By: Jorge Ganoza, Metallurgist.IntroductionFor many years, the concentration and recovery of valuable mineral at low cost in the grinding circuit and flotation circuit have been practiced at coarse size, especially when the mineralogy indicates the presence of minerals of good floatability and liberation. The information is variable, and most of the studies were focused on the flotation circuit and the fineness of grind. If the mineralogy and liberation is understood, the floatability will too, because several factors play an important role in flotation of valuable minerals.In order to get the benefits of a coarse grinding circuit is important to understand how to grind coarser and recovering the coarser particles. The potential benefit can be obtained quickly at relatively low cost. The application requires the control of the final ground product size and the variations should be reduced as much as possible to get an appropriate feed to the existing recovery technologies. The grinding circuit usually includes a ball mill operating in closed circuit with the hydrocyclones [1].The concept of recovering minerals in the grinding circuit or from the discharge of grinding mills is not new. Gold concentrators have utilized jigs and flotation cells while lead-zinc operations have employed unit cells to recover coarse particles of galena and silver minerals from the discharge of grinding mills or hydrocyclone underflow. The technology of coarse particle flotation increases the upper limit of the particle size of flotation feed size and can reduce energy consumption, which is especially suitable for the disposal of tailings by the presence of coarse particles [2].In recent years, many coarse particle flotation technologies have been applied, but the equipment is usually of variable design and always taking as a reference the designs made by manufacturers with many years of experience. Based on the properties of minerals and their difficulty in flotation, the flotation equipment such as unit cell and flash flotation cell have been used to treat base metals and precious metals. They are the reference to design new flotation cells to float coarse particles.Particle AttachmentThe flotation process is a very effective method to separate valuable minerals from the gangue during treatment in mineral processing operations. The process consists of a series of consecutive sub-processes that include bubble-particle collision, attachment and stability of a bubble-particle. After bubble-particle collision, the particle is attached to the bubble surface and a bubble-particle aggregate is formed in the process. The aggregate is transported to the froth phase. The bubble-particle attachment and detachment are critical sub-processes for successful flotation. The bubble-particle collision is the principal sub-process of flotation that has a significant effect on the flotation and recovery [3].The attachment efficiency (Ea) and detachment efficiency (Ed) quantify the sub-processes of attachment and detachment, respectively. The bubble–particle attachment and detachment sub-processes have been relatively unexplored because they are complex processes and are controlled by the surface chemistry and physicochemical aspects of particles and air bubbles. The detachment process is controlled by the hydrodynamic of flotation cells. In general, Ea increases when the particle size is smaller and the contact angle is larger. In the same way, Ea decreases with increasing the particle size and bubble size, but increases with increasing particle contact angle [3]. This point is important at the moment of performing the flotation of coarse particles. In addition, Ed increases by increasing the turbulence. The collision efficiency increases by increasing the bubble size.Before a bubble–particle attachment can take place, a particle has to collide with a bubble, taking a distance of separation at which, the surface forces start to operate. The determination of the bubble-particle collision implies the evaluation of the forces that motivate a particle to deviate from the fluid streamlines near the bubble surface and collide with a bubble. The determination of bubble–particle collision involves the evaluation of forces that cause a particle to deviate in its trajectory from fluid streamlines near the bubble surface and collide with a bubble. Figure 1 shows four bubble–particle collision mechanisms, inertia (a), gravity (b), interception (c), and Brownian diffusion (d). The coarse lines represent the movement of the particles and the fine lines the movement of the fluid. The mechanism of inertial collision is more probable for coarse and dense particles that are not able to follow fluid streamlines and tend to move along a straight trajectory. If the density of particles is greater than that of the surrounding fluid, the particles could have a certain settling velocity and then, their trajectory