Dealing with Fire Scale

Fire scale or fire stain is a reddish purple toned 'bloom' or 'stain' that appears on silver/copper alloys such as sterling silver when they are heated in the presence of oxygen. It even occurs in gold alloys with high copper contents. As it generally appears in the form of blotchy patches following abrasive polishing it is seen as a blemish which destroys the clear reflectivity of finished silver surfaces.

It is usually observed best when the silver object is placed on a piece of white paper as then it shows up well. Sometimes it is chosen as a surface finish in order to emphasize texture as when a silversmith like Lois Betteridge uses fire scale to accentuate the visual depth of the hammer marks on a surface by polishing off the tops of the resulting texture leaving them a lighter grey than the reddish tones of the fire scale in the recesses.

Fire scale is caused by oxygen combining with copper present in the silver-copper alloy. A sterling silver alloy contains 925 parts out of a thousand parts silver and 75 parts out of a thousand parts copper. When the alloy is exposed to oxygen in the air at higher temperatures during annealing or soldering procedures the copper at the surface is converted to Cu2O, cuprous oxide which has a reddish colour to it and then to cupric oxide CuO which is black. It is however not just the metal at the very surface that is affected but deep inside the metal as well.

Silver has the ability to absorb oxygen at high temperatures and conduct the oxygen to the interior of the metal itself where it can bond with copper atoms present thus causing deep fire scale to occur. Alloys such as the 'reticulation silver' alloy with 800/1000ths parts silver are especially susceptible to oxidation. In alloys of lower silver content this effect is limited by the fine grained eutectic structure that occurs in them. In low gold content silver/gold/copper alloys fire scale also occurs as the silver in the alloy conducts the oxygen to the interior of the metal where it forms Cu2O. Sterling silver is the standard alloy in North America and is well known for it's tendency to form fire scale during soldering and annealing operations.

Industry Approaches

Industry deals with the problem of fire scale in a number of ways, most of which eliminate oxygen from the surface of the metal during a heating procedure. The solutions used include heating in a hydrogen atmosphere where copper oxides present can actually be converted back to copper or in ammonia atmospheres. Production line work is sometimes placed in titanium jigs on a moving conveyer belt which enters a hydrogen, nitrogen or ammonia gas atmosphere where heating for soldering occurs. The object then cools down in the same atmosphere and emerges at the other end soldered and ready to be removed form the jig-with no fire scale.

Other approaches involve the use of fluxes sprayed through the flame while it is used on a piece or (according to an Australian silversmith I met in Sydney) by having a precise mixture of gasses which prevents fire scale and makes flux unnecessary though I am not sure why there would be no fire scale if the flame is moved around in use.

Sometimes if a minimal number of solderings are to be done then a protective coating is placed on the work so that oxygen cannot reach the surface during heating. The protective coatings are usually based on boric acid which is deposited evenly over the work. One production shop I worked in in Toronto used a large wire french fry basket in which all the items to be soldered were placed. This was held suspended in a rapidly boiling solution of boric acid and some borax mixed with water. After the items had heated up the basket was removed and while tossing gently the items were dried in the basket over a large flame. They were then dumped out on the soldering board for setting up to solder. Of some interest in the boric acid approach is that where the borax based soldering flux is used on the same piece fire scale is likely to occur.

Another approach used in industry is to make a solution of boric acid, some borax and solvent alcohol. Because of the potential for accident this approach is not recommended. The water based approach is much safer and should be used if possible. In the alcohol based procedure a container with a lid is chosen. Unbreakable ceramic or heavy plastic is recommended as the danger of fire by accidental breakage of a glass container is very high. A halfway mark is made on one side of the container and the other side of the container is marked in thirds. Boric acid powder or crystals (from the drug store) is poured in until the halfway mark is reached. Then borax powder is added until the top third mark is reached (These proportions would be good for the water based approach as well.). Solvent alcohol is added until the container is almost full. The mixture is shaken to thoroughly mix it and then the work is dipped in it, quickly placed on the soldering block and set alight. The alcohol burns with a greenish flame and the heat of it's burning deposits an even, thin protective layer over the surface of the metal.

