Communication—An Approach to Measuring Local Electrochemical Impedance for Monitoring Cathodic Disbondment of Coatings

An electrochemically integrated multi-electrode array has been used for monitoring and visualizing the cathodic disbondment of defective coatings by measuring local electrochemical impedance. Compared with the conventional electrochemical impedance and local current measurement approaches, this new approach signiﬁcantly enhances the sensitivity of detecting the propagation of coating disbondment by eliminating the effects of the dominating low impedance regions, such as those that arise at coating defects, and thus increases the visibility of higher impedance regions deep in the disbonded coating. Furthermore, it facilitates the probing of electrode processes and mechanisms in selected local electrode regions.

Cathodic disbondment is a major form of electrochemically induced coating failure that frequently takes place at the metal/coating interface on cathodically protected steel infrastructure such as pipelines.Extensive research over the past decades has developed good understanding of the phenomenon, 1-2 however currently there is no technique that can be used to perform in-situ monitoring of its occurrence in the field.Traditional methods of evaluating cathodic disbondment of pipeline coatings are based on ex-situ visual inspection of excavated pipes.Electrochemical techniques such as conventional electrochemical impedance spectroscopy (EIS), [3][4][5][6][7] localized electrochemical impedance spectroscopy (LEIS), [8][9][10][11] scanning kelvin probe and scanning vibrating electrode techniques (SVET) [12][13][14][15][16] have been employed to measure coating disbondment in the laboratory; however, there are significant obstacles for these techniques being practically used to monitor in-situ cathodic disbondment of thick pipeline coatings (e.g.1000 μm in thickness).A thick coating would 'shield' the current from reaching the disbonded area, especially far away from the original defect, and therefore the measurements are more likely to be dominated by the lower impedance present at the coating defect areas.Under these conditions, little information can be obtained from higher impedance regions deep in the disbonded area.Indeed, in a previous study the authors have found that conventional EIS loses sensitivity in detecting cathodic disbondment propagation due to such limitations. 17The electrode array 18 is a method that has been applied to measure local direct currents for evaluating the cathodic disbondment of defective thin coatings (<100 micron) by Le Thu et al. 19 and Wang et al. 20 However, there is little evidence to show that direct current mapping is sensitive enough to detect coating disbondment, especially at its propagation stage.This is a concern because a resistive coating film could 'shield' the direct current from flowing into the disbonded coating area.Here we describe a new approach to measuring coating disbondment based on local AC impedance measurement using the electrode array and assess the viability of different approaches.Previously Kong et al. 21measured the EIS of individually selected steel electrodes in an electrode array; however the purpose of their measurement was for assessing the degradation of intact coatings (100 micron), not for monitoring coating disbondment.

Experimental
Fig. 1 shows the multi-electrode array sensor and the experimental setup employed to perform local EIS measurements.The dimension of the sensor is 25.3 mm × 25.3 mm consisting 100 closely packed but isolated square shaped carbon steel electrodes (2.44 mm × 2.44 mm) embedded in epoxy resin.The gaps between neighboring electrodes were kept small (0.10 ± 0.05 mm).The area ratio of steel to isolating resin is 595/45.After being ground using SiC grit paper, the surface of the sensor was coated with a transparent polyester coating (Barnes products Pty.Ltd.) with a dry film thickness of 1000 ± 20 μm.An artificial defect of 5 mm diameter was made at the center of the coated sensor surface to simulate coating damage.The sensor was installed in a plastic electrochemical cell filled with aqueous solution of 3 wt% NaCl, with a three electrode cell configurations shown in Fig. 1.In order to measure a local impedance, the terminal of a selected electrode (WE1) was connected to the channel 1 of a VMP3 potentiostat (Bio-Logic Science Instruments) via a manual switcher, while the remaining 99 coupled electrodes were connected to channel 2 of the potentiostat.The VMP3 was also used to apply an excessive cathodic protection potential of −1.40 V Ag/AgCl to the electrode array.Overall EIS and local impedance were measured under the same CP potential with a perturbation potential of ±10 mV in the frequency range of 100 KHz to 300 mHz.Local impedance amplitude (lZl 300 mHz ) data were arranged in the form of a 10 by 10 data matrix for plotting the impedance maps.For measuring the direct current maps, the CP potential of −1.4 V Ag/AgCl was applied to all electrodes by the potentiostat using a typical three electrode setup.A zero resistance ammeter was internally connected between working electrodes to measure direct CP current flowing to each electrode.More details on the experimental and data analysis methods can be found elsewhere. 18,22

