Countries with extractive economies are often under the risk of carrying out practices of low efficiency and/or high environmental risk. For example, producers of gold and silver still rely in cyanidation to leach valuable metals from ores . Even though mining companies have a robust set of strategies for the handling and disposal of cyanide, the high toxicity of this compound is an environmental and social concern. In spite of the auspicious use of alternative leaching agents such as thiosulfate or thiourea , , , , , these “greener” agents are significantly less favorable than cyanide for complexing noble metal ions (Table 1 shows that, thermodynamically, cyanide is by far the best complexing ligand for gold and silver) , , , . Such favorability may be enhanced using larger quantities of thio-compounds and/or additional chemicals, which may in turn render the reaction not as friendly as originally intended .
More recently, however, there have been two major changes in the rules followed in mineral processing: (1) the global trend for developing and using more sustainable, environmentally friendly procedures, and (2) the depletion of rich, easy-to-process ores, which forces the processing of refractory minerals (i.e. minerals impervious to traditional procedures due to their high complexity and/or their low concentration of valuable metals) , . In response to these changes, there is a current interest in the development and optimization of novel methods for processing minerals. In general, these efforts align rather well with the principles of green chemistry . For example, the safety principle (“inherently safer chemistry for the prevention of accidents”) is observed in the transition from pyrometallurgical methods (high-temperature processes such as roasting, which may yield gaseous products difficult to manage) to hydrometallurgical methods (low-temperature processes in solution such as leaching, which yield aqueous products more easily handled) . In addition, the efforts for searching alternative leaching agents align well with the principle of less hazardous chemical synthesis (“the design of synthetic routes should use and generate substances as innocuous as possible”) Other principles will be discussed towards the end of the manuscript, with particular attention directed towards the optimization of cyanide usage, especially considering the case of refractory minerals.
At a molecular level, the processing of refractory minerals is challenging because there are several scenarios which may take place (Fig. 1). Assuming that the valuable metal is present as a mineral embedded into a host mineral (for example, a silver-containing mineral embedded into pyrite, an iron sulfide), the first challenge is to expose the valuable minerals to facilitate the leaching process. This is achieved by crushing and milling the samples, through energy-intensive strategies which are known collectively as comminution. Since comminution strategies are demanding not only energetically, but also economically, the size at which the valuable mineral is more exposed (more liberated) is often established on economic grounds. The final particle size for mineral processing is then a mixture of valuable minerals liberated, partially liberated and still embedded in the sample. The different scenarios which may happen during leaching are shown in Fig. 1, starting with the “ideal” leaching reaction (A), where a valuable mineral is exposed and easily reacts with the leachant. For example, in the case of silver and silver sulfide the reactions are:
Figure 1 shows also other scenarios, such as when unexposed valuable minerals are embedded in a host mineral which is impervious to cyanide (B), or when such host mineral reacts with cyanide, whether to form valueless complexes (C) or to destroy the ion (D). If the mineral is permeable to cyanide there will be a case in which diffusion becomes relevant (E), which will be discussed further below. Finally, it is possible that the desired products are re-adsorbed on the surface of a mineral (F), a process known as preg-robbing and that is often seen when the sample has carbon or hydrocarbonaceous solids. In the description shown in Fig. 1 it has been assumed that the valuable mineral is isolated from the host mineral and that it easily reacts with a leachant. However, it is possible that the valuable metal is atomically dispersed in a host mineral (in a similar way as a dopant) and/or that the valuable element is forming a highly stable mineral, which is not as reactive as the pure metal or the sulfide. In these extreme cases, the leaching process can only be possible through aggressive chemistries which completely destroy (dissolve) the mineral .
The scenario shown in Fig. 1(E) is somehow representative of a regular operation of mineral processing: the valuable metal is not liberated (due to limitations in comminution or to atomic-level distribution in the mineral) but the host mineral is permeable to the leachant and the leaching process can occur. In this scenario, the process of leaching will take place through a series of steps which include (i) adsorption of leaching agent on the surface of the host mineral, (ii) diffusion of the leaching agent through the host mineral towards the area where the valuable metal is present, (iii) formation of the complex metal – leachant, (iv) diffusion of the valuable complex through the host mineral, and (v) desorption of the valuable complex into the liquid phase . Kinetically speaking, processes (i) and (v) are often fast, as they can be facilitated by various means during processing (stirring, percolation, etc.). The limiting processes are often via steps (ii) and/or (iv) (diffusion through the host mineral) or via step (iii) (the formation of a soluble complex). These two possibilities are well represented in the so-called shrinking core model, which assumes that a reactant (the leaching agent) needs to pass across a forming layer of by-products, whose thickness increases as the original particle (the core) shrinks during the reaction , . Even though this model has been developed for gas-solid systems, it has been broadly employed by the hydrometallurgical community because the data from leaching can be plotted in equations corresponding to a diffusion-controlled process (Eq. 3) or a reaction-controlled process (Eq. 4):
where t is the reaction time, kd is the kinetic constant for a diffusion-controlled process, kr is the kinetic constant for a reaction-controlled process and X is the fraction of metal leached , . Then, if a leaching process is followed over time (amount of metal leached over time), it is possible to fit the data into the forms of Eqs. 3 or 4 to know which step is limiting. Diffusion-controlled processes can be optimized by reducing the path travelled by the leaching agent, which is often done by fine-milling the sample. They can also be optimized by modifying/destroying the by-product layer, which is particularly important when this layer is impervious to the leaching agent. In this regard, here we present the results of optimization of silver extraction and cyanide consumption using ultrasound (to facilitate the removal of by-product layers) and of an alkaline pretreatment (likely to remove elements which interfere with the leaching process). Through these examples, we demonstrate that the low efficiency of cyanide consumption (close to a striking 2%) is not improved with fine milling, but that ultrasound-assisted leaching and the alkaline pretreatment can lead to an improve in cyanide consumption and, simultaneously, in silver extraction.
Description of mineral samples
The mineral concentrate employed can be defined as a silver-containing polymetallic sulfide. Characterization using polarization microscopy showed that it was composed mainly of pyrite-based minerals (60%), spharelite (10%) and gangues (30%). Through atomic absorption (AA) essays it was determined that the sample had as main metals iron (20.9%) and manganese (8.3%), with smaller quantities of lead (1.64%), arsenic (1.54%), zinc (1.33%) and antimony (0.63%). Silver content in the sample is 47 oz/ton (~1330 g/ton). According to their granulometry, the sample was found to consist of particles greater than 80 μm (mesh 170, 29%), particles between 53 and 80 μm (mesh 170–270, 32%) and particles smaller than 53 μm (mesh 270, 39%). Experiments investigating the effect of milling (Section “The effect of fine milling in the efficiency of silver extraction and cyanide consumption”) employed only the particles greater than 80 μm, which were used with and without ball milling (until all particles were smaller than 53 μm). Other experiments used the sample without sieving. Samples were rinsed with water prior to leaching in order to decrease the amount of contaminants and remnants from previous procedures.
Caution: Sodium cyanide, NaCN, is a dangerous chemical which poses health and environmental risks for laboratory workers and personnel. It has to be handled with extreme caution and using the appropriate personal protection equipment (safety glasses, gloves and mask). Experiments must be carried out in a hood. Cyanide solutions can be considered safe if confined to containers at pH values above 10, since at lower pH values there is a risk to produce HCN, which is a very poisonous gas. Solutions and materials in contact with cyanide should be oxidized with a strong oxidizer to guarantee the decomposition of cyanide to cyanate (CNO−) first and to CO2 later. Here, bleach was employed as oxidizer, always in a basic medium.
Leaching experiments were conducted using 50 g of sample in 200 mL of a NaCN solution (8 g NaCN/L) at room temperature. The slurry (~25% solids) was agitated with overhead stirrers at a speed of 100 RPM. Leaching was kept at room temperature (22°C). The solution was kept at pH = 13 at all times by monitoring the pH with a Thermo Orion reader (Orion Star A 324) and adding dropwise a 0.1 M NaOH solution to adjust the pH when needed.
The concentration of cyanide was kept constant after determining periodically the amount of cyanide consumed and adding such amount to the slurry. The amount of free cyanide in solution is calculated by titrating 1 mL of leaching solution with AgNO3 (2 g/L), using rodamine as indicator (a change in color from orange to pink was indicative of the equivalence point). The amount of silver leached was determined using atomic absorption spectroscopy (Perkin Elmer Analyst). It is important to note that the calculation of consumed cyanide cannot be related to the amount of silver leached, since [CN]consumed is the sum of cyanide complexing silver, [CN]Ag, and cyanide participating in side, uncontrolled reactions:
Then, it is possible to calculate the efficiency in the consumption of cyanide from the ratio [CN]Ag/[CN]undesired. This ratio is shown multiplied by 100 to represent the percentage of efficiency in cyanide consumption and will be linked to the atom economy principle of green chemistry (“synthetic methods should maximize the incorporation of chemicals in the final product”).
