Electrochemical and related processes at surface conductive diamond–solution interfaces

phys. stat. sol. (a) 199, No. 1, 49–55 (2003) / DOI 10.1002/pssa.200303817

© 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Electrochemical and related processes at surface conductive diamond–solution interfaces

Arnab Chatterjee1, Richard G. Compton1, John S. Foord*, 1, Mineo Hiramatsu2, and Frank Marken3

1 University of Oxford, Physical and Theoretical Chemistry Laboratory, South Parks Road, Oxford OX1 3QZ, United Kingdom

2 Department of Electrical and Electronic Engineering, Meijo University, Shiogamagachi Tempaku, Nagoya 468-8502, Japan

3 Department of Chemistry, Loughborough University, Loughborough, Leicestershire LE11 3TU, United Kingdom

Received 20 March 2003, accepted 31 July 2003 Published online 15 September 2003

PACS 68.08.De, 73.25.+i, 82.45.Jm, 82.45.Vp

The electrochemical activity of surface conductive diamond in aqueous media has been investigated for two redox systems, the reduction of Ru (NH3)6

3+ to Ru (NH3)62+ and the reduction of Ag+ to Ag. Although

electrochemical activity is observed, the responses are not as well defined as at conventional boron-doped electrodes, and in particular, poor long term stability is exhibited. This is attributed to a degradation of the electronic properties of the surface conductive phase as a result of the electrochemical reactions them-selves, which tend to disrupt chemically the surface hydrogen layer required for surface conductivity. Electroless deposition of Ag is also observed, as a result of the unusual electronic and chemical properties of surface conducting diamond.


1 Introduction There is great interest in the application of diamond electrodes in electrochemistry since their properties make it a preferred choice in a range of applications [1–3]. The chemical inertness and strength of the material makes it suitable for harsh environments, whilst it displays high current efficiencies for the total oxidation of organic pollutants in aqueous media. Low background currents confer high sensitivity for electrochemical sensors, and a very wide electrochemical potential window allows access to potentials, which are impossible using many other materials. Most diamond electrochemistry has been performed using highly boron-doped diamond (BDD) mate-rial, although nanocrystalline diamond has also been used [4, 5]. Surface conductive diamond (SCD) is another form with potential applications in electrochemical sensors, since it would then allow direct integration with other high quality diamond device forms, which tend to be based on this material [6–8]. The electrochemical properties of surface conductive diamond are largely uncharacterised, although Ramesham [9] has noted the increased electrochemical stability, which hydrogenated diamond layers can bring about when deposited on other substrates. In the present work, we begin the more detailed investi-gation of electrochemical processes at surface conductive diamond interfaces. 2 Experimental Chemicals were purchased from Aldrich in the purest form available or analytical grade and used without further purification. Aqueous solutions were prepared in deionised and filtered water with a resistivity of at least 18 M Ωcm. All solutions were de-aerated with argon for at least 10 min prior to experiments. Electrochemical measurements were carried out with a conventional three-

* Corresponding author: e-mail: [email protected], Tel.: + 44 (0) 1865 275967

50 A. Chatterjee et al.: Electrochemical and related processes


electrode arrangement controlled by a potentiostat (µ-Autolab, Eco Chemie, The Netherlands), with a diamond working electrode, a platinum coil counter electrode and a standard calomel reference electrode (SCE). Experiments were carried out on “optical grade” white diamond thin films grown to a thickness of 100 microns by micro-wave CVD, with a random grain orientation and size of 5–20 microns. Hydro-genation, to produce the surface conductive properties, was achieved in a microwave hydrogen plasma (1.5 kW power, 600 °C platen temperature, 45 mbar, 200 sccm), after which the electrodes were exposed to the laboratory atmosphere for around 24 hrs prior to use. Electrical contacts were made via evaporated gold films. The back face of the CVD wafer and the front contact were protected from the electrolyte solution by insulating lacquer or physical separation, so that only an active SCD electrode area of ap-proximately 5 mm × 5 mm was exposed to the electrolyte solution. No measurable electrochemical re-sponses were observed unless surface conductivity was produced by surface hydrogenation, suggesting that no spurious electrical signals were present in the mounting arrangements used. 3 Results and discussion 3.1 Electrochemical reduction of Ru (NH3)6