deviates from fluid streamlines. This deviation may cause particles to collide with the bubble surface. The collision of particles with the surface of the bubble for interception is due to a flow that transports particles along the fluid streamlines. The particles come in contact with the surface of the bubble because their finite size. The bubble-particle collision by Brownian diffusion is significant for very tiny particles that moves randomly in the fluid [4]. The bubble-particle collision can occur by the individual mechanisms mentioned or could be the result of two or more of these mechanisms.Limit of Particle Size The degree of hydrophobicity can be expressed by the contact angle, which is the angle in the triphasic line of contact between the mineral, the aqueous phase and the air bubble. It is accepted that the higher the contact angle of a mineral surface, the more easily it becomes wetted with the air, therefore more hydrophobic. The hydrophobicity or contact angle of the particles depends on the type and distribution of species present on the mineral surface. Recovery decreases with increasing particle size due to detachment, and decreases with a small particle size due to inefficient collision [5].There is an upper size limit for floatable particles. The balance of forces acting on the particle and the bubble will determine the stability of the assemblage. Coarse particles, either of a single type or composite, can detach from the surface of the bubble [6]. After attaching, two conditions are necessary for flotation: stability and floatability of the aggregate.The predominant forces are gravitational and capillary forces. The maximum liftable particle size is different from the maximum floatable particle size, the first is obtained under static conditions while the second is influenced by a dynamic state. Some results with quartz particles indicate that a higher contact angle is required to lift large particles. For example, a 3.4 mm particle can only be lifted by a 1.8 mm bubble, at a constant upward velocity of 20 µm/s, and if the contact angle of advancing water with the particles is at least 80º. The size of the particles that can rise decreases as the diameter of the bubble decreases. The upward velocity is also important and the size of the particle that can be lifted decreases as the upward velocity increases with the acceleration [5]. Figure 2 shows the influence of the contact angle for a bubble to transport the maximum quartz particle size respect a bubble size.Although high agitation/turbulence in the flotation cell increases the frequency of particle and bubble collision, and therefore the possibility of flotation and recovery, too much turbulence is detrimental as coarser particles can be detached from the bubbles when the force of detachment that can be simply represented by the centrifugal force and gravity become larger than the bubble-particle adhesion force, which is considered proportional to the hydrophobicity of the particle or the contact angle [7]. Considering that the centrifugal force and gravity. are proportional to the mass of the particle, the detachment force increases with the diameter of the particles and the density of the particles, and consequently the possibility of flotation and recovery will decrease in the flotation cell. After attachment, two conditions are necessary for flotation: stability and floatability of the bubble-particle aggregate.Figure 3 shows the recovery curves versus particle size for a particular copper ore along a bank of flotation cells [7]. These curves present a maximum recovery at approximately 100 µm and a decrease in recovery for smaller and larger particles, attributed to the low bubble-particle collision efficiency and high detachment of the collected minerals. The shape of these curves is common to all mechanical flotation cells, regardless of feed size and types of minerals floated, although maximum particle size and recovery can be increased or decreased by mineralogical and density factors of the minerals floated. Figure 4 shows results of increased bubble-particle collision/attachment of different types of minerals.Gold Flotation Due to its high density and malleability, gold tends to flatten out during grinding, and surfaces can become coated or encrusted with gangue particles and iron coatings from liners or steel balls. Although flattened particles have a larger surface area than spherical particles, the detrimental effects of surface degradation can have a significant impact on floatability [8]. Such flattened gold particles have been found to have very rough flat surfaces, and the higher the roughness, the less hydrophobic the gold particles become, thus hindering bubble attachment and floatability. For this reason, gravity concentration is considered ahead of flotation where fine particles of free gold or associated with sulphides can be recovered.Loss of free gold