Elsevier

Progress in Organic Coatings

Volume 88, November 2015, Pages 23-31
Progress in Organic Coatings

Electrochemical impedance spectroscopy as a tool to measure cathodic disbondment on coated steel surfaces: Capabilities and limitations

https://doi.org/10.1016/j.porgcoat.2015.06.010Get rights and content

Highlights

  • EIS has been evaluated as a tool for monitoring cathodic disbondment of coatings.

  • EIS was found to be capable of monitoring the early stages of cathodic disbondment.

  • EIS was found to lose sensitivity with the propagation of coating disbondment.

  • Localized EIS measurement is necessary for effective monitoring of coating cathodic disbondment.

Abstract

The disbondment of protective organic coatings under excessive cathodic protection potentials is a widely reported coating failure mechanism. Traditional methods of evaluating cathodic disbondment are based on ex situ visual inspection of coated metal surfaces after being exposed to standard cathodic disbondment testing conditions for a long period of time. Although electrochemical impedance spectroscopy (EIS) has been employed as an effective means of evaluating various anti-corrosion properties of organic coatings; its application for assessing the cathodic disbondment resistance of coatings has not been sufficiently exploited. This paper reports an experimental study aimed at developing EIS into a tool for in situ measurement and monitoring of cathodic disbondment of coatings. A clear correlation between EIS parameters and the disbonded coating areas has been confirmed upon short term exposure of epoxy-coated steel electrodes to cathodic disbondment conditions; however the degree of this correlation was found to decrease with the extension of exposure duration. This observation suggests that EIS loses its sensitivity with the propagation of coating disbondment, and that in order to achieve quantitative determination of the coating cathodic disbondment localized EIS measurements are required to measure the parameters related to local disbonded areas.

Introduction

Cathodic disbondment is a mode of coating failure that frequently occurs in metal structures such as oil and gas pipelines, which are usually protected by organic coatings in conjunction with cathodic protection (CP). It is generally believed that cathodic disbondment of pipeline coatings is due to a CP induced strong alkaline environment at the coating defects that damages the bonding between coating and metal surfaces [1], [2]. Traditional methods of evaluating cathodic disbondment of pipeline coatings in industry are based on the excavation and visual inspection of the pipeline surface. Current laboratory assessment of cathodic disbondment resistance (CDR) of coatings is based on ex situ visual inspection using standard laboratory testing methods such as AS4352, ASTM G8-96 and ISO 15711 [3], [4], [5]. Since CDR tests in sand or soil environmental conditions can take a long time, standard laboratory tests usually involve accelerated testing in a corrosive solution which aims to simulate the most aggressive field test condition. For instance according to the Australian Standard AS4352, a constant current of 3 ± 0.02 mA is applied to a coated steel specimen with an artificial defect of 6 mm diameter immersed in a 3 wt.% NaCl solution. After 28 days of accelerated immersion test, the specimen is removed from the test cell and the coating disbonded area is estimated by stripping the coating from the metal surface using a knife followed by visual inspection of the test specimen [3].

There are a number of limitations associated with the current cathodic disbondment test methodologies. Firstly, these accelerated tests are operating at CP current densities significantly higher than those experienced in practical pipelines; therefore results from these tests may not be representative of long term coating performance under normal pipeline operating conditions. On the other hand it was reported that a high variation in test results could be found from different test samples [6], [7]. For instance Holub et al. [6] investigated the performance of coatings under various standard cathodic disbondment test conditions, and noted major issues in test reproducibility and results inconsistency. These are believed to be due to differences in parameters and procedures specified in different CDR test standards. For instance different CDR test standards could apply different test temperatures, different electrolyte type or concentrations, different test duration, and different CP techniques (e.g. CP under either constant potential or constant current). Indeed Smith et al. [8] noted that testing under high temperature conditions may not be a suitable accelerated testing method. In addition, current CDR test methods are designed for laboratory testing and hence are not transferrable to in situ monitoring of the disbondment of pipeline coatings.

In situ measurement and monitoring of the initiation and propagation of cathodic disbondment in field conditions would be desirable, however this would require techniques that are able to measure the dynamic and localized changes of parameters associated with these processes. Over the past decades the advent of advanced microscopic and scanning probe techniques, such as the scanning Kelvin probe technique [9], [10], [11], [12], has facilitated substantial research aimed at measuring localized coating disbondment occurring at the interface between coating and the metal surface; however these scanning probes are laboratory based. They are not readily applied in practical testing of thick pipeline coatings as they require the scanning of a coated surface by positioning micro-sized sensors extremely close to the surface (e.g. 100 μm) [13].

