Stockholm 26 april 1996

 

Corrosion growth of surface pores on copper

 

Erik Levlin

Water Resources Engineering, Royal Institute of Technology

100 44 Stockholm

 

Content

 

1        Background  1

2        First prerequisite; Good conditions for local corrosion  1

2.1     Theory for pitting corrosion  1

2.2     Local corrosion of copper  2

3        Second prerequisite; Pore is anodic compared to surrounding surface  4

5        Experimental investigation  5

6        Conclusions  6

7        References  6

 

1          Background

 

 

 

Figure 1. Sketch of a surface pore on copper that may corrode and grow into a hole.

 

 

High level nuclear waste is planned to be restored in copper canisters embedded in bentonite clay in the bed rock at a depth of 500 m. One potential risk for the safety is occurrence of surface pores growing by corrosion (figure 1) into holes penetrating the canisters. On most metals pores will cause crevice corrosion. Oxygen diffusion is obstructed to the pore, making the surrounding surface cathodic and the pore anodic.

 

This report deals with possible mechanisms for preventing or promoting corrosion of surface pores on copper. For a pore to grow where must be both good conditions for local corrosion and that corrosion inside the pore is favoured compared to corrosion of the surrounding surface. Differences in corrosion conditions between the inside of a pore and the surrounding surface may cause crevice or local corrosion. If corrosion inside the pore is favoured the pore will be anode and grove and may, if it grows big enough, penetrate the canister. Otherwise, if corrosion of the surrounding surface is favoured the canister will be protected against perforation.

 

2          First prerequisite; Good conditions for local corrosion

 

As corrosion growth of a pore is a type of local corrosion there one part of the metal surface is anodic and corrode while an other part i cathodic, good conditions for local corrosion is a first prerequisite for corrosion growth of a pore.

 

2.1       Theory for pitting corrosion

 

The theory for pitting corrosion, which is one form of local corrosion, has been described in another SKB report /1/. Figure 2 shows a schematically potential pH diagram and figure 3 a schematically potential log current density diagram illustrating the criteria for pitting corrosion. In the diagram there is an area of corrosion at lower pH-levels. If the corrosion potential is above the protection potential a local acidification due to hydration of dissolved copper ions makes the pH-level to decrease and move to the left into the area of corrosion, resulting in pitting or local corrosion. This implies that the metal have a corrosion potential there a local decrease in pH moves the local conditions into the area of corrosion before the pH reaches the line of minimum pH.

 

 

 

Figure 2. A schematic potential pH diagram illustrating criteria for pitting corrosion /1/.

 

 

Figure 3. A schematic potential log current density diagram illustrating criteria for pitting corrosion.

 

 

 

 

The protection potential, is the potential above which a pit is able to growth. The pitting potential, which is the potential there a pit is initiated, lies above the protection potential. However, if a pit has been initiated when the potential has been increased above the pitting potential, the potential must be decreased down below the protection potential to stop the pit from growing and repassivate the surface. The protection potential is the same for all form of local corrosion including crevice corrosion as well as pitting. Therefore one important prerequisite for corrosion growth of a pore is that the corrosion potential is above the protection potential.

 

The pH-level inside the active pit lies to the left of the border of the corrosion area, since hydration of the dissolved metal ions gradually decreases then the pH moves into the corrosion area. The border determined by the solubility of the metal hydroxide is equal to 10-6 M. Due to resistance in the solution the ohmic drop causes the potential inside the pit to decrease until it lies at the bottom of the corrosion area. The potential of the surrounding surface must therefore lie above the protection potential, which lies above the potential inside the active pit.

 

The good electric conductivity of clays promotes local corrosion. Since the corrosion current for the anodic dissolution inside the pit is supplied by the cathodic reaction on the surface area around it, there is a corrosion current going from the surrounding metal surface into the pit. The cathode reaction will depend on the rate of oxygen diffusion to the surface. At good electric conductivity a large cathode area can supply some few pores, which can grow faster and deeper, thus increasing the risk or perforating the canister. The total amount of corrosion current can be calculated from the supply of oxygen from the bentonite, but how deep a pore will grow depends on how many pores the corrosion current is divided.

 

2.2       Local corrosion of copper

 

 

          [Cl] = 10-3 M                                                   [Cl] = 10-1 M.

Figure 4. Potential-pH diagram with conditions for protection of copper /2/. Line O is the potential for oxygen reduction and H is the potential for hydrogen evolution. The black spot in the area for general corrosion (marked with diagonal lines) marks the pH and potential inside an active pit.

 

 

Figure 4 shows two potential-pH diagram for copper in water with chloride concentration 0.001 M and 0.1 M /2/. The protection potential is determined by the extension of the area of corrosion in these potential-pH-diagram. In both diagrams the protection potential lies above line H, which is the electrode potential for Hydrogen evolution. In aerobic conditions the corrosion potential lies between line O and H and in anaerobic conditions below line H. Local corrosion and growth of pores may only occur in aerobic conditions. In the case of copper canister the environment is aerobic during the first 30 to 100 year, until all oxygen in the bentonite has been consumed through oxidation of pyrite. When the bentonite have become anaerobic corrosion growth of pores will not occur.

