- Email: [email protected]
Electrochemical electrolyte spreading studies of the protective properties of ultra-thin films on zinc galvanized steel
Surface & Coatings Technology 228 (2013) 286–295
Contents lists available at SciVerse ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Electrochemical electrolyte spreading studies of the protective properties of ultra-thin films on zinc galvanized steel R. Posner a,⁎, N. Fink a, M. Wolpers b, G. Grundmeier c a b c
Max-Planck-Institut für Eisenforschung GmbH, Department of Interface Chemistry and Surface Engineering, Max-Planck-Str. 1, 40237 Düsseldorf, Germany Henkel AG & Co. KGaA, Henkelstr. 67, 40589 Düsseldorf, Germany University of Paderborn, Department of Technical and Macromolecular Chemistry, Warburger Str. 100, 33098 Paderborn, Germany
a r t i c l e
i n f o
Article history: Received 5 February 2013 Accepted in revised form 18 April 2013 Available online 25 April 2013 Keywords: Conversion coating Reactive electrolyte spreading Accelerated corrosion testing Cathodic delamination Zinc galvanized steel
a b s t r a c t Reactive electrolyte spreading along the surfaces of different conversion films on zinc galvanized steel in humid air was monitored visually and with a height-regulated scanning Kelvin Probe. Electrochemical impedance spectroscopy and current density-potential curves revealed that decelerated spreading kinetics are connected with increasing pore resistances of the pre-treatment layers and decreasing oxygen reduction current densities in the electron transfer controlled potential region. After a few days the progress ranking of electrolyte spreading along uncoated conversion films reflected the progress tendencies of cathodic delamination observed on epoxy coated conversion layers after long-time exposure to the same corrosive environment. Such correlation was not discovered for pre-treatment films that do not provide relevant electrochemical barrier properties. The results suggest that oxygen reduction driven electrolyte wetting is an option for accelerated performance testing of anticorrosive ultra-thin films on metal substrates that can be subject to cathodic delamination. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Adequate characterization of polymer coated metal compounds modified by adhesion promoting and corrosion protective thin films remains a challenge due to the limited availability of analytical approaches for buried interfaces. The residual electrochemical activity of the pre-treated substrate surface, barrier properties of the organic coating and its adhesion as well as molecular and morphologic characteristics of the organic/inorganic interface structure synergistically determine the corrosion resistance [1–6]. Recently launched two-step pretreatment procedures such as the deposition of amorphous iron oxide prior to the formation of Zr- or Ti-based barrier layers on zinc were developed to offer superior performance for specific applications. However, they also result in complex interface designs which require even more detailed understanding and careful analysis [7,8]. In general, standardized salt spray exposure and long-term weathering tests have to be passed to achieve approval for the use of conversion chemistry based protection systems e.g. at car bodies or building facades [9]. As such procedures are time-consuming, industrial research will especially benefit from short-term and cost-efficient routines to quickly review progress in daily product development. In this context, the comparison of ion transport processes at polymer/metal interfaces seems to offer promising ⁎ Corresponding author at: Henkel AG & Co. KGaA, Henkelstrasse 67, 40589 Düsseldorf, Germany. Tel.: +49 211 797 2374; fax: +49 2133 537379. E-mail address: [email protected] (R. Posner). 0257-8972/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.surfcoat.2013.04.042
options for accelerated testing and interface characterization. It was observed that the ion mobility sensitively responds to the electrochemical activity of the substrate material and its oxide properties [10–12], to the permeability of the polymer with respect to water and oxygen [13,14], to paint adhesion as well as to mechanical properties of its macromolecular network [13–15]. Moreover, previous publications reported that transport mechanisms detected at polymer covered zinc and steel substrates resemble those observed on uncoated Zn and Fe surfaces during the exposure to humid air [10,11]. Ion convection starts from electrolyte droplets deposited/formed on metal surfaces or from electrolyte covered defects in the polymer coating that extend to the metal substrate. It is initiated by the reduction of atmospheric oxygen and the formation of hydroxide species in the periphery of the electrolyte reservoir. A flow of cations from the reservoir centre, where active dissolution of the base material occurs, toward the cathodic sites of the local galvanic cell then ensures charge compensation. This promotes an extension of the oxygen reduction zone. For droplet spreading along the substrate surface such finding was explained by a reduction of the oxide/electrolyte interface tension with increasing pH above the isoelectric point of the oxide due to OH formation [16,17]. Cation ingress into polymer/oxide/metal interface sections adjacent to an electrolyte covered coating defect occurs preferentially on zinc, exclusively on iron and is commonly regarded as an important part of the cathodic delamination mechanism [18–23]. Fig. 1 provides a schematic illustration of relevant ion transport processes for metal substrates with and without organic coating. In-situ monitoring of electrode potentials at electrolyte/metal and polymer/metal interfaces
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
287
Fig. 1. Schematic illustration of electrolyte transport mechanisms on zinc and iron substrates in humid air (relative humidity: typically > 90%). a) Reactive electrolyte spreading along the bare metal surfaces, starting at a highly viscous droplet of NaCl electrolyte. b) Cathodic delamination along a polymer/oxide/metal interface, starting at a coating defect covered with NaCl electrolyte. Please be aware that the O2 and H2O activity in the polymer phase will be significantly lower than in humid air.
by the Scanning Kelvin Probe (SKP) and subsequent investigation of elemental distributions resulting from ion transport reflect established state-of-the-art approaches for analysis [10–12,18–23]. Stability rankings for different interface designs on iron and steel samples were indeed accessible by cathodic delamination studies [13,14], but occurred to be problematic for zinc substrates so far [14,24]. The present publication consequently focuses on the introduction of an analogous approach for conversion chemistry pre-treated zinc electro galvanized steel. Electrolyte spreading was initiated on pre-treatment layers with different barrier properties. The results were compared to the characteristics of polarization curves as well as impedance spectra recorded on these anticorrosive films and correlated with cathodic delamination rankings obtained
from ion transport studies at epoxy/conversion layer/zinc interfaces. These data will demonstrate the practicability as well as the limits of the electrolyte spreading approach for accelerated performance testing of Zr-/Ti-based conversion chemistry layers. 2. Experimental 2.1. Sample preparation Zinc electro galvanized steel sheets with a coating thickness of approx. 7.5 μm were supplied by voestalpine Stahl Linz GmbH (Linz, Austria). After alkaline spray cleaning with a solution that contained 2% Ridoline 1250 BR and 0.2% Ridosol 1270 (from Henkel AG & Co.
Fig. 2. SEM images of sample surfaces used for experiments discussed in the present publication. a) and b): Zinc electro galvanized steel. c) and d): Zinc electro galvanized steel pre-treated according to procedure (a).
288
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
Fig. 3. SEM images of sample surfaces used for experiments discussed in the present publication. a) and b): Zinc electro galvanized steel pre-treated according to procedure (b). c) and d): Zinc electro galvanized steel pre-treated according to procedure (d).
KGaA, Düsseldorf, Germany) for 15 s at 50 °C and a pressure of 1 bar, they were tap water, then DI water rinsed and dried in warm air. Subsequent pre-treatment with conversion chemistry (patented by Henkel AG & Co. KGaA, Düsseldorf, Germany) was performed according to one of the following procedures:
layer mainly consisted of Ti-oxide species and of zinc phosphate embedded in an organic network [25]. (c) With an aqueous solution of a composition similar to that described for procedure (b) except that the titanium component was partially substituted by a fluoric acid of zirconium. Roll coating led to an amorphous organic-inorganic hybrid film with a coating weight of ~ 7 mg/m 2 for Ti and of ~ 3 mg/m 2 for Zr. The conversion layer mainly consisted of Ti-oxide species, of Zr-oxide and of zinc phosphate embedded in an organic network [25]. (d) A two-step pre-treatment approach including procedure (a) followed by a rinse step with DI water and subsequently by procedure (b).
(a) With an aqueous, acidic solution for electroless iron deposition based on a Fe2+-containing inorganic compound, a reduction agent, and chemical mediators as major components. 10 s of sample immersion at 50 °C resulted in the precipitation of porous and amorphous deposits of iron, of Fe(II)-, and of Fe(III)-compounds at a coating weight of ~40 mg/m2 for Fe and of ~10 mg/m2 for P [7]. (b) With an aqueous solution containing a fluoric acid of titanium, a derivative of a polymeric vinyl phenol and a phosphate containing inorganic reagent as major components. Roll coating at a speed of 60 m/min and a quench pressure of 1 kN at room temperature resulted in an amorphous organic-inorganic hybrid film with a Ti coating weight of ~ 10 mg/m 2. The conversion
After rinsing with DI water and drying in a nitrogen stream, the conversion chemistry pre-treated samples were covered with a ~ 90 μm thick layer of a hot-curing two-component epoxy resin provided by Henkel AG & Co. KGaA, Düsseldorf, Germany. An area of ~ 1 × 1 cm was omitted that was supposed to be used as an
Table 1 Overview of the elemental composition detected by EDS on bare zinc electro galvanized steel and on zinc electro galvanized steel pre-treated according to procedure (a), (b) or (d).
Table 2 Overview of the elemental composition detected by XPS on bare zinc electro galvanized steel and on zinc electro galvanized steel pre-treated according to procedure (a), (b) or (d).
