Elsevier

Electrochimica Acta

Volume 78, 1 September 2012, Pages 65-74
Electrochimica Acta

Influence of fluoride content on the barrier layer formation and titanium dissolution in ethylene glycol–water electrolytes

https://doi.org/10.1016/j.electacta.2012.05.093Get rights and content

Abstract

In the present paper, the initial stages of anodic film growth on Ti in water-containing ethylene glycol electrolyte with the addition of fluoride (0.015–0.17 M) were investigated using electrochemical and surface analytical techniques. Steady-state current–potential curves and electrochemical impedance spectra as depending on potential and fluoride content point to two parallel reaction pathways – film growth/dissolution and titanium dissolution through the film. The chemical composition of the anodic films in electrolytes with different fluoride content was estimated by X-ray photoelectron spectroscopy (XPS). XPS analyses revealed the presence of a non-stoichiometric oxide containing mainly Ti4+ and a certain amount of Ti3+, with a certain degree of hydroxylation. A kinetic model of the process is proposed and its parameters are estimated by quantitative comparison with the current–potential and EIS data. The apparent reaction orders of the individual steps with respect to fluoride are also estimated. The main features of the XPS data are also reproduced by the model.

Highlights

► Initial stages of Ti oxidation in ethylene glycol containing water and fluoride investigated. ► Electrochemical techniques point to two parallel reactions – film growth and Ti dissolution. ► A partly hydroxylated oxide containing mainly Ti4+ and Ti3+ is formed. ► A kinetic model of the process reproduces well both electrochemical and XPS data. ► Apparent reaction orders of individual steps with respect to fluoride estimated.

Introduction

The electrochemical formation of self-ordered arrays of TiO2 nanotubes in fluoride-containing electrolytes has attracted great interest due to their applications in sensing [1], [2], [3], [4], photocatalysis [5], [6], solar cells [7], batteries [8] and biology/medicine [9]. Several comprehensive reviews that appeared in the last decade [10], [11], [12], [13] have demonstrated that the benefit of electrochemically grown TiO2 is in the ability to fabricate nanotube arrays of different shape, pore size, length, and wall thickness by varying process parameters such as the applied voltage, electrolyte composition, the pH of the electrolyte, and process duration. Recently, smooth titania nanotubes with large thicknesses (up to several hundreds of micrometers) have been obtained in fluoride-containing electrolytes based on organic solvents such as ethylene-glycol [10], [11], [13], key parameters for the kinetics of their growth being the fluoride and water content of the electrolyte [14], [15], [16].

Several model approaches have been advanced to explain generation and growth of nanopores and nanotubes during anodic oxidation of Al, Ti and other valve metals [17], [18], [19], [20], [21], [22], [23]. They attribute the pore generation and propagation to either the field-assisted dissolution of metal cations at the barrier oxide film/electrolyte interface [17], to the local breakdown (or thinning) of the initial barrier film [18], [19], [20], or to the growth-induced plasticity of the oxide resulting in a material flow from the barrier film towards the regions of pore growth. The so-called flow model of growth of nanoporous and nanotubular oxides [22] assumes the flow of material from the barrier film toward the pore walls. Based on such considerations, a mathematical model for the influence of anodizing conditions on the morphology of porous films has been developed [23]. Very recently, a criterion for film instability leading to the formation of ordered porous morphology has been advanced [24]. On the other hand, much less attention has been paid to the process of nanopore initiation during the growth of the barrier layer [25], even if recent reports suggest that porous structures are formed on Ti already at comparatively low potentials [26].

To account for both the film growth and underlying valve metal dissolution under anodic polarization in fluoride-containing electrolytes, a quantitative model has recently been proposed by some of us [13], [27], [28], [29], [30], [31]. It has been found to reproduce successfully both the steady-state current vs. potential curves and impedance spectra of Nb and W for a range of potentials [28], [29], [30] and to be able to predict the oxidation state of the cation at the oxide/electrolyte interface [31]. Based on these results, we believe that nanopore initiation could be due to the subdivision of the surface into zones of growth/chemical dissolution and zones of electrochemical cation ejection, resulting in the perturbation of the film/electrolyte interface [30]. To quantify such a concept, it is of paramount importance to obtain reliable estimates of the kinetic parameters of individual reaction steps at this interface. The present paper is a step towards this goal. It presents comprehensive electrochemical data concerning the influence of fluoride content in an ethylene glycol–water electrolyte on the growth and dissolution of anodic titanium oxide films at comparatively low potentials (0.1–5 V). First, voltammetric and electrochemical impedance spectroscopic results in a wide potential range for Ti in ethylene glycol–water solution containing various amounts of fluoride (0.015–0.17 M) are presented and discussed. Second, the adaptation of the kinetic model to the titanium system is thoroughly described and the adopted computational procedure to obtain estimates of the kinetic parameters is discussed. Third, the parameter values are considered in view of the proposed concept of pore initiation and the contemporary level of knowledge about the specific nanotubular structure. Finally, some limitations of the approach are outlined and plans for further research are presented in some detail.

