An adaptation of:Pourbaix (Eh-pH or EpH) Diagrams and Archaeological Corrosion of BronzesW. Thomas Chase Michael NotisAnd Arthur Pelton

BUMA VI, Beijing, Sep. 15, 2006

Cross-sectional sample from tiger from Xingan, Dayangzhou, in bright and dark field, showing typical delta-removal corrosion along with cuprite and redeposited copper.

AcknowledgementsRubin-Ladd FoundationDepartment of Conservation and Scientific Research, Freer Gallery of Art and Arthur M. Sackler Gallery, Smithsonian InstitutionArthur D. Pelton, Thermfact LTD., 447 Berwick, Mount-Royal, Qubec

Pourbaix DiagramsAlso called Potential-pH, Eh-pH, or EpH Diagramsx = pH (acidity-alkalinity)y = Eh (oxidation potential)

MUST BE CALCULATED FOR:Specific compoundsSpecified temperatureSpecified concentration of reactants

Modern computer methods help with the solution of many simultaneous equations; skill and judgement are also necessary.

Pourbaix diagram for the Cu-Cl-H2 O ternary system at 25C;Cl- = 355 ppm.

Pourbaix Stabilityof WaterRegion of thermodynamic stability of water under a pressure of 1 atmosphere at 25 C.

Pourbaix oxidizing and pHAcid, alkaline, oxidizing, and reducing areas on the Pourbaix diagram.

Pourbaix Diagrams EnvironmentsFrom Garrells and Christ Mineral Equilibria at Low Temperature and Pressure

Combined 11.1 and 11,2

Pourbaix Cu-Water (Redrawn)

Pourbaix copperTheoretical conditions of corrosion, immunity, and passivation of copper at 25 C.

Pourbaix Diagram-Sn

Tin - Immunity

Wang Dongning

Pourbaix diagram for the Cu-Cl-H2 O ternary system at 25C;Cl- = 355 ppm.

Pourbaix vs Pelton

Fig 4 with labels

Pelton Figure 4

Pelton Figure 4c

Pelton - Figure 4d

Comparison Diagrams 4

Pelton Diagrams 1,4,7,10

Pelton Figures 7,8,9

Pelton Figure 22

Pelton Figures 13, 16, 19, 22

Pelton Figures 19, 20, 21

Pelton Figure 22

Figure 22 + Environments

Fig 18

A reproduction of the American Liberty Bell (made in 1976) outside Union Station, Washington, D.C., showing outdoor corrosion of bell metal.

Fig 22 & 4

Wang Quanyu PixFigure 6.20 from Wang Quanyu, Deterioration of Jin Bronzes, showing corrosion and different types of secondary metallic copper; W denotes width of the image.

20% tin Bronze-Xingan DingX1000in SEM

Xingan fang-ding alpha and delta corrosion bright field

Xingan fang-ding alpha and delta corrosion dark field

Conclusions - IUsing Cu and Sn together to calculate the Pourbaix diagrams gives a better picture.Intermetallics improve the picture.Stability of SnO2 throughout the diagrams explains some of the corrosion we see.

Conclusions - IIStability of Cu with SnO2 in reducing environments explains preferential corrosion of the delta phase. Delta will corrode (and copper redeposit) in reducing conditions. Alpha will corrode in oxidizing conditions.Computer calculation of Pourbaix diagrams for specialized multicomponent systems is a powerful tool to promote understanding of corrosion behavior and to aid in conservation.

