AgCn Recovery

Question:

What is state of the art for "treatment" of silver cyanide rinse waters? We are installing a silver plating line with the following characteristics: 4 oz/gal (30 g/liter) of silver cyanide, 16 oz/gal (120 g/liter) of potassium cyanide, 75F (24C) operating temperature, no air agitation, and estimated dragout of 1 gph or 3.8 liters/hr.

One option is that I could capture the dragout in double counterflow dragout recovery tanks, plate out the silver and pass the rinse waters onto alkaline chlorination. However, I would rather set up a six-stage counterflow rinse scheme recirculating at 10 gph through a reverse osmosis (RO) system. The first recovery rinse would already be at 10% of bath concentration. If the RO could concentrate that by a factor of 10, then I'd have reconstituted the bath allowing me to return the dragout to whence it came. P.T.

Answer:

If you already have a wastewater alkaline chlorination system for the oxidation of cyanides, I would recommend that you implement your first option with these modifications. Instead of two counterflow recovery tanks, I would increase it to four recovery tanks and on a periodic and batch basis flow to an off-line plate-out tank. After you have plated as much of the silver from the rinse water, the rinse water would then be discharged to the alkaline chlorination system. Also, I would recommend two counterflow flowing rinse tanks after the recovery tanks for final rinsing with the very small flow going to alkaline chlorination.

If you do not have an alkaline chlorination system already in place, then I would recommend that you pursue your second option, basically a closed loop system for silver cyanide plating. With six counterflow recovery rinses, the concentration of silver in the last rinse will be less than 0.05 mg/liter while the cyanide will be less than 0.1 mg/liter cyanide indicating superior rinsing.

Your greatest challenge is to "make room" for the recovery dragout since your plating tank is operating at room temperature and will have essentially no significant evaporation. While it is true that the RO unit should be able to concentrate its reject stream 10 times to the same strength as the plating bath, and the reject flow rate equals the dragout rate of 1 gph, the tank also receives dragin from its preceeding rinse tank. Therefore, you will likely need to "make room" for the recovered dragout solution. My recommendation is to use an off-line evaporation tank in order to "make room" for the recovered dragout. You will likely need to remove 2-3 gph of plating solution out of its tank, evaporate 1 gph, and then return the remaining solution back to the plating tank. It takes about 9,500 Btus to evaporate 1 gal of water.

While this system will essentially return all (about 99.9%) of the silver and cyanide back into the plating tank, keep in mind that it will also return all contaminants as well. These contaminants include carbonates from the breakdown of cyanide, breakdown compounds of any organic additives and contaminants in the water supply. It is critical that the water entering your dragout recovery tanks (a makeup rate of 1 gph is needed) be deionized water either through a separate RO system or exchangeable ion exchange columns. Also, you may want the rinse preceeding the plating tank to be fed by deionized water to minimize the dragin of water supply contaminants.

Lastly, while great advances have been made in recent years for the application of membrane separation technology, such as RO, in the finishing industry, be sure to thoroughly evaluate the type of membrane and its ability to withstand the plating solution's chemistry as well as cleaning frequency. You will be greatly disappointed if the RO works, but requires cleaning too frequently due to membrane fouling from contaminants such as carbonates. Collection, treatment and disposal of this cleaning solution is another consideration.