Relative amount [%] →
C
O
Zn
P
Fe
Ti
F
Relative amount [%] →
C
O
Zn
P
Fe
Ti
Mn
N
F
Zn Zn + pre-treatment (a) Zn + pre-treatment (b) Zn + pre-treatment (d)
3.0 3.9 16.9 15.2
2.2 6.4 7.4 10.5
94.8 81.5 68.4 61.4
– 1.9 1.1 3.0
– 6.3 – 3.8
– – 1.0 0.9
– – 4.7 5.2
Zn Zn + pre-treatment (a) Zn + pre-treatment (b) Zn + pre-treatment (d)
12.9 24.3 59.8 52.9
56.4 51.6 26.2 30.1
30.7 18.5 1.5 2.5
– 1.9 1.5 1.5
– 3.7 – 0.2
– – 3.0 3.6
– – 1.0 1.4
– – 1.3 1.1
– – 5.7 7.0
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
289
Fig. 4. Photography of a set of epoxy coated, conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel samples taken after the exposure to humid air. Cathodic delamination was initiated after an artificially prepared coating defect was covered with 0.5 molar NaCl solution of high viscosity. a) Progress of cathodic delamination after one month of exposure. Two black lines observed on the zinc reference sample refer to the front of macroscopic polymer de-adhesion (on the left) and to the front of interfacial electrolyte transport (on the right). On all the other samples, both fronts are equivalent. b) Progress of cathodic delamination after two months of exposure. The electrolyte front is indicated by solid black lines.
artificial coating defect. In this area, also the conversion coating was removed by grinding. The polymer was hardened for 1 h at 120 °C applying a pressure of 50 g/cm 2 [10]. For the weathering experiments, the defect area of the samples was covered with highly viscous 0.5 molar NaCl or 0.5 molar KBr solution. To ensure high viscosity, ~ 3% of agar was added to the liquid prior to heating to ~ 70 °C. Subsequent cooling then resulted in an almost ‘solid’ electrolyte solution. Weathering exposure was performed in humid air (with a relative humidity of ≥95%) at room temperature. All used chemicals and solvents were of analytical grade. 2.2. Electrochemical and surface analytical methods Scanning Kelvin Probe (SKP) line scans were recorded with a height-regulated, custom-made SKP apparatus in humid air (with a relative humidity of ≥ 95%) at room temperature. Detected Volta potential differences were converted to the standard hydrogen electrode (SHE) scale after calibration against Cu/CuSO4 [10].
Electrochemical measurements in three-electrode arrangement were done with a Gamry FAS2 Femtostat, an Ag/AgCl/3 molar KCl reference electrode, a Pt counter electrode, and chloride-free borate buffer solution with a pH of 8.4. Electrochemical impedance spectroscopy (EIS) experiments were performed in a frequency range of 0.1 Hz to 100 kHz with an amplitude of 10 mV at open circuit potential of the investigated sample system. X-ray photoelectron spectroscopy (XPS) analysis was conducted with a Quantum 2000 (Physical Electronics, Chanhassen, MN/USA). Spectra were measured with a spot size of 100 μm × 100 μm at 45° using monochromated Al Kα radiation, 23.9 eV pass energy, and a step size of 0.2 eV for high resolution spectra. Time-of-flight secondary ion mass spectrometry (ToF-SIMS) measurements were executed with a PHI TRIFT CE (Physical Electronics, Chanhassen, MN/USA) applying a gallium ion gun at an acceleration voltage of 15 kV and a spot size of 100 μm × 100 μm. Scanning electron microscopy (SEM) images were performed with a LEO 1550VP Field Emission SEM apparatus (Leo Elektronenmikroskopie
290
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
Fig. 5. Photography of a set of epoxy coated, conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel samples taken after 21 days of exposure to humid air. Cathodic delamination was initiated after an artificially prepared coating defect was covered with 0.5 molar NaCl solution of high viscosity. Two black lines observed on the zinc reference sample refer to the front of macroscopic polymer de-adhesion (on the left) and to the front of interfacial electrolyte transport (on the right). On all the other samples, both fronts are equivalent.
features. Figs. 2 and 3 present images recorded on bare zinc electro galvanized steel (Fig. 2a–b), on zinc electro galvanized steel pre-treated according to procedure (a) (see Fig. 2c–d), procedure (b) (see Fig. 3a–b), or procedure (d) (see Fig. 3c–d). Non-pre-treated surfaces exhibit structures of stacked and terraced zinc plates. They are grouped in differently oriented domains, which arise from crystallographically preferred growth directions of electro deposited Zn crystallization cores on grains of the underlying steel surface. Sharp and well-defined edges of the zinc plates with characteristic angles resulted (see Fig. 2a and b) [27,28]. Immersion in the solution for electroless iron deposition led to coverage of the Zn plates, which visually also manifests in grading of their edges. However, vacancies along the plate boundaries partially remain. Their diameters seem to depend on the size of interstice volumes between differently oriented plate domains (see Fig. 2c and d). In contrast, pre-treatment procedure (b) causes visually lesspronounced modifications of the sample surfaces. Deposits resulting from conversion processes are hardly verifiable even at high magnification in Fig. 3b, but the edges at the boundaries of the Zn plates are slightly glazed. This can be interpreted as an indicator for the presence of a very thin conversion film that occurs to be almost transparent when investigated by SEM. Fig. 3c and d illustrate the appearance of the substrate surface after the two-step pre-treatment (d) was performed. Deposits on the zinc plates are clearly visible and more pronounced than those observed in Fig. 2c and d. The original morphology of the Zn surface is still identifiable; however, the grading effect resulting from procedure (a) seems to be amplified. Tables 1 and 2 provide an overview of elements detected on pre-treated and non-pre-treated zinc electro galvanized steel. The Zn portion is mainly attributed to the base material, because it occurs to be reduced after the deposition of a conversion layer. Compared to the zinc reference surface, a strong increase of the carbon percentage refers to the formation of an organic network after the samples were processed according to procedure (b). Iron is indeed verifiable if procedure (a) was part of the pre-treatment process. However, its amount is reduced after sample immersion in conversion bath (b), presumably partly due to coverage with organic and inorganic Ti-, P- and Mn-species and partly due to re-dissolution of Fe at low pH. The relative percentage values obtained with EDS may not be directly compared to those resulting from XPS analysis, because XPS is by far more surface sensitive with a penetration depth of only a few nanometers compared to several micrometers for EDS. The XPS data consequently indicate an enrichment of trace elements near the surface of the deposited conversion layers. Therefore, Tables 1 and 2 confirm that almost all major ingredients of pre-treatment solutions (a) and (b) described in chapter 2 are verifiable on the samples, as well. Moreover, Figs. 2 and 3 clearly show that the different pre-treatment approaches also result in a visually distinguishable surface appearance of zinc electro galvanized steel. 3.2. Weathering of and electrolyte spreading on conversion layers
GmbH, Oberkochen, Germany) using an InLens detector in high vacuum mode. Energy-dispersive X-ray spectroscopy (EDS) analysis was made with an INCA-7426 detector from Oxford Instruments (Oxford, UK) that was connected to the SEM machine. Both SEM images and EDS spectra were recorded with an electron high tension voltage of 10 kV and a working distance of 10 mm. 3. Results and discussion 3.1. Characterization of conversion chemistry layers The pre-treatment procedures described in chapter 2 resulted in the deposition of conversion chemistry layers on the substrate surfaces with a thickness of typically some ten nanometers [26]. Visual inspection by electron microscopy revealed characteristic surface
In order to investigate their corrosion resistance, conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel sheets were exposed to humid air after they were coated with an epoxy layer. A defect zone remained uncoated and was filled up with highly viscous 0.5 molar NaCl solution to initiate cathodic delamination. Fig. 1b reflects a schematic cross-section view of the selected sample composition. Photos were taken after one and two months to reveal the progress of electrolyte transport along the epoxy/ZnO/Zn- and the epoxy/conversion layer/Zn-interfaces. Fig. 4 presents the resulting images. After one month, a characteristic stability ranking was established, which did not change during the following weeks (see Fig. 4b). Visible delamination became manifest in the formation of blisters in the coating and in electrolyte droplets evolving in the vicinity of
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
291
Fig. 6. Photography of conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel sheets taken after the exposure to humid air. Reactive electrolyte spreading was initiated after a droplet of highly viscous 0.5 molar NaCl solution was deposited on the samples. a) Progress of electrolyte spreading after one day of exposure. b) Progress of electrolyte spreading after 16 days of exposure.