The leaching experiments outlined in Section “Leaching methods” were replicated with the leaching vase under sonication, using a conventional sonication bath (Branson Bransonics 220). Due to sonication, the temperature of the solution increased from room temperature (~20°C) to 60°C. Even though there may be a thermal effect together with the ultrasonic effect, these two effects were not decoupled in the work presented here. Further details can be found in Ref. .
Chemical pretreatments of the mineral
Before leaching, the sample was exposed to a solution of sodium hydroxide (5 g/L) at 70°C during 4 h. After this pretreatment, the sample was filtered and washed until the effluent had a pH below 11. Then, the pretreated sample was dried in an oven at 100°C and after drying it was leached following the procedures described in Section “Leaching methods”.
Results and discussion
The effect of fine milling in the efficiency of silver extraction and cyanide consumption
Figure 2 shows the results of silver extraction (grams of Ag leached/ton of mineral) during cyanide leaching for both the as-received and the fine-milled samples. As reference, the total amount of silver present in the sample (1330 g Ag/ton of mineral) is shown. Then, it can be observed that the as-received sample has an efficiency of silver leaching of around 25%, which is increased to ~40% upon fine milling. In both cases the trend of silver extraction is asymptotic, in spite of the fact that the experimental procedure provides a constant cyanide concentration throughout the experiment, which demonstrates the refractory nature of the sample under study. As mentioned in the Introduction, it is possible to fit the data corresponding to silver extraction into the equations of the so-called shrinking core model (Eqs. 3 and 4) and obtain information regarding the rate-limiting step of the process. The results are shown in Fig. 2 and provide evidence of a diffusion-controlled leaching process. This indicates that the greater efficiency in silver extraction for finer particles is not because of an enhanced liberation, but because of an easier diffusion through the host mineral, either from the fact that particles are fractured during milling or from the fact that cyanide diffuses through shorter distances in finer particles.
The other parameter of interest is the consumption of cyanide, which is plotted in Fig. 3, together with the data for silver extraction to show the good correlation between these two variables. Milling also increases the consumption of cyanide (from ~20 to ~30 kg NaCN per ton of sample), which is expected as fine-milling originates an increase of surface area in the sample, which in turn may promote side reactions involving cyanide. Indeed, it is important to notice that these values are large with respect to the theoretical value, 1.2 kg NaCN/ton of sample, calculated assuming that the total amount of silver (1330 g Ag/ton) is leached as Ag(CN)2−. The correlation between the trends of silver extraction and cyanide consumption indicates that these two reactions are part of the same process (it will be shown below that this is not always the case). Because of this correlation, the efficiency of cyanide consumption (defined in Eq. 5) is nearly constant, for both granulometries close to 2%.
Regardless of the amount of silver extracted, it is important to determine the reason for the halt of silver extraction after 6 h. The fact that the process is controlled by diffusion in both granulometries suggests that diffusion itself is stopped, which would be the case if passivating layers are grown as leaching proceeds. This suggestion is supported by the detection of changes in the surface composition during leaching (these studies constitute an additional publication). Then, strategies directed towards the breaking of this layer can improve the leaching process, as demonstrated in the next sub-section.
Ultrasound-assisted leaching as an option for increasing the extraction of silver
Figure 4 shows the extraction of silver and the consumption of cyanide during the regular leaching and with the leaching assisted with sonication. These experiments used the sample without sieving. The results for the mineral under regular leaching are very similar for those obtained for the coarse mineral in the previous subsection under the same timeframe (first 400 min, approximately). However, at longer reaction times (considered to match the times needed for the ultrasound-assisted process) the consumption of cyanide continues, even though the amount of silver remains constant. This indicates that a different cyanide-consuming process starts to take place, but it will not be discussed as it is beyond the purpose of this study.
The amount of silver extracted increases when the leaching is conducted under ultrasound, although it requires longer times for reaching the maximum. The fact that during the first 100 min both the amount of silver extracted and the consumption of cyanide are similar with and without sonication indicates that initially the ultrasound does not affect the progress of the reaction. Then, the effect of ultrasound is only noticeable at latter times, likely when the mineral starts forming passivating layers. Experiments reported previously by our group showed that sonication is effective only during leaching (leaching of a mineral previously exposed to sonication for 1 day did not show differences in silver extraction), indicating that ultrasound operates over surface by-products and not over the original sample . These facts support the proposal that sonication helps by inhibiting the passivation of the sample. The trend of cyanide consumption is not correlated to the amount of silver extracted towards the end of the leaching, likely because the minerals/elements which consume cyanide have been depleted and/or passivated. This fact allows for an increase in the efficiency of silver extraction from 2 to 3% (Fig. 4). The different stages of the leaching process observed during ultrasound-assisted leaching are subject of a different publication. Here, it is important to highlight that the efficiency of both silver extraction and cyanide consumption can be enhanced by using methods able to disrupt the formation of a passivating layer. Another option is to modify the behavior of the sample not during leaching, but through a previous stage, where elements and/or minerals can be modified to facilitate the leaching process. This option involves a chemical treatment before leaching, which is shown in the final set of results.
