To appreciate the development of industrial silver chloride processing it is helpful to outline some of its general properties. Since more comprehensive reviews are available, only basic concepts will be covered here.
Silver chloride is a white, curdy, crystalline solid compound. In the lab, it is prepared from a solution of AgNO3 by precipitation with NaCl solution. It is roughly 75% silver by weight. Silver chloride melts at a temperature of 455° C and boils at 1550° C. It is one of the few chloride compounds known to be very insoluble in water. Only PbCl2 and Hg2CL2 share this unique property, with PbCl2 being soluble in hot water. Thus, selective precipitation of silver from a mixed metal ion solution is an ideal method. Upon precipitating, silver chloride tends to agglomerate into larger and larger masses. This is enhanced by mild agitation. For improving solid settling rates and obtaining clear resultant filtrate solutions from liquid/solid separations, this agglomeration enhancement principle is important.
Electrochemically, AgCl is quite noble. The standard potential relative to hydrogen is 0.2223 V for the following reaction.
AgCl (S) + e- → Ag° + Cl-…………………………………………………(1)
As such, it may be reduced by redox couples with many less noble materials. This knowledge has been adapted by the battery industry. A prime example of this is the use of AgCl-Mg batteries for submarine torpedoes. Here the redox couple and the resultant electrical energy necessary to drive the propeller is started by immersion in seawater, an excellent electrolyte.
Also, because of limited solubility and favorable electrochemical properties, silver chloride finds use as a material in the fabrication of electrochemical reference electrodes. Compared to the calomel or hydrogen electrode, silver chloride is the preferred choice for most reference electrode applications.
Perhaps the greatest use of silver chloride exploits its photochemical properties. This is directly related to the ability of light energy to readily reduce silver chloride to silver metal. When exposed to light, silver chloride turns violet at first and finally black; it is decomposed to its elements. This is represented as:
2AgCl (S) + light → 2Ag° + Cl2……………………………………………………….(2)
As such, the silver chlorides, bromides and iodides present in film form the basis for photography.
Silver Chloride Reduction
In the realm of silver chloride chemistry, many reaction paths allow reduction to silver. These encompass hydro-, pyro- and electrometallurgical technologies as well as some more exotic treatment schemes. At the Sunshine facility, many potential systems have been evaluated and several have been implemented, each with increasing plant scale success. What follows is a description of these systems with a discussion of their relevant chemistry.
In the initial testwork done before the plant was started up in 1985, reduction of the silver chloride by metal powder was chosen as the best process. Of the known metal candidates, powdered iron was selected to be the reducing agent. It carried with it many favorable characteristics including ready availability, low cost, minimal environmental impact, apparent ease of control, and some removal of excess iron by magnetic separation. The basic chemistry can be described as:
2AgCl(S) + Fe° → 2Ag° + FeCl2(Aq) E =+0.8536 V…………………………………………(17)
In the process design it was decided to use a continuous flow system for the reduction process, rather than a batch system. This allowed for smaller process equipment, but eventually compromised the effectiveness of the process.
As with the precipitation system, the reduction system was composed of several CSTRs in series. In the original configuration, there were 3 of these reactors in succession. (Three more were added later in an attempt to lengthen residence times and conceivably raise the silver chloride conversion.) These were fabricated from fiberglass and each had a capacity of 190 liters. Stirring in each vessel was accomplished by a 0.25 kilowatt axial propeller. These reduction CSTRs were positioned so that they overflowed sequentially into each other by gravity. In practice, silver chloride slurry was pumped into the first CSTR using the material handling configuration described earlier for supplying the belt filter. Both iron powder and sulfuric acid were then added to this vessel to start the reduction reaction. Iron powder addition was controlled by monitoring the reaction with platinum redox probes immersed in the reactor. It was supplied to the first CSTR by way of a plastic feeder tube coming from a screw fed powder bin. The acid addition was controlled by pH probes immersed in the reactors. It was supplied as concentrated sulfuric acid from a day storage tank and kept the reactor pH below 3.0.
Primarily due to materials handling difficulties, the control of the continuous system was challenging even though the corresponding theory and lab testing had justified the process. In practice, system maintenance was cumbersome. The process control probes were subject to breakage and wear by the slurry. The iron powder chute plugged easily due to water vapor. The acid line plugged due to extremely low flows. Finally, prill formation was a major difficulty and made the material handling of the slurry frustrating.
Chemically, the system did work, though not up to standards achieved on the bench scale. Initially it was thought that a pH of 3 to 4 would be sufficient to prevent ferric hydroxide formation. However, in practice, a pH of 1.8 had to be maintained to achieve a filterable slurry. This was apparently due to the formation of some ferric hydroxide. The lowered pH in turn tended to consume iron, and a vicious circle of iron dissolution ensued. The sulfuric acid added to this chloride ion system resulted in a solution of hydrochloric acid. At this pH, the in-situ hydrochloric acid reacted with the metallic silver produced to reform silver chloride. Thus, overall conversion was lowered because of back reaction. The result was a system fighting itself. The silver sponge product ran as high as 15% silver chloride and 15% iron and proved almost impossible to dry and cast into anodes.
