Journal of Electroanalytical Chemistry 657 (2011) 46–53

Contents lists available at ScienceDirect

Journal of Electroanalytical Chemistry

journal homepage: www.elsevier .com/locate / je lechem

Scanning electrochemical microscopy at thermal sprayed anti-corrosioncoatings: Effect of thermal spraying on heterogeneous electron transfer kinetics

Lee Johnson a, Akbar Niaz b, Adrian Boatwright a, K.T. Voisey b, Darren A. Walsh a,⇑a School of Chemistry, The University of Nottingham, Nottingham NG7 2RD, UKb Faculty of Engineering, The University of Nottingham, Nottingham NG7 2RD, UK

a r t i c l e i n f o a b s t r a c t

Article history:Received 5 August 2010Received in revised form 9 March 2011Accepted 10 March 2011Available online 17 March 2011

Keywords:Thermal sprayingElectron transfer kineticsUltramicroelectrodeSuperalloySECM

1572-6657/$ - see front matter � 2011 Elsevier B.V. Adoi:10.1016/j.jelechem.2011.03.009

⇑ Corresponding author. Tel.: +44 115 951 3437; faE-mail address: darren.walsh@nottingham.ac.uk (D

The effect of thermal spraying on the electrochemical activity of an anti-corrosion superalloy was studiedquantitatively using scanning electrochemical microscopy (SECM). The superalloy used was Inconel 625(a Ni base superalloy) and thin coatings of the alloy were formed on mild steel using high velocityoxy-fuel (HVOF) thermal spraying. The kinetics of electron transfer (ET) across the Inconel 625coating/electrolyte interface were studied using SECM using ferrocenemethanol as the redox mediator.For comparison, the kinetics of ET across stainless steel/electrolyte and bulk wrought Inconel 625/electrolyte interfaces were also studied using SECM. The standard heterogeneous ET rate constant, k�,for ferrocenemethanol reduction at stainless steel was 1.0 ± 0.5 � 10�3 cm s�1, compared to2.6 ± 1.8 � 10�2 cm s�1 at the wrought Inconel 625 surface. However, at the HVOF-sprayed Inconel 625surface, the kinetics of ET varied across the surface and k� for ferrocenemethanol reduction rangedbetween �2.2 � 10�4 cm s�1 and �2.6 � 10�3 cm s�1. These results clearly demonstrate that SECM canbe used to quantify the effect of thermal spraying on the electrochemical properties of Inconel 625and that thermal spraying results in an electrochemically-heterogeneous surface.

� 2011 Elsevier B.V. All rights reserved.

1. Introduction

Thermal sprayed corrosion resistant coatings are used in a widevariety of industries to enhance the lifetime of engineering compo-nents [1]. A number of thermal spraying methods are availableincluding arc spraying, detonation gun spraying, low pressure plas-ma spraying and high velocity flame spraying [2]. The introductionof high velocity oxygen fuel (HVOF) spraying was one of the mostsignificant developments in thermal spray technology and involvesusing a supersonic flame-jet to spray a feedstock powder throughan expansion nozzle onto a substrate surface [3]. The jet acceler-ates and melts the powder particles, which deform upon impactwith the substrate and adhere by mechanical interlocking. HVOFprocesses use lower temperatures and higher velocities than otherthermal spray processes, which results in more compact and betterquality coatings than are obtained using many other thermal sprayprocesses. A range of coatings have been formed using HVOFspraying, including metal carbide [4], cermet [5], ceramic [6] andpolymer coatings [7].

One of the drawbacks of thermally sprayed coatings is that theyoften do not offer the same level of corrosion resistance as the cor-responding bulk alloys. Optimum coating performance is onlyachieved if the coating is free of interconnected pores and inclu-

ll rights reserved.

x: +44 115 951 3562..A. Walsh).

sions as these are potential sources of localized attack [8]. How-ever, HVOF coating microstructures are dominated by inter-particle (splat) boundaries, which contain pores and inclusionsthat are often depleted of alloy elements [9]. As a result, the passiv-ating films that form on these surfaces may not cover the metalsurface uniformly resulting in less than optimum corrosion resis-tance. Therefore, visualisation of the uniformity of the coating sur-face reactivity, as well as measurement of the corrosion resistanceof these coatings, is important for optimising their performance[10].