Alkaline pretreatment as an option for increasing the efficiency of leaching
Figure 5 shows that the levels of silver extraction increase to ~700 and 1000 g/L for samples pretreated in an alkaline solution at 70°C and concentrations of 5 and 10 g/L NaOH, respectively. Pretreatment of the minerals at temperatures lower than 70°C did not show any improvement with respect to the leaching without pretreatment (data not shown). Interestingly, the asymptotic behavior of silver extraction data is similar to the one observed for the mineral without pretreatment: a maximum is observed in both cases after 400 min, marking a difference with respect to the time needed for ultrasound-assisted leaching. This can be explained by assuming that while in the latter case sonication needs time to disrupt the formation of a passivating layer, the chemical pretreatment does not postpone the formation of such layer, but allows for an easier, more efficient extraction process where diffusion is not limiting the reaction. In agreement with this view, the data corresponding to silver extraction cannot be fitted to neither the functions for diffusion-controlled nor reaction-controlled processes. Further supporting the notion that the pretreatment modifies the composition of the samples and originates different behaviors during leaching, the consumption of cyanide does not increase significantly during leaching. Thus, it seems clear that the pretreatment may be attacking and removing some minerals interfering with the cyanidation process. It is important to highlight that, according to the data in Table 1, this hydroxide pretreatment is not expected to involve the formation of the complex Ag(OH)2−, as it is not favorable thermodynamically [K = 103.6 vs. K = 1020.1 for the formation of Ag(CN)2− and K = 1012.8 for the formation of Ag(OH)(CN)−]. Figure 5 shows that the efficiency of cyanide consumption is initially 10% and later decreases to 5%, a value that is still large compared to the 2% seen during regular leaching. The fact that the efficiency in consumption of cyanide decreases over time indicates that the outer layers of the sample are the ones that are more significantly modified during the pretreatment, suggesting that larger pretreatment times and/or the use of harsher conditions of temperature and/or NaOH concentration may render the reaction more efficient towards the use of cyanide. These alternatives were beyond the purpose of our research.
Challenges and opportunities related to cyanide leaching
Table 2 presents a summary of the results reported in the previous section for an easier comparison of the processes considered here. With an efficiency of cyanide consumption of only ~2%, silver leaching of the mineral under study can be hardly considered a “green” chemical process, especially considering a compound with such high toxicity. The three strategies considered in this work (fine milling, ultrasound assistance and chemical pretreatment) are able to increase the amount of silver extracted, but only the latter two can simultaneously increase the efficiency in cyanide consumption. As indicated in Section “Experimental section”, the efficiency in cyanide consumption can be related to the atom economy principle of green chemistry, since it evaluates the amount of cyanide that is incorporated into the final product as the complex Ag(CN)2−(ac). Considering this principle, the most convenient, “greener” strategies will be the assistance with ultrasound and the chemical pretreatments. However, it is important to consider additional green principle principles and also views from an industrial point to provide a complete evaluation of each strategy. Table 3 presents an evaluation of the strategies outlined here in critical aspects such as feasibility for implementation and consumption of energy and resources, which will be further discussed for each strategy.
Even though this procedure does not increase the efficiency of cyanide consumption, it is attractive because it can increase the amount of silver extracted using technologies already available for mining companies. The fact that the equipment needed for fine milling is well known at industrial level is a great driving force towards implementation. Even more, most industrial settings have already milling units already installed and it would be only matter of extending the energy and time for the comminution process. Knowing that milling is an energy-intensive process, it does not follow the design for energy efficiency principle (“energy requirements should be minimized”). However, this strategy may be considered attractive in industrial settings because it does not require additional chemicals (prevention principle: “preventing waste is better than treating waste”) and does not require the use of additional substances (safer solvents/auxiliaries principle: “auxiliary substances to facilitate the reaction, including solvents, should be minimized”).