Next, zinc powder was tried as a reducing agent. Early on it was considered as a likely candidate for reduction. However, due to its higher price, it was dismissed from initial plant use. However, in 1988, research work was initiated to prove its worth as a replacement for the troublesome iron reduction method. The bench scale tests were encouraging and soon a design fell into place. The chemistry can best be described as:
2AgCl (S) + Zn° → 2Ag° + ZnCl2 (Aq) E = +1.2074 V……………………………………………(18)
In practice, the iron used before was quite easy to replace with zinc. As the reaction could again be controlled by platinum redox probes, essentially all existing reduction system equipment in the plant could be used without modification. This simplified the process changeover.
The use of zinc offered several noticeable improvements over iron powder. First, since it was more reactive, fewer tanks in series were found to be needed. This allowed quicker throughput with less equipment. Next, the use of fewer tanks in series alleviated the problem of back reaction of silver to silver chloride. Lab testwork had shown that silver left in a residual chloride ion solution, such as that found in the reduction process, would to some degree reform silver chloride. However, this reaction was very slow. Thus, removal of several reactor vessels in series shortened residence times to the degree that back reactions were minimized. Moreover, zinc consistently enabled realization of higher conversions of silver chloride to silver when compared to iron. Additionally, unlike iron, no acid addition was necessary during the reduction step to control pH. This conveniently eliminated the need for both pH probes and acid addition lines in the configuration. Lastly, the zinc reaction with silver chloride gave a much more distinctive endpoint relative to the use of iron.
Unfortunately, one particular problem noticed early on with zinc was the occasional occurrence of prills. As with the use of iron metal, these marble sized agglomerated particles could be a persistent problem totally disrupting production. It was found that high percent AgCl solids slurry in the reactor vessel was the main culprit. Soon it was found that a useful indication of percent solids in the reaction vessels was the temperature of the first reduction CSTR. As the reduction reaction is quite exothermic, and since it is the sole source of heat in the vessel, there is a direct correlation between how much solid is reacting and the vessel temperature. Thus, from experience, it was learned that keeping the first CSTR below approximately 45° C kept the solids percentage to a point that essentially no prills were formed. Thus, the flow of AgCl feed, and its relative dilution, were directly used by the operators to control the temperature and ultimately, to suppress the formation of prills.
Despite the success of the use of zinc as a reducing agent, this was not without its negative qualities. It still only converted around 97% of the AgCl to silver. It was capable of producing quantities of hydrogen gas when charged to the acidic solutions in the system. (In particular, this gas occasionally caused great excitement due to some loud, but relatively harmless, ignitions!). It was also reducing enough in character to potentially promote the formation of arsine gas in remotely occurring circumstances. In addition, it was a relatively costly reagent. It was capable of cementing out of solution other impurity ions that damage silver quality. Lastly, to achieve a high conversion of AgCl required an amount of zinc well above the stoichiometric value. This was mostly due to the side reactions of hydrogen formation, impurity ion cementation and zinc hydrolysis. Thus, research was continued to find a more suitable reaction scheme.
In 1990, bench scale testing was completed on a new process to replace metal powder reduction of silver chloride. In essence, the system revolves around the use of an alkaline hydrometallurgical reduction scenario. While this concept has been tried before this particular system is distinct in itself because it relies initially on carbonate rather than hydroxide as a source of alkalinity and because it is done on a production scale.
Accordingly, pilot scale testing was conducted to confirm the lab data. The performance of test was positive, funds were appropriated and the new process was installed.
The reagents used in the process chemistry are soda ash and dextrose. Heating to a temperature of 85-95° C is required for completion of the reaction. The perceived reactions are:
Na2CO3 (S) + H2O → NaHCO3 (Aq) + NaOH (Aq)………………………………(19)
AgCl (S) + NaOH (Aq) → AgOH (S) + NaCL (Aq)………………………………….(20)
2AgOH(s) → Ag2O(S) + H2O……………………………………………………………..(21)
12Ag2O (S) + C6H12O6 (Aq) → 24Ag° + 6CO2 (G) + 6H2O…………………..(22)
In plant practice clean AgCl solids from the horizontal belt filter are mixed in batch in a glass lined tank with water. The appropriate amount of dextrose is added and allowed to dissolve while the slurry mixes. Then, the soda ash is added carefully as a solid. As it dissolves and reacts, its exothermic nature provides some of the heat necessary to drive the reaction. To supplement this steam is used to bring the reaction to temperature. Essentially, if the stoichiometry is correct, the reaction is complete at this point. If necessary a platinum redox probe can be used to indicate the endpoint. From this process a high quality silver sponge of above 99% purity was readily produced.
Overall, compared to iron or zinc powder reduction, this new method offers several distinct advantages. The reagent cost is much lower and the product purity is much higher. The system is much less labor intensive due to batch operation. Lastly, the reagents are much less harmful to the plant, the operators and the environment.
As a technical sidelight to these three methods of reduction, the formation of “prills” in the early stages of plant startup resulted in the need for some alternate method of processing this material. Otherwise, as nothing else could be done with them, they would accumulate, requiring the costly practice of bulk storage. As the nature of these large AgCl spheres effectively prohibited metallic reduction, some means had to be developed to break them apart. This was accomplished in a small high shear mixing vessel. Affectionately known as the “prill buster”, this device sufficiently pureed the prills into a fine slurry. This solution could then be reduced in batch with the chosen metal powder and a portable platinum redox meter by an operator.