Scanning electrochemical microscopy (SECM) is an extremelypowerful tool for obtaining spatially resolved reactivity informa-tion from surfaces. SECM is an electrochemical scanning probetechnique that uses the faradaic reaction occurring between anultramicroelectrode (UME) and a substrate surface as the analyticalsignal. The position of the UME (the SECM tip) is controlled usingpiezoelectric positioners and it can be moved perpendicular or par-allel to the substrate to obtain information about the surface con-ductivity or reactivity [11]. SECM has been used to visualise thelocal electrochemical activity of materials such as iron [12], stain-less steel [13,14], anodized aluminium [15–19] and titanium oxide[20–22].

SECM was recently used to visualise the electrochemical activ-ity of thin coatings of Inconel 625, a Ni base superalloy that is usedas a corrosion-resistant material, during ferrocenemethanol(FcOH+) reduction [23]. In this paper, we describe the use of SECM

A

B µ

Fig. 1. (A) Optical microscope image of the surface of wrought Inconel 625. (B) SEMimage of the surface of HVOF-sprayed Inconel 625 coatings on mild steel substrates.The black and white arrows show HVOF thermal-spray formed splats and splatboundaries, respectively.

Fig. 2. Cyclic voltammograms obtained in 1 mM FcOH in 0.1 M K2SO4 at (A) a 2 mm diamsteel. All CVs were recorded at 10, 20, 30, 40 and 50 mV s�1 and the potential limits wereThe dotted line in D shows the voltammogram obtained at stainless steel in blank 0.1 M

L. Johnson et al. / Journal of Electroanalytical Chemistry 657 (2011) 46–53 47

to quantify the effect of the HVOF process on the electron transfer(ET) reactivity of HVOF-sprayed Inconel 625 towards FcOH+ reduc-tion. Furthermore, we compare the electrochemical activity of anHVOF-sprayed Inconel 625 coating with that of a stainless steelsubstrate and a bulk wrought Inconel 625 substrate. SECM feed-back approach curve experiments were used to determine the rateof ET across the interface between each substrate and an aqueouselectrolyte. This analysis reveals that the HVOF process introducesheterogeneity to the surface of Inconel 625, which can be quanti-fied by variations in the standard heterogeneous ET rate constant,k�, measured at different locations on the coating surface.

2. Experimental

2.1. Materials and apparatus

Inconel 625 powder (particle size 20–53 lm) was obtainedfrom William Rowland Ltd. (Sheffield, UK). Wrought Inconel 625was from Special Metals Corporation (Hereford, UK) and mildand stainless steel substrates were from Smith Metals (Bedford-shire, UK). Thin coatings (approximately 300 lm thick) of Inconel625 were formed on mild steel substrates using a Met-Jet II liquidfuel HVOF system (Metallization Limited, Dudley, UK) using themethod described previously [23]. Pt wire (25 lm diameter) wasfrom Goodfellow (Huntingdon, UK) and all chemicals were ob-tained from Sigma Aldrich and were used as received. Electro-chemical experiments were performed using a Model 910BScanning Electrochemical Microscope from CH Instruments Inc.(Austin, TX). Electrolyte solutions were prepared using Milli-Qdeionised water (18.2 MX).

2.2. Electrode fabrication and electrochemical measurements

Cyclic voltammetry was performed at a 2 mm diameter Pt diskelectrode and metal substrates sealed in non-conducting resinwith exposed areas of �1 cm2. SECM tips were prepared by heatsealing a 25 lm diameter Pt wire into a quartz capillary tube andpulling the capillary using a laser micropipette puller (ModelP2000, Sutter Instrument Co., Novato, CA) under vacuum

eter Pt disk, (B) wrought Inconel 625, (C) an Inconel 625 coating and (D) stainless(A) 0.0–0.6 V, (B) �0.2 to 0.6 V, (C) �0.2 to 0.6 V and (D) �0.4 to 1.2 V, respectively.K2SO4 at 10 mV s�1.