the defect zone. The respective front of cathodic delamination is highlighted by black lines in Fig. 4. They indicate that corrosive degradation proceeded slightly inhomogeneous. However, considering the integral coating area affected by cathodic undermining, it is clear that the untreated zinc sample delaminates fastest, followed by zinc pre-treated with solution (c). Samples (a) and (d) that were both immersed in conversion chemistry bath (a) do not only exhibit a darker surface, but also show almost equivalent performance. Based on Fig. 4, procedure (b) seems to be superior to the other pre-treatment procedures, but this finding was not generally verifiable. Fig. 5 presents a photo of another polymer coated sample batch that was subject to corrosive delamination for 21 days. A similar ranking was observed, but sample “zinc + pre-treatment (b)” did not show the best performance in this case. For a better differentiation, the formation of blisters as a result of the delamination process may be used an additional distinguishing feature: Electrolyte transport resulted in very few, but large blisters on sample (b) compared to a pattern of numerous little ones on samples (a) and (d). A different type of electrolyte transport experiment was performed with non epoxy coated, but conversion chemistry pre-treated zinc surfaces. Again, defect zones free of conversion layers were covered with highly viscous droplets of sodium chloride solution doped with agar (for the composition of the sample see the schematic in Fig. 6). Exposed to humid air, these droplets started to spread along the surfaces of the adjacent pre-treatment films. Fig. 6 presents two photos taken after one and sixteen days. The wetting progress on the bare Zn surface is clearly visible by eye and can be identified as a dark grey area on the left next to a brighter, non-wetted surface zone on the right. Such electrolyte spreading process on zinc results in an
alkalization of the liquid/solid interface and is determined by O2 reduction kinetics, because no spreading occurs if the driving forces for oxygen reduction are minimized by a decreased atmospheric O2 partial pressure [11]. Therefore, the progress of electrolyte transport in humid air is expected to depend on the electrochemical activity of the substrate surfaces. Fig. 6 shows that liquid spreading occurred on the conversion chemistry pre-treated samples, as well. Moreover, a characteristic ranking resulted, which is also confirmed by Fig. 7: Zinc electro galvanized steel processed with solution (a) seems to induce relative fast spreading kinetics, as the position of the electrolyte front is relatively close to that on bare Zn. The sample pre-treated according to procedures (b) and (d) exhibited best performance and therefore slowest spreading kinetics, whereas the surface processed with pre-treatment (c) occurred to be electrochemically more active. Adjacent to the defect, voluminous corrosion products formed in the electrolyte wetting zone (see, in particular, Fig. 7). It was shown that they mainly consist of oxide, hydroxide, and carbonate species of zinc [16,21,23]. Such deposits are due to the amphoteric properties of Zn at high pH values, which are achieved in the oxygen reduction zone at and near the front of electrolyte wetting. Zinc takes part in a dissolution/precipitation mechanism and reacts with atmospheric CO2 dissolved in the electrolyte film [16,23,29]. Fig. 7 shows that shape, size, and distribution of deposited particles seem to vary between the samples. However, no qualitative or quantitative correlations between corrosion products and the compositions of the conversion layers were verifiable, presumably due to the complexity of the degradation processes. Therefore, further investigations focused on the kinetics and mechanisms of the electrolyte transport.
292
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
Fig. 7. Photography of conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel sheets taken after four days of exposure to humid air. Reactive electrolyte spreading was initiated after a droplet of highly viscous 0.5 molar NaCl solution was deposited on the samples.
Fig. 8 presents SKP line scans of Volta potential differences recorded during the initial stage of the reactive electrolyte wetting experiment reflected by Fig. 7. For all samples, the geometry of the potential profiles exhibit characteristics that are typically associated with cathodic delamination. A potential of −600 mVSHE or below is detected in sample areas affected by electrolyte wetting due to ongoing oxygen reduction kinetics [11,21–23,29]. These zones are located at the left adjacent to the electrolyte droplet covered reservoir zone. Dry sample areas possess more anodic potentials of −500 mVSHE or higher, because O2 reduction is kinetically inhibited. This results in steady state conditions at high anodic overpotentials [18–23]. In fact, the potential
difference (ΔE) between areas of inhibited and ongoing oxygen reduction varies between 400 mV (sample “zinc + pre-treatment (a)” and “zinc + pre-treatment (c)”) and 700 mV (for sample “zinc + pre-treatment (d)). As a consequence, sigmoid potential profiles are detected and the lateral translation of their inflection point is associated with the progress of electrolyte wetting [18–23]. This means that reactive electrolyte transport proceeds by more than 12500 μm during the initial 2.75 h on bare zinc and only about 1300 μm during the first 20.5 h on the sample “zinc + pre-treatment (d)”. As expected, Fig. 8 indeed shows that all applied conversion chemistry treatment procedures distinctly reduced the potential of the zinc substrates [30]. In contrast, no obvious correlation between ΔE and the speed of electrolyte wetting is observed although ΔE is commonly expected to reflect the driving force for electrochemically driven ion transport/cathodic delamination on Fe and Zn substrates [18–23]. This especially applies to sample “zinc + pre-treatment (d)”, which exhibits exceptionally low spreading kinetics combined with a large ΔE value during the first day of exposure to humid air. After 4 days, however, inhibition of the electrolyte transport released and its progress is almost equivalent to that of sample “zinc + pre-treatment (b)” (compare Fig. 7 with Fig. 8). Such behavior was typically observed on one of five differently pre-treated substrates of a sample series. It indicates that a ranking should not be defined within the initiation period of the electrolyte spreading process. Except for sample “zinc + pre-treatment (d)”, the results ranking obtained from Fig. 8 after 1230 min exactly reflects the conclusions drawn from analysis of Fig. 7. This confirms that visual inspection is sufficient in the present case to adequately assess the kinetics of electrolyte transport processes along the substrate surfaces, especially as it was carefully checked before that visual detection of the electrolyte front generally tracks with SKP detection. Moreover, the ranking obtained from liquid spreading on non-polymer coated, but conversion chemistry pre-treated samples correlates with that observed after weathering of epoxy/conversion layer/zinc interfaces (compare Figs. 4 and 5) with one major exception: Samples pre-treated according to procedure (a) fail in the electrolyte spreading experiment (see also Fig. 1a), but perform well in the cathodic delamination test (see also Fig. 1b). This implies that the influence of some of the parameters which determine the stability of polymer/conversion layer/metal interfaces is predictable based on short-time surface wetting tests. To check whether this finding is associated with equivalent mechanisms of cathodic delamination and electrolyte transport on conversion layer coated zinc substrates, ion profiles were recorded on the substrate surfaces after the wetting experiments. Fig. 9 exemplarily presents two of them obtained by elemental ToF-SIMS surface analysis on samples pre-treated by procedure (a) and (c). In general, it is known that preferentially cations of the defect electrolyte are transported into the cathodic delamination or electrolyte wetting zone [10–12,16–23,29]. As sodium chloride is a common contaminant on most surfaces, it can complicate the differentiation between salt amounts resulting from ion transport processes and signals arising from the background contamination [10,11]. Therefore, the wetting experiments analyzed in Fig. 9 were performed with KBr solution. After liquid spreading was stopped, ToF-SIMS line scans were recorded along the path of the electrolyte transport. In both diagrams, potassium as well as bromide were verifiable in the electrolyte reservoir zone, whereas almost only K + was detectable in the area of electrolyte spreading. It is unclear whether the reduced potassium signal between x ≈ 14 mm and x ≈ 18 mm in Fig. 9a results from a local anode following up a local cathode at the electrolyte front, as it was supposed by Fürbeth et al. [21–23], or may be simply due to a locally inhomogeneous distribution of corrosion precipitates. Anyhow, Fig. 9b confirms that neither K +, nor Br - were present in nonwetted surface sections. In general, the ToF-SIMS data are in agreement with the assumption that reactive spreading along conversion
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
293
Fig. 8. SKP potential profiles recorded on the samples shown in Fig. 7 during the initial stage of the electrolyte spreading experiment. Ion transport proceeds from the left to the right.
chemistry layers is mainly determined by electrochemical processes [11,16,29]. 3.3. Electrochemical activity & barrier properties of conversion chemistry layers To better understand the electrochemical properties of conversion chemistry pre-treated zinc samples, polarization curves were recorded employing a mildly alkaline borate buffer. Such electrolyte was used before to adequately simulate the chloride deficient, alkaline environment at cathodically delaminated epoxy/conversion layer/zinc interfaces [30]. Fig. 10 presents Tafel plots for bare Zn as well as for substrates that were subject to pre-treatment procedures (a), (b) or (d). The graphs show that the free corrosion potentials/open circuit potentials for zinc and sample “Zn + pre-treatment (a)” with - 0.61 VSHE and - 0.59 VSHE are similar. The same applies to the samples processed according to procedure (b) and (d), because their open circuit potential is approx. - 0.78 VSHE. In agreement with previous reports and with the SKP data presented in Fig. 8, the conversion chemistry treatment resulted in a cathodic potential shift of at least 160 mV [30]. Moreover, the recorded current densities are reduced compared to bare zinc except for sample “Zn + pre-treatment (a)” below approx. - 0.75 VSHE. As oxygen reduction kinetics are supposed to determine the progress of electrolyte spreading along conversion layer covered zinc surfaces, Fig. 11 shows cyclovoltammograms for the relevant potential range. The data confirm that very similar current densities were obtained for samples “Zn + pre-treatment (b)” and “Zn + pre-treatment (d)”. Contrary, the Zn-sample reflects the electrochemically most active substrate, whereas sample “Zn + pre-treatment (a)” shows rather intermediate reactivity and consequently possesses intermediate current densities. In general, this ranking is in agreement with the observations made during the electrolyte spreading experiment presented in Figs. 6, 7 and 8. It also correlates with the Tafel plots presented in Fig. 10. Nevertheless,
the performance of substrates pre-treated according to procedure (a) is generally not easy to estimate. These samples perform either clearly better than Zn (cathodic delamination tests, see Figs. 4–5 and cyclovoltammograms, see Fig. 11) or rather equivalent to Zn samples (reactive electrolyte spreading experiments, see Figs. 6–7 and the Tafel plots in Fig. 10). Similar applies to the barrier properties of surfaces that were processed according to procedure (a). Fig. 12 presents Bode plots recorded on bare Zn as well as on substrates that were subject to pre-treatment (a), (b) or (d). Comparison of the impedance at low frequencies, which typically reflect the pore resistance at the electrolyte/substrate interface, reveals lowest values for Zn (≈2.5 kΩ · cm2 at 0.1 Hz) and pre-treatment (a) processed samples (≈6.3 kΩ · cm2 at 0.1 Hz). In contrast, the resistance for substrates processed according to procedures (b) and (d) is distinctly higher and almost equivalent at low frequencies. It varies around 50 kΩ · cm2 at 0.1 Hz (see Fig. 12a). The corresponding phase curves for the systems “Zn + pre-treatment (b)” and “Zn + pre-treatment (d)” shown in Fig. 12b are mainly capacitatively dominated at higher frequencies, because the modulus of the phase values is at or above 80° between 100 Hz and ≈ 30 kHz. In contrast, the graphs for Zn and “Zn + pre-treatment (a)” rather reflect bell-shaped curves which are typically obtained on bare metal surfaces that are not covered by a layer providing pronounced barrier properties [13]. It means that the resistance of type (b) and type (d) pre-treatment coatings is higher than for type (a) pre-treated zinc surfaces. This finding correlates with reduced oxygen reduction kinetics (see Fig. 11), cathodically shifted open circuit potentials detected in the Tafel plots (see Fig. 10), and decelerated reactive electrolyte spreading along the conversion layer surfaces (see Figs. 6 and 7). Moreover, it indicates that the stabilizing effects of the pre-treatment approach (a) observed in Figs. 4 and 5 presumably result from contributions to e.g. adhesion promotion of covering polymer coatings rather than from a direct inhibition of O2
294
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
Fig. 11. Investigation of oxygen reduction kinetics on conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel. Polarization curves recorded at a rate of 1 mV/s from the open circuit potential (OCP) to −0.1 V vs. OCP and back to OCP are shown.
