Environ. Sci. Technol. 2005, 39, 3849-3854
Iron(VI) and Iron(V) Oxidation of Copper(I) Cyanide V I R E N D E R K . S H A R M A , * ,† CHRISTOPHER R. BURNETT,† RIA A. YNGARD,† AND DIANE E. CABELLI‡ Chemistry Department, Florida Institute of Technology, 150 West University Boulevard, Melbourne, Florida 32901, and Department of Chemistry, Brookhaven National Laboratory, Long Island, Upton, New York 11973
Copper(I) cyanide (Cu(CN)43-) in the gold mine industry presents the biggest concern in cyanide management because it is much more stable than free cyanide. Cu(CN)43- is highly toxic to aquatic life; therefore, environmentally friendly techniques are required for the removal of Cu(CN)43from gold mine effluent. The oxidation of Cu(CN)43- by iron(VI) (FeVIO42-, Fe(VI)) and iron(V) (FeVO43-, Fe(V)) was studied using stopped-flow and premix pulse radiolysis techniques. The stoichiometry with Fe(VI) was determined to be 5HFeO4- + Cu(CN)43- + 8H2O f 5Fe(OH)3 + Cu2+ + 4CNO- +3/2O2 + 6OH-. The rate law for the oxidation of Cu(CN)43- by Fe(VI) was found to be first-order with each reactant. The rates decreased with increasing pH and were mostly related to a decrease in concentration of reactive protonated Fe(VI) species, HFeO4-. A mechanism is proposed that agrees with the observed reaction stoichiometry and rate law. The rate constant for the oxidation of Cu(CN)43- by Fe(V) was determined at pH 12.0 as 1.35 ( 0.02 × 107 M-1 s-1, which is approximately 3 orders of magnitude larger than Fe(VI). Results indicate that Fe(VI) is highly efficient for removal of cyanides in gold mill effluent.
After the extraction of gold from ores, cyanides are leached into the environment as effluents and as solid mine tailing. Each year, more than one billion tons of gold ore are leached with cyanide. There is an increasing risk to the environment from spills such as those at Baia Mare (Romania), Kumtor (Kyrgyzstan), Omai (Guyana), and Summitville (Colorado) (4, 5). Effective treatment of effluent must take place to prevent water contamination. Various treatment procedures, such as adsorption, complexation, and oxidation are known for treating cyanides (1, 6-10). The procedures other than oxidation give highly concentrated products in which toxic cyanides still exist. A treatment method for destruction of the cyanides is still needed before discharging the products into the environment. The addition of oxidizing chemicals is the most popular procedure to destroy cyanides. Chlorine, hypochlorite, hydrogen peroxide, ozone, sulfur dioxide, and ammonium bisulfite are common oxidants for cyanide treatment. Alkaline chlorination, although adequate, has many disadvantages, such as high chemical costs, formation of cyanogens, chloride contamination, and incomplete decomposition of some metal cyanide complexes (1). Hydrogen peroxide is successful in oxidizing cyanides, but it is not effective for SCN-. Ozone has been studied extensively for the destruction of cyanide (7-9). The pH of the cyanide solution must be above pH 11 to avoid the formation of hydrocyanic gas. Ozone is decomposed by hydroxide ion; therefore, ozonation becomes less efficient at pH higher than 11. Both sulfur dioxide and bisulfite procedures have been applied to treat cyanide. These procedures are most efficient at pH 9, and lime is added for pH control of the reactions. Gypsum is thus generated, which causes sludge generation. In recent years, we have examined the oxidations of cyanide (HCN and CN-) and SCN- by iron(VI) (Fe(VI), FeVIO42-) (11-13). Iron(VI) is a superior oxidant (14-16), and destruction of cyanides can be accomplished in seconds to minutes with the formation of less harmful products (eqs 1 and 2).
2HFeO4- + 3HCN + OH- f 2Fe(OH)3 + 3CNO-
(1)
4HFeO4- + SCN- + 5H2O f 4Fe(OH)3 + SO42- +
Introduction
CNO- + O2 + 2 OH- (2)
Cyanide is used or produced in several types of industry that include gas production, metal plating, pharmaceutical, and mining (1, 2). Speciation of cyanide determines its degree of toxicity, where cyanide is usually classified into four categories (1): (1) free cyanide (HCN, CN-), (2) readily soluble (NaCN, KCN) or relatively insoluble electrically neutral cyanide complexes (e.g. Zn(CN)2, Cd(CN)2), (3) weak acid dissociable cyanides (CNWAD; relatively unstable complexes of cyanide with transition metals such as Cd, Cu, Ni, and Zn), which dissociate under neutral or mildly acidic conditions, and (4) strong acid dissociable cyanides (CNSAD; strong complexes with metals such as Fe, Co, Ag, and Au) that are dissociable under extreme acidic conditions. Thiocyanate (SCN-) is considered in its own separate category, although it is a WAD. The free cyanide form is the most toxic form, whereas ironcyanide complexes, Fe(CN)63- and Fe(CN)62-, are relatively less toxic (3). Gold mining is one of the largest industries for cyanide consumption due to the high affinity of gold with cyanide.
Experimental Section
* Corresponding author phone: (321)674-7310; fax: 321-674-8951; e-mail:
[email protected]. † Florida Institute of Technology. ‡ Brookhaven National Laboratory.
Materials. All chemicals (Sigma, Aldrich) were reagent grade or better and were used without further purification. Solutions were prepared with distilled water that was passed through an 18.2-MΩ Milli-Q water purification system. Iron(VI) as
10.1021/es048196g CCC: $30.25 Published on Web 04/13/2005
2005 American Chemical Society
Gold is leached from the cyanide complex by applying soluble copper, which forms stronger complexes with cyanide than gold. The resulting copper cyanide complexes are highly toxic to aquatic life and are problematic because they are much more stable than free cyanide. The presence of copper cyanide complexes in the gold industry thus presents the biggest concern in cyanide management. In the present work, we have sought the destruction of the copper(I) cyanide complex (Cu(CN)43-) by Fe(VI). The rates of oxidation of Cu(CN)43- by Fe(VI) were determined as a function of pH (9.9-12.5) and temperature (5-35 °C). The reactivity of iron(V) (FeVO43-, Fe(V)) with Cu(CN)43- was also studied at pH 12.0 to compare its reactivity with Fe(VI). We have also evaluated Fe(VI) as an oxidant for removing cyanides from synthetic gold mill wastewater.
VOL. 39, NO. 10, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3849
the potassium salt (K2FeO4) was prepared according to the method of Thompson et al. (17). The purity of K2FeO4 was >98%. A copper(I) cyanide salt (K3Cu(CN)4) was prepared (>98.8% purity) by a method described by Izatt et al. (18), in which CuCN powder was dissolved in a solution of KCN at a molar ratio of 1:3.1 M. K3Cu(CN)4 was separated from the solution by precipitating with the addition of methanol, followed by filtration. After separation, the salt was dried for 24 h in a vacuum desiccator. Fe(VI) solutions were prepared by adding solid samples of K2FeO4 to 0.005 M phosphate/0.001 M borate at pH 9.0. Phosphate served as a complexing agent for Fe(III), which otherwise rapidly precipitates as a hydroxide that would interfere with the optical monitoring of the reactions and also accelerate the spontaneous decomposition of Fe(VI). Dissolved Fe(VI) concentrations were determined by measuring the absorbance at a wavelength of 510 nm in an Agilent Technologies 8453 UV-visible spectrophotometer equipped with a Hewlett-Packard Vectra computer with Agilent ChemStation software. A molar absorption coefficient (510nm ) 1150 M-1 cm-1) was used for the calculation of [FeO42-] at pH 9.0 (19). Methods. Stoichiometric and removal experiments were carried out by mixing equal volumes (10-2 L) of Fe(VI) and reactant solutions. The copper(I) cyanide concentrations before and after mixing with Fe(VI) were determined using Waters Alliance high performance liquid chromatography (HPLC). A Waters ion analysis method M-304 was followed. The eluent was a 5 × 10-3 M Low UV Pic A solution that consisted of 23% acetonitrile by volume. The eluent was pumped through a Waters Alliance 2690 separations module at a flow rate of 1.0 × 10-3 L/min. Both the eluent and the injected sample passed through the Nova-Pak C18 column (150 mm × 3.9 mm) to a Waters 996 photodiode array detector. Data were collected at 214 nm, where copper(I) cyanide had a maximum absorbance. Cyanide concentrations were determined before and after mixing with Fe(VI) using an ion-selective electrode (ISE) method. An Orion cyanide ion-selective electrode (model number 94-06) was used with an Orion model 620A pH/mV digital potentiometer. A distillation method was used to determine total cyanide in which hydrogen cyanide (HCN) was liberated from an acidified sample by distillation and purging with air. The HCN gas was collected by passing it through a sodium hydroxide scrubbing solution. Cyanide as free cyanide in the scrubbing solution was measured by the ISE method. Cyanate (CNO-) in the studies was determined with HPLC using a Waters Alliance 2690 separations module inline with a Waters 432 conductivity detector. A mobile phase solution of 2% borate/gluconate buffer (pH 8.5) was pumped through a Waters anion column IC-PAK Anion HR anionexchange column at a flow rate of 1.0 × 10-3 L/min. Stopped-Flow Kinetics. Kinetic experiments of Fe(VI) reaction with copper(I) cyanide were carried out using an Applied Photophysics SX-18MV stopped-flow spectrophotometer. For all experiments, the stop-flow apparatus was set for single mixing. In the experiments, Fe(VI) solutions at pH 9.0 were mixed in a 1:1 volume ratio with copper cyanide solutions in 0.01 M phosphate buffers at the desired pH. Kinetic studies were carried out under pseudo-first-order conditions. The concentrations of copper(I) cyanide were kept in excess of Fe(VI) by at least 1 order of magnitude. Fe(VI) absorbance at 510 nm was followed as a function of time to determine rate constants. An Endocal circulating water bath was used to maintain the temperature to (0.1 °C. The reaction system was thermostated for 10 min after every refill of the drive syringes. We were able to determine the rates of the reaction at 15 °C only at pH g 9.9. The reactions were too rapid at pH values lower than 9.9. 3850
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 10, 2005
Premix Pulse Radiolysis. A premix pulse radiolysis apparatus was used to study the reaction of Fe(V) with Cu(CN)43-. The reaction of Fe(VI) with copper(I) cyanide was fast enough to interfere with the measurement of Fe(V) reactions using standard pulse radiolysis in which the reactants are mixed and held in a reservoir from which the pulsing cell is automatically filled. The experiments were therefore carried out using the BNL 2-MeV van de Graff accelerator, which was computer-interfaced with a premixing apparatus consisting of three Hamilton precision liquid dispensor (PDL II) units, electronic controls, and an optical cell (2-cm light path), located in the path of the electron beam. The Fe(VI) solution (0.001 M borate/0.005 M phosphate; pH 9.0) in one syringe was mixed with the Cu(CN)43- solution (0.10 M phosphate buffer; 0.7 M ethanol) of the other syringe. Both solutions were saturated with nitrous oxide; [N2O] ) 0.026 M. The mixed solution was promptly injected into the optical cell and exposed to the ionizing pulse. Typically, the time between mixing and pulsing was
