It is now well documented that trace amounts of silver, gold, and platinum group metals occur in certain geothermal brine deposits. A significant body of literature has been generated over the last twenty years which describes attempts to quantify the frequency and levels of these precious metals in the brines.
There are numerous reports of the presence of precious metals in the silicious deposits around geothermal wells. Ellis and Mahon report 400 ppm silver and 10 ppm gold in the deposits around Steamboat Springs, Nevada. The same authors report up to 6% silver in silicious deposits around Salton Sea wells.
The common technique used by us and others to prepare brine for fire assay has been to boil it down to elemental salt. Not only would volatiles be lost, but adsorption on the walls of the vessel could reduce the measured levels of the metal. Wall adsorption can be minimized by the use of Teflon vessels. Additionally the salt, depending on its exact composition, is a hard, intractable solid which may melt in a grinder.
For our work we have arbitrarily divided brines into three basic types: oilfield, hot silicious geothermal, and cool geothermal with groundwater intrusion. We realize that this division is arbitrary and, in fact, a continuum exists. However, it is convenient to discuss these types. Most of the work in the literature describes hot silicious geothermal, while our work has been with oil-field and cool geothermal brine with groundwater intrusion.
The other type of brine which we have examined is cool geothermal brine with ground¬water intrusion. An area has been identified in the Carson Sink in Nevada where artesian water in certain wells contain 2-5 mg/l platinum and 1-3 mg/l gold. These wells have a total dissolved solids content of less than 1%, so AAS and ICP can be used for analysis. These results were substantiated by wet fire assay in our laboratory and by chemical precipitation and isolation of the gold and platinum.
It is evident that precious metal atoms in brine are highly complexed. Thio- and thio- arseno-complexes of gold have been described earlier. We now propose that organo or carbonyl complexes may also exist.
The technique used to determine efficiency of removal of platinum was to fire assay the brine before and after treatment. The fire assay value for platinum for the brine used in a series of tests was 1.5 mg/l. This brine was titrated, using a silver specific ion electrode with sodium sulfide after adjusting the pH up to 7.0. Alternatively sodium dimethyl dithiocarbamate was used as a precipitant. After treatment with the precipitant the brine was filtered through 0.7 m glass filter media.
Laboratory work indicates that gold precipites from the brine spontaneously on surfaces. We have not been able to determine whether the reaction is pH dependent and/or is dependent on the oxidation-reduction (redox) potential. The work indicates that to precipitate platinum the following minimum sequence is necessary: pH down, redox down, pH up, sulfide precipitation, filtration.
The removal of precious metals from geological brines involves, in part, the same steps used in wastewater treatment by the CPU-4000. Additional steps of one or more alternation of the redox potential and possible multiple change of pH may be necessary. A variant of the CPU—4000, consisting of two modules, each controlled by a microprocessor (which communicate with each other) has been constructed.
