris, D. Drazic, A.R. Despic, Electrochim. Acta 4 (1961) 325. [23] C.Q. Cui, Jim Y. Lee, J. Electrochem. Soc. 141 (1994) 2030. [24] C.Q. Cui, Jim Y. Lee, Electrochim. Acta 40 (1995) 1653. [25] E. Gomez, C. Muller, W.G. Proud, E. Valles, J. Appl. Electrochem. 22 (1992) 872. [26] R. Orinakova, A. Turonova, D. Kladekova, M. Galova, R.M. Smith, J. Appl. Electrochem. 36 (2006) 957. [27] B. Bozzini, C. Mele, V. Romanello, Mater. Corros. 58 (2007) 362. [28] T. Iwasita, F.C. Nart, in: H. Gerischer, C.W. Tobias (Eds.), Advances in Electrochemical Science and Engineering, vol. 4, VCH, Weinheim (D), 1995. [29] M. Ito, M. Nakamura, Faraday Discuss. 121 (2002) 71. [30] C.F. Baes, R.E. Mesmer, The Hydrolysis of Cations, J. Wiley & Sons, New York, 1976 . [31] P.L. Cavallotti, B. Bozzini, L. Nobili, G. Zangari, Electrochim. Acta 39 (1994) 1123. [32] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, Part B, J. Wiley & Sons Inc., New York, 1997.