Pourbaix diagramfor lead

Lead on fig 17 & 22

*Pourbaix (Eh-pH or EpH) Diagrams and Archaeological corrosion of BronzesW. Thomas Chase and Michael NotisBUMA VI, Beijing, Sep. 15, 2006DRAFT 4 9/9/06 12:05Slide 1Anyone who would attempt to tell you about Pourbaix diagrams, especially new Pourbaix diagrams, in 15 minutes is a fool! So let's get started.*Slide 2The background I'm using, a section of the bronze tiger excavated at Xingan, shows typical delta-removal corrosion and redeposited copper.*Slide 3Supporters. The actual work was done by Arthur Pelton of Thermfact Ltd. of Montral. I recommend you look at his full report, which will be available on the Lehigh Archaeometallurgy Laboratory website, and Ill give you details in the last slide.*Slide 6Pourbaix diagrams use pH for the horizontal x-axis and Eh or oxidation potential for the vertical y-axis. To do the calculations one must specify possible reactants and solid phases, temperatures, and reactant concentrations. With many reactants the computations become complex. Modern computer methods and a skilled operator with good judgment are necessary. *Slide 16Here's a diagram from Pourbaix of the copper-chlorine-water system at 355 ppm. You can see areas of stability of copper, cuprite, tenorite, nantokite, and atacamite/paratacamite. Contours for different concentrations of copper in solution complicate this diagram.*Slide 7Here's a Pourbaix diagram showing the stable region of water. Below line a, water dissociates to hydrogen. Above line b, water dissociates to oxygen. The region we're interested in is between lines a and b, and these lines are shown on all the subsequent diagrams.*Slide 8Regions within the Pourbaix diagram. Here is the neutral point. The four regions around the neutral point are oxidizing and acidic, reducing and acidic, reducing and alkaline, and oxidizing and alkaline. How do these relate to natural environments?*Slide 9Environments are denoted on the diagram to the left, and actual environments (hundreds of measurements from a variety of sources) are plotted on the right. *Slide 10If we look at these two diagrams together, we can see the numbers of environments detected at various Eh-pH levels. Here's outdoor rain, with more acid rain up to the left. Marine open-water environments are here. Burial environments in waterlogged soils are here, and highly-reducing marine muds are here. We'll see this point cloud again.*Slide 11Here's the Pourbaix diagram for copper-water. Copper is stable in reducing environments. As things get more oxidizing, we encounter cuprite and tenorite. Above neutral pH, copper dissolves. These actions are summarized on -*Slide 12Which shows areas of corrosion, passivation, and immunity. We have done a lot of looking at various copper diagrams and combined them with looking at the tin diagram, shown as -*Slide 13You can see that tin is not stable anywhere within the region of stability of water. If we look at -*Slide 14You can see that tin passivates within the region of stability of water. We know it forms tin oxide, SnO2, and this can be a pretty good protective layer.*Slide 15The question of how intermetallics affect Pourbaix diagrams was first addressed by Wang Dongning in an appendix in her PhD thesis at Lehigh University. Here are two diagrams of the copper-tin-water system; the one on the right includes thermodynamic data for the epsilon, Cu3Sn phase. Areas of stability of copper and cuprite become larger with the inclusion of epsilon. Tin oxide can be seen throughout the diagram, but we did not remark on this at the time. This work seemed quite promising. In discussion with Dr. Notis, he pointed out that it had been done with FactSage software, which was fairly complicated. He suggested that we hire someone experienced with FactSage to draw the diagrams, and so we contracted with Arthur Pelton.*Slide 16Here's a diagram from Pourbaix of the copper-chlorine-water system at 355 ppm. You can see areas of stability of copper, cuprite, tenorite, nantokite, and atacamite/paratacamite. Contours for different concentrations of copper in solution complicate this diagram.*Slide 17Compare this with Pelton's diagram which we have seen before. Pelton didn't include Cu2O3 hydrate, so this line doesn't appear. However the rest of the diagram is quite similar, except that the epsilon phase is shown, which is only stable below the lower limit of stability of water. SnO2 is also prominent. I quote from Pelton's report: "It can be seen that solid SnO2 is the stable Sn-containing phase everywhere between the PO2 = 1 and PH2 = 1 lines in all figures except in a couple of cases at very low pH and under very reducing conditions. Tin is present in metallic form in the compound Cu3Sn (s) only below the PH2 = 1 line." We'll return to SnO2 behavior later.*Slide 5Solid phases are shown here. Many of the diagrams you will see simply have the chemical notation, but where we have been able to do it we have used a consistent color. Solid metal is orange, copper plus tin oxide is yellow, nantokite (copper (I) chloride) is light gray, the atacamite/paratacamite field (copper hydroxychloride) is yellow-green, cuprite (copper (I) oxide) is red, and tenorite (copper (II) oxide) is dark gray. Areas of dissolution are shown in light blue. It is interesting that tin oxide, SnO2, shows up everywhere on the diagram. We'll return to this later.