Search

The branched pitting corrosion often occurs on copper tubes used in heat-exchanger units in air-conditioners and leads to leakage of heat-exchanger fluid for several months. Figure 1 shows a photograph of this unique localized corrosion formed by galvanostatic polarization in the laboratory. The morphology is characterized by the fractal structure in which pits of several ten micro-m diameter are formed at copper surface and smaller pores are developed on the pit walls randomly. This localized corrosion has been reported with some names: ant-nest corrosion, formicary corrosion, unusual corrosion, branched pits and pinhole corrosion [1]. It’s well known that the occurrence of this localized corrosion is related to trace phosphorus in copper and presence of low molecular weight carboxylic acid such as formic acid. However, the origin to form branched structure has not been clarified. We reported an impedance method to analyze the branched (fractal) pore electrode with transmission line model (TLM) [2, 3]. Impedance spectrum of simple cylinder pore electrode on Nyquist plane normally describes straight-line of 45 degree slope in the high frequency range. In the case of electrode with fractal pores, the slope of straight-line decreases depending on the fractal dimension [3]. In this study, we employed this impedance analysis to investigate the initiation and growth of branched pores in ant-nest corrosion of copper. References: [1] D. M. Bastidas, V. M. La Iglesia, E. Cano, S. Fajardo, and J. M. Bastidas, J. Electrochem. Soc., 155, C578 (2008). [2] M. Itagaki, S. Suzuki, I. Shitanda, K. Watanabe, H. Nakazawa, J. Power Sources, 161, 415 (2007). [3] M. Itagaki, Y. Hatada, I. Shitanda, K. Watanabe, Electrochim. Acta, 55, 6255 (2010). Figure 1

Refrigerant leakage due to the occurrence of formicary corrosion, also called ant nest corrosion, has been serious problem in copper piping used for the heat exchangers 1, 2). Because the ant nest corrosion begins at small pit and progresses rapidly in the depth direction, it is difficult to detect the occurrence of the ant nest corrosion. Notoya 2) reported that the progress of the ant nest corrosion in the depth direction was related to the repetition of the proportional reaction composed of copper and divalent carboxylic acid copper complex. In the present study, the corrosion potential measurement was performed to investigate the influence of proportional reaction composed of copper and divalent copper ions on the progress of the ant nest corrosion. A silicon sheet and two copper plates were used to simulate the pore environment on the ant nest corrosion. The small hole was drilled on the silicon sheet and copper plates. The diameter of the holes was 1 mm. The silicon sheet was sandwiched between copper plates, and the small hole penetrated into the silicon sheet and coper plates. This hole, which simulates the pore on ant nest corrosion, was filled with an aqueous solution of 100 ppm formic acid. In this case, the upper and lower copper plates were defined as working electrode 1 and 2 (WE1 and WE2), respectively. The corrosion potential measurement of the WE2 was carried out during the galvanostatic polarization of the WE1. A phosphor-deoxidized copper plates were used as the WE1 and WE2, and the electrochemical measurement was performed by a four-electrode system using a dual potentiostat. The corrosion potential of the WE2 was measured during the applying the constant current of 20 μA to the WE1. The corrosion potential of WE2 was shifted to noble potential because of the dissolution of the divalent copper ions on the WE1. This implies that the cathodic reaction rate of WE2 was increased due to the increase of the divalent copper ions concentration near the electrode interface of WE2, leading to the repetition of the proportional reaction composed of copper and divalent copper ions. References: [1] S. Yamauchi, K. Nagata, S. Sato, M. Shimono, Sumitomo light metal technical reports, 52, 1-7 (1984). [2] T. Notoya, T. Hamamoto, K. Kawano, Sumitomo light metal technical reports, 30, 123-128 (1989).