Fig. 9. ToF-SIMS line scans recorded on conversion chemistry pre-treated zinc electro galvanized steel after electrolyte spreading occurred along the substrate surfaces during their exposure to humid air (for the experimental setup please see the schematic presented in Fig. 5). To initiate reactive wetting, a droplet of highly viscous 0.5 molar KBr solution was deposited on the samples. Electrolyte spreading proceeded from the left to the right. a) K+ and Br- distribution detected after three days of exposure on a sample pre-treated according to procedure (a). b) K+ and Br- distribution detected after three days of exposure on a sample pre-treated according to procedure (c).
reduction processes on the Zn substrates. In general, this is in line with the fact that the iron oxide deposits are not supposed to form dense oxide barrier layers such as anodized Al-oxide on aluminum. Comparison of performance tendencies obtained by EIS, potentiodynamic measurements and the progress of electrolyte spreading with the kinetics of cathodic delamination implies certain, but also limited predictability of polymer/conversion layer/zinc interface stabilities. It shows that simple, quick and cost-efficient wetting experiments can be a useful option to roughly estimate the corrosion protective properties of conversion chemistry based pre-treatment approaches. No sophisticated equipment such as a potentiostat is necessary and just a sample chamber with a humid atmosphere would be needed. To reliably identify outlier in the performance tendencies obtained from the exposure of a sample series, however, reproduction experiments are definitely required. For the systems investigated in the present study, the results of short-term electrolyte spreading anticipate long-term cathodic delamination experiments. This does not seem to be an adequate strategy for pre-treatments that do not result in a distinct reduction of the electrochemical activity of the substrate, such as systems that generate iron oxide precipitates on a zinc surface. In this context, previous studies reported that tailored iron oxide morphologies indeed specifically determine parameters such as the bonding of adhered organic species. It was shown that cathodic delamination along the polymer coated Fe-oxide surfaces accelerated with decreasing surface hydrophilicity, which in turn was connected to reduced polymer/substrate adhesion forces. In contrast, a different ranking resulted for reactive electrolyte spreading experiments along non-polymer coated iron oxide surfaces [12]. This is similar to what was observed on samples that were subject to pre-treatment (a) (compare with Figs. 4–5 and 6–7). Although a direct comparison between the study of Wielant et al. [12] and the present publication cannot be drawn based on the available data, such findings may nevertheless illustrate that adhesion forces between a polymer layer and the metal substrate can determine cathodic delamination kinetics if the metal surface is not additionally covered by a thin film that contributes relevant barrier properties. 4. Conclusion
Fig. 10. Tafel plots resulting from polarization curves recorded (from cathodic to anodic potentials at a rate of 1 mV/s) on conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel sheets.
The corrosion resistance and the electrochemical properties of zinc surfaces pre-treated with different conversion chemistry solutions were investigated. Iron/iron oxide layers as well as organic/inorganic
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
295
pre-treatment layers with barrier properties as low as those of the native oxide film on Zn. In the present study, deposits of iron oxide hardly decelerated electrolyte spreading, but distinctly increased the epoxy/conversion layer/zinc interface stability as indicated by cathodic delamination studies. It was discussed that this may be attributed to, for example, improved adhesion of the polymer on Fe-oxide rather than to a direct inhibition of O2 reduction processes at the Fe-oxide covered Zn surface.
Acknowledgments The financial support of the Christian-Doppler-Research Association (Vienna, Austria), voestalpine Stahl Linz GmbH (Linz, Austria), and Henkel AG & Co. KGaA (Düsseldorf, Germany) is gratefully acknowledged. The authors would like to thank K.-H. Stellnberger and M. Fleischanderl for helpful discussions.
References [1] [2] [3] [4] [5] [6] [7]
[8]
[9]
[10] Fig. 12. Bode plots resulting from EIS measurements that were performed on conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel sheets.
[11] [12]
hybrid films based on precipitates of zinc phosphate and Zr-/Ti-oxide embedded in an organic network were formed. An increased resistance of the conversion coatings as measured by means of EIS at low frequencies correlated with reduced oxygen reduction current densities, with cathodically shifted open circuit potentials and with decelerated electrolyte spreading kinetics along the sample surfaces. It was shown that electrolytic wetting of the pre-treatment layers, which starts from an electrolyte reservoir where active dissolution of Zn takes place, is mainly determined by electrochemical oxygen reduction processes. This was proven by the preferential transport of cations to the front of wetting, which is also characteristic for corrosive delamination of organic films from metals covered by semi conductive oxide layers. Moreover, the kinetics of short-term electrolyte spreading processes reflected those obtained from long-term cathodic delamination experiments of epoxy coated conversion films. This indicated that cost-efficient wetting experiments can be a complementary tool for accelerated corrosion testing of surface modified zinc substrates. However, the approach did not work for thin
[13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]
T. Prosek, A. Nazarov, J. Stoulil, D. Thierry, Corros. Sci. 61 (2012) 92. A. Nazarov, D. Thierry, P. Volovitch, K. Ogle, Surf. Interface Anal. 43 (2011) 1286. A.J. Kinloch, Adhesion and Adhesives, Chapman & Hall, London, 1987. A. Rubina, R. Oltra, B. Vuillemin, K. Ogle, Electrochim. Acta 53 (2008) 6484. G. Williams, H.N. McMurray, M.J. Loveridge, Electrochim. Acta 55 (2010) 1740. K.N. Allahar, B.R. Hinderliter, D.E. Tallman, G.P. Bierwagen, J. Electrochem. Soc. 155 (2008) F201. K. Hackbarth, M. Wolpers, W. Lorenz, P. Kuhm, K. Meagher, C. Rosenkranz, M. Roth, R. Wark, G. Sanchis Otero, E. Wilke, “Metallisierende Vorbehandlung von Zinkoberflächen“, European patent no. EP 2 292 808 A1. N. Kanani, Chemische Vernicklung: Nickel-Phosphor-Schichten. Herstellung, Eigenschaften, Anwendungen. Ein Handbuch für Theorie und Praxis, E.G. Leuze Press, Bad Saulgau/Germany, ISBN: 3874802299, 2007. In: S.D. Carmer, B.S. Covino (Eds.), ASM Handbook, Volume 13A; Corrosion: Fundamentals, Testing, and ProtectionASM International, Materials Park OH/USA, ISBN: 0871707055, 2003. R. Posner, K. Wapner, M. Stratmann, G. Grundmeier, Electrochim. Acta 54 (2009) 891. R. Posner, T. Titz, K. Wapner, M. Stratmann, G. Grundmeier, Electrochim. Acta 54 (2009) 900. J. Wielant, R. Posner, R. Hausbrand, G. Grundmeier, H. Terryn, Corros. Sci. 51 (2009) 1664. R. Posner, K. Wapner, S. Amthor, K.J. Roschmann, G. Grundmeier, Corros. Sci. 52 (2010) 37. R. Posner, M. Marazita, S. Amthor, K.J. Roschmann, Corros. Sci. 52 (2010) 750. R. Posner, G. Giza, R. Vlasak, G. Grundmeier, Electrochim. Acta 54 (2009) 4837. Z.Y. Chen, D. Persson, C. Leygraf, Corros. Sci. 50 (2008) 111. Z.Y. Chen, D. Persson, A. Nazarov, S. Zakipour, D. Thierry, C. Leygraf, J. Electrochem. Soc. 152 (2005) B342. A. Leng, H. Streckel, M. Stratmann, Corros. Sci. 41 (1999) 547. A. Leng, H. Streckel, M. Stratmann, Corros. Sci. 41 (1999) 579. A. Leng, H. Streckel, K. Hofmann, M. Stratmann, Corros. Sci. 41 (1999) 599. W. Fürbeth, M. Stratmann, Corros. Sci. 43 (2001) 207. W. Fürbeth, M. Stratmann, Corros. Sci. 43 (2001) 229. W. Fürbeth, M. Stratmann, Corros. Sci. 43 (2001) 243. R. Posner, G. Giza, M. Marazita, G. Grundmeier, Corros. Sci. 52 (2010) 1838. S.E. Dolan, “Composition and process for treating metals“, US patent no. US5427632. S. Adhikari, K.A. Unocic, Y. Zhai, G.S. Frankel, J. Zimmerman, W. Fristad, Electrochim. Acta 56 (2011) 1912. F.C. Walsh, J. Phys. D Appl. Phys. 24 (1991) 217. D. Vasilakopoulos, M. Bouroushian, M. Spyrdlis, Electrochim. Acta 54 (2009) 2509. A.K. Neufeld, I.S. Cole, A.M. Bond, S.A. Furman, Corros. Sci. 44 (2002) 555. G. Klimow, N. Fink, G. Grundmeier, Electrochim. Acta 53 (2007) 1290.