3+ (aq) The electrochemical reduction of Ru (NH3)63+

involves a simple one electron transfer process

Ru (NH3)63+ (aq) + e–

→ Ru (NH3)6

2+ (aq)

and this (and the reverse process) is a common system to demonstrate the occurrence of electrochemical activity at diamond electrodes. Typical results are shown in Fig. 1 for 10 mM Ru complex in 0.1M KCl at scan rates of 50 mV s–1. Results shown in Fig. 1a for a “fresh” SCD electrode are compared to those observed using a conventional boron-doped diamond in Fig. 1b. The latter displays good voltammetric responses with peak separation and positions corresponding to rapid electron transfer kinetics and peak current densities anticipated for an electrode area roughly matching the geometric area. The SCD elec-trode also shows corresponding peaks which track the concentration of the Ru compound in solution, confirming that the SCD electrode is active electrochemically. Furthermore the relative peak currents of the SCD and BDD electrodes correspond roughly to the differing geometric areas of the electrodes used, again indicative of the fact that most of the electrode area is involved in both cases. However the electro-chemical characteristics show an appreciable current background, with ohmic characteristics, and the peak separations are significantly larger, around 200 mV for SCD compared to 70 mV for BDD elec-trodes, suggesting that the electrode kinetics may be significantly slower.

Fig. 1 Cyclic voltammograms for 0.01M Ru(NH)6Cl3 in 1M KCl at 0.05 V s–1 scan rate, a) new SCD electrode, b) BDD electrode, c) used SCD electrode, d) more heavily used SCD electrode.

phys. stat. sol. (a) 199, No. 1 (2003) / www.physica-status-solidi.com 51


The main difference between the SCD and BDD electrodes was found in reproducibility and stability. Whereas the latter display an extremely high degree of reproducibility and stability, this was not true of the former, which showed quite a variation in performance for apparently similar hydogenation condi-tions. It is well known that the precise electrical properties of surface conductive diamond can be quite variable, and it is apparent that this can significantly influence the electrochemistry observed. Secondly the electrochemistry observed showed a significance dependence on past history and electrode usage, with typical data being shown in Fig. 1c–d, and after degradation in performance, this could only be restored if the electrode was removed from solution and rehydrogenated. Studies reported elsewhere which utilise SCD diamond as the gate electrode in an ion sensitive field effect transistor structure (ISFET) observe stable operation provided no current is passed through the electrode-solution interface as a result of operation of the ISFET outside the “potential window” pertaining to the electrolyte used [10, 11]. It thus would appear that it is the electrochemical process, rather than immersion in the electro-lyte per se, which is responsible for degradation. It is known that most diamond electrodes become some what oxidised during use, so it is probably this oxidation process which results in the disruption of the

Fig. 2 a) Wide scan XPS spectra of elec-trode before and after electrochemical depositon of Ag, b) XPS spectra of C 1s region before and after heavy electroless deposition of Ag from 1 M AgNO3, c) High resolution XPS spectrum of the C 1s region after heavy electroless deposition of Ag.Peak fits show the elemental diamond peak at 285 eV, and chemically shifted compo-nents at +1.3 eV (hydoxyl, ether functional-ities) and +4.2 eV (carboxylic acid type functionalities).



