The Direct Leaching of Nickel Sulfide Flotation Concentrates – A Historic and State-of-the-Art Review Part III: Laboratory Investigations into Atmospheric Leach Processes

ABSTRACT

The following review is Part III of a series concerned with the direct hydrometallurgical processing of nickel sulfide flotation concentrates. In the first part of this series, piloted leaching processes and commercial nickel sulfide operations that employed direct hydrometallurgical processing were comprehensively reviewed. In the second part of this series, laboratory investigations into pressure leaching of nickel sulfide concentrates were critically analyzed. In Part III of this series, laboratory investigations into the leaching of nickeliferous sulfide concentrates at ambient pressure are reviewed, and the challenges and research opportunities in the direct leaching of nickel sulfide flotation concentrates are summarized. The majority of the published studies on atmospheric leaching of nickel sulfide flotation concentrates have focused on leaching in chloride media due to the faster leaching kinetics in chloride lixiviants relative to sulfate media; bioleaching due to the perceived environmental advantages over other leaching systems; and pyrometallurgical pre-treatments to render refractory sulfide minerals more amenable to subsequent leaching.

KEYWORDS:

• Nickel
• cobalt
• pentlandite
• nickel sulfides
• hydrometallurgy

1. Introduction

The extraction of nickel from sulfidic resources was described in Part I where it was shown that direct hydrometallurgical processing of nickel sulfide flotation concentrates has several distinct advantages over pyrometallurgy. In the first part of this review, a survey of piloted technologies for direct leaching of nickel sulfide concentrates and commercial nickel sulfide operations that have implemented hydrometallurgical processing was described. Additionally, the geology and mineralogy of nickel sulfide deposits and typical beneficiation practices were summarized. In Part II of this series, laboratory investigations into the pressure leaching of nickeliferous sulfide flotation concentrates were critically reviewed and summarized. The purpose of Part III of this series is to critically review and summarize research activities relating to the direct hydrometallurgical processing of nickel sulfide flotation concentrates at ambient pressure. A distinction is made among the different leaching processes based on the type of lixiviant used to achieve base metal dissolution from the concentrate and whether concentrate pre-treatment by mechanical activation, roasting, or sonication to render the material more amenable to subsequent leaching has been used. Following on from an examination of research activities concerning the hydrometallurgy of nickel sulfide concentrates at ambient pressure, challenges and research opportunities relating to nickel sulfide concentrate hydrometallurgy are then discussed.

2. Laboratory investigations into direct leaching of nickel sulfide concentrates

2.1. Ammoniacal lixiviants

The ability of ammonia (NH₃) to selectively dissolve Ni, Cu, and Co from sulfidic materials (concentrate, matte, and mixed sulfide precipitate) underpins the Sherritt-Gordon process, an ammoniacal pressure oxidation process that was described in Part I of this review and that has seen commercial deployment at two locations (Canada and Australia). Presumably, ammoniacal leaching at ambient pressure would result in lower capital cost relative to a pressure leach process due to simpler equipment requirements. However, the volatility of ammonia is likely to pose challenges to any atmospheric leach process due to loss of lixiviant (Dutrizac 1981; Nicol 2018) and impose limits on operating temperature and lixiviant concentration for an ambient pressure leaching process, impacting dissolution kinetics. The ammoniacal leaching of nickel sulfide concentrates at ambient pressure has been explored by several researchers, which are described in this section and are summarized in Table 1.

Table 1. Summary of laboratory investigations carried out on ammoniacal leaching of nickel sulfide concentrates at ambient pressure.

Reference	Nickel sulfide mineralogy	Nickel grade (wt% Ni)	Concentrate feed particle size	Mode of leaching	Leaching conditions
Aaltonen et al. (1980)	Pentlandite, mackinawite^a	5.4%	86.5% under 74 µm	Stirred tank leaching	2 M NH3, 0.2 M (NH4)2SO4, 50°C 1 atm O2 pressure, up to 2 days leaching time, solids conc. 10 g/100 mL
Welham et al. (2008)	Pentlandite, violarite, and pyrrhotite	14.6%	Not stated	Not stated	Leaching with 40 g/L (NH4)2CO3 +20 g/L NH3 at ambient temperature & pressure. Concentrate also cured with NaClO prior to leaching.
Muzawazi and Petersen (2015)	Pentlandite	3.17%	P97 45 µm, coated onto 5–10 mm granite particles for column leaching	Column leaching and stirred tank leaching	Stirred tank leaching − 25 - 60°C 0.5–4 M NH4OH, 0.5–4 M (NH4)2CO3, 120 mL/min air flow, up to 5 days leaching time Column leaching − 25 - 40°C 0.5 – 4 M NH4OH, 0.25–4 M (NH4)2CO3, 80 mL/min air flow, up to 42 days leaching time

^aMackinawite - (Fe, Ni)S_0.9.

Aaltonen et al. (1980) studied the dissolution of a bulk Ni-Cu concentrate grading 5.4% Ni and 1.5% Cu in sulfuric acid and ammoniacal solutions at ambient pressure, the concentrate being produced from flotation of a serpentinitic low-grade Ni-Cu ore. Only a single leach test was performed under fixed conditions (refer to Table 1) where it was observed that Ni and Cu dissolution from the concentrate was considerably faster relative to sulfuric acid leaching. During ammonia leaching, un-leached particles of pentlandite were observed to be coated by a thick layer of iron oxides, which would have slowed down dissolution kinetics. Aaltonen et al. (1980) suggested that performing ammonia leaching under more intensive stirring conditions would improve dissolution kinetics by continuously removing precipitated iron oxides from the sulfide mineral surface as they form and by providing better dispersion of oxygen gas into the leach pulp.

Welham, Johnston, and Sutcliffe (2008) patented a process for leaching of base metal oxide and sulfide ores and concentrates which involved curing the ore or concentrate with either an oxidizing or reducing agent or metal-complexing ligand followed by leaching in an aqueous solution of ammonium carbonate and ammonia at ambient pressure via heap leaching. Three examples were given in the patent of the application of this technique at laboratory scale (0.25 g sample mass) to the treatment of nickel sulfide bearing material (pentlandite, concentrate, and matte). A nickel sulfide concentrate was leached with an ammoniacal solution (refer to Table 1) at ambient pressure for 168 hours, resulting in 17% of the Ni being leached. The same concentrate was cured for 24 hours with a 25 g/L solution of sodium hypochlorite (NaClO) adjusted to pH 4 (enough was added to dampen the powder) prior to leaching using an ammoniacal solution for 4 hours which resulted in 97.4% of the Ni being leached. Similar results were achieved when leaching nickel matte, pentlandite, and a range of other sulfide materials. No surface or mineralogical characterization was carried out to explain this phenomenon though the inventors speculated that curing with an oxidizing agent resulted in oxidation of the sulfide minerals, which were subsequently readily dissolved in ammoniacal ammonium carbonate. For carbonaceous ores, the oxidizing agents also served the purpose of destroying any organic matter that would have consumed lixiviant.

Muzawazi and Petersen (2015) studied the stirred tank and column leaching of a low-grade PGM bearing Cu-Ni sulfide concentrate grading 3.17% Ni and 2.02% Cu in an aqueous ammonia-ammonium carbonate lixiviant using air as the oxidant. Higher ammonia concentrations (up to 4 M) and leaching temperatures (up to 60°C during stirred tank leaching, 40°C for column leaching) favored greater dissolution rates and extents for Ni and Cu. Dissolution kinetics were considerably faster during stirred tank leaching relative to column leaching, with Cu and Ni extractions reaching completion after 3 days of stirred tank leaching at 60°C and 4 M NH₃ plus 4 M (NH₄)₂CO₃ (pH 9.35) at a pulp density of 2% w/v. Muzawazi and Petersen (2015) also measured the ammonia losses over a 4-day period during stirred tank and column leaching and found that ammonia losses ranged from 7% to 10% for stirred tank leaching and 6–16% during column leaching. Ammonia losses were found to increase slightly with increasing temperature and markedly with increasing NH₃ concentration in solution. Muzawazi and Petersen (2015) suggested that this not only represented an obvious loss of reagent but that the associated drop in pH due to NH₃ evaporation could shift the pH of the system outside of the stability window of Cu and Ni-ammine complexes in solution, which would presumably impact leaching extents.

2.2. Atmospheric sulfuric acid leaching

The leaching of nickel concentrates in an acid sulfate medium at atmospheric pressure would offer considerable capital and operating cost savings versus leaching at high pressure and temperature; however, there are only a very limited number of laboratory investigations concerning the leaching of nickel concentrates at ambient pressure in sulfuric acid media, which are summarized in Table 2. Starting with Corrans and Scholtz (1976), an investigation into the leaching of a high-grade pentlandite concentrate under simulated heap leaching conditions using an acid ferric sulfate [Fe₂(SO₄)₃] solution as the oxidant was carried out. The study aimed to elucidate the mechanism of Ni dissolution from pentlandite in acid ferric sulfate media; higher temperature, ferric ion concentration, and oxygen partial pressure resulted in higher dissolution rates of Ni. Corrans and Scholtz (1976) suggested that pentlandite dissolution was an electrochemical process, and cathodic processes (Fe³⁺ and O₂ reduction) were the rate limiting step. The chemical leaching rate of pentlandite was found to be dependent on the rate of diffusion of ferric ions to the surface (mass transport limited) and on the rate of adsorption of oxygen on the mineral surface (chemical control). The dependence of base metal sulfide dissolution on oxygen partial pressure has also been reported by Schneerson, Leshch, and Frumina (1966), Peter, Hansen, and Wadsworth (1973), Lawson, Cheng, and Lee (1992), Filippou, Konduru, and Demopoulos (1997) and Kniss, Naboichenko, and Zhukov (2013). The dependence of pentlandite dissolution on oxygen partial pressure was found by Corrans and Scholtz (1976) to agree with the findings of Schneerson, Leshch, and Frumina (1966), who observed that the rate of pentlandite dissolution had a half-order dependence on oxygen pressure during autoclave leaching of a mixed sulfide concentrate containing pyrrhotite and chalcopyrite in addition to pentlandite.

Table 2. Summary of laboratory investigations carried out on sulfuric acid leaching of nickel sulfide concentrates at atmospheric pressure.

Reference	Nickel sulfide mineralogy	Nickel grade (wt% Ni)	Concentrate feed particle size	Leaching conditions
Corrans and Scholtz (1976)	Pentlandite	30.6%	−74 + 44 µm	30 & 80°C 0.01 & 0.25 M Fe(III), 2 × 10^−4 & 2 M H+ (as sulfuric acid), solids conc. 100 g/L, 0 & 400 rev/min, PO2 1% & 100%
Aaltonen et al. (1980)	Pentlandite and mackinawite	5.4%	86.5% under 74 µm	pH 2, 3.5 & 5, 25 – 95°C up to 5 days leaching time, 21% O2 (air) and 100% O2 as oxidant
Kuzeci, Li, and Kammel (1989)	Pentlandite	6.3%	Not specified, concentrate reground in a Netzsch mill using 2 mm steel grinding media	20 g/L H2SO4, 22–23 g/L Fe^3+, 60–95°C solids conc. 20 g/L
Bredenhann and Van Vuuren (1999)	Millerite	50.2%	Not specified	60–90°C 5–45 g/L NaNO3, pH 0 sulfuric acid solution, 500 rev/min, up to 20 hours leaching time, solids conc. not specified
Lu and Lawson (1998), Lu, Jeffrey, and Lawson (2000)	Pentlandite	18.5%	P80 20 µm	60 – 96°C sulfuric acid solution (pH < 0.8), 0–2 M NaCl, 0 & 0.28 M Fe(III), solids conc. 40 g/L, 10 hour leaching time, 1 atm O2 pressure

Bredenhann and Van Vuuren (1999) studied the oxidative dissolution of a high-grade millerite (NiS) concentrate in sulfuric acid media using sodium nitrate (NaNO₃), which was found to be a better oxidant than ferric sulfate; higher temperatures and additions of nitrate (up to 30 g/L NaNO₃) were reported to increase the dissolution of Ni from millerite. Nickel leaching from millerite was reported to be inhibited by the formation of elemental sulfur on the particle surface, which caused the dissolution mechanism to change from a surface chemical reaction-controlled process during the initial stages of leaching to a diffusion-controlled mechanism.

Aaltonen et al. (1980) studied the dissolution of the same Ni-Cu concentrate previously described (Section 2.1) in sulfuric acid. Greater nickel and copper dissolution extents were favored by lower pH and higher temperatures, however Mg dissolution also increased with temperature and higher acid concentrations. The sulfuric acid concentration also influenced the degree of sulfur oxidation to sulfate; decreasing acid concentration resulted in 60% of the sulfur being oxidized to sulfate whilst at pH 2, 40% of the sulfur was oxidized to sulfate. A four-fold increase in leaching rate of pentlandite was observed when pure oxygen gas was used instead of air, which was suggested to be due to faster adsorption of oxygen onto the mineral surface. As stated previously, this dependence of leaching rate on oxygen partial pressure has been observed by several other researchers in the field (Corrans and Scholtz 1976; Filippou, Konduru, and Demopoulos 1997; Kniss, Naboichenko, and Zhukov 2013; Peter, Hansen, and Wadsworth 1973; Schneerson, Leshch, and Frumina 1966). Leaching of the concentrate at pH 3.5 and 80°C under an oxygen atmosphere for five days resulted in the extraction of approximately 80% of the contained Ni. Characterization of the residues produced from sulfuric acid leaching showed that mackinawite was more readily leached than pentlandite. Silica gel formation was observed during leaching at pH 3.5, and this material was reported to coat the surfaces of unreacted sulfide minerals, resulting in a reduced rate of leaching. Iron precipitation occurred during leaching with higher acid concentrations resulting in the formation of a basic ferrous sulfate precipitate [FeSO_4x(OH)_y] whilst hematite formation occurred at lower acidities. Nickel diffusion into silicate and iron oxide phases was found with the Ni content in the former being 0.1–0.4% and the iron oxides typically contained 0.5–1.5% Ni. The mineralogy of these newly formed phases was not specified by the authors of this study.

Kuzeci, Li, and Kammel (1989) conducted both pressure oxidation and atmospheric acid ferric sulfate leaching on a pentlandite concentrate grading 6.3% Ni. The findings from pressure oxidation have been described in the second part of this series. Leaching of the as-received concentrate in 20 g/L H₂SO₄ plus 23 g/L Fe(III) at 95°C for 120 min resulted in less than 20% extraction of the contained Ni and Co in the concentrate, whilst regrinding of the concentrate in an attrition mill for 80 min prior to leaching resulted in extraction of approximately 70% and 50% of the Ni and Co, respectively, under identical leaching conditions. Attrition grinding resulted in an increase in the surface area of the concentrate, which was found to be 6.6–7 m²/g versus 0.72 m²/g in the as-received concentrate, as measured by BET. Presumably, the increase in surface area contributes to an increased reactivity during leaching though there may be other possible mechanisms at play such as surface oxidation and amorphization as seen in several other studies described in Section 2.6.1. Fine grinding and leaching of the concentrate at ambient pressure resulted in similar leaching rates to oxygen pressure leaching of the as-received material. The study by Kuzeci, Li, and Kammel (1989) demonstrated that in the absence of fine grinding, the dissolution rate of pentlandite in acidic ferric sulfate is very slow, presumably due to passivation of the particle surface by elemental sulfur.

Lu and Lawson (1998), Lu, Jeffrey, and Lawson (2000) investigated the dissolution of pentlandite in oxygenated sulfuric acid solutions containing dissolved chloride (added as sodium chloride). The concentrate used in the study was subjected to attrition grinding for 30 min to produce a feed for leaching which had a P₈₀ of 20 µm. Leaching of the ground concentrate in H₂SO₄ (pH <0.8) at 85°C for 10 hours under 1 atm oxygen pressure resulted in only 70% extraction of Ni from the concentrate. Leaching the same concentrate in 0.4 M H₂SO₄ plus 1 M NaCl under identical conditions resulted in 96% of the Ni being leached in a 10-hour period. Chloride is believed to enhance Ni dissolution from pentlandite by promoting the formation of a porous, crystalline layer of elemental sulfur during oxidation, which allows for diffusion of reactants through the product layer to the mineral surface. In the absence of chloride, un-leached pentlandite particles were reported to be coated by an amorphous sulfur product layer, which retarded leaching. At a minimum, 0.5 M NaCl was found to be sufficient to enhance Ni dissolution from pentlandite; further addition of chloride above 0.5 M NaCl did not improve leaching kinetics, and addition of 2 M NaCl resulted in slightly lower dissolution kinetics, which was suggested to be due to reduced solubility of O₂ with increasing ionic strength. An enhancement in base metal sulfide dissolution and formation of porous, crystalline elemental sulfur has also been observed during the leaching of copper mattes and chalcopyrite in oxygenated sulfuric acid solutions containing dissolved chloride (Lawson, Cheng, and Lