Environ. Sci. Technol. 2002, 36, 840-845
Photochemical Changes in Cyanide Speciation in Drainage from a Precious Metal Ore Heap CRAIG A. JOHNSON,* REINHARD W. LEINZ, DAVID J. GRIMES, AND ROBERT O. RYE U.S. Geological Survey, Box 25046, Denver, Colorado 80225
In drainage from an inactive ore heap at a former gold mine, the speciation of cyanide and the concentrations of several metals were found to follow diurnal cycles. Concentrations of the hexacyanoferrate complex, iron, manganese, and ammonium were higher at night than during the day, whereas weak-acid-dissociable cyanide, silver, gold, copper, nitrite, and pH displayed the reverse behavior. The changes in cyanide speciation, iron, and trace metals can be explained by photodissociation of iron and cobalt cyanocomplexes as the solutions emerged from the heap into sunlight-exposed channels. At midday, environmentally significant concentrations of free cyanide were produced in a matter of minutes, causing trace copper, silver, and gold to be mobilized as cyanocomplexes from solids. Whether rapid photodissociation is a general phenomenon common to other sites will be important to determine in reaching a general understanding of the environmental risks posed by routine or accidental water discharges from precious metal mining facilities.
Introduction The environmental risks of cyanide releases from precious metal mining facilities are receiving increasing scrutiny partly as a consequence of large spills such as those that have occurred in recent years at Baia Mare, Romania, at Kumtor, Kyrgyzstan, at Omai, Guyana, and at Summitville, CO (1-6). The cyanide anion (CN-) is a key hydrometallurgical agent that allows gold and other precious metals to be recovered from ores in a highly efficient manner. The environmental concern stems from the potential toxicity to aquatic life. Cyanide can be lethal to sensitive fish species at concentrations as low as 0.1 mg L-1 (7); several recent spills have had high enough cyanide concentrations, up to several tens of mg L-1, to threaten ecosystems in local streams and rivers (1-6). Because cyanide leaching continues to be the leading technology for recovering gold from ores (8), routine or accidental solution releases from precious metal mining facilities will undoubtedly continue to cause environmental concerns related to their cyanide contents. The environmental impact of cyanide in discharged process solution depends on the speciation of cyanide in the solution and on the behavior of the different species in receiving waters. Free cyanide, which is predominantly CNor HCNaq above or below the pKa of 9.3, respectively, is the most toxic of the cyanide species. An important attenuation mechanism for free cyanide is volatilization due to the high * Corresponding author phone: (303)236-7935; fax: (303)236-4930; e-mail:
[email protected]. 840 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 5, 2002
vapor pressure of HCNaq, which is favored over CN- at the circumneutral pH characteristic of most surface waters. Cyanide also forms stable complexes with many metals. These are traditionally classified according to their stabilities as weak (cyanocomplexes of Ag, Cd, Cr, Cu, Hg, Mn, Ni, Zn) or strong (cyanocomplexes of Au, Co, Fe, Mo, W, Re, Pt-group elements), although this classification scheme is probably oversimplified (9). In general, the risk that cyanometallic complexes pose to aquatic life arises from their dissociation to form toxic free cyanide. For the strong complexes, the dissociation constants that have been measured are small enough and the dissociation kinetics slow enough to suggest that the resulting free cyanide would be too low in concentration to be of environmental significance (10). For the cyanocomplexes of Fe, Co, and Cr, dissociation can be enhanced by photochemical reactions (9). The photochemical enhancement is considered by most workers to be insufficient to produce environmentally significant free cyanide (10), although the reverse conclusion has been reached in an experimental study (11). An important attenuation mechanism for metal-complexed cyanide is dissociation and volatilization of the resulting free cyanide, but other attenuation pathways are also possible including adsorption onto iron oxides, organic matter, or clays and precipitation of cyanometallic compounds. Cyanides are also subject to oxidation by either abiotic or biological reactions that ultimately produce carbonate and ammonium or nitrate (10). Despite the large amount of research that has been carried out on cyanide attenuation reactions (9, 10), their rates remain difficult to predict for complex natural waters and can be difficult to determine empirically due to problems associated with sampling and analysis (12-14). Our objective in this paper is to report empirical observations of cyanide behavior at an inactive gold mine where a rinsed ore heap was draining cyanide-bearing water to a holding pond. The data, which were obtained using improved sampling and analytical methods, reveal regular diurnal cycles in cyanide speciation and in other chemical constituents. The cycles are the first empirical evidence that photochemical reactions can in fact influence the toxicity of actual effluents.
Experimental Section The study site is at the inactive Standard Hill mine near Mohave, CA where gold was recovered by heap leach cyanidation. The exhausted ore heap contains 1.5 × 109 kg of material stacked to a height of 24 m. Ores were composed of felsic subvolcanic rocks with quartz and calcite veins in which gold was present mainly as the native metal with sparse sulfide minerals. Cyanide application was carried out by dripping a dilute cyanide solution onto the top surface of the heap. The practice was halted many years prior to this study, and process waters were then recirculated through the heap with no cyanide amendment, with the effect that dissolved cyanide concentrations declined over time. After 5 years of rinsing, the heap was permitted to drain to a holding pond. Samples for this study were collected after 5 months of draining when the effluent flow from the heap as a whole was 0.6-0.8 L s-1. Effluent was collected from drain pipes that emerge from the toe of the heap, from an open receiving channel downstream of the heap, and from the holding pond into which the channel feeds (Figure 1). The open channel collects flow from several pipes that together drain the entire heap. Effluents from different drains are chemically similar but not identical. For this study, samples were taken at two of 10.1021/es011064s Not subject to U.S. copyright. Publ. 2002 Am. Chem.Soc. Published on Web 01/24/2002
FIGURE 1. Schematic map view of the study site showing the outline of the ore heap, the locations of drains, the layout of the open channels, and the effluent holding pond. A and B are representative drains that are discussed in the text. the higher flow drains that were known from earlier analytical work to be representative. The open channel is underlain by impermeable plastic that is covered by sand- to gravel-sized sediment derived mainly from the heap surface by sheet flow during precipitation events or by wind action. Effluent flowing in the channel was clear in appearance and averaged a few cm in depth across the ∼50 cm channel width. Samples were collected just before sunrise, at midday, and at dusk on a sunny October day in 1999 and at 4 h intervals through a sunny day-night-day cycle the following July. Waters were filtered (0.45 µm) into Nalgene bottles that had been wrapped
in aluminum foil to exclude sunlight and were immediately transferred to a cooler. pH was measured at the time of collection. Total cyanide, meaning the sum of free cyanide (CNplus HCNaq) and all cyanide occurring as metal complexes, was analyzed using a modification of the USEPA reflux distillation method (15). Without modification, this method is known to give incomplete yields for cyanide complexed with Au, Co, Pd, Pt, or Ru (14, 16, 17). Our modifications were to extend the boiling time to 8 h and to expose the boiling sample to UV radiation. The changes were made to obtain complete yields for cyanide complexed with Co (14), which is a metal that can be present at significant concentrations in cyanidation solutions (18). Whether recoveries were improved for Au, Pt, Pd, or Ru cyanocomplexes is unknown, but this is of no consequence for the present study because Au was low enough (Table 1) and the Pt-group elements were probably low enough (see below) that the quantity of cyanide that could have been complexed with these metals is negligible. Weak-acid-dissociable (WAD) cyanide, meaning the sum of free cyanide and cyanide contained in weaker complexes with Ag, Cd, Cr, Cu, Hg, Mn, Ni, or Zn, was measured using the picric acid method ((19), with unpublished modifications by P. Emsbo). Our previous experience had been that the best way to ensure accurate results for WAD cyanide is to forego chemical preservatives in favor of immediate analysis (14). Accordingly, the measurements were carried out within
TABLE 1. Dissolved Constituents (mg L-1) in Representative Solutionsa drain pipe constituent Na K Ca Mg Ba SO4 Cl F SiO2 Al Mn Zn Fe Pb Cu Cd Co Cr Au Ag Ni W Mo As Sb Se U V Re NO3 NO2 NH4 SCN CNO CNWAD CNtotal Fe(CN)6 pH a
predawn
2400
0.18 7.56
midday 1100 3.9 390 24 0.0006 780 850 0.9 35
