About Chemical Innovation - Subscription Information
January 2001
Vol. 31, No. 1, pp. 49–50.
Chemscripts

Table of Contents

Silver removal and recovery from effluent streams

The conventional technology for recovering silver from aqueous effluent streams is based on electrodeposition. For example, starting from a 5 g/L solution, silver is deposited on plate cathodes in two separate processing steps. Final concentrations <100 ppm are difficult to obtain; low current efficiencies, long processing times, and brittle deposits are typical problems.

In view of tightening regulations that specify lower effluent concentrations for silver—upper limits of 20 ppm are under discussion in the European Union—more powerful removal technologies will be required.

In addition, an advanced technology that achieves the silver removal in a one-step unit operation, along with lower investment and utility costs, should be of interest to processors of silver-containing streams. Our company has developed such a technology, and we describe it as it is applied to photographic effluents.

The process

The underlying principle is the relative “nobility” of metallic silver. A low reduction potential is required to reduce silver ion to the metallic state, and it is similarly easy to induce its crystallization and precipitation from the liquid phase.

Figure 1. Pilot plant plug-flow pipe reactor.
Figure 1. Pilot plant plug-flow pipe reactor.
The aqueous stream is heated to 100–150 °C in a pressurized reactor with a maximum operating pressure of ~10 bar (Figure 1). Before this step, an inexpensive, proprietary reducing agent is dissolved or slurried in the liquid phase.

When heated to the desired temperature, silver is reduced

to the metallic state, and a slurry of small crystals is formed. The residence time and temperature are adjusted to obtain the desired final silver concentration. Finally, the slurry is cooled to <100 °C and depressurized. The heat is recovered and used to warm the next batch of influent.

The cooled slurry, from which the silver particles settle out quickly, is passed through a simple filter. The filter cake is dried and melted to produce the desired silver bullion. Final silver concentrations of 20 ppm can easily be attained in the effluent.

Our company proposes to use a non-backmixing reactor system, as distinct from the standard continuously stirred tank reactor. We use an inexpensive plug-flow pipe reactor with controlled heat transfer and stationary mixing elements. With this system and a starting silver concentration of 2.5 g/L in the influent, the concentration can be reduced to 20 ppm in ~60 min at 120 °C after controlled heating to the operating temperature in 35 min.

The purity of the silver cake, and the eventual metal bullion, depends on the following factors:

  • More noble metals. Under the given conditions, only gold and, to some extent, copper will be reduced as well and contaminate the silver bullion.
  • Less noble metals. Metals such as iron are not reduced to the metallic state. Careful process control prevents their precipitation as hydroxides or hydrates into the silver phase, so a pure filter cake is produced. In the example given, the original iron concentration was ~5 g/L, virtually all of which remained in solution.
  • Organic matter. Often organic matter, such as metal-complexing agents, is present in these streams. Upon heating, it may disintegrate or precipitate as well. Normally, it is combusted in the subsequent silver melting step and does not affect the quality of the product.

Investment and operating costs

The processing setup is simple. It consists of a pump, the pipe reactor, a depressurizing valve, and a filter unit. For the plug-flow pipe reactor, a simple engineering system is chosen: Using a tube–shell heat exchanger design, several straight pipe segments are connected to obtain a reactor element with a length of, say, 100 m (e.g., 10 pipes of 10 m each). Three such units would be used for heating, maintaining temperature, and cooling. The heating–cooling control of the three units is integrated.

Pipe diameter and length depend on the processing requirements (volume per hour), and thus determine the capital investment. In any case, the investment and operating costs will be considerably lower than that for the conventional electrolysis cells. For the 2.5 g/L stream in the example, we estimated the total variable costs at U.S. $1/m3.

State of the technology

The development work on the process was carried out on a pilot scale: A pipe reactor of up to 250 m in length and operating temperatures of 200 °C is available. The process can be scaled up without problems to volumes of several cubic meters per hour. The process can be optimized in the pilot plant for the specific requirements of different streams. Based on the information generated, turnkey plants can be delivered.


Volker G. Aurich is general manager and Joep I. A. Koene is project manager at CES B.V. (Wildekamp 1B, 6704 AT Wageningen, The Netherlands; +31-317-421114; fax +31-317-424716; ces@bart.nl).
Return to Top || Table of Contents