48 L. Johnson et al. / Journal of Electroanalytical Chemistry 657 (2011) 46–53

(�10�2 mbar), as described previously [24]. Two tips are obtainedusing this method, each of which has a long narrow taper at theend where the Pt wire radius is reduced significantly from its ori-ginal value. The Pt disks were exposed by polishing gently on amicroelectrode beveller (Model BV-10 from Sutter Instruments,Novato, CA) and the tip was then sharpened on the beveller to re-duce the Rg ratio of the tip (Rg = rg/a, where rg is the radius of theinsulating glass sheath and a is the Pt disk radius) to approximately10. The tip geometry was characterised using cyclic voltammetryand negative feedback SECM approach curve experiments in1 mM ferrocenemethanol (FcOH) in aqueous 0.1 M K2SO4 and thetips used in this study were approximately 2.0 lm in radius. Priorto use, all substrates were polished using 1 lm diamond polishingpaste and, when required, electrical contact was made to the backof the substrate using brass wire. SECM experiments were per-formed by placing a drop of electrolyte (0.1 M K2SO4), containing1 mM FcOH as the redox mediator, onto the substrate surface. APt wire counter electrode and Ag|AgCl reference electrode werethen placed into the drop of electrolyte and the SECM tip enteredthe drop from above. During SECM measurements the tip/substratedistance was estimated as follows: a series of negative feedbackSECM approach curves at insulating substrates were recordedusing the SECM tip immersed in 1 mM FcOH. Fitting these experi-mental feedback approach curves to the theoretical responses forpure negative feedback [25] allowed us to determine the distanceof closest approach of the tip to the substrate. When performing ki-netic measurements using the SECM, the tip was purposefully andgently crashed into the substrate. With the assumption that thedistance of closest approach of the tip to the substrate was similarat each sample substrate, we could then calculate the distance be-tween the tip and the substrate, which was then used in the kineticanalysis. It is important to note that contact between the tip andthe region of interest on the substrate was avoided as this resultedin activation of the surface. Therefore, the tip/substrate distancewas determined adjacent to the region of interest (approximately25 lm away) and the tip was then moved into position. We esti-mate conservatively that the error in our estimation of L using thismethod is approximately ±0.1 and this error was accounted for inour kinetic analysis.

2.3. Surface profilometry and scanning electron microscopy

Before conducting SECM experiments the surface profile of allsamples was measured using a Talysurf CLI 1000-3D profilometer(Taylor Hobson Ltd., UK). An area of 5 � 5 mm was scanned usinga 2 lm radius stylus at 500 lm s�1. The average surface roughnesswas 0.043 lm, 0.048 lm and 0.068 lm for stainless steel, wroughtInconel 625 and the Inconel 625 coating, respectively. The surfaceof each material was characterised using a Phillips XL30 ScanningElectron Microscope. Prior to imaging by SEM, the substrate sur-faces were etched to reveal microstructural features. The stainlesssteel and Inconel 625-coated steel sample were etched with 2% ni-tal (2% nitric acid in ethanol) and wrought Inconel 625 was etchedwith aqua regia (1:3 concentrated HNO3/concentrated HCl) for 10–15 s. The etched surfaces were then rinsed thoroughly with deion-ised water and dried in air.

Fig. 3. (A) Experimental SECM feedback approach curves (solid lines) obtained at astainless steel surface using a 2 lm radius Pt SECM tip. The tip potential was 0.6 Vand the substrate potential was (from bottom to top) 0.0, �0.10, �0.20, �0.30 and�0.35 V vs. Ag|AgCl. The open markers show the theoretical responses generatedusing Eq. (1) with j = 0.021, 0.044, 0.12, 0.31 and 0.66, respectively. (B) Tafel plotsof k vs. overpotential, g, obtained using the data shown in A.

3. Results and discussion

3.1. Microstructural examination of Inconel 625 coatings

Fig. 1A and B shows images of the surface of wrought Inconel625 and the surface of the HVOF-sprayed Inconel 625 on mild steelsubstrates. The wrought alloy (Fig. 1A) consists of a single face-centred cubic (fcc) c-phase solid solution with Mo and Nb rich car-

bides at the grain boundaries [26,27]. Fig. 1 shows that the surfacemorphology of the coated sample (Fig. 1B) was very different fromthat of the wrought alloy (Fig. 1A). Approximately disk-shapedsplats (black arrow in Fig. 1B) were observed on the surface ofthe coated sample, as is common for HVOF-sprayed coatings of thismaterial [23]. The black areas in Fig. 1B are the splat boundaries(white arrow). During the HVOF process some oxidation of thehot alloy particles occurs, which depletes the outer regions of theparticles of chromium and these depleted regions are predomi-nantly found at the splat boundaries [26].

