ACTA ACCLA, December 2013

Recipe for common bronze disease treatment found on several websites is very wrong

Merrill Gibson

Bronze coin of Constantius II showing signs of active bronze disease (the blue spot on the Emperor's neck). Without treatment, the corrosion may worsen causing damage similar to that seen on the reverse. Photograph by Merrill Gibson.

Successfully treating coins that have bronze disease is difficult. Simple methods (such as soaking in distilled water and then drying) are unsuccessful in a large number of cases, and most other treatments involve using chemicals (e.g. benzotriazole) which carry an element of danger or are expensive or difficult to obtain. An exception is the treatment of bronze disease with sodium sesquicarbonate, a non-dangerous chemical which can purportedly (although in reality cannot) be easily made using household chemicals. I have lately noticed that the "recipe" for the production of sodium sesquicarbonate from household chemicals that is given on many websites will produce a solution that is much different than the authors of these websites believe, and could quite possibly lead to harmful consequences for a treated coin.

Furthermore, these websites all seem to have copied the same erroneous recipe, or, more likely, copied each other.

The solution that the instructions on all these websites intend to produce, and the solution that is recommended by scientifically rigorous authors such as Plenderleith and Werner, 1971 [ 1 ] is 5% by weight sodium sesquicarbonate in water. The recipe that is given by the erroneous websites is to take 10.6 grams of sodium carbonate, add it to 8.4 grams of sodium bicarbonate (the common "baking soda"), and mix with 100 ml of water. It is easy to see where this recipe comes from. The molar weight of anhydrous sodium carbonate (also known as soda ash) is 106 and the molar weight of sodium bicarbonate is 84, so if you add 10.6 grams of anhydrous sodium carbonate to 8.4 grams of sodium bicarbonate and add water to a volume of 100 ml, you will produce a 1 molar sodium sesquicarbonate solution (sodium sesquicarbonate, in water, is just an addition compound of one sodium carbonate "stuck together with" one sodium bicarbonate molecule). However, 1 molar sodium sesquicarbonate is NOT a 5% solution - it is much, much stronger, well over three times as strong. Somehow, somebody misinterpreted 1 molar as 5% by weight, and people have been reproducing this error over and over.

Another big problem with some of the aforementioned websites is that they recommend using washing soda for the sodium carbonate in the above formula. Washing soda is a powdered household chemical found in many grocery stores; however it is NOT ANHYDROUS sodium carbonate. Actually, it is impossible to figure out the correct weight of washing soda to use when trying to make sodium sesquicarbonate from washing soda and baking soda. This is because sodium carbonate is found in several forms, different hydrates of the compound (that is, with several different numbers of water molecules attached to each Na2CO3 molecule). Sodium carbonate exists in hydrates with 0 (anhydrous), 1, 2, 5, and 10 water molecules per Na2CO3 molecule. If you don't know which form of sodium carbonate you are dealing with, it is impossible to measure out the right amount. The Merck Index, a standard chemistry reference, states that the term "washing soda" refers to sodium carbonate decahydrate. In sodium carbonate decahydrate, ten molecules of water are attached to each molecule of Na2CO3. However, as the Merck Index also states, sodium carbonate decahydrate rapidly effloresces on exposure to air. This means that it rapidly loses water, transforming into a lower hydrate. Because of the way washing soda is sold in stores, typically in cardboard boxes, it can be expected that a large portion of the material in the box is in an unknown state of hydration (although none of it would be in the anhydrous state, which ABSORBS water, as also noted in the Merck Index).

Several of the erroneous websites give warnings that "a 5% solution will remove any patina" or may remove a coin's patina, so they give recipes for producing a 2% solution, which are just their 5% recipes scaled down to 2/5th strength. However, even their "2%" recipes produces a solution which is in fact still significantly stronger than a true 5% by weight solution (if soda ash is used) and has a large amount of unreacted sodium bicarbonate in the resulting solution (if washing soda is used). It is quite possible that the reason that the website authors have encountered patina harm from sodium sesquicarbonate solutions is that the solutions they used were much different than what they believed them to be.

A true 5% by weight solution of sodium sesquicarbonate in water is what should be used for bronze disease treatment. This is the solution that has been investigated by the British Museum and found to be effective and safe.

I believe that by far and away the best way to make a 5% by weight sodium sesquicarbonate solution is to simply buy powdered sodium sesquicarbonate and then, using whatever weight units are handy, add a convenient quantity of sodium sesquicarbonate to 19 times its weight of distilled water. For example, add 5 grams of powdered sodium sesquicarbonate to 95 ml of water. Sodium sesquicarbonate is fairly easily acquired - for example, on US eBay, there are currently a couple of sellers selling it. Several chemical supply houses also carry it.

Thanks to Mike Connor of the Ancient Coin Club of Los Angeles for providing the information on sodium carbonate hydrates from the Merck Index, and for posting this article on the ACCLA website.

Merrill Gibson, Apollo Numismatics


[ 1 ] Plenderleith, J. and Werner, A. E. A. 1971. The Preservation of Antiquities and Works of Art. Oxford University Press.


ACTA ACCLA edited by Michael J. Connor.