surface conductive properties as a result of changes in the electron affinity levels of the semiconducting diamond electrode. 3.2 Reduction of Ag+ (aq) The reduction of Ag+ to form metallic Ag has been studied in detail on BDD diamond electrodes, to investigate redox and adsorption properties, so this represents another use-ful system in order to make a comparison between BDD and SCD electrodes [12]. Again reduction and oxidation currents could be observed at SCD electrodes in dilute Ag+ solutions, which responded to changes in concentration, suggesting that reductive deposition to form Ag deposits, and their subsequent electrochemical stripping was occurring. Secondary electron microscopy (SEM) data revealed the pres-ence of small particulates on the diamond surface after immersion of the SCD electrode at cathodic po-tentials, as is typically observed in similar experiments with BDD, and XPS analysis shown in Fig. 2 indicated the presence of silver on the diamond surface at concentrations of approximately 5 at%. It is clear therefore that Ag+ can be electrochemically reduced to adsorbed Ag at SCD surfaces. Cyclic voltammetric data is presented in Fig. 3 to compare BDD and SCD electrodes. The data using BDD is similar to that reported previously [12], with the onset of cathodic reduction to adsorbed silver occurring at potentials around 0.3 V vs. SCE as the electrode is scanned negatively from high positive potentials, and the reverse process, the electrochemical stripping of Ag, being observed around 0.5 V vs. SCE as the electrode is scanned anodically. On SCD, the voltammetric data is much less well-defined and changes significantly during subsequent scans. Again the picture is one of an electrochemically ac-tive material, but one which is relatively unstable. Although only electrochemical deposition of silver was observed in dilute Ag+ solutions, the results changed dramatically in more concentrated solution (1 M AgNO3) where simple immersion of an SCD electrode caused the gradual plating out of Ag deposits on the electrode surface, even if it was electri-cally isolated. SEM micrographs of the deposits produced in this way are illustrated in Fig. 4. On areas of the diamond surface which appeared extremely smooth in SEM analysis, only small particulates were observed especially near grain boundaries, similar to that seen in electrochemical deposition. However in

Fig. 3 a) Cyclic voltammogram from 0.1 M AgNO3 onto SCD, showing first three scans, b) Cyclic Voltammogram from 0.1 M AgNO3 onto BDD. Scan rate = 50 mV s–1.



Fig. 4 SEM micrographs of a) clean SCD surface, b) high resolution picture of clean SCD diamond fac-ets, c) filamentous silver deposits, d) nucleation and growth of plate-like silver deposits, e) high resolution picture in area of high silver coverage, f) high resolution picture of the growing silver plates.

rougher areas, silver with a filamentous appearance developed, which at varying resolutions took on a fractal-like structure. This is suggestive of a mechanism often seen during electrodeposition when diffu-sion-limited irreversible attachment of growth species occurs at sites defined by the growing film [13]. The so-called box counting method [14] was used to calculate the fractal dimension, which characterises this chaotic form of growth, as illustrated in Fig. 5, where an array of 2 dimensional boxes is used to cover the object of interest, and the number of boxes which cover part of it is evaluated as a function of the size of the box. Using this method, the fractal dimension was calculated at 1.58. Finally after prolonged exposure to the solution for 3–4 hours, a dense array of silver platelets devel-oped in regions where the highest amounts of deposition occurred. Subsequent XPS analysis of the C 1s region (Fig. 2b) showed at this point that the C 1s signal had virtually completely disappeared from the



a) b)

Fig. 5 Examples of the grids used for the box-counting method for fractal analysis.