Electrochemical impedance spectroscopy (EIS) is a technique that has been used for decades to measure and monitor the degradation of organic coatings by quantitatively measuring the resistances and capacitances of coated electrodes in an electrochemical cell [14], [15], [16], [17], [18], [19], [20]. The Nyquist plots of coated metals have characteristic behavior dependent on the state of the coating, thereby making it possible to follow the penetration of an electrolyte into the coating and to detect the initiation of corrosion at the metal/coating interface [21]. For instance when the electrolyte permeates the coating and reaches the metal surface, the typical single capacitive arc response of an high impedance intact coating would be replaced by one or more semi-circles and a Warburg impedance (typified by a line with the slop of 45̊) that indicate the existence of different electrode processes with various time constants as well as a diffusion process on the coated metal surface. Analysis of impedance data can be used to quantitatively determine both resistances and capacitances of a coated electrode surface in an electrochemical cell if an appropriate equivalent circuit is used to help with the physical interpretation of the impedance data. Each element of the equivalent circuit models a specific function of the electrode/electrolyte interface.

Over the past decades many different equivalent circuits have been proposed to interpret the complex impedance behavior of coated metals. The most commonly used equivalent circuit for describing a degraded organic coating on a metallic substrate is shown in Fig. 1a [20]. In this circuit Ru (or Rs) is the resistance of the electrolyte or the uncompensated resistance between the working electrode (WE) and reference electrode (RE); Cc is the capacitance of the organic coating; Rpo is the pore resistance of the coating and Rch is the charge transfer resistance (or polarization resistance) of the metal substrate beneath the coating. The Rch has been related to disbonded area under the coating through an equation, Rch = Rcho/Ad, where Ad is the disbonded area under the coating and Rcho, represents the area specific magnitude of this parameter for bare metal [20]. Also, if coating disbondment occurs, water could reach the metal surface and a double layer forms. Therefore, changes in Cdl can be related to the disbonded area of the coating. Based on this assumption the Cdl is also related to the area (wet area) under the coating through the equation, Cdl = CdloAd, where Ad is the disbonded area under the coating and Cdlo is the area specific magnitude of parameter for bare metal [20].

Another attempt of studying coating disbondment using EIS was made by Hirayama and Haruyama [22] who proposed the model shown in Fig. 1b to fit the EIS data from degraded coatings containing pores. In this model a parallel circuit for coating pores is combined with the circuit for the pore-free part of a coated steel substrate. In this model Rsol is the solution resistance, Cf and Rf are the capacitance and ionic resistance of the coating, respectively; Cdl and Rc represent the double-layer capacitance and the charge transfer resistance at the substrate/coating interface, respectively; and Zw represents the Warburg impedance. For the second part of the circuit, RP represents the total resistance for ionic migration through the pores and Cdlp and Rcp are the total double-layer capacitance and the total charge transfer resistance at substrate/solution interface at the bottom of the pores. Although a coating may actually consist of many pores, in this model they are represented by a single pore having the total cross sectional area equal to the sum of all of the individual pore areas. Hirayama and Haruyama [22] developed the break-point frequency method to relate the elements of this equivalent circuit to the disbonded area under the coating. The break point frequency, fb, is defined as the frequency at which the phase angle equals 45̊. The fb was related to the circuit elements through equations: fb = 1/2 π Rf Cc = fb0D, where D = Ad/A = fb/fb0, D is the disbondment ratio and A is the sample surface area, Ad is disbonded area and fb0 corresponds to the break point frequency for totally disbonded coating which depends only on the coating parameters through, fb0 = 1/2 π ɛ ɛ0 ρ, where ɛ is the dielectric constant of the coating and ρ is the specific ionic resistance of the coating. Although this work suggested that disbondment area can be estimated successfully using the breakpoint frequency, some criticism of this approach has also been reported. For example, Mansfeld and Tsai [23] consider that the main assumption in breakpoint frequency method, i.e. the coating properties (ɛ and ρ) do not change with exposure time, is incorrect and thus this method is not a valid approach to determining disbondment area. They suggest that, due to coating water up-take with an increase in exposure time, additional conductive paths and defects develop in the coating so ɛ increases while ρ decreases.