Section snippets

Experimental

Pure Ti foils (99.9%, Goodfellow) with an exposed area of 4 cm2 were used as working electrodes. Prior to anodic oxidation, the samples were chemically polished in a 1:3 mixture of HF (40 wt%) and HNO3 (65 wt%) until a mirror finish was obtained. The electrochemical measurements were carried out using a three-electrode configuration featuring a Pt mesh as a counter electrode and Ag/AgCl/3 M KCl electrode (E = 0.200 V vs. SHE) as a reference. All the potentials are quoted vs. this kind of reference

Electrochemical measurements

The steady-state current vs. potential curves for Ti in ethylene glycol (EG)–0.6 M H2O solution containing different concentrations of NH4F are shown in Fig. 1. The current at constant potential is found to increase with fluoride content, indicating an increase of the rate of the anodic oxidation process. This increase is more pronounced at potentials higher than 1.0 V. In general, the curves consist of two segments – a relatively fast increase at low potentials up to 1 V and a much more gradual

Physical model

In the present experiments, a mixed-valence oxide is present on the titanium surface already at potentials slightly anodic than the open-circuit potential. In this film, Ti(III) and Ti(IV) positions coexist in the cation sub-lattice, together with a certain concentration of Ti(IV) vacancies. It is also assumed that a significant concentration of oxygen vacancies exists in the anion sub-lattice, so that the barrier film growth could take place at the inner interface, the oxygen migrating inwards

Conclusions

In the present paper, an investigation of the effect of fluoride addition on anodic film growth and dissolution on Ti in water-containing EG electrolyte using electrochemical and surface analytical techniques, as well as calculations on the basis of quantitative kinetic model, are reported. The following main conclusions can be drawn from the obtained results:

  • Higher fluoride content in the electrolyte leads to a higher current density and to a certain extent larger barrier film thickness at

Acknowledgment

The financial support of the by the National Science Fund, Bulgarian Ministry of Education and Science, under contract DDVU-02-103 “Nanoporous anodic oxides as new generation of optically active and catalytic materials (NOXOAC, 2010-2013)” is gratefully acknowledged.

References (57)

  • K.S. Raja et al.

    Electrochemistry Communications

    (2007)
  • W. Wei et al.

    Electrochemistry Communications

    (2010)
  • G.D. Sulka et al.

    Electrochimica Acta

    (2010)
  • G.K. Mor et al.

    Solar Energy Materials and Solar Cells

    (2006)
  • J.M. Macak et al.

    Current Opinion in Solid State and Materials Science

    (2007)
  • K.S. Raja et al.

    Electrochimica Acta

    (2005)
  • M. Bojinov et al.

    Electrochimica Acta

    (2003)
  • V. Karastoyanov et al.

    Materials Chemistry and Physics

    (2008)
  • I. Betova et al.

    Electrochimica Acta

    (2010)
  • I. Frateur et al.

    Journal of Electroanalytical Chemistry

    (2000)
  • M. Curioni et al.

    Electrochimica Acta

    (2010)
  • R. Giovanardi et al.

    Electrochimica Acta

    (2011)
  • M. Bojinov

    Electrochimica Acta

    (1997)
  • J.T. Mayer et al.

    Journal of Electron Spectroscopy and Related Phenomena

    (1995)
  • E.L. Bullock et al.

    Surface Science

    (1996)
  • E. McCafferty et al.

    Applied Surface Science

    (1999)
  • I. Milosev et al.

    Biomaterials

    (2000)
  • C.E.B. Marino et al.

    Thin Solid Films

    (2004)
  • Z. Xia et al.

    Electrochemistry Communications

    (2007)
  • I. Milosev et al.

    Electrochimica Acta

    (2008)
  • T.K. Sham et al.

    Chemical Physics Letters

    (1979)
  • W. Göpel et al.

    Surface Science

    (1984)
  • R.P. Antony et al.

    Material Chemistry and Physics

    (2012)
  • B. Yang et al.

    Material Chemistry and Physics

    (2011)
  • H.H. Huang

    Biomaterials

    (2002)
  • H. Habazaki et al.

    Electrochemistry Communications

    (2007)
  • A. Valota et al.

    Electrochimica Acta

    (2009)
  • K.E. Healy et al.

    Journal of Colloid and Interface Science

    (1992)
  • Cited by (62)

    View all citing articles on Scopus
    View full text