grains in the slimes (<5 µm) is the main cause of free gold losses in flotation plants. Efficiency drops off rapidly for particles below 5 to 7 µm, and the problem is often overgrinding, which is also a problem in a regrind circuit. It should be indicated that the production of gold slimes and losses are common for primary grinding circuits and regrinding circuits that are integrated with flotation cleaning circuits [8, 9]. Figure 5 shows the recovery of fine gold in a flotation circuit.Free gold can be recovered very effectively by flotation, although it is more commonly recovered together with sulphides (e.g. chalcopyrite, galena), where the gold is intimately associated with the sulphides as fine grains without free gold (in solid solution), or occurs with hydrophobic sulphides with not economic value. The most common gold-bearing sulfides are pyrite, arsenopyrite and, to a lesser extent, pyrrhotite [10]. Gold flotation from sulphide-free ores containing very low concentrations of free gold is difficult due to the low mass of material reported in the concentrate and the high density of gold (19,300 kg/m3). For example, 0.005% (50 g/t) Au would be a very high gold grade, compared with grades >0.5% for most copper or lead ores treated by flotation. This results in a very poor froth stability, lower recovery, and/or concentrate grade [10].Free or native gold particles can vary in size, ranging from large particles to very fine disseminated gold particles that may be associated in a complex sulphide matrix. Typically, free gold particles of several hundred microns or larger are efficiently recovered by gravity concentration. This technique is successful due to the large specific gravity difference between gold (15.0 to 19.3) and gangue minerals (2.7). When gold is associated with a sulphide matrix, historically, most industrial plants consider size reduction to have adequate liberation in a size ranging from 50% to 80% less than 75 µm. Finer grinding might not be economical, unless the ore has a high gold content [11].Figures 6 and 7 show the results of flotation tests performed on a free gold ore from Africa. In the test shown in Figure 6 no collector was added, only frother, while in the test in Figure 7, collector and frother were added. The grind size for both tests was 52.5% <75 µm, which is a coarse size, in order to have a wide range of gold particle sizes in the flotation feed [11]. Smaller gold particles float much faster than larger particles and also give higher equilibrium recoveries. This is true with or without the addition of a collector. The recovery of gold particles below 150 µm in many plants is often surprisingly good whether or not a collector is added.The problem with the recovery of floating gold flakes is twofold. First, there is an upper limit to the size of particles on which flotation can occur, set primarily by the physical limitations of the bubble in lifting coarse particles. The upper limit of para floatation rarely exceeds 500 µm and is usually below 300 µm. Optimal flotation size is typically in the range of 10 to 100 µm. Second, the flotation of coarse particles become more significant if gold flake is present during cleaner flotation stage, where the cells are designed for small particle processing [12]. Figure 8 shows the gold flakes found in the tailings of a cleaning flotation circuit.MineralogyThe mineralogy is one of the main factors to be evaluated to determine if it is advisable and feasible to perform the flotation at a coarse size, for which the degree of liberation must be determined. The liberation of valuable minerals from gangue is achieved by size reduction, which consists of crushing and grinding to such a size that the product is a mixture of relatively clean particles of valuable minerals and gangue [13]. One objective of comminution is the liberation at the coarsest possible particle size. If that objective is achieved, then not only energy is saved, it is possible to reduce the production of fines. In this way, the flotation process becomes easier and cheaper to operate. If high grade products are required, then good liberation is essential.To perform the mineralogical study, a study by QEMSCAN (Quantitative Evaluation of Minerals by Scanning Electron Microscopy) or MLA (Mineral Liberation Analyzer) is recommended. The potential liberation of valuable minerals can be determined by characterizing grain sizes of the minerals present in the ore. This can be achieved by reducing the drill core samples to a relatively coarse size (typically around 600 µm) to preserve the texture of the samples, including grain size, association, and shape [13, 14].Figure 9 shows images from an MLA study. The presence of liberated chalcopyrite and pyrite particles is shown in a coarse size fraction, -4000+1000 µm. The presence of free sulphides in a coarse size fraction suggests that when milling the ore, a coarse flotation