Attempts have also been made to use EIS for studying the cathodic disbondment of coatings after coated electrodes are exposed to solution under CP [24], [25]. This is usually conducted by carrying out EIS measurements of pre-disbonded coatings at open circuit potential (OCP). For instance Kendig et al. [24] measured the EIS data from fusion bonded epoxy with an inserted macro-defect. Samples were exposed to a cathodic potential of −1.44 V versus saturated calomel electrode (SCE) for a duration of 4 days before EIS measurements were carried out at open circuit potential, immediately after stopping the cathodic polarization voltage. They proposed an equivalent circuit shown in Fig. 1c for analyzing EIS data. In this circuit Cdli and Rdi are double layer capacitance and charge transfer resistance values at the bare metal surface in defect area as well as under the disbonded coating. When the ohmic resistance under the film (Rsi) is relatively large, due to a small gap defining the disbondment, then the result may be modeled by a transmission line as shown in Fig. 1c. Although this equivalent circuit appears to be reasonable in modeling an electrode surface with a disbonded coating, however no further detailed results from EIS analysis using this equivalent circuit was reported. This study was not able to find a correlation between EIS parameters and coating disbonded areas. This is probably not surprising because the measurement and EIS parameter determination, such as Cdli and Rdi, from an extensive disbondment area, would be extremely difficult if the ohmic resistances under the film (Rsi) are relatively large. This would be the situation in case of a small gap disbondment, and therefore the impedance data would be dominated by the coating defect area outside the disbondment areas.

Papavinasam et al. [25] also evaluated the cathodic disbondment behavior of various types of pipeline coatings of different thicknesses and containing defects using EIS data measured at OCP. The likely correlation between the rates of changes in coating capacitance with the rates of coating disbondment was investigated, however they highlighted that quantitative correlation between coating capacitance and coating disbondment could not be established. Attempts were also made to correlate other EIS parameters with coating performance but no correlation was found for any type of the examined coatings.

It should be noted that the EIS measurements of coating cathodic disbondment described thus far are all carried out under open-circuit potential. The behavior of EIS measurements when a cathodic protection potential is present is critical for in situ measurement of cathodic disbondment. The only published report on EIS measurement of cathodic disbondment under CP potential was by Margarit et al. [26] who studied the impedance behavior of steel electrodes coated with defected fusion bonded epoxy coatings. The authors asserted that a decrease in the impedance values during exposure time was attributable to the coating disbondment, although a quantitative relationship between the delaminated area and the electrochemical parameters could not be established. Raghunathan [27], in a dissertation, reported the results of EIS measurement at CP potential of −1.5 V (versus SCE) for the prediction of disbondment behavior of defected fusion bonded epoxy. Based on data analysis using a general coated metal circuit, Raghunathan et al. found that the double layer capacitance values derived from EIS data increased with the increase in the disbondment area during first 10 days of exposure, suggesting a correlation between the capacitance of the electrode and coating disbondment. However, the capacitance value decreased or plateaued afterwards. Therefore the author was able to establish a relationship between disbonded area and double layer capacitance for the first 8 days of exposure [27], even though no clear understanding of the behavior was presented.

EIS measurements on coated multi-electrodes array, often referred to as the wire beam electrode [28], under open circuit potential to monitor and estimate the coating disbondment have also received some interest recently. In a study by Wang et al. [29], a multi-electrodes array (consisting of 121 mini-electrode) coated with a 95 μm red alkyd coating with an artificial defect was exposed to an aggressive solution for 88 days. The mini-electrodes during the EIS measurement were connected to each other to simulate a single continuous electrode. The evolution of disbonded area of coating with time was calculated using the equation Ad = Cdl/Cdl0. On the other hand Le Thu et al. [30] studied the cathodic disbondment of a 60 μm epoxy coated multi electrodes array containing a macro-defect using EIS measurement at open circuit potential. The authors reported difficulties in estimating the disbonded area as the results are dominated by the naked metal from the defect area.

This work is designed to investigate the capabilities and possible limitations of the EIS method in measuring and monitoring cathodic disbondment of organic coatings. The aim is to develop EIS into a practical and reliable tool for measuring cathodic disbondment of defective coatings under simulated pipeline CP conditions.