3.2. Cyclic voltammetry at Pt, Inconel 625 and stainless steel

Fig. 2 shows cyclic voltammograms recorded at a range of scanrates in 1 mM FcOH at Pt, wrought Inconel 625, HVOF-sprayedInconel 625 and stainless steel substrates. At the Pt substrate, theresponse was essentially reversible; the ratio of the anodic andcathodic peak currents, ip,a/ip,c was approximately 1, the peak-to-peak separation, DEp was approximately 60 mV, the peak poten-tials, Ep,a and Ep,c, were independent of the scan rate, t, and ipwas proportional to t1/2. However, the CVs obtained at the wrought

L. Johnson et al. / Journal of Electroanalytical Chemistry 657 (2011) 46–53 49

Inconel 625 substrate (Fig. 2B), the Inconel 625 coating (Fig. 2C)and at the stainless steel surface (Fig. 2D) were all markedly differ-ent to that obtained at Pt. DEp for FcOH oxidation and reductionwas greater than 500 mV in each case and increased as t increased,as expected for kinetically-controlled heterogeneous ET reactions.In the case of the stainless steel surface, the peak due to reductionof FcOH oxidation was observed at approximately �0.1 V and the

Fig. 4. SECM image obtained at a HVOF-sprayed Inconel 625 coating on mild steel by rectip was held at 0.6 V to drive the oxidation of FcOH at a diffusion-controlled rate, the subThe electrolyte was 0.1 M K2SO4, which contained 1 mM FcOH. (A) and (B) show an SECMthe SECM image overlaid on the optical image. Labels a–d indicates positions at which ecolours in this figure, the reader is referred to the web version of this paper.)

peak for FcOH oxidation was observed at approximately 0.7 V.However, the oxidation peak was obscured by the presence of alarge peak centred at approximately 1.1 V. Unlike the peaks dueto FcOH oxidation and reduction observed at each substrate sur-face, ip for this peak was proportional to t, suggesting that it wasdue to a surface process. To confirm that this was a surface oxida-tion peak, a CV was recorded at the stainless steel surface in blank

ording iT as a 2 lm radius Pt SECM tip was scanned in x–y plane at 100 lm s�1. Thestrate was at open circuit potential and the tip-substrate distance was 3 lm (1.5L).and optical image of the same region, respectively. (C) shows a composite image ofxperimental SECM feedback approach curves were obtained. (For interpretation to

50 L. Johnson et al. / Journal of Electroanalytical Chemistry 657 (2011) 46–53

K2SO4 and this is shown by the dashed line in Fig. 2D. The presenceof this peak in the CV obtained in the blank electrolyte clearlyshows that the peak was due to oxidation of the stainless steelsurface.

3.3. Measuring the Kinetics of Heterogeneous Electron Transfer usingSECM

The kinetics of ET across the substrate/electrolyte interface canbe measured using SECM feedback approach curves recorded whenthe tip process is diffusion limited [28]. The rate constant for het-erogeneous ET across the substrate/electrolyte interface, k, can beobtained by fitting an experimental SECM feedback approach curveto the following equation [29]:

NiTðL;Rg ;jÞ � NicondT ðLþ 1

j;RgÞ

þ NiinsT ðL;RgÞ � 1

ð1þ 2:47R0:31g LjÞð1þ L0:006Rgþ0:113j�0:0236Rgþ0:91Þ

ð1Þ

where NiT is the normalised tip current, L is the normalised tip/sub-strate distance (d/a, where d is the tip-substrate distance) and j isthe dimensionless rate constant for heterogeneous electron transfer(j = ka/D). Nicond

T and NiinsT are the normalised tip currents at a con-

ductor and an insulator, respectively, and expressions for thesequantities can be found in the original paper by Cornut and Lefrou[30]. By fitting experimental approach curves to theoretical curvesgenerated for various values of j, the rate constant, k, for heteroge-neous ET across the substrate/electrolyte interface can be deter-mined. In this study, the Pt SECM tip was held at 0.6 V to drivethe diffusion-limited oxidation of FcOH to FcOH+ and a series offeedback approach curves were recorded at each substrate at arange of substrate potentials [31]. For example, Fig. 3A (solid lines)shows a series of experimental approach curves recorded at a stain-less steel substrate as the substrate potential was systematically al-tered. The increase in iT in each curve as the substrate potential wasmade more negative shows that the rate of reduction of FcOH+ toFcOH at the substrate increased as the substrate potential became

Fig. 5. Experimental SECM feedback approach curves (solid lines) obtained at aHVOF-sprayed Inconel 625 coating on mild steel using a 2 lm radius Pt SECM tip.The tip potential was 0.6 V and the substrate was at open circuit potential. Curvesa–d were obtained at the positions shown in Fig. 4A. The open markers show thetheoretical responses generated using Eq. (1) with (from top to bottom) japp = 0.65,0.29, 0.079 and 0.034, respectively.