could be obtained, but with some competition due to collector adsorption.A QEMSCAN study allows minerals to be classified into different groups according to the percentage that a mineral of interest occupies within an area of the studied section. Thus, there are totally free particles when they occupy ≥ 95% of the area, liberated when it is ≥ 80%, and not liberated when it is < 80% of the area. Non-liberated grains can be classified into groups according to their mineralogical association as binary or complex [13,15]. It may also indicate the need for fine regrinding to minimize the displacement of unwanted minerals in the final concentrate, this is an analysis of liberation of valuable particles [15].Figure 10 shows QEMSCAN images in the size fraction -300+150 µm, noting that most of the copper minerals in the particles are associated with gangue and the proportion of copper minerals in these particles is at least 30% in the concentrate and less in the tailings. After collision, air bubbles disperse on the surface of a hydrophobic particle until the equilibrium contact angle is reached, but for a composite of particles the final contact angle in the hydrophobic zone is less than the contact angle in equilibrium because the bubble perimeter is pulled towards the hydrophobic-hydrophilic interface, which results in a very weak bubble-particle adhesion at a coarse size greater than 200 µm [7].By determining the appropriate grind and possible size of liberation, we can know if there might be any mechanical entrainment issues of gangue minerals in the froth. There are several factors that influence mechanical entrainment, one of them being particle size. This is usually evaluated using a recovery curve at different sizes, and comparing the results with the average particle size evaluated. In the analysis, the particles reported by mechanical entrainment show a greater displacement when the particle size is reduced [16]. The flotation at a coarse size can partially reduce the presence of gangue in the concentrate, but particles with very little liberation can also be reported in the froth, which can complicate the design and/or operation of the regrind and cleaning circuit. Flotation CellThe flotation of coarse particles has been practiced since the last century and is not a new topic, since the first studies were carried out in the 1930’s [17]. There are two flotation cells that have shown their effectiveness in the process, one is the Unit Flotation Cell and the other is the Flash Flotation Cell (SkimAir®). Based on the efficiency of these cells, other manufacturers have tried to develop other cells, but without reaching the efficiency of the models mentioned above. For the vast majority of minerals, recovery is good for particle sizes between 50 and 150 µm, and outside of this range, it decreases. It is desirable that good recoveries can be obtained at sizes greater than 150 µm, and it is advisable to consider a size somewhat coarser than 200 µm [18]. For which it is necessary to use the appropriate flotation cell, as well as to know if the mineralogy is suitable, and make a good selection of flotation reagents.It has traditionally been found that economic benefits can be obtained by using the unitary flotation cell in the grinding circuit. This cell was specially designed for the recovery of floatable minerals at coarse size as soon as the ore is liberated; together with the finer sizes. Although many ores require fine grinding for maximum recovery, most ores liberate a large percentage of the valuable minerals during the pass through the grinding circuit. By placing a unit flotation cell between the grinding unit and the classifier, it is possible to rapidly recover free valuable minerals from coarse gangue at low cost [19]. Figure 11 shows the installation and operation of this cell in the primary grinding circuit.The unit cell has demonstrated that it can operate at high pulp densities and is specially designed to operate under this condition, so it is practical to use the unit cell in almost any grinding circuit and in almost all cases, it is necessary to supply water to the discharge of the unit cell to obtain the desired dilution at the classifier overflow. This means that the unit cell can be installed in the grinding circuit without altering the mill control or drastically changing the grinding area layout. Figure 12 shows a more detailed arrangement of the unit flotation cell in the grinding-classification circuit. One of the variations adopted by operations was to install two mechanical cells in the primary grinding mill discharge to effect rapid flotation at coarse size based on unit cell operation. The Lake Dufault copper-zinc operation used this design some time ago and the rod mill discharge at 37% minus 200 mesh size was diluted to 57.5% to feed two 100 cubic foot Denver mechanical cells [20]. In this circuit, about 30% of the copper was