Section snippets

Experimental

Pipeline steel plates (API X65) of 100 mm by 100 mm were used in this work for making working electrodes. The steel is a low carbon steel with approximately 0.14% C, 1.3% Mn, 0.23% Si, 0.01% P, 0.03% S and trace levels of Al, Nb, Cu, Cr, Ni and Ti. The steel surface preparation was based on the SSPC-SP 10/NACE No. 2 standard, with all oil, grease, dirt, mill scale, rust, corrosion products, oxides, paint or other foreign matter completely removed from the surface by abrasive blasting. The coating

Results and discussion

The key to developing EIS into a method for in situ measurement and monitoring of cathodic disbondment of coatings is a clear correlation between EIS parameters and the disbonded coating areas after various periods of exposure of coated steel electrodes to cathodic disbondment conditions. In this work, the coating disbondment area was determined through XPS measurement while EIS measurements under CP potential were performed using conventional EIS technique.

Concluding remarks

EIS measurement under CP potential is able to detect and monitor the initiation and early propagation of cathodic disbondment of coatings. A clear correlation between electrochemical parameters and coating disbondment area was confirmed; however this correlation is true only for a relatively short period of exposure. This is probably due to difficulties in accurately measuring impedance from deep coating disbondment areas. The results suggest that localized EIS measurements under disbonded

Acknowledgment

This work was funded by the Energy Pipelines CRC, supported through the Australian Government's Cooperative Research Centers Program. The funding and in-kind support from the APGA RSC is gratefully acknowledged. The authors wish to thank Bruce R. W. Hinton, Alan Bryson, Brian Martin, Bruce Ackland, and, Geoff Cope for their experienced advice, support and comments on our research.

References (34)

  • Y. Tan

    An overview of techniques for characterizing inhomogeneities in organic surface films and underfilm localized corrosion

    Prog. Org. Coat.

    (2013)
  • H. Bi et al.

    Cathodic disbonding of an unpigmented epoxy coating on mild steel under semi- and full-immersion conditions

    Corros. Sci.

    (2011)
  • S. Feliu et al.

    An interpretation of electrical impedance diagrams for painted galvanized steel

    Prog. Org. Coat.

    (1989)
  • Australian Standard AS4352

    Test for Coating Resistance to Cathodic Disbonding

    (1995)
  • ASTM G8-96

    Standard Test Method for Cathodic Disbonding of Pipeline Coatings

    (1996)
  • International Standard ISO15711:2003(E)

    Paints and Varnishes – Determination of Resistance to Cathodic Disbonding of Coatings Exposed to Sea Water

    (2003)
  • J. Holub et al.

    Analysis of CDT methods and factors affecting cathodic disbondment

  • Cited by (32)

    • The effect of pulse current cathodic protection on cathodic disbondment of epoxy coatings

      2022, Progress in Organic Coatings
      Citation Excerpt :

      It should be pointed out that since the sampling frequency of the electrochemical workstation is capped at 1000 Hz, the pulse frequency has not been further increased in this work, and the results also show that when the pulse frequency is 500 Hz, PCCP can hardly alleviate the cathodic delamination process. It should be noted that EIS can only provide average electrochemical information on the electrode surface [27,28], and some researchers have expressed that the extended cathodic delamination area might not be detected because the collected EIS data is dominated by defect area with low impedance [25,29]. Here, the WBE method was introduced in this section, which can offer local electrochemical information as indicated in our previous work [16].

    • Corrosion behavior of zirconium-pretreated/epoxy-coated mild steel: New approach for determination of cathodic disbondment resistance by electrochemical impedance spectroscopy

      2021, Journal of Alloys and Compounds
      Citation Excerpt :

      Research efforts over the past few years on control and prevention of organic coating destruction have been mainly carried out by electrochemical impedance spectroscopy (EIS). Mahdavi et al. [22] lowered the amount of cathodic disbondment of the epoxy coating by increasing the charge transfer resistance obtained from EIS data. Naderi et al. [23] revealed that impedance data for epoxy films enriched with anticorrosion pigments.

    • A critical look at interpretation of electrochemical impedance spectra of sol-gel coated aluminium

      2021, Electrochimica Acta
      Citation Excerpt :

      Electrochemical impedance spectroscopy (EIS), on the other hand, is a useful tool for studying corrosion resistance of coated metals and follow the evolution of coatings properties, water uptake and development of corrosion activity during immersion [30–35]. EIS was successfully applied in conjunction with accelerated testing including a combination of DC and AC measurements (AC/DC/AC) [36,37], and to understand the kinetics of cathodic disbondment of organic coatings [38]. In literature, however, there is a large discrepancy of results concerning the interpretation of the impedance spectra of coated aluminium alloys by organic and hybrid coatings.

    View all citing articles on Scopus
    View full text