Historical Ceramics and Electrolysis: An Experiment

A few weeks ago, I conducted an experiment using electrolytic cleaning to rid stubborn chunks of iron corrosion from ceramic fragments. These ceramic sherds were recovered by Tim Bennett and his family at the historic Warner Pioneer Homestead (20LV334), from Feature 19. Due to post-depositional processes (presumably from being in the ground for a hundred years with oxidizing soils and iron artifacts), these potsherds were encrusted with ferric (iron) concretions, which made it impossible to mend the sherds together with others from the same vessels (i.e. cross-mending). Cross-mending ceramic sherds is useful for determining the forms of vessels and their full decorations, which can be functionally and temporally diagnostic. The ferric concretions that were encrusted on these sherds were impossible to detach with typical artifact cleaning procedures.

For my purposes here, I’ll very briefly summarize the electrolysis process (I posted a lab manual for the process on my blog, located here). In short, electrolysis is the process by which ferrous artifacts are cleared of ferric corrosion via electrical current. Electrolysis is useful for determining an iron artifact’s past function. Since iron oxidizes, electrolysis also helps conserve artifacts by preventing further corrosion. The artifact (or in chemist’s terms, the cathode) is connected to a battery’s negative charge. To ensure an efficient electrical current, the iron artifact is Dremeled in the places where the negatively charged clips will be connected. The battery’s positive charge is connected to a steel rod or mesh, which is called the anode. The cathode and anode are placed in a bath with distilled water and baking soda (which acts as an electrolyte); the battery’s electrical charge essentially attracts the oxidation from the artifact to the anode, aided by the baking soda electrolytes. The end result is an iron artifact that no longer has ferric corrosion on its surface. In order for the electrolytic cleaning process to work, there needs to be a strong connection from the iron artifact’s intact/corrosion-free surface to the negative charge, to facilitate a strong electrical current.

At the beginning of this electrolysis experiment, I decided that I would test three hypotheses. The first hypothesis is that the battery’s electrical current would not be able to flow through the ceramic well enough. According to the American Ceramic Society, there are trace amounts of iron in all earthenware and stoneware clay bodies (I can’t seem to find a source that can tell me exactly how much is in each paste category, but there seems to be many variables that alter the amounts). Some ceramic glazes also contain amounts of iron, all of which vary depending on the glaze’s chemical composition and intended color. Being that the sherds were all plain whitewares and hard-paste porcelains with lead-based glazes, I feared that there wouldn’t be enough iron in the pastes and glazes to ensure a strong enough connection through which the current could flow. My second hypothesis was that the sherds would not be able to withstand the electrical current at all, and that they would fall apart soon after the voltage was increased. Only one of the sherds exhibited any kind of decoration (gilded bands running along the rim); I was concerned that any overglaze decoration would flake away during the process. Finally, my third hypothesis was that the electrolysis process would be successful, and that the electrolysis process could be used by historical archaeologists in the future to clear ceramics of ferric concretions.

Before I began the process, I gave each sherd an identification number. Giving them arbitrary numbers helped me monitor each sherd’s overall progress in the electrolysis bath more closely. I also kept track of which cathode clips were attached to which sherd, and I kept detailed notes during the whole experiment. Sherd #s 1-5 are all from lead-glazed whiteware vessels, and sherd # 6 was made from hard-paste porcelain. Sherd # 2 has two overglaze gilt bands that run along the rim, which unfortunately can’t be seen very well in the photo above (for descriptions of what these paste and glaze categories mean, the Maryland Archaeological Conservation Lab does a nicejob). Since our electrolysis bath at CMU is only equipped with four cathode wires, I put only four sherds in the bath at a time. As I stated previously, the iron artifacts that I clean electrolytically are always Dremeled first in spots where the cathode wires can be connected to the artifact’s original surface under the corrosion, which facilitates a strong electrical connection. Since the sherds were only covered with ferric concretions in small areas, the Dremeling step was not necessary. It did occur to me that a gentle Dremeling on the concretions themselves would efficiently get rid of them, but a grinding stone bit or a steel brush bit of a Dremel would certainly scratch the paste and glaze surfaces, thus furthering the damage on the sherds.

Since I really did not want my second hypothesis to occur, I altered the typical electrolysis process for cleaning iron artifacts to be slightly gentler on these ceramic fragments. As a general rule of thumb, the amount of voltage/amperage needed to clean iron artifacts in an efficient amount of time is calculated as one amperage per every two square centimeters of the artifact. Additionally, the higher the voltage/amperage, the faster the ferric corrosion repels from the artifact’s core. Since I didn’t want to zap these sherds into oblivion, I refused to follow my own guidelines. Instead, I kept the amperage low at first (5V/0.5A), and I increased the amperage gradually once I knew that it was safe to do so. Since I never leave the electrolysis running overnight (the process needs to be monitored closely), I needed to disconnect the battery wires and take the sherds out of the bath until I could tend to them again. After successfully cleaning iron artifacts, they need to be “stabilized” by letting them simmer in distilled water for 2-3 hours, and then baked in an oven at 200 degrees Fahrenheit for an additional 2-3 hours. After they’re baked and cooled down, they are covered with a microcrystalline wax to ensure that they don’t oxidize again. Since the ceramics are obviously not totally made out of iron, I skipped these last steps.

Fortunately, my first and second hypotheses didn’t work out, and my third hypothesis was deemed successful. Collectively, after about 18 hours in the electrolysis bath at 10V/2A, the ferric concretions could easily be wiped away from the glazed surfaces and hard-paste porcelain. The concretions on the whiteware clay bodies were slightly more stubborn, although I could still persuade them to come off by gently picking at them with a dental pick. The gilt decoration on sherd # 2, fortunately, did not disintegrate. Furthermore, unlike full iron artifacts, leaving the sherds out to dry overnight did not make the corrosion any worse. The ferric concretions flaked away during this process, but the ferrous staining remained. To remedy this, I soaked them in white vinegar for a few hours while periodically giving them a gentle scrub with a toothbrush; however, no changes were made.

Before electrolysis

After electrolysis

To conclude, these results suggest that electrolysis can be used to clean ferric concretions from historical ceramics. The process is especially useful for vessel fragments, since the absence of concretions make the cross-mending and reconstruction of full vessels possible. Furthermore, the electrolysis process did not damage the surface treatment or decoration on the sherds in the sample. Before using electrolysis as a standardized practice, however, more testing needs to be done on ceramic fragments with other forms of decoration to make sure that the electrolysis process does not harm other types of surface treatment. Additionally, the samples need to be monitored to make sure that the ferric staining does not grow worse (especially since the sherds were not “stabilized” afterwards).

For more background information about the Warner Pioneer Homestead, implore you to check out Tim’s blog. I am indebted to both Tim Bennett and Sarah Surface-Evans, who originally came up with the idea and let me take control of the experiment.