recovered with a grade of 24%. The flotation time was 6 minutes.The concentrate obtained by the unit flotation cell is a final concentrate that can be sent to the thickening and filtration circuit. The tailings from the unit flotation cell can be sent to the classifier, either a spiral or a hydrocyclone.It is important to indicate that laboratory evaluations are important to estimate recoveries and behavior of a mechanical cell to evaluate the kinetics of coarse flotation. This idea is of considerable importance if processing a polymetallic ore such as lead-zinc, since, if lead flotation is done to coarse size, the zinc circuit should be processed to coarse size as well. For example, a feed with a K80 of 390 µm was evaluated in the laboratory, and the results indicated that a size range of -300+150 µm could recover zinc by about 90% [21], but it was necessary to add an adequate amount of collector since coarse particles require a higher dosage of collector than fine particles. See Figure 14.In the early 1980’s, the Finnish Company Outokumpu introduced the Flash Flotation Cell called SkimAir. Tests were performed at Hammaslathi concentrator. The idea was to place a mechanical flotation cell in the grinding-classification circuit, its feed was the coarse discharge of the classifier, which could have free particles that could be floated and avoid unnecessary regrinding [22]. The tailings from the OK cell were reground, and the concentrate obtained was a final concentrate. Its application allowed to improve the recovery of the valuable minerals, improve the flotation kinetics, reduce the volume of equipment in the flotation circuit, avoid losses due to the generation of slimes, and facilitate the filtration of the concentrate to reduce the moisture content. Figure 15 shows a typical arrangement of a flash cell in a grinding circuit.The main design difference between the SkimAir flash flotation cell and conventional flotation cells is the conical discharge. The bottom discharge and slurry inlet point are designed to handle a coarse feed such as the hydrocyclone underflow. This coarse material is short-circuited at the bottom discharge, thereby preventing the material interfering with flotation, where it could lead to higher pulp density in the cell and preventing the particles rising to the surface. The top has a lower pulp density of 40–50% solids compared to the bottom discharge with a density of 60–70% solids, and is sometimes diverted to the grinding unit discharge. The bottom discharge from the flash flotation cell is reground. In the cell, there is a variation in pulp density that helps to manage the different grinding and flash flotation water requirements [24].Flash flotation has many applications in lead, copper-gold and gold flotation circuits in order to recover free gold and gold associated with sulphides such as pyrite or copper minerals, in high grade concentrates that could be sold as copper concentrates or processed in situ using centrifugal concentrators and cyanide leaching [25].The flash flotation cell can be used in a grinding circuit to obtain a final concentrate, or it can be installed as part of the rougher flotation circuit, whose concentrate is cleaned in a mechanical cell. It can also work with gravity concentrators [26]. It is necessary to perform concentration tests and determine the mineralogy of the ore to be processed. Figure 17 shows the froth of a flash flotation cell in operation, part of the level control is also observed.It has also been suggested that the flash flotation cell can be used to treat polymetallic ores such as lead-zinc where the selectivity usually offered by coarse galena with respect to sphalerite allows less displacement of sphalerite. Likewise, its installation is recommended when the flotation of particles larger than 250 µm is feasible [28]. The flash flotation cell can treat coarse particles in the range of 600 to 1,000 µm [29].It is known that many copper deposits contain significant gold contents, which give significant additional value to the copper concentrate. When there is a loss of gold due to excessive grinding, the installation of a Flash flotation cell has been a real and practical solution. An important example is the Freeport Indonesia Concentrator, where there was a positive effect after installing Flash flotation cells in the grinding circuit because free gold particles with a size of 150 µm were observed in the rougher tailings. The grinding-classification circuit with flash flotation reported a K80 of 180 µm and operated with hydrocyclones inclined at 45 degrees. Results were good, increasing the processing capacity from 15,000 to 17,000 t/d, and gold recovery increased by 13% by avoiding overgrinding [30]. Figure 18 schematically shows the installation of the flash flotation cell in Freeport Indonesia. Table 1 shows the results of the operation prior and after installing the Flash flotation cells.ConclusionsFlotation of coarse particles has been practiced for many years, being successful when the mineralogy of the ore to be floated is favorable, showing good liberation and a not very complex mineralogical association. Likewise, the use of flotation cells such as the unit flotation cell and the flash flotation cell have been an important means to perform the recovery and flotation of coarse particles of base metals and gold bearing minerals.References1. Maron, R., Sepulveda, J., Jordens, A., O’Keefe, C., Walqui, H. Coarser Grinding: Economic Benefits and Enabling Technologies. 