Results and Discussion
Fig. 2 shows typical maps of local impedance amplitude (|Z| at 300 mHz ) and a direct current map measured after different periods of exposure of the sensor to the test solution under CP potential of −1.40 V Ag/AgCl or −0.95 V Ag/AgCl .It is clearly shown in maps (a)-(f) that, under a CP potential of −1.40 V Ag/AgCl , the impedance of electrodes surrounding the defect area continuously decreased (to less than 10 5 ohm) over the 624 hours exposure period.These low impedance areas expanded with the increasing exposure time, while electrodes located far away the defect area maintained a high impedance of larger than 10 7 ohm after 624 hours.These maps clearly indicate coating disbondment due to permeation of the test solution along the disbonded coating/metal interface gap rather than absorption of the solution by the coating.After 624 hours, as shown in Fig. 2f, the majority of electrodes on the sensor were disbonded.Direct current maps measured at −1.40 V Ag/AgCl (not shown here) also show similar coating disbondment processes and behavior.However, when the CP potential was reduced from −1.40 V Ag/AgCl to −0.95 V Ag/AgCl , as shown in Fig. 2g and Fig. 2h, the impedance map still clearly shows the disbonded area, while the direct current map, on the other hand, lost sensitivity and this coating disbonded area was not visible, as seen in Fig. 2h.This may explain a result reported by Le Thu et al. 19 that, in a previous   attempt to measure coating disbondment using array electrodes (coating thickness 60 μm) under a CP potential of −1.5 V vs.SCE , no significant coating disbondment was observed on direct current maps over a 336 hour exposure period. 19This is clearly a major limitation of the direct current measurement technique given that a CP potential of −0.95 V Ag/AgCl is close to industry standard CP criteria for practical, coated pipeline.
In order to further understand the mechanism of the processes occurring under disbonded coatings, local impedance measurements of selected areas were also performed using the electrode array.As shown in Figs.3a-3c, conventional overall EIS (in Nyquist plots) measured by connecting all electrodes of the coated sensor only showed EIS behavior with a single time constant and a small impedance value that is believed to be dominated by the low impedance defect area.Although a drop in impedance was observed after 120 hours (Fig. 3b), which should be due to coating disbondment, no further significant change in impedance is observed with further extension of exposure time.The EIS plots in Fig. 3b and Fig. 3c appear very similar, indicating that the conventional EIS does not recognize the coating disbondment after 120 hour exposure.This is in agreement with a finding we reported previously, suggesting a limitation of using traditional EIS for studying the coating disbondment process and its mechanisms. 17his limitation could be overcome when the electrodes located at the defect area are excluded from EIS measurement.As shown in Figs.3d-3f, EIS of the coating disbondment area (excluding the electrodes located at the defect area) exhibits very different EIS characteristics at different stages of exposure.Fig. 3d presents a typical capacitive behavior suggesting that none of the coated electrodes were disbonded at the first hour of exposure.After 120 hours exposure (Fig. 3e) the Nyquist plot showed two time constants.The one at higher frequency is believed to be related to coating impedance and the one at lower frequency is most likely related to the diffusion of the electrolyte to the metal under the disbonded coating.After 624 hours exposure (Fig. 3f) two time constants are recognizable in the Nyquist plot and the diameter of the first semi-circle decreased compare with that in Fig. 3e.Although correlating these EIS characteristics to electrochemical and diffusion processes that may be occurring under the disbonded area needs more detailed work, these results nevertheless confirm that, selectively measuring local EIS over specific electrode areas can facilitate the study of electrode processes occurring over that electrode area, e.g.under a disbonded coating.

Summary
Local electrochemical impedance measurements using an electrode array sensor have shown significantly improved sensitivity for monitoring the propagation of cathodic disbondment of defective coatings compared with the conventional overall electrochemical impedance and local current measurements approaches.This new approach also provides the opportunity of eliminating the effects of the low impedance coating defect regions on the visibility of higher impedance regions deep in the disbond coating, facilitating the probing of electrode processes and mechanisms in selected regions of heterogeneous electrode surfaces.

Figure 1 .
Figure 1.Schematic diagram of the experimental setup for local impedance measurement under cathodic protection.

Figure 2 .
Figure 2. Typical maps of impedance amplitude (|Z| at 300 mHz ) and direct currents measured over a coated sensor after various periods of exposure and under different CP potential.

Figure 3 .
Figure 3.Comparison of overall EIS (a-c) and local area EIS (d-f) measured after various periods of exposure.