Contents lists available at SciVerse ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Electrochemical electrolyte spreading studies of the protective properties of ultra-thin films on zinc galvanized steel R. Posner a,⁎, N. Fink a, M. Wolpers b, G. Grundmeier c a b c
Max-Planck-Institut für Eisenforschung GmbH, Department of Interface Chemistry and Surface Engineering, Max-Planck-Str. 1, 40237 Düsseldorf, Germany Henkel AG & Co. KGaA, Henkelstr. 67, 40589 Düsseldorf, Germany University of Paderborn, Department of Technical and Macromolecular Chemistry, Warburger Str. 100, 33098 Paderborn, Germany
a r t i c l e
i n f o
Article history: Received 5 February 2013 Accepted in revised form 18 April 2013 Available online 25 April 2013 Keywords: Conversion coating Reactive electrolyte spreading Accelerated corrosion testing Cathodic delamination Zinc galvanized steel
a b s t r a c t Reactive electrolyte spreading along the surfaces of different conversion films on zinc galvanized steel in humid air was monitored visually and with a height-regulated scanning Kelvin Probe. Electrochemical impedance spectroscopy and current density-potential curves revealed that decelerated spreading kinetics are connected with increasing pore resistances of the pre-treatment layers and decreasing oxygen reduction current densities in the electron transfer controlled potential region. After a few days the progress ranking of electrolyte spreading along uncoated conversion films reflected the progress tendencies of cathodic delamination observed on epoxy coated conversion layers after long-time exposure to the same corrosive environment. Such correlation was not discovered for pre-treatment films that do not provide relevant electrochemical barrier properties. The results suggest that oxygen reduction driven electrolyte wetting is an option for accelerated performance testing of anticorrosive ultra-thin films on metal substrates that can be subject to cathodic delamination. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Adequate characterization of polymer coated metal compounds modified by adhesion promoting and corrosion protective thin films remains a challenge due to the limited availability of analytical approaches for buried interfaces. The residual electrochemical activity of the pre-treated substrate surface, barrier properties of the organic coating and its adhesion as well as molecular and morphologic characteristics of the organic/inorganic interface structure synergistically determine the corrosion resistance [1–6]. Recently launched two-step pretreatment procedures such as the deposition of amorphous iron oxide prior to the formation of Zr- or Ti-based barrier layers on zinc were developed to offer superior performance for specific applications. However, they also result in complex interface designs which require even more detailed understanding and careful analysis [7,8]. In general, standardized salt spray exposure and long-term weathering tests have to be passed to achieve approval for the use of conversion chemistry based protection systems e.g. at car bodies or building facades [9]. As such procedures are time-consuming, industrial research will especially benefit from short-term and cost-efficient routines to quickly review progress in daily product development. In this context, the comparison of ion transport processes at polymer/metal interfaces seems to offer promising ⁎ Corresponding author at: Henkel AG & Co. KGaA, Henkelstrasse 67, 40589 Düsseldorf, Germany. Tel.: +49 211 797 2374; fax: +49 2133 537379. E-mail address: [email protected] (R. Posner). 0257-8972/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.surfcoat.2013.04.042
options for accelerated testing and interface characterization. It was observed that the ion mobility sensitively responds to the electrochemical activity of the substrate material and its oxide properties [10–12], to the permeability of the polymer with respect to water and oxygen [13,14], to paint adhesion as well as to mechanical properties of its macromolecular network [13–15]. Moreover, previous publications reported that transport mechanisms detected at polymer covered zinc and steel substrates resemble those observed on uncoated Zn and Fe surfaces during the exposure to humid air [10,11]. Ion convection starts from electrolyte droplets deposited/formed on metal surfaces or from electrolyte covered defects in the polymer coating that extend to the metal substrate. It is initiated by the reduction of atmospheric oxygen and the formation of hydroxide species in the periphery of the electrolyte reservoir. A flow of cations from the reservoir centre, where active dissolution of the base material occurs, toward the cathodic sites of the local galvanic cell then ensures charge compensation. This promotes an extension of the oxygen reduction zone. For droplet spreading along the substrate surface such finding was explained by a reduction of the oxide/electrolyte interface tension with increasing pH above the isoelectric point of the oxide due to OH formation [16,17]. Cation ingress into polymer/oxide/metal interface sections adjacent to an electrolyte covered coating defect occurs preferentially on zinc, exclusively on iron and is commonly regarded as an important part of the cathodic delamination mechanism [18–23]. Fig. 1 provides a schematic illustration of relevant ion transport processes for metal substrates with and without organic coating. In-situ monitoring of electrode potentials at electrolyte/metal and polymer/metal interfaces
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
287
Fig. 1. Schematic illustration of electrolyte transport mechanisms on zinc and iron substrates in humid air (relative humidity: typically > 90%). a) Reactive electrolyte spreading along the bare metal surfaces, starting at a highly viscous droplet of NaCl electrolyte. b) Cathodic delamination along a polymer/oxide/metal interface, starting at a coating defect covered with NaCl electrolyte. Please be aware that the O2 and H2O activity in the polymer phase will be significantly lower than in humid air.
by the Scanning Kelvin Probe (SKP) and subsequent investigation of elemental distributions resulting from ion transport reflect established state-of-the-art approaches for analysis [10–12,18–23]. Stability rankings for different interface designs on iron and steel samples were indeed accessible by cathodic delamination studies [13,14], but occurred to be problematic for zinc substrates so far [14,24]. The present publication consequently focuses on the introduction of an analogous approach for conversion chemistry pre-treated zinc electro galvanized steel. Electrolyte spreading was initiated on pre-treatment layers with different barrier properties. The results were compared to the characteristics of polarization curves as well as impedance spectra recorded on these anticorrosive films and correlated with cathodic delamination rankings obtained
from ion transport studies at epoxy/conversion layer/zinc interfaces. These data will demonstrate the practicability as well as the limits of the electrolyte spreading approach for accelerated performance testing of Zr-/Ti-based conversion chemistry layers. 2. Experimental 2.1. Sample preparation Zinc electro galvanized steel sheets with a coating thickness of approx. 7.5 μm were supplied by voestalpine Stahl Linz GmbH (Linz, Austria). After alkaline spray cleaning with a solution that contained 2% Ridoline 1250 BR and 0.2% Ridosol 1270 (from Henkel AG & Co.
Fig. 2. SEM images of sample surfaces used for experiments discussed in the present publication. a) and b): Zinc electro galvanized steel. c) and d): Zinc electro galvanized steel pre-treated according to procedure (a).
288
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
Fig. 3. SEM images of sample surfaces used for experiments discussed in the present publication. a) and b): Zinc electro galvanized steel pre-treated according to procedure (b). c) and d): Zinc electro galvanized steel pre-treated according to procedure (d).
KGaA, Düsseldorf, Germany) for 15 s at 50 °C and a pressure of 1 bar, they were tap water, then DI water rinsed and dried in warm air. Subsequent pre-treatment with conversion chemistry (patented by Henkel AG & Co. KGaA, Düsseldorf, Germany) was performed according to one of the following procedures:
layer mainly consisted of Ti-oxide species and of zinc phosphate embedded in an organic network [25]. (c) With an aqueous solution of a composition similar to that described for procedure (b) except that the titanium component was partially substituted by a fluoric acid of zirconium. Roll coating led to an amorphous organic-inorganic hybrid film with a coating weight of ~ 7 mg/m 2 for Ti and of ~ 3 mg/m 2 for Zr. The conversion layer mainly consisted of Ti-oxide species, of Zr-oxide and of zinc phosphate embedded in an organic network [25]. (d) A two-step pre-treatment approach including procedure (a) followed by a rinse step with DI water and subsequently by procedure (b).
(a) With an aqueous, acidic solution for electroless iron deposition based on a Fe2+-containing inorganic compound, a reduction agent, and chemical mediators as major components. 10 s of sample immersion at 50 °C resulted in the precipitation of porous and amorphous deposits of iron, of Fe(II)-, and of Fe(III)-compounds at a coating weight of ~40 mg/m2 for Fe and of ~10 mg/m2 for P [7]. (b) With an aqueous solution containing a fluoric acid of titanium, a derivative of a polymeric vinyl phenol and a phosphate containing inorganic reagent as major components. Roll coating at a speed of 60 m/min and a quench pressure of 1 kN at room temperature resulted in an amorphous organic-inorganic hybrid film with a Ti coating weight of ~ 10 mg/m 2. The conversion
After rinsing with DI water and drying in a nitrogen stream, the conversion chemistry pre-treated samples were covered with a ~ 90 μm thick layer of a hot-curing two-component epoxy resin provided by Henkel AG & Co. KGaA, Düsseldorf, Germany. An area of ~ 1 × 1 cm was omitted that was supposed to be used as an
Table 1 Overview of the elemental composition detected by EDS on bare zinc electro galvanized steel and on zinc electro galvanized steel pre-treated according to procedure (a), (b) or (d).
Table 2 Overview of the elemental composition detected by XPS on bare zinc electro galvanized steel and on zinc electro galvanized steel pre-treated according to procedure (a), (b) or (d).
Relative amount [%] →
C
O
Zn
P
Fe
Ti
F
Relative amount [%] →
C
O
Zn
P
Fe
Ti
Mn
N
F
Zn Zn + pre-treatment (a) Zn + pre-treatment (b) Zn + pre-treatment (d)
3.0 3.9 16.9 15.2
2.2 6.4 7.4 10.5
94.8 81.5 68.4 61.4
– 1.9 1.1 3.0
– 6.3 – 3.8
– – 1.0 0.9
– – 4.7 5.2
Zn Zn + pre-treatment (a) Zn + pre-treatment (b) Zn + pre-treatment (d)
12.9 24.3 59.8 52.9
56.4 51.6 26.2 30.1
30.7 18.5 1.5 2.5
– 1.9 1.5 1.5
– 3.7 – 0.2
– – 3.0 3.6
– – 1.0 1.4
– – 1.3 1.1
– – 5.7 7.0
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
289
Fig. 4. Photography of a set of epoxy coated, conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel samples taken after the exposure to humid air. Cathodic delamination was initiated after an artificially prepared coating defect was covered with 0.5 molar NaCl solution of high viscosity. a) Progress of cathodic delamination after one month of exposure. Two black lines observed on the zinc reference sample refer to the front of macroscopic polymer de-adhesion (on the left) and to the front of interfacial electrolyte transport (on the right). On all the other samples, both fronts are equivalent. b) Progress of cathodic delamination after two months of exposure. The electrolyte front is indicated by solid black lines.
artificial coating defect. In this area, also the conversion coating was removed by grinding. The polymer was hardened for 1 h at 120 °C applying a pressure of 50 g/cm 2 [10]. For the weathering experiments, the defect area of the samples was covered with highly viscous 0.5 molar NaCl or 0.5 molar KBr solution. To ensure high viscosity, ~ 3% of agar was added to the liquid prior to heating to ~ 70 °C. Subsequent cooling then resulted in an almost ‘solid’ electrolyte solution. Weathering exposure was performed in humid air (with a relative humidity of ≥95%) at room temperature. All used chemicals and solvents were of analytical grade. 2.2. Electrochemical and surface analytical methods Scanning Kelvin Probe (SKP) line scans were recorded with a height-regulated, custom-made SKP apparatus in humid air (with a relative humidity of ≥ 95%) at room temperature. Detected Volta potential differences were converted to the standard hydrogen electrode (SHE) scale after calibration against Cu/CuSO4 [10].
Electrochemical measurements in three-electrode arrangement were done with a Gamry FAS2 Femtostat, an Ag/AgCl/3 molar KCl reference electrode, a Pt counter electrode, and chloride-free borate buffer solution with a pH of 8.4. Electrochemical impedance spectroscopy (EIS) experiments were performed in a frequency range of 0.1 Hz to 100 kHz with an amplitude of 10 mV at open circuit potential of the investigated sample system. X-ray photoelectron spectroscopy (XPS) analysis was conducted with a Quantum 2000 (Physical Electronics, Chanhassen, MN/USA). Spectra were measured with a spot size of 100 μm × 100 μm at 45° using monochromated Al Kα radiation, 23.9 eV pass energy, and a step size of 0.2 eV for high resolution spectra. Time-of-flight secondary ion mass spectrometry (ToF-SIMS) measurements were executed with a PHI TRIFT CE (Physical Electronics, Chanhassen, MN/USA) applying a gallium ion gun at an acceleration voltage of 15 kV and a spot size of 100 μm × 100 μm. Scanning electron microscopy (SEM) images were performed with a LEO 1550VP Field Emission SEM apparatus (Leo Elektronenmikroskopie
290
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
Fig. 5. Photography of a set of epoxy coated, conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel samples taken after 21 days of exposure to humid air. Cathodic delamination was initiated after an artificially prepared coating defect was covered with 0.5 molar NaCl solution of high viscosity. Two black lines observed on the zinc reference sample refer to the front of macroscopic polymer de-adhesion (on the left) and to the front of interfacial electrolyte transport (on the right). On all the other samples, both fronts are equivalent.
features. Figs. 2 and 3 present images recorded on bare zinc electro galvanized steel (Fig. 2a–b), on zinc electro galvanized steel pre-treated according to procedure (a) (see Fig. 2c–d), procedure (b) (see Fig. 3a–b), or procedure (d) (see Fig. 3c–d). Non-pre-treated surfaces exhibit structures of stacked and terraced zinc plates. They are grouped in differently oriented domains, which arise from crystallographically preferred growth directions of electro deposited Zn crystallization cores on grains of the underlying steel surface. Sharp and well-defined edges of the zinc plates with characteristic angles resulted (see Fig. 2a and b) [27,28]. Immersion in the solution for electroless iron deposition led to coverage of the Zn plates, which visually also manifests in grading of their edges. However, vacancies along the plate boundaries partially remain. Their diameters seem to depend on the size of interstice volumes between differently oriented plate domains (see Fig. 2c and d). In contrast, pre-treatment procedure (b) causes visually lesspronounced modifications of the sample surfaces. Deposits resulting from conversion processes are hardly verifiable even at high magnification in Fig. 3b, but the edges at the boundaries of the Zn plates are slightly glazed. This can be interpreted as an indicator for the presence of a very thin conversion film that occurs to be almost transparent when investigated by SEM. Fig. 3c and d illustrate the appearance of the substrate surface after the two-step pre-treatment (d) was performed. Deposits on the zinc plates are clearly visible and more pronounced than those observed in Fig. 2c and d. The original morphology of the Zn surface is still identifiable; however, the grading effect resulting from procedure (a) seems to be amplified. Tables 1 and 2 provide an overview of elements detected on pre-treated and non-pre-treated zinc electro galvanized steel. The Zn portion is mainly attributed to the base material, because it occurs to be reduced after the deposition of a conversion layer. Compared to the zinc reference surface, a strong increase of the carbon percentage refers to the formation of an organic network after the samples were processed according to procedure (b). Iron is indeed verifiable if procedure (a) was part of the pre-treatment process. However, its amount is reduced after sample immersion in conversion bath (b), presumably partly due to coverage with organic and inorganic Ti-, P- and Mn-species and partly due to re-dissolution of Fe at low pH. The relative percentage values obtained with EDS may not be directly compared to those resulting from XPS analysis, because XPS is by far more surface sensitive with a penetration depth of only a few nanometers compared to several micrometers for EDS. The XPS data consequently indicate an enrichment of trace elements near the surface of the deposited conversion layers. Therefore, Tables 1 and 2 confirm that almost all major ingredients of pre-treatment solutions (a) and (b) described in chapter 2 are verifiable on the samples, as well. Moreover, Figs. 2 and 3 clearly show that the different pre-treatment approaches also result in a visually distinguishable surface appearance of zinc electro galvanized steel. 3.2. Weathering of and electrolyte spreading on conversion layers
GmbH, Oberkochen, Germany) using an InLens detector in high vacuum mode. Energy-dispersive X-ray spectroscopy (EDS) analysis was made with an INCA-7426 detector from Oxford Instruments (Oxford, UK) that was connected to the SEM machine. Both SEM images and EDS spectra were recorded with an electron high tension voltage of 10 kV and a working distance of 10 mm. 3. Results and discussion 3.1. Characterization of conversion chemistry layers The pre-treatment procedures described in chapter 2 resulted in the deposition of conversion chemistry layers on the substrate surfaces with a thickness of typically some ten nanometers [26]. Visual inspection by electron microscopy revealed characteristic surface
In order to investigate their corrosion resistance, conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel sheets were exposed to humid air after they were coated with an epoxy layer. A defect zone remained uncoated and was filled up with highly viscous 0.5 molar NaCl solution to initiate cathodic delamination. Fig. 1b reflects a schematic cross-section view of the selected sample composition. Photos were taken after one and two months to reveal the progress of electrolyte transport along the epoxy/ZnO/Zn- and the epoxy/conversion layer/Zn-interfaces. Fig. 4 presents the resulting images. After one month, a characteristic stability ranking was established, which did not change during the following weeks (see Fig. 4b). Visible delamination became manifest in the formation of blisters in the coating and in electrolyte droplets evolving in the vicinity of
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
291
Fig. 6. Photography of conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel sheets taken after the exposure to humid air. Reactive electrolyte spreading was initiated after a droplet of highly viscous 0.5 molar NaCl solution was deposited on the samples. a) Progress of electrolyte spreading after one day of exposure. b) Progress of electrolyte spreading after 16 days of exposure.
the defect zone. The respective front of cathodic delamination is highlighted by black lines in Fig. 4. They indicate that corrosive degradation proceeded slightly inhomogeneous. However, considering the integral coating area affected by cathodic undermining, it is clear that the untreated zinc sample delaminates fastest, followed by zinc pre-treated with solution (c). Samples (a) and (d) that were both immersed in conversion chemistry bath (a) do not only exhibit a darker surface, but also show almost equivalent performance. Based on Fig. 4, procedure (b) seems to be superior to the other pre-treatment procedures, but this finding was not generally verifiable. Fig. 5 presents a photo of another polymer coated sample batch that was subject to corrosive delamination for 21 days. A similar ranking was observed, but sample “zinc + pre-treatment (b)” did not show the best performance in this case. For a better differentiation, the formation of blisters as a result of the delamination process may be used an additional distinguishing feature: Electrolyte transport resulted in very few, but large blisters on sample (b) compared to a pattern of numerous little ones on samples (a) and (d). A different type of electrolyte transport experiment was performed with non epoxy coated, but conversion chemistry pre-treated zinc surfaces. Again, defect zones free of conversion layers were covered with highly viscous droplets of sodium chloride solution doped with agar (for the composition of the sample see the schematic in Fig. 6). Exposed to humid air, these droplets started to spread along the surfaces of the adjacent pre-treatment films. Fig. 6 presents two photos taken after one and sixteen days. The wetting progress on the bare Zn surface is clearly visible by eye and can be identified as a dark grey area on the left next to a brighter, non-wetted surface zone on the right. Such electrolyte spreading process on zinc results in an
alkalization of the liquid/solid interface and is determined by O2 reduction kinetics, because no spreading occurs if the driving forces for oxygen reduction are minimized by a decreased atmospheric O2 partial pressure [11]. Therefore, the progress of electrolyte transport in humid air is expected to depend on the electrochemical activity of the substrate surfaces. Fig. 6 shows that liquid spreading occurred on the conversion chemistry pre-treated samples, as well. Moreover, a characteristic ranking resulted, which is also confirmed by Fig. 7: Zinc electro galvanized steel processed with solution (a) seems to induce relative fast spreading kinetics, as the position of the electrolyte front is relatively close to that on bare Zn. The sample pre-treated according to procedures (b) and (d) exhibited best performance and therefore slowest spreading kinetics, whereas the surface processed with pre-treatment (c) occurred to be electrochemically more active. Adjacent to the defect, voluminous corrosion products formed in the electrolyte wetting zone (see, in particular, Fig. 7). It was shown that they mainly consist of oxide, hydroxide, and carbonate species of zinc [16,21,23]. Such deposits are due to the amphoteric properties of Zn at high pH values, which are achieved in the oxygen reduction zone at and near the front of electrolyte wetting. Zinc takes part in a dissolution/precipitation mechanism and reacts with atmospheric CO2 dissolved in the electrolyte film [16,23,29]. Fig. 7 shows that shape, size, and distribution of deposited particles seem to vary between the samples. However, no qualitative or quantitative correlations between corrosion products and the compositions of the conversion layers were verifiable, presumably due to the complexity of the degradation processes. Therefore, further investigations focused on the kinetics and mechanisms of the electrolyte transport.