4th International Seminar on Operational Excellence in Mining. April, 2019.2. Hengtong, L., Dongxia, F., Yang, D., Xiaoyong, Z., Yunong, X. Coarse Particle Flotation Technology and its Application. Conservation and Utilization of Mineral Resources, pp. 129-137. February. 2022.3. Darabi, H., Koleini, J., Deglonb, D., Rezai, B., Abdollahy, M. Investigation of Bubble-Particle Attachment, Detachment and Collection Efficiencies in a Mechanical Flotation Cell. Powder Technology (375), pp 109-123. 2020.4. Miettinen, T., Ralston, J., Fornasiero, D. The Limits of Fine Particle Flotation. Minerals Engineering (23), pp. 420–437. 2010.5. Gontijo, C., Fornasiero, D., Ralston, J. The Limits of Fine and Coarse Particle Flotation. The Canadian Journal of Chemical Engineering. Vol 85, pp. 739-747. 2007.6. Crawford, R. and J. Ralston. The Influence of Particle Size and Contact Angle in Mineral Flotation. International Journal of Mineral Processing. Vol 23, 1, pp. 1-24. 1988.7. Fornasiero, D., Filipov, L. O. Innovation in the Flotation of Fine and Coarse Particles. Journal of Physics: Conf. Series 879 012002, 2017.8. Adams, M. Gold Ore Processing – Project Development and Operations. Elsevier, 2016.9. Chryssoulis, S.L., Dimov, S.S. Optimized Conditions for Selective Gold Flotation by ToF-SIMS and ToF-LIMS. Applied Surface Science 231–232, pp. 265–268. 2004.10. Marsden, J. The Chemistry of Gold Extraction. Published by the Society for Mining, Metallurgy, and Exploration, Inc. 2009.11. Klimpel, L.L. Industrial Experiences in the Evaluation of Various Flotation Reagent Schemes for the Recovery of Gold. Minerals and Metallurgical Processing. Vol. 16. No 1. February 1989.12. Knipe, W., Chryssoulis, S.L., Clements, B. Flaky Gold: Problems with Recovery and Mineralogical Quantification. JOM, pp 58-62. July 2004. 13. Wills, B., Finch, J. Will’s Mineral Processing Technology. Eighth Edition. Elsevier. 2016.14. Tungpalan, K., Wightman, E., Manlapig, E., Keeney, L. The influence of Veins on Mineral Liberation as Described by Random Masking Simulation. Minerals Engineering (100), pp.109–114. 2017.15. Grammatikopoulus, T. High-Definition Mineralogy by QEMSCAN: From Exploration to Processing. SGS Canada. Presentation. 2018.16. Wang, L. Peng, Y., Runge, K., Bradshaw, D. A review of Entrainment: Mechanisms, Contributing Factors and Modelling in Flotation. Minerals Engineering (70), pp.77–91. 2015.17. Anderson, C., Dunne, R., Uhrie, J. Mineral Processing and Extractive Metallurgy. 100 Years of Innovation. SME. 2014.18. Goodbody, A. Par for the Coarse. Mining Magazine, pp 10-19. September 2017.19. Denver Equipment Index. Second Edition. Volume I. 1947.20. Williams, A.J. Flotation of Base Metals from Grinding Mill Discharges. Paper No 1. The 16th Annual Meeting of the Canadian Mineral Processor Division of the Canadian Institute of Mining and Metallurgy. January, 1984. 21. Fosu, S.,Pring, A., Skinner, W., Zanin, M. Characterization of Coarse Composite Sphalerite Particles with Respect to Flotation. Minerals Engineering (71), pp.105–112. 2015.22. Kallioinen, J., Tarvainen, M. Flotation as Part of Grinding Classification Circuits. The Canadian Mineral Processor Division of the CIM Annual Operator’s Conference. 1984.23. Tracy, S., Gillis, J. Flotation in Grinding Circuits. SME Meeting, Littleton, Colorado. Preprint Number 86-316. September, 1986.24. Mackinnon, S., Yan, D., Dunne, R. The Interaction of Flash Flotation with Closed Circuit Grinding. Minerals Engineering (16), pp. 1149–1160. 2003.25. Lakshmanan, V., Roy, R., Ramachandran, V., Innovative Process Development in Metallurgical Industry - Concept to Commission. Springer. 2016.26. Outokumpu. Technology Mineral Processing – Presentation. Flash Flotation + SkimAir®. August, 2000.27. Lamberg, P., Bernal., L. Flash Flotation - From Pilot Size to Full Size Installation at Esperanza. Procemin 2008.28. Newcombe, B., Bradshaw, D., Wightman, E. Flash Flotation and the Plight of Coarse Particles. Minerals Engineering (34), pp. 1–10. 2012. 29. Gorain, B.K. Innovations and Breakthrough in Mineral Processing that Have Shaped the Existing Mining Industry. XXVIII IMPC, Canada, pp. 1-45. September, 2016. 30. McCulloch Jr, W. Flash Flotation for Improved Gold Recovery at Freeport Indonesia. Minerals and Metallurgical Processing, pp. 144-148. August, 1990.