*Slide 18Here's Figure 4 again, larger. When I discussed the work with Pelton I emphasized that alpha and delta were the two phases which we see in Chinese bronzes. Pelton redrew Figure 4 with delta as Figure 4c*Slide 19And it is identical to Figure 4. He also redrew at with pure copper and tin as Figure 4d*Slide 20And this is almost identical, except that some things are happening around the PH2 = 1 line.*Slide 21He also redrew it using epsilon at 40 degrees Celsius. Here there's a little difference in the field boundaries of the chlorides, but the diagram is substantially the same.*Slide 22Let's look at the effect of increasing chloride concentration. You can see here the log of the molality for chloride is minus three, minus two, minus one, and zero red, black, green, and blue. The chloride fields get larger as chloride concentration rises.*Slide 23If we hold the chloride molality at a log of minus one and decrease the concentrations of the miscellaneous species in solution, areas of solubility of copper increase. The nantokite field disappears in favor of the CuCl-2 field, and the atacamite/paratacamite field gets negligibly small.*Slide 24Heres a diagram of the copper-tin-carbon dioxide-water system, showing the solid phases; the only difference is the azurite/malachite field which is dark green. While Pourbaix shows separate fields for azurite and malachite, Pelton found disparities in the available thermodynamic data and concluded that one or the other, or both, could be stable and showed them as a combined field.*Slide 25Increased concentration of CO2 enlarges the field of stability of malachite.*Slide 26Decreased concentration of other reactants increases the areas of solubility. In more dilute solutions, solubility of copper takes place more readily.*Slide 27Here's the areas of stability for fairly concentrated solutions with a lot of CO2. If we superimpose the point cloud of environments -*Slide 28We see that most of the environments are clear of the tenorite field. You have dissolution, malachite, cuprite, copper plus tin oxide. This is exactly what we see on Chinese bronzes! Most of the Pourbaix diagrams which I have seen show the tenorite area extending well into the area of environments, but we very rarely see tenorite on actual bronzes. Its my opinion that the Pourbaix diagrams we see here reflect actual corrosion conditions better than those we have seen before.*Slide 29What may happen in an actual case? The bronze probably begins its existence in an environment rather like figure 18, with low concentrations of reactants and a little CO2. In normal pH the bronze is going to corrode, copper is going to disappear and tin oxide is going to form.*Slide 30This is what we actually see on an outdoor bell. The high-copper centers of the dendrites are corroded away, and the delta remains, probably with a coating of tin oxide on it. Well leave aside the question of sulfates for the moment.*Slide 31After burial, exclusion of air from the environment, and the accumulation of chlorides, one gets to a condition more like Figures 4 plus 22. Here you can see that corrosion in oxidizing conditions is going to lead to either azurite-malachite or the atacamites. As things get more reducing we enter an area of stability of nantokite or cuprite and eventually tin oxide with stable copper.**Slide 33In this plate from Wang Quanyu's book, one can see typical delta- removal corrosion, along with redeposited copper.*Slide 34Heres an SEM photograph of a 20% tin fang-ding from Xingan to remind you of the usual cored alpha plus alpha-delta eutectoid structure.*Slide 35A bright-field metallograph shows delta-preservation on the right and alpha-preservation on the left. A huge area of redeposited copper is present on the far left.*Slide 36In dark field, the differences between the two areas are even more striking. One can see tin oxide in the area on the right and cuprite (probably mixed with tin oxide) in the area on the left. One needs to do some deep thinking about exactly what's happening here, and these new Pourbaix diagrams will be very helpful.**Slide 38Conclusions:Looking at copper and tin together gives us a better picture of the thermodynamics of corroded bronzes.Taking the intermetallic compounds into account improves the picture.Realizing that SnO2 is the stable tin compound throughout the region of water stability helps explain the permeation of SnO2 throughout corroded bronze structures, often in different forms and mixtures with other corrosion products..*The stability of copper with SnO2 in reducing environments explains preferential corrosion of the delta phase in archaeological bronzes. Alpha will corrode in more oxidizing environments, but delta will corrode in reducing ones along with corrosion and redeposition of copper.Computer calculation of Pourbaix diagrams for specialized multicomponent systems is a powerful tool to promote understanding of corrosion behavior and to aid in conservation.Armed with these improved diagrams, we can look further at questions such as electrochemical kinetics and ion migration in archaeological corrosion.*Slide 39In closing let me repeat that Pelton's full report along with a version of this PowerPoint presentation are available on Lehigh's website at http://www.lehigh.edu/~inarcmet.Thank you very much.**