more negative. The markers in Fig. 3A show the best fits for eachcurve obtained by fitting the experimental curves to Eq. (1) and,from these fits, a j value was obtained for the substrate reactionat each substrate potential. Using this data, Tafel plots of log k(k = j�D/a) vs. the overpotential, g (g = E � E�0, where E�0 is the for-mal potential of the FcOH/FcOH+ redox couple) were constructed.Fig. 3B shows a typical Tafel plot obtained using the fits shown inFig. 3A. Good linearity was observed in the Tafel plot indicating thatvalid kinetic measurements could be performed at this surfaceusing SECM. Extrapolation of the Tafel plot to zero overpotentialyielded a k� value of 1.0 ± 0.5 � 10�3 cm s�1 for FcOH+ reductionat the stainless steel substrate. The transfer coefficient, a, deter-mined from the slope of the Tafel slope was 0.23, which is signifi-cantly lower than the value of approximately 0.5 expected for asimple outer-sphere ET reaction. However, this a value is close tothat reported recently by Mirkin and co-workers for theRuðNH3Þ3þ=2þ

6 redox couple at a stainless steel surface (a = 0.15)[14]. Notably, the k� value measured for FcOH+ reduction atstainless steel is orders of magnitude lower than the k� values of

Fig. 6. (A) Experimental approach curves (solid lines) obtained at a wrought Inconel625 surface using a 2 lm radius Pt SECM tip. The tip potential was 0.6 V and thesubstrate potential was (from bottom to top) 0.10, 0.0, �0.10 and �0.20 V vs.Ag|AgCl. The open markers show the theoretical responses generated using Eq. (1)with j = 0.015, 0.028, 0.064 and 0.11, respectively. (B) Tafel plots of k vs.overpotential, g, obtained using the data shown in A.

L. Johnson et al. / Journal of Electroanalytical Chemistry 657 (2011) 46–53 51

between 0.2 and 2.0 cm s�1 determined using carbon and Pt elec-trodes [32–34]. However, it is close to that determined by Mirkinand co-workers for RuðNH3Þ3þ=2þ

6 oxidation at stainless steel(�10�3 cm s�1). It is highly likely that, as in the case ofRuðNH3Þ3þ=2þ

6 oxidation at stainless steel, both k� and a are affectedby the presence of a passive film on the substrate surface [14].

3.4. Visualising heterogeneity in HVOF-sprayed Inconel 625 surfacesusing SECM

HVOF-sprayed Inconel 625 surfaces were imaged in SECM feed-back mode using a 2 lm radius SECM tip and Fig. 4A shows a typ-ical SECM image of the surface obtained by recording iT as the tipwas scanned across the surface. iT varied across the sample surfaceas was observed at thermal sprayed Inconel 625 coatings previ-ously using larger SECM tips (12.5 lm) [23]. Notably, using thesesmaller (2 lm) radius SECM tip, ring-like features could beobserved in the surface and this is highlighted at position b inFig. 4A. Comparison of an optical image of a HVOF-sprayed Inconel625 surface (Fig. 4B) with the SECM image shown in Fig. 4A clearlyillustrates that the ring-like features observed in the SECM imagingexperiment corresponded to the splat boundaries observed usingoptical microscopy. To confirm further that the ring-like featuresin the SECM image corresponded to the splat boundaries, the SECMimage was overlaid on top of the photograph of the sprayed surfaceand the resulting composite image is shown in Fig. 4C. This clearlyshows that the regions of higher electrochemical activity withinthe SECM image (brown features) matched the splat boundaries

Fig. 7. (A) SECM image obtained at a HVOF-sprayed Inconel 625 coating on mild steel byThe tip was held at 0.6 V to drive the oxidation of FcOH at a diffusion-controlled rate, t(1.5L). The electrolyte was 0.1 M K2SO4, which contained 1 mM FcOH. Labels in (A) indicaarea shown in (A). Experimental SECM feedback approach curves (solid lines) obtainedpotential was 0.6 V and the substrate potential was (from bottom to top) 0.1, 0.0, �0.1 ausing Eq. (1) with (C) j = 0.021, 0.048, 0.096, 0.18, and (D) j = 0.015, 0.023, 0.070, 0.15, rweb version of this paper.)

in the optical image, while the featureless areas of the surface (inthe photograph) corresponded to regions of low electrochemicalactivity. Therefore, from our data, it does appear that the splatboundaries are the most active areas for FcOH+ reduction on thesprayed surface and that, in general, these boundaries wereapproximately 10–25 lm in diameter.