Again, where regular flux has to be applied for soldering parts together fire scale will have a tendency to appear. If one is just annealing a single time, as when a highly polished object is repaired the polish will be almost completely retained on the metal. One might expect that one could prepolish pieces before assembly and perform multiple solderings without affecting the polish but this is not really possible as the fire scale preventative is not a hundred percent effective and after several heatings there tends to be an accumulation of fire scale such that the polishing time required is the same as when no preventative was used. In the Industry therefore this approach is almost always used only for repairs or when there are only one or two soldering steps involved. If by chance the alcohol in the container catches fire the solution is very simple: calmly put the lid back on which extinguishes the flames. A panicked reaction to flames in the container which knocks it over can result in a severe fire.

A further hazard note on an alcohol based fire scale preventative solution is that alcohol fumes are toxic. While this would make little difference for the occasional jewellery scale object larger objects and great surface areas treated in this manner give off a lot of fumes. A teacher of mine who was working on a large liturgical piece which was regularly swabbed with a similar solution went to the doctor for his annual checkup to be questioned as to whether he had a drinking problem (which he strenuously denied). It turned out that the alcohol fumes were affecting his liver and the addition of an alcohol fume mask to his workshop (and one would hope better ventilation) solved the problem.

Sometimes objects are bombed or electrostripped, that is function as an anode usually with high current densities in a stripping bath of some kind, often cyanide compounds in industry, sometimes a 1-2% nitric acid mixture, sometimes mixtures of other chemicals.

A common solution chosen for production pieces is to electroplate the pre-polished surfaces with silver or with rhodium which has a hard, dead chrome-like color. It also has an unfortunate tendency when heated, (as during a repair) to burn to a hard and nasty black color which requires polishing to remove it. Silver plating is fine for smaller items like brooches but on surfaces where wear is a problem the fire scale will eventually emerge from beneath the plate again. For silver an individual polishing of each piece to remove fire scale would require too much labour and raise the unit cost to the producer over what is competitive in the industry in North America. This does not mean that hand made work need attempt to be competitive with such mass produced items, as it often addresses a slightly different market sector.

Small Studio Approaches

There are then several approaches to fire scale. One can remove it by abrasion (polishing, sand-blasting, filing, emerying) and remove all metal to below the level of the fire scale. This is time-consuming and can be wasteful of material and if polishing is hurried and careless can lead to a loss of detail or crispness in the piece. If polishing is chosen as a solution the best approach is to begin with bristle brushes on the polishing machine and use a lot of tripoli or other cutting compound, as it is the compound that does the work, not the brush. This can with practice be controlled by guiding the path of the bristles so that little detail and a minimum of metal is removed during polishing. Polishing demands a high degree of concentration to understand it's control factors. Traditional polishing apprenticeships lasted from three to five years; not that this is the best way to learn, only that there is obviously a lot to learn. Brushes combined with thoughtful use of other buffs can offer sharp edges and minimally disturbed detail even though the fire scale has been removed.

The surface can be protected from oxygen in some manner. While some people have built appropriate gas atmosphere kilns this is beyond the interests and skills of most individuals in the field. In medieval times and still in many parts of the world work is soldered and heated on a charcoal forge into which air is blown through tubes or pumped from bellows or with a motorized fan (for example a vacuum cleaner) which causes the heat to rise by supplying oxygen. This system produces a highly reducing (low oxygen or oxygen devouring) atmosphere which limits oxidation somewhat. The most basic thing you can do in a small workshop to limit oxidation is to use a slightly gassy flame (more gas than air) which produces a local reducing atmosphere on the surface where the unburned gas present seeks to bond with available oxygen thus preventing it from entering the surface of the metal.