spectrum, indicating that most of the surface had become covered in silver, in contrast to the situation seen in Fig. 2a following electrochemical deposition of Ag nuclei, where very little of the surface is covered with silver. High resolution scans of the C 1s region after the electroless deposition indicated significant intensity shifted by 1.5–4.5 eV to higher binding energies of the C 1s diamond peak (Fig. 2c), indicative of a high degree of oxidation, in the small percentage of the diamond surface where electroless deposition had not occurred. Electroless deposition was not observed at BDD electrodes, so it appears a feature of the SCD mate-rial. As noted elsewhere [6, 7], SCD material is unusual in two respects. Firstly the material normally contains a high degree of hydrogen, for example along grain boundaries, since the route to produce the surface conductive state is to expose a CVD diamond film to a hydrogen plasma. Secondly the material is unusual in that it has a negative electron affinity, so that the valence band maximum is sufficiently close to the vacuum level that it can reduce adsorbates in immediate vicinity of the surface, such as H+ species. It is therefore unsurprising that Ag+ can be reduced to Ag silver since the standard redox poten-tial of this couple (+0.80 V) is significantly more positive than that for H+/H2. The electroless reduction of chemical species at the surface of SCD is not normally a self-sustaining process however, since the bulk-surface electron transfer process results in upward band bending which essentially lowers the va-lence band with respect to the acceptor level of the surface species. This does not seem to be occurring so rapidly in the present case since substantial amounts of Ag become deposited at the surface, implying that some other charge transfer process must be occurring to neutralise the positive charge which other-wise would develop in the diamond. In view of the presence of hydrogen in the diamond phase, the most likely explanation is that this becomes oxidised to H+ which diffuses into the aqueous phase, so that the diamond is in essence acting as a catalyst by shuttling electrons from the hydrogen to the Ag+ species near the diamond surface. It should be noted that previous electrochemical studies have also suggested a role for sub-surface hydrogen in controlling the observed electrochemical properties of CVD diamond [15]. 4 Concluding summary The work described above has examined the electrochemical properties of surface conductive diamond (SCD) material. Electrochemical activity is detected in two facile electro-chemical systems in aqueous media, the reduction of Ru(NH3)6

3+ and the reduction of Ag+. However the cyclic voltammetric data recorded is not as well-defined as at conventional boron-doped electrodes, and in particular exhibit poor long term stability. Changes in surface conductive properties do not occur if the SCD surface is simply immersed in aqueous media, so it appears that it is the actual electrochemical process itself which brings about degradation in electrical properties. Electroless deposition of Ag is detected at the surface of SCD material, as a result of the unique electronic and chemical properties of SCD diamond.



References

[1] M. C. Granger, J. S. Xu, J. W. Strojek, and G. M. Swain, Anal. Chim. Acta 397, 145 (1999). [2] J. C. Angus, H. B. Martin, U. Landau, Y. E. Evstefeeva, B. Miller, and N. Vinokur, New Diam. Front. Carbon

Technol. 9, 175 (1999). [3] Yu. V. Pleskov, in: Advances in Electrochemical Science and Engineering, edited by R. Alkire (Wiley,

New York, 2003). [4] Q. Chen, D. M. Gruen, A. R. Krauss, T. D. Corrigan, M. Witek, and G. M. Swain, J. Electrochem. Soc. 148,

E44 (2001). [5] L. C. Hian, K. J. Grehan, R. G. Compton, J. S. Foord, and F. Marken, J. Electrochem. Soc. 150, E59 (2003). [6] F. Maier, M. Riedel, B. Mantel, J. Ristein, and L. Ley, Phys. Rev. Lett. 85, 3472 (2000). [7] J. S. Foord, C. H. Lau, M. Hiramatsu, R. B. Jackman, C. E. Nebel, and P. Bergonzo, Diam. Rel. Mater. 11, 856

(2002). [8] J. A. Garrido, C. E. Nebel, R. Todt, G. Rosel, M. C. Amann, M. Stutzmann, E. Snidero, and P. Bergonzo, Appl.

Phys. Lett. 82, 988 (2003). [9] R. Ramesham and M. F. Rose, Thin Sol. Films 315, 222 (1998). [10] T. Ogawa, Y. Araki, M. Tachiki, and H. Kawarada, New Diam. Front. Carbon Technol. 10, 58 (2000). [11] Y. Araki, T. Sakai, H. Umezawa, M. Tachiki, and H. Kawarada Chem. Sens. 17 (Suppl. A), 70 (2001). [12] A. J. Saterlay, F. Marken, J. S. Foord, and R. G. Compton, Talanta 53, 403 (2000). [13] L. M. Sander, Nature 322, 789 (1986). [14] R. C. Hilborn, Chaos and Nonlinear Dynamics, (OUP, New York, 1994). [15] G. Pastor-Moreno and D. J. Riley, Electrochem. Comm. 4, 218 (2002).

Documents

Electrochemical and related processes at surface conductive diamond–solution interfaces