Prior to performing kinetic measurements on Inconel 625 sur-faces, it is important to confirm that the variation in current re-sponse observed at the surface (Fig. 4A) was due to localisedchanges in the kinetics of heterogeneous electron transfer andnot simple topographical effects. Therefore, SECM feedback ap-proach curves were performed at positions of apparent high activ-ity and low activity at open circuit potential and Fig. 5 showsresponses obtained at positions a–d in Fig. 4A. In all cases, iT de-creased at small L values. However, at positions a and b, whichwere located at regions of higher activity in the SECM image, somepositive feedback was observed in the feedback approach curves atapproximately 0.75 < L < 5. Positions c and d were regions of lowactivity in the SECM image and the experimental feedback ap-proach curves obtained at these positions gave predominantly neg-ative feedback. The radius of the area of the substrate thatparticipates in the feedback loop during SECM experiments isr ffi a + 1.5 d [28]. Therefore, given the relative dimensions of theactive regions identified using SECM imaging and the SECM tip ra-dius, it should be possible to perform local kinetic measurementsusing SECM feedback experiments. Apparent rate constants forheterogeneous ET across the substrate/electrolyte interface (kapp)were obtained at positions a–d by fitting the SECM feedback

recording iT as a 2 lm radius Pt SECM tip was scanned in x–y plane at 100 lm s�1.he substrate was at open circuit potential and the tip-substrate distance was 3 lmte positions at which kinetic measurements were obtained. (B) Optical image of theat (C) an active spot and (D) an inactive spot of the surface shown in (A). The tipnd �0.2 V vs. Ag|AgCl. The open markers show the theoretical responses generatedespectively. (For interpretation to colours in this figure, the reader is referred to the

52 L. Johnson et al. / Journal of Electroanalytical Chemistry 657 (2011) 46–53

approach curves to Eq. (1). kapp was determined to be 1.7 ± 0.1 �10�1 cm s�1, 7.4 ± 1.0 � 10�2 cm s�1, 2.0 ± 0.3 � 10�2 cm s�1 and8.7 ± 2.7 � 10�3 for positions a–d, respectively. These data clearlyindicate that the kinetics of heterogeneous electron transfer atthe HVOF-sprayed Inconel 625 surface varied across the surfaceand confirm that the active areas of the surface predominantlylie at the splat boundaries formed during HVOF spraying.

3.5. Measuring the spatial variation in kinetics of heterogeneouselectron transfer at HVOF-sprayed Inconel 625 surfaces using SECM

One of the goals of our study is to compare the electrochemicalactivity of wrought Inconel 625 with that of the thermal sprayedsurface as previous work has suggested a performance gap be-tween the materials; it is generally acknowledged that the corro-sion resistance of the sprayed material is lower than that of thewrought material [35]. SECM feedback approach curves were re-corded at wrought Inconel 625 and a series of fitted SECM feedbackapproach curves and a Tafel plot obtained at this surface are shownin Fig. 6. Good linearity of the Tafel plot was also obtained for thedata obtained at the wrought Inconel 625. The transfer coefficient,a, was determined as 0.12 from the slope of the Tafel plot for thewrought Inconel 625 and extrapolation to zero overpotential

Fig. 8. Tafel plots of k vs. overpotential, g, obtained using the data shown in Table 1.(A) Tafel plots for the data obtained at positions a–c. (B) Tafel plots for the dataobtained at positions d–f.

yielded a k� value for FcOH+ reduction of 2.6 ± 1.8 � 10�2 cm s�1.It is important to note that feedback-mode SECM images of thewrought Inconel surface do not reveal any features indicating thatthe surface of wrought Inconel 625 is electrochemically homoge-neous [23].

To measure k� for FcOH+ reduction at thermal sprayed Inconel625, kinetic measurements were performed at a number of loca-tions on an Inconel 625 coating surface by first recording a feed-back-mode SECM image of the surface, which is shown in Fig. 7Aalong with an optical image of the same region (Fig. 7B). SECMfeedback approach curves were recorded at a range of substratepotentials at each of the regions labelled a–f in Fig. 7A and repre-sentative approach curves obtained at an active and inactive spotare shown Fig. 7C and D, respectively, along with the best-fit the-oretical feedback approach curves generated using Eq. (1). Tafelplots of log k vs. g were constructed for each of the data sets ob-tained at positions a–f on the thermal sprayed surface and theseare shown in Fig. 8. k� values determined for FcOH+ reduction fromthe Tafel plots at each location are shown in Table 1, as are the jvalues at each potential. The average k� value at the ‘‘active’’ sites(a, b and c) on the sprayed surface was 1.3 ± 0.1 � 10�3 cm s�1,which was higher than that measured at the inactive sites d, eand f (3.7 ± 0.2 � 10�4 cm s�1). The variation in the electrochemicalactivity of the surface towards FcOH+ reduction across small areasof the surface is illustrated clearly by comparing the activities atpositions b and d, which were located on the boundary and centreof a single splat, respectively; k� at the point marked b was signif-icantly higher than at the point marked d.

While our analysis reveals that the HVOF spraying processundoubtedly introduces local variations in electrochemical activityof the coatings (at least during FcOH+ reduction) and the sprayformed splat boundaries are regions of a high electrochemicalactivity, our data suggest that the effects revealed using SECMimaging are not responsible for the performance gap betweenthe sprayed and wrought superalloys. The lower electrochemicalactivity of the sprayed alloy’s surface towards FcOH+ reductionthan that of the bulk alloy does not correlate with the poorer cor-rosion resistance of the sprayed material. However, it is very unli-kely that a simple correlation between the electrochemical activityof a thermal sprayed coating for FcOH+ reduction and its corrosionresistance under normal operating conditions is possible but ouranalysis does reveal new insights into the effects of thermal spray-ing on the electrochemical homogeneity of metal coatings. Thereduction in k� at the sprayed surface compared to the wroughtmaterial’s surface may be due to formation of localised oxides dur-ing thermal spraying, which differ from those formed on the bulkalloy’s surface. One of the goals of our future research will be to re-duce the surface heterogeneity introduced during the sprayingprocess and to determine the effects of such factors on the perfor-

Table 1j values determined from fitted SECM feedback approach curves as a function ofpotential at the locations marked a–f in Fig. 7. The k� values were determined fromTafel plots of log k vs. overpotential.

a b c d e f

j (0.10 V) 0.093 0.030 0.021 0.015 0.027 0.015j (0.05 V) 0.12 0.040 0.032 0.020 0.023 0.020j (0.00 V) 0.12 0.044 0.048 0.023 0.035 0.024j (�0.05 V) 0.14 0.060 0.067 0.037 0.048 0.027j (�0.10 V) 0.14 0.10 0.096 0.070 0.042 0.060j (�0.15 V) 0.18 0.13 0.13 0.098 0.047 0.078j (�0.20 V) 0.23 0.16 0.18 0.15 0.060 0.13j (�0.25 V) 0.30 0.22 0.25 0.22 0.11 0.20104 k�/

cm s�126 ± 5 6.2 ± 0.9 4.4 ± 1.7 2.2 ± 0.9 6.5 ± 2.3 2.2 ± 1.1

L. Johnson et al. / Journal of Electroanalytical Chemistry 657 (2011) 46–53 53

mance of these materials. As our results here show, SECM will be avery useful tool for quantifying such modifications.

4. Conclusions

In this study, SECM has been used to measure the rate of elec-tron transfer across the interface between a thermal sprayedanti-corrosion superalloy (Inconel 625) coating and an aqueouselectrolyte. The rate of electron transfer from the coating to a redoxspecies in solution was measured at discrete locations on thesprayed surface and the rate varied across the surface. This datacorrelated very well that obtained using SECM imaging and dem-onstrated that the sprayed surface was electrochemically hetero-geneous. Specifically, SECM imaging revealed that the regions ofhighest electrochemical activity for FcOH+ reduction are the splatboundaries formed during the spraying process. In addition, thedata obtained from the sprayed surface was compared with dataobtained from samples of the bulk wrought superalloy and stain-less steel. Our analysis reveals that, while the thermal sprayingprocess introduces significant heterogeneity to the superalloy sur-face, the electrochemical activity of the coating for FcOH+ reduc-tion is lower than that of the bulk, wrought superalloy.

Acknowledgements

We thank the University of Nottingham Research Strategy Fundfor funding. DW also thanks the EPSRC for funding this workthrough the Science and Innovation Award (Driving Innovation inChemistry and Chemical Engineering: Grant EP/D501229/1).

References

[1] J.R. Davis, Handbook of Thermal Spray Technology, ASM International,Materials Park, OH, USA, 2004.

[2] J.H. He, J.M. Schoenung, Mater. Sci. Eng., A 336 (2002) 274.[3] P. Fauchais, A. Vardelle, B. Dussoubs, J. Therm. Spray Technol. 10 (2001) 44.