292
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
Fig. 7. Photography of conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel sheets taken after four days of exposure to humid air. Reactive electrolyte spreading was initiated after a droplet of highly viscous 0.5 molar NaCl solution was deposited on the samples.
Fig. 8 presents SKP line scans of Volta potential differences recorded during the initial stage of the reactive electrolyte wetting experiment reflected by Fig. 7. For all samples, the geometry of the potential profiles exhibit characteristics that are typically associated with cathodic delamination. A potential of −600 mVSHE or below is detected in sample areas affected by electrolyte wetting due to ongoing oxygen reduction kinetics [11,21–23,29]. These zones are located at the left adjacent to the electrolyte droplet covered reservoir zone. Dry sample areas possess more anodic potentials of −500 mVSHE or higher, because O2 reduction is kinetically inhibited. This results in steady state conditions at high anodic overpotentials [18–23]. In fact, the potential
difference (ΔE) between areas of inhibited and ongoing oxygen reduction varies between 400 mV (sample “zinc + pre-treatment (a)” and “zinc + pre-treatment (c)”) and 700 mV (for sample “zinc + pre-treatment (d)). As a consequence, sigmoid potential profiles are detected and the lateral translation of their inflection point is associated with the progress of electrolyte wetting [18–23]. This means that reactive electrolyte transport proceeds by more than 12500 μm during the initial 2.75 h on bare zinc and only about 1300 μm during the first 20.5 h on the sample “zinc + pre-treatment (d)”. As expected, Fig. 8 indeed shows that all applied conversion chemistry treatment procedures distinctly reduced the potential of the zinc substrates [30]. In contrast, no obvious correlation between ΔE and the speed of electrolyte wetting is observed although ΔE is commonly expected to reflect the driving force for electrochemically driven ion transport/cathodic delamination on Fe and Zn substrates [18–23]. This especially applies to sample “zinc + pre-treatment (d)”, which exhibits exceptionally low spreading kinetics combined with a large ΔE value during the first day of exposure to humid air. After 4 days, however, inhibition of the electrolyte transport released and its progress is almost equivalent to that of sample “zinc + pre-treatment (b)” (compare Fig. 7 with Fig. 8). Such behavior was typically observed on one of five differently pre-treated substrates of a sample series. It indicates that a ranking should not be defined within the initiation period of the electrolyte spreading process. Except for sample “zinc + pre-treatment (d)”, the results ranking obtained from Fig. 8 after 1230 min exactly reflects the conclusions drawn from analysis of Fig. 7. This confirms that visual inspection is sufficient in the present case to adequately assess the kinetics of electrolyte transport processes along the substrate surfaces, especially as it was carefully checked before that visual detection of the electrolyte front generally tracks with SKP detection. Moreover, the ranking obtained from liquid spreading on non-polymer coated, but conversion chemistry pre-treated samples correlates with that observed after weathering of epoxy/conversion layer/zinc interfaces (compare Figs. 4 and 5) with one major exception: Samples pre-treated according to procedure (a) fail in the electrolyte spreading experiment (see also Fig. 1a), but perform well in the cathodic delamination test (see also Fig. 1b). This implies that the influence of some of the parameters which determine the stability of polymer/conversion layer/metal interfaces is predictable based on short-time surface wetting tests. To check whether this finding is associated with equivalent mechanisms of cathodic delamination and electrolyte transport on conversion layer coated zinc substrates, ion profiles were recorded on the substrate surfaces after the wetting experiments. Fig. 9 exemplarily presents two of them obtained by elemental ToF-SIMS surface analysis on samples pre-treated by procedure (a) and (c). In general, it is known that preferentially cations of the defect electrolyte are transported into the cathodic delamination or electrolyte wetting zone [10–12,16–23,29]. As sodium chloride is a common contaminant on most surfaces, it can complicate the differentiation between salt amounts resulting from ion transport processes and signals arising from the background contamination [10,11]. Therefore, the wetting experiments analyzed in Fig. 9 were performed with KBr solution. After liquid spreading was stopped, ToF-SIMS line scans were recorded along the path of the electrolyte transport. In both diagrams, potassium as well as bromide were verifiable in the electrolyte reservoir zone, whereas almost only K + was detectable in the area of electrolyte spreading. It is unclear whether the reduced potassium signal between x ≈ 14 mm and x ≈ 18 mm in Fig. 9a results from a local anode following up a local cathode at the electrolyte front, as it was supposed by Fürbeth et al. [21–23], or may be simply due to a locally inhomogeneous distribution of corrosion precipitates. Anyhow, Fig. 9b confirms that neither K +, nor Br - were present in nonwetted surface sections. In general, the ToF-SIMS data are in agreement with the assumption that reactive spreading along conversion
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
293
Fig. 8. SKP potential profiles recorded on the samples shown in Fig. 7 during the initial stage of the electrolyte spreading experiment. Ion transport proceeds from the left to the right.
chemistry layers is mainly determined by electrochemical processes [11,16,29]. 3.3. Electrochemical activity & barrier properties of conversion chemistry layers To better understand the electrochemical properties of conversion chemistry pre-treated zinc samples, polarization curves were recorded employing a mildly alkaline borate buffer. Such electrolyte was used before to adequately simulate the chloride deficient, alkaline environment at cathodically delaminated epoxy/conversion layer/zinc interfaces [30]. Fig. 10 presents Tafel plots for bare Zn as well as for substrates that were subject to pre-treatment procedures (a), (b) or (d). The graphs show that the free corrosion potentials/open circuit potentials for zinc and sample “Zn + pre-treatment (a)” with - 0.61 VSHE and - 0.59 VSHE are similar. The same applies to the samples processed according to procedure (b) and (d), because their open circuit potential is approx. - 0.78 VSHE. In agreement with previous reports and with the SKP data presented in Fig. 8, the conversion chemistry treatment resulted in a cathodic potential shift of at least 160 mV [30]. Moreover, the recorded current densities are reduced compared to bare zinc except for sample “Zn + pre-treatment (a)” below approx. - 0.75 VSHE. As oxygen reduction kinetics are supposed to determine the progress of electrolyte spreading along conversion layer covered zinc surfaces, Fig. 11 shows cyclovoltammograms for the relevant potential range. The data confirm that very similar current densities were obtained for samples “Zn + pre-treatment (b)” and “Zn + pre-treatment (d)”. Contrary, the Zn-sample reflects the electrochemically most active substrate, whereas sample “Zn + pre-treatment (a)” shows rather intermediate reactivity and consequently possesses intermediate current densities. In general, this ranking is in agreement with the observations made during the electrolyte spreading experiment presented in Figs. 6, 7 and 8. It also correlates with the Tafel plots presented in Fig. 10. Nevertheless,
the performance of substrates pre-treated according to procedure (a) is generally not easy to estimate. These samples perform either clearly better than Zn (cathodic delamination tests, see Figs. 4–5 and cyclovoltammograms, see Fig. 11) or rather equivalent to Zn samples (reactive electrolyte spreading experiments, see Figs. 6–7 and the Tafel plots in Fig. 10). Similar applies to the barrier properties of surfaces that were processed according to procedure (a). Fig. 12 presents Bode plots recorded on bare Zn as well as on substrates that were subject to pre-treatment (a), (b) or (d). Comparison of the impedance at low frequencies, which typically reflect the pore resistance at the electrolyte/substrate interface, reveals lowest values for Zn (≈2.5 kΩ · cm2 at 0.1 Hz) and pre-treatment (a) processed samples (≈6.3 kΩ · cm2 at 0.1 Hz). In contrast, the resistance for substrates processed according to procedures (b) and (d) is distinctly higher and almost equivalent at low frequencies. It varies around 50 kΩ · cm2 at 0.1 Hz (see Fig. 12a). The corresponding phase curves for the systems “Zn + pre-treatment (b)” and “Zn + pre-treatment (d)” shown in Fig. 12b are mainly capacitatively dominated at higher frequencies, because the modulus of the phase values is at or above 80° between 100 Hz and ≈ 30 kHz. In contrast, the graphs for Zn and “Zn + pre-treatment (a)” rather reflect bell-shaped curves which are typically obtained on bare metal surfaces that are not covered by a layer providing pronounced barrier properties [13]. It means that the resistance of type (b) and type (d) pre-treatment coatings is higher than for type (a) pre-treated zinc surfaces. This finding correlates with reduced oxygen reduction kinetics (see Fig. 11), cathodically shifted open circuit potentials detected in the Tafel plots (see Fig. 10), and decelerated reactive electrolyte spreading along the conversion layer surfaces (see Figs. 6 and 7). Moreover, it indicates that the stabilizing effects of the pre-treatment approach (a) observed in Figs. 4 and 5 presumably result from contributions to e.g. adhesion promotion of covering polymer coatings rather than from a direct inhibition of O2
294
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
Fig. 11. Investigation of oxygen reduction kinetics on conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel. Polarization curves recorded at a rate of 1 mV/s from the open circuit potential (OCP) to −0.1 V vs. OCP and back to OCP are shown.