This is one of the reasons that goldsmiths in the west traditionally soldered on charcoal blocks. It was a way of obtaining a reducing atmosphere-the charcoal present burns forming CO2 and removes oxygen from the vicinity of the metal surface. It's burning also heats the metal and provides a high ambient temperature which then allows easy local heating of areas for soldering. Charcoal blocks are however expensive and it is difficult to completely envelop a piece with such a reducing flame during a heating operation so at best one can in this way limit the degree to which fire scale occurs. Usually a boric acid based mixture similar to those described above is the most appropriate solution for a small workshop. Alcohol based solutions work well but are a severe fire hazard and their fumes may present problems. A water based solution is safer.

A choice often used in art schools is to 'bring up the fine silver', a misleading term for depletion silvering. This involves the repeated oxidation of surface or near surface copper in the alloy by heating until the surface discolors or the flame as it leaves the object turns yellow (a temperature indicator of about 800oF) followed by a pickling in an acid solution that attacks copper oxides preferentially to silver and copper thus removing the copper from the surface and leaving the fine silver behind. Many such pickling solutions are used, from vinegar and salt to the jeweller's standard pickle, Sparex®, which is sodium bisulfate and is often sold as 'swimming pool acid'.

The effect is to give the surface a dead white covering which although attractive tarnishes in time and is therefore not an easily maintained permanent finish. The surface may be brass brushed with a little soapy water in between each heating and acid pickling sequence and this renders a smooth, bright, satin finish. If a high polished surface is desired the metal is prepolished (and usually one can finish the polishing and eliminate the fire scale completely during this stage-the heating and pickling may be extra work) and then the surface is depletion silvered as described above. Then it may be very lightly buffed with a new buff with only a faint trace of rouge on it (so as not to polish off the fine silver and expose the fire scale beneath) or burnished in a burnishing tumbler with steel shot. Like an industrial electroplating it can be worn off and the fire scale beneath appear again in the worn areas.

There are also various less complex versions of stripping used by goldsmiths. The use of a fume hood is essential in any procedure involving chemicals. Several involve the use of an acid with an oxidizing agent. A sodium dichromate pickle with a low percentage of sulfuric acid has been mentioned as a fire scale remover, though I have not used it for that. It's action would be similar to a 50/50 mixture of sparex pickle and hydrogen peroxide solution from the drug store, a mixture which is usually used to remove the copper flashing from brass. I am unsure about any efficacy as a fire scale remover. Another is a nitric acid dip which tends to attack solder seams first [nitric acid to water 1:5. Always add acid to water]. Another method reported by Doug Zaruba, also one I've not yet tried is to use a thick aluminum pot and simmer a lye solution in it. After a polishing operation that still leaves blotches of fire scale on the surface the object is suspended in the pot for some time and the residual fire scale is removed by the solution in the presence of the aluminum. The pot eventually dissolves, hence the suggestion for using a thick one.

In Conclusion

There are then several approaches to fire scale: avoidance ( creating oxygen free conditions during heating), abrasive removal ( removal by polishing or other abrasive material removal), chemical removal (removal of the surface of the metal to remove fire scale by bombing or stripping) and covering up (plating or depletion silvering by oxidation of surface and near surface copper and removal of the oxidized copper by acids to cover the fire scale over with fine silver or in some cases; plated with rhodium).

Personally I deal with sterling's potential for fire scale in two ways, depending upon the potential wear on the surface. If possible I brass brush my work which is often pins and depletion silver it, choosing that as my finish, usually using burnished edges as a frame of highlights for the object. If a silver piece is subject to wear (a ring for example) then it must be polished and I usually use bristle brushes, and then larger cotton buffs to smooth the surface and finally if necessary flex shaft brushes and buffs. Most jewellery objects can be polished to remove fire scale using bristle brushes and lots of compound in 5-10 minutes. Remember it is the compound that does the work, not the buff. If it is a single soldering or just a few then I will use a boric acid fire scale preventative solution on the work. It depends on which method looks like the shortest in time and trouble for the particular piece.