[4] C. Verdon, A. Karimi, J.L. Martin, Mater. Sci. Eng. A-Struct. Mater. Prop.Microstruct. Process. 246 (1998) 11.

[5] D. Toma, W. Brandl, G. Marginean, Surf. Coat. Technol. 138 (2001) 149.[6] A. Kulkarni, J. Gutleber, S. Sampath, A. Goland, W.B. Lindquist, H. Herman, A.J.

Allen, B. Dowd, Mater. Sci. Eng. A-Struct. Mater. Prop. Microstruct. Process. 369(2004) 124.

[7] E. Petrovicova, R. Knight, L.S. Schadler, T.E. Twardowski, J. Appl. Polym. Sci. 78(2000) 2272.

[8] G. Bolelli, L. Lusvarghi, R. Giovanardi, Surf. Coat. Technol. 202 (2008) 4793.[9] A.J. Sturgeon, J. Mater. Sci. 32 (1997) 863.

[10] K. Fushimi, K.A. Lill, H. Habazaki, Electrochim. Acta 52 (2007) 4246.[11] P. Sun, F. Ois, O. Laforge, M.V. Mirkin, Phys. Chem. Chem. Phys. 9 (2006) 802.[12] C. Gabrielli, S. Joiret, M. Keddam, H. Perrot, N. Portail, P. Rousseau, V. Vivier,

Electrochim. Acta 52 (2007) 7706.[13] D.O. Wipf, Colloids Surf. A 93 (1994) 251.[14] P. Sun, Z. Liu, H. Yu, M.V. Mirkin, Langmuir 24 (2008) 9941.[15] I. Serebrennikova, H.S. White, Electrochem. Solid-State Lett. 4 (2001) B4.[16] M.B. Jensen, A. Guerard, D.E. Tallman, G.P. Bierwagenb, J. Electrochem. Soc. 155

(2008) C324.[17] J.C. Seegmiller, D.A. Buttry, J. Electrochem. Soc. 150 (2003) B413.[18] A.M. Simoes, D. Battocchi, D.E. Tallman, G.P. Bierwagen, Corros. Sci. 49 (2007)

3838.[19] A. Davoodi, J. Pan, C. Leygraf, S. Norgren, Electrochim. Acta 52 (2007) 7697.[20] R. Zhu, C. Nowierski, Z. Ding, J.J. Noel, D.W. Shoesmith, Chem. Mater. 19 (2007)

2533.[21] S.B. Basame, H.S. White, J. Phys. Chem. 99 (1995) 16430.[22] S.B. Basame, H.S. White, Anal. Chem. 71 (1999) 3166.[23] D.A. Walsh, L.E. Li, M.S. Bakare, K.T. Voisey, Electrochim. Acta 54 (2009) 4647.[24] A.W. Taylor, F.L. Qiu, J.P. Hu, P. Licence, D.A. Walsh, J. Phys. Chem. B 112 (2008)

13292.[25] R. Cornut, C. Lefrou, J. Electroanal. Chem. 608 (2007) 59.[26] N. Ahmed, M.S. Bakare, D.G. McCartney, K.T. Voisey, Surf. Coat. Technol. 204

(2010) 2294.[27] V. Shankar, B.S. Rao, K. Mannan, J. Nucl. Mater. 288 (2001) 222.[28] M.V. Mirkin, Determination of electrode kinetics, in: C.G. Zoski (Ed.),

Handbook of Electrochemistry, Elsevier, 2007, p. 639.[29] C. Lefrou, R. Cornut, Chem. Phys. Chem. 11 (2010) 547.[30] R. Cornut, C. Lefrou, J. Electroanal. Chem. 621 (2008) 178.[31] S. Bollo, S. Finger, J.C. Sturm, L.J. Nunez-Vergara, J.A. Squella, Electrochim. Acta

52 (2007) 4892.[32] W.J. Miao, Z.F. Ding, A.J. Bard, J. Phys. Chem. B 106 (2002) 1392.[33] C. Bourdillon, C. Demaille, J. Moiroux, J.M. Saveant, J. Am. Chem. Soc. 117

(1995) 11499.[34] A.P. O’Mullane, J. Zhang, A. Brajter-Toth, A.M. Bond, Anal. Chem. 80 (2008)

4614.[35] M.S. Bakare, K.T. Voisey, M.J. Roe, D.G. McCartney, Appl. Surf. Sci. 257 (2010)

786.