Fig. 9. ToF-SIMS line scans recorded on conversion chemistry pre-treated zinc electro galvanized steel after electrolyte spreading occurred along the substrate surfaces during their exposure to humid air (for the experimental setup please see the schematic presented in Fig. 5). To initiate reactive wetting, a droplet of highly viscous 0.5 molar KBr solution was deposited on the samples. Electrolyte spreading proceeded from the left to the right. a) K+ and Br- distribution detected after three days of exposure on a sample pre-treated according to procedure (a). b) K+ and Br- distribution detected after three days of exposure on a sample pre-treated according to procedure (c).
reduction processes on the Zn substrates. In general, this is in line with the fact that the iron oxide deposits are not supposed to form dense oxide barrier layers such as anodized Al-oxide on aluminum. Comparison of performance tendencies obtained by EIS, potentiodynamic measurements and the progress of electrolyte spreading with the kinetics of cathodic delamination implies certain, but also limited predictability of polymer/conversion layer/zinc interface stabilities. It shows that simple, quick and cost-efficient wetting experiments can be a useful option to roughly estimate the corrosion protective properties of conversion chemistry based pre-treatment approaches. No sophisticated equipment such as a potentiostat is necessary and just a sample chamber with a humid atmosphere would be needed. To reliably identify outlier in the performance tendencies obtained from the exposure of a sample series, however, reproduction experiments are definitely required. For the systems investigated in the present study, the results of short-term electrolyte spreading anticipate long-term cathodic delamination experiments. This does not seem to be an adequate strategy for pre-treatments that do not result in a distinct reduction of the electrochemical activity of the substrate, such as systems that generate iron oxide precipitates on a zinc surface. In this context, previous studies reported that tailored iron oxide morphologies indeed specifically determine parameters such as the bonding of adhered organic species. It was shown that cathodic delamination along the polymer coated Fe-oxide surfaces accelerated with decreasing surface hydrophilicity, which in turn was connected to reduced polymer/substrate adhesion forces. In contrast, a different ranking resulted for reactive electrolyte spreading experiments along non-polymer coated iron oxide surfaces [12]. This is similar to what was observed on samples that were subject to pre-treatment (a) (compare with Figs. 4–5 and 6–7). Although a direct comparison between the study of Wielant et al. [12] and the present publication cannot be drawn based on the available data, such findings may nevertheless illustrate that adhesion forces between a polymer layer and the metal substrate can determine cathodic delamination kinetics if the metal surface is not additionally covered by a thin film that contributes relevant barrier properties. 4. Conclusion
Fig. 10. Tafel plots resulting from polarization curves recorded (from cathodic to anodic potentials at a rate of 1 mV/s) on conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel sheets.
The corrosion resistance and the electrochemical properties of zinc surfaces pre-treated with different conversion chemistry solutions were investigated. Iron/iron oxide layers as well as organic/inorganic
R. Posner et al. / Surface & Coatings Technology 228 (2013) 286–295
295
pre-treatment layers with barrier properties as low as those of the native oxide film on Zn. In the present study, deposits of iron oxide hardly decelerated electrolyte spreading, but distinctly increased the epoxy/conversion layer/zinc interface stability as indicated by cathodic delamination studies. It was discussed that this may be attributed to, for example, improved adhesion of the polymer on Fe-oxide rather than to a direct inhibition of O2 reduction processes at the Fe-oxide covered Zn surface.
Acknowledgments The financial support of the Christian-Doppler-Research Association (Vienna, Austria), voestalpine Stahl Linz GmbH (Linz, Austria), and Henkel AG & Co. KGaA (Düsseldorf, Germany) is gratefully acknowledged. The authors would like to thank K.-H. Stellnberger and M. Fleischanderl for helpful discussions.
References [1] [2] [3] [4] [5] [6] [7]
[8]
[9]
[10] Fig. 12. Bode plots resulting from EIS measurements that were performed on conversion chemistry pre-treated and non-pre-treated zinc electro galvanized steel sheets.
[11] [12]
hybrid films based on precipitates of zinc phosphate and Zr-/Ti-oxide embedded in an organic network were formed. An increased resistance of the conversion coatings as measured by means of EIS at low frequencies correlated with reduced oxygen reduction current densities, with cathodically shifted open circuit potentials and with decelerated electrolyte spreading kinetics along the sample surfaces. It was shown that electrolytic wetting of the pre-treatment layers, which starts from an electrolyte reservoir where active dissolution of Zn takes place, is mainly determined by electrochemical oxygen reduction processes. This was proven by the preferential transport of cations to the front of wetting, which is also characteristic for corrosive delamination of organic films from metals covered by semi conductive oxide layers. Moreover, the kinetics of short-term electrolyte spreading processes reflected those obtained from long-term cathodic delamination experiments of epoxy coated conversion films. This indicated that cost-efficient wetting experiments can be a complementary tool for accelerated corrosion testing of surface modified zinc substrates. However, the approach did not work for thin
[13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]
T. Prosek, A. Nazarov, J. Stoulil, D. Thierry, Corros. Sci. 61 (2012) 92. A. Nazarov, D. Thierry, P. Volovitch, K. Ogle, Surf. Interface Anal. 43 (2011) 1286. A.J. Kinloch, Adhesion and Adhesives, Chapman & Hall, London, 1987. A. Rubina, R. Oltra, B. Vuillemin, K. Ogle, Electrochim. Acta 53 (2008) 6484. G. Williams, H.N. McMurray, M.J. Loveridge, Electrochim. Acta 55 (2010) 1740. K.N. Allahar, B.R. Hinderliter, D.E. Tallman, G.P. Bierwagen, J. Electrochem. Soc. 155 (2008) F201. K. Hackbarth, M. Wolpers, W. Lorenz, P. Kuhm, K. Meagher, C. Rosenkranz, M. Roth, R. Wark, G. Sanchis Otero, E. Wilke, “Metallisierende Vorbehandlung von Zinkoberflächen“, European patent no. EP 2 292 808 A1. N. Kanani, Chemische Vernicklung: Nickel-Phosphor-Schichten. Herstellung, Eigenschaften, Anwendungen. Ein Handbuch für Theorie und Praxis, E.G. Leuze Press, Bad Saulgau/Germany, ISBN: 3874802299, 2007. In: S.D. Carmer, B.S. Covino (Eds.), ASM Handbook, Volume 13A; Corrosion: Fundamentals, Testing, and ProtectionASM International, Materials Park OH/USA, ISBN: 0871707055, 2003. R. Posner, K. Wapner, M. Stratmann, G. Grundmeier, Electrochim. Acta 54 (2009) 891. R. Posner, T. Titz, K. Wapner, M. Stratmann, G. Grundmeier, Electrochim. Acta 54 (2009) 900. J. Wielant, R. Posner, R. Hausbrand, G. Grundmeier, H. Terryn, Corros. Sci. 51 (2009) 1664. R. Posner, K. Wapner, S. Amthor, K.J. Roschmann, G. Grundmeier, Corros. Sci. 52 (2010) 37. R. Posner, M. Marazita, S. Amthor, K.J. Roschmann, Corros. Sci. 52 (2010) 750. R. Posner, G. Giza, R. Vlasak, G. Grundmeier, Electrochim. Acta 54 (2009) 4837. Z.Y. Chen, D. Persson, C. Leygraf, Corros. Sci. 50 (2008) 111. Z.Y. Chen, D. Persson, A. Nazarov, S. Zakipour, D. Thierry, C. Leygraf, J. Electrochem. Soc. 152 (2005) B342. A. Leng, H. Streckel, M. Stratmann, Corros. Sci. 41 (1999) 547. A. Leng, H. Streckel, M. Stratmann, Corros. Sci. 41 (1999) 579. A. Leng, H. Streckel, K. Hofmann, M. Stratmann, Corros. Sci. 41 (1999) 599. W. Fürbeth, M. Stratmann, Corros. Sci. 43 (2001) 207. W. Fürbeth, M. Stratmann, Corros. Sci. 43 (2001) 229. W. Fürbeth, M. Stratmann, Corros. Sci. 43 (2001) 243. R. Posner, G. Giza, M. Marazita, G. Grundmeier, Corros. Sci. 52 (2010) 1838. S.E. Dolan, “Composition and process for treating metals“, US patent no. US5427632. S. Adhikari, K.A. Unocic, Y. Zhai, G.S. Frankel, J. Zimmerman, W. Fristad, Electrochim. Acta 56 (2011) 1912. F.C. Walsh, J. Phys. D Appl. Phys. 24 (1991) 217. D. Vasilakopoulos, M. Bouroushian, M. Spyrdlis, Electrochim. Acta 54 (2009) 2509. A.K. Neufeld, I.S. Cole, A.M. Bond, S.A. Furman, Corros. Sci. 44 (2002) 555. G. Klimow, N. Fink, G. Grundmeier, Electrochim. Acta 53 (2007) 1290.