An advantage seen over fine milling and the use of chemical pretreatments is that ultrasound-assisted leaching does not require additional steps, since sonication is applied during the regular leaching. This fact sets this strategy along the principles of prevention of waste and safer solvents/auxiliaries. The main drawback of this strategy is that sonication at industrial level is not widespread and it may represent an uncertain technique which carries significant risks not only on technical grounds, but also on economic and even safety grounds.
This strategy also increases the efficiency in cyanide consumption, particularly during the first stages of leaching, making pretreatment the best option from the perspective employed here for the atom economy principle. However, this strategy rises several concerns which needs to be considered towards implementation. First, it requires an additional step in the metallurgical operation, after comminution and before leaching, which will involve additional investment and/or expenses. The most critical considerations, however, are from the point of view of the green chemistry principles, since the consideration of an additional step, which requires heating, presents conflicts with respect to the principles or prevention of waste, design for energy efficiency and safer solvents/auxiliaries. The latter principle takes, in the case of mining, an unusual dimension, since reducing the amount of fresh water required is as important as reducing the use of chemicals. In the context of mining, the minimization of the solvent (water) is of the utmost importance to avoid social problems, especially in areas of water scarcity and/or in areas where mining and agriculture coexist.
Thus, this discussion has intended to present a complete view of three strategies for improving leaching processes using cyanide. These (and other) strategies must be considered in a case-by-case basis, since different minerals can feature a different behavior. If a strategy results in an actual decrease in cyanide consumption, this outcome will need to be weighed with respect to the consumption of energy and resources of such strategy to decide whether it is worth to consider investigating the scaling-up. For example, the increase in the mass of reagents employed during pretreatment (e.g. the mass of the acid employed in a pretreatment) will need to be contrasted to the decrease in the mass of cyanide consumed, although it will almost always work as a rule that a decrease in cyanide consumption will be worth the increase in any other chemical. An exception to this statement will be water, which is currently considered not only a solvent, but also a precious resource.
Summary and conclusions
The leaching of silver from a refractory mineral has been shown as a complex process where several different mechanisms can take place, where silver extraction itself can be considered a diffusion-controlled process, characterized by a low efficiency in silver extraction. Fine milling, a usual strategy to improve the extraction of silver at industrial level, is shown to achieve that purpose but at the cost of increasing also the consumption of cyanide. Ultrasound-assisted leaching was shown to be efficient for increasing the amount of silver extracted by interrupting the formation of passivating layers during leaching, and chemical pretreatments prior to leaching were shown to significantly increase the efficiency of both silver extraction and cyanide consumption. From the point of view of green chemistry, however, the use of chemical pretreatments is not very attractive as it requires the use of additional processes, chemicals, energy and water. The use of sonication is more advantageous, but has as main challenge the evaluation at larger scale and the lack of familiarity in the mining sector towards this technology.
Instituto Superior Tecnológico Tecsup, is acknowledged for facilitating the use of wet chemistry and instrumental analysis laboratories. In particular, Ms. Yorsel Mayhua, Prof. Marixa Zegarra and Prof. Jorge Castillo (Tecsup) are thanked for their support during the research. C. G. acknowledges the support from Universidad de Ingeniería y Tecnología, UTEC. A. A. and C. S. participated in this research under the undergraduate program Vivir la Ingeniería at UTEC.
P. T. Anastas, J. C. Warner. Green Chemistry: Theory and Practice, Oxford University Press, New York, NY (1998). Google Scholar
C. Gamarra, A. Alarcon, T. Canchanya, M. Garcia, G. Pareja, C. Segura, J. C. F. Rodriguez-Reyes. Invest. Apl. Inov. 10, 14 (2016). Google Scholar
About the article
Published Online: 2018-04-12
Published in Print: 2018-07-26
Author contributions: The manuscript was written by J. C. F. Rodriguez-Reyes, with contributions of all other authors. A. Alarcon and C. Segura conducted the experiments, under the supervision of C. Gamarra and J. C. F. Rodriguez-Reyes. All authors worked in data analysis and have given approval to the final version of the manuscript.
Funding sources: This research was supported by the Phosagro/UNESCO/IUPAC Partnership in Green Chemistry for Life (Contract 4500245048) and by Peru´s National Council for Science, Technology and Technological Innovation (CienciActiva-CONCYTEC) and the British Embassy in Lima (contracts 154-2015 and 002-2016).
Citation Information: Pure and Applied Chemistry, Volume 90, Issue 7, Pages 1109–1120, ISSN (Online) 1365-3075, ISSN (Print) 0033-4545, DOI: https://doi.org/10.1515/pac-2017-0904.http://creativecommons.org/licenses/by-nc-nd/4.0/.