 

Increasing chloride ion concentration promotes local corrosion. Due to formation of chloride complexes the area of corrosion increases with increasing chloride concentration, making the protection potential to decrease. The protection potential, marked in the diagrams, is about +450 mV for 0.1 M and +300 mV for 0.001 M.

 

       log(H2CO3)tot

 

Figure 5. Stability domain of CuO and CuOH(CO3)0.5 in aerated water as function of H2(CO3)tot and pH at 10oC and 50oC /3/.

 

 

Increasing bicarbonate concentration may prevents local corrosion. Figure 5 shows stability domain of copper oxide CuO and basic copper carbonate CuOH(CO3)0.5 in aerated water as function of H2(CO3)tot and pH at 10oC and 50oC /3/. If the carbonate content is higher than 5 mg/l (10oC) or 100 mg/l (50oC) basic copper carbonate is formed:

 

              2Cu + 2HCO3 + 2H2O ® Cu2(OH)2CO3 + 2H+ + 4e                 (1)

 

However, the acidification is at pH-level 6.4 reduced by the bicarbonate:

 

              HCO3 + H+ ®  + H2CO3                       pka=6.4                            (2)

 

The result can be demonstrated in a formula there the species carrying the corrosion current are marked vertically /5/:

 

 

 

(3)

That the lowest concentration for formation of basic copper carbonate increases at higher temperature from 5 mg/l at 10oC to 100 mg/l at 50oC indicates that the preventive effect of carbonate decreases at increasing temperatures. The copper canisters for high level nuclear waste will have a temperature of 90oC. The border of the corrosion are was transferred from pH-level 5.8 to pH-level 6.3 at increasing the temperature from 10oC to 50oC. If this is extrapolated to 90oC the border will be at pH-level 6.8. This is above pH-level 6.4, there the bicarbonate ion is buffering, and the preventive effect of carbonate ions may be reduced.

 

Increasing sulphate ion concentration promotes local corrosion. It the corrosion current is carried by sulphate ion acidification due to accumulation of sulphuric acid will occur at anodic sites. This can be demonstrated in a formula for formation of basic copper sulphate Cu(OH)0.5(SO4)0.25, there the species carrying the corrosion current are marked vertically /5/:

 

 

 

(4)

3          Second prerequisite; Pore is anodic compared to surrounding surface

 

The second prerequisite for corrosion growth of pores is that corrosion inside the pore is favoured compared to corrosion of the surrounding surface. Otherwise, if corrosion of the surrounding surface is favoured compared to corrosion inside the pore, the canister will be protected against corrosion growth of pores.

 

 

 

Figure 6. Sketch illustrating obstructed diffusion of copper ions out of a pore compared to diffusion through a diffusion boundary layer, for copper exposed to water solution.

 

 

One mechanism that favours corrosion on the surrounding surface is that diffusion of dissolved copper ions from inside a pore is obstructed compared to diffusion from the surrounding surface, thus creating a higher copper ion concentration inside the pore than outside. The higher copper ion concentration prevents anodic dissolution of copper inside the pore, thus making the pore cathodic compared to the surrounding surface, there the anodic dissolution increases. This difference in diffusion conditions between the inside of a pore and the surrounding surface is largest when the copper as in figure 6 is exposed to a water solution. Copper ions from the surrounding surface may diffuse a short distance through the diffusion boundary layer, before they are transported away from the surface by convection. A copper ion from the inside of a pore has a much longer way to diffuse, out through the pore and across the diffusion boundary layer. In the case of copper canister embedded in bentonite clay, there is no convection in the clay and therefore no diffusion boundary layer. This mechanism for preventing corrosion growth of pores will thus not be very strong.

 

 

Figure 7. Sketch of the case there bentonite clay not have penetrated into the pore.

 

 

One mechanism that favours corrosion inside the pore is obstructed diffusion of oxygen into the pore, making the surrounding surface cathodic. Another mechanism is buffering by the bentonite clay only on the surrounding surface. If the bentonite clay as in figure 7 not have penetrated into the pore only the surface outside the pore will be in contact with the bentonite clay. Due to the swelling of the bentonite it may penetrate almost all pores, but there is allways a risk of getting no penetrated pores. With few no penetrated pores, corroding pores can be supplied with corrosion current from a large cathode area and thus grow deeper. Precipitation of copper oxide will produce hydroxide ions. On the surrounding surface the produced hydroxide ions will be taken up by the clay and exchanged to other ions, thus preventing any decrease in pH-level on the surrounding surface. The bentonite clay has a pH-level of about 9 and a large buffering capacity. The difference in pH-level between the inside of the pore and the surrounding surface may make the pore anodic and promote corrosion growth of the pore.If the corrosion current is carried by chloride ions into the pore, there will be an accumulation of hydrochloric acid inside the pore.

 

 

 

(5)

Mattsson and Fredriksson /6/ have measured corrosion currents in an artificial pit (figure 10) covered by a crust of basic copper carbonate, basic copper sulphate, basic copper chloride Cu(OH)1.5Cl0.5 and without a crust of corrosion products. With a crust of basic copper sulphate the corrosion current was higher than without crust. The corrosion current was smallest with a crust of basic copper carbonate, followed by a crust of basic copper chloride, both with a smaller corrosion current than without crust. The conclusion can be made that if basic copper sulphate is precipitated inside a pore, the pore will be anodic compared to the surrounding surface and grow. However, if basic copper carbonate is precipitated in the pore, the pore will be cathodic compared to the surrounding surface and be protected against corrosion growth. Without crust of corrosion products on the surface, copper in bentonite, which have the pH-level 9, will be covered by a film of copper oxide.

 

That precipitation of copper chloride promote corrosion growth can be indicated that copper (I) chloride has been in corrosion pits on copper /7/. Figure 8 shows a corrosion pit with a layer of copper chloride at the bottom of the pit. The pit may be caused by precipitation of copper chloride on the surface, but the copper chloride in the pit can just as well be precipitated due to high chloride concentration in the pit, caused by chloride ions beeing transported to the pit by the corrosion current.

 

 

 

Figure 8. Sketch of a corrosion pit on copper by Lucey /7/.

 

5          Experimental investigation

 

Figure 9 shows a sketch of an experiment for studying corrosion growth of pores in bentonite clay. This experiment has similarities to earlier laboratory experiments on aeration cell corrosion of cast iron in soil /8,9/ and to the experimental artificial pit used by Mattsson and Fredriksson /6/ (figure 10). In this experiment the copper surface surrounding the pore is separated from the copper surface inside the pore. This makes it possible to measure the corrosion current passing between the pore and the surrounding surface and from the direction of the corrosion current decide whether the pore is the anode or the cathode. The artificial pore anode is copper electrode inside a plastic chamber embedded in the bentonite. From the chamber there is a hole out to the bentonite. The surrounding surface is represented by a copper electrode embedded in the bentonite. The corrosion current is measured with a zero resistance ampere meter. The corrosion potential of the electrodes can also be measured with a reference electrode

 

 

 

Figure 9. Sketch of an experiment for studying corrosion growth of pores.

 

Figure 10. Sketch of an experimental artificial pit used by Mattsson and Fredriksson /6/.

 

6          Conclusions

 

Local corrosion and growth of pores is possible in aerobic conditions during the first 30 to 100 year, until all oxygen in the bentonite has been consumed through oxidation of pyrite. Whether a pore will grow or not depends on the combined effect of different preventing and promoting factors.

 

Factors preventing corrosion growth of pores:

     High carbonate content.
Precipitation of basic copper carbonate will prevent local corrosion, but the preventive effect may be reduced at higher temperatures.

     Obstructed diffusion of dissolved copper ions from the pore.
More preventive in water solutions than in bentonite clay.

 

Factors promoting corrosion growth of pores:

     High electric conductivity of the bentonite clay.
Determines the size of the cathode area and thus how deep the pore can grow.

     High chloride content.
The protection potential decreases and copper chloride plays a role in the mechanism for pitting corrosion.

     High sulphate content.
Precipitation of basic copper sulphate in the pore is most dangerous.

     Buffering only of the surrounding surface by the bentonite clay.

 

To estimate the risk of getting corrosion growth of pores in aerobic conditions is difficult. However, the risk can be examined through a laboratory experiment.

 

 

7          References

 

1.    Levlin, E: Corrosion of copper in anaerobic clay. Prerequisites for pitting and whiskers formation. SKB Projekt Inkapsling, Projekt PM PPM 95-3420-09, Stockholm (1995). Report

 

2.    Pourbaix, M: Applications of electrochemistry in corrosion science and in prac­tice. 5th International Congress of Metallic Corrosion, Tokyo, pp. 17-48 (1972)

 

3.    Mattsson, E: Counteraction of pitting in copper water pipes by bicarbonate dosing. Werkstoffe und Korrosion, Vol. 39, pp. 499-503 (1988)

 

4.    Lindman, E K and Mattsson, E: Försurningens inverkan på korrosionen i vattenledningsrör av koppar. Kol-Häls-Miljöprojektet, Statens Vattenfallsverk, Teknisk report 44, (1982)

 

5.    Levlin, E: Corrosion of water pipes due to acidification of natural waters. 10th Scandinavian Corrosion Congress, Stockholm, pp. 421-425 (1986)

 

6.    Mattsson, E and Fredriksson, A-M: Pitting corrosion in copper tubes – Cause of corrosion and counter-measures. British Corrosion Journ. Vol. 3, No. 9, pp. 246-257 (1968)

 

7.    von Lucey, V F: Lochkorrosion von kupfer in trinkwasser. Werkstoffe und Korrosion, Vol. 26, No. 3, pp. 185-191 (1975)

 

8.    Levlin, E: Corrosion of water pipe systems due to acidification of soil and groundwater, Doctors dissertation in Applied Electrochemistry and Corrosion Science, Royal Institute of Technology, Stockholm, TRITA-TEK 1992:01, ISBN 91-7170-094-3, (1992)

 

9.    Levlin, E: Corrosion of underground structures due to acidification: laboratory investigation. British Corrosion Journ. Vol. 26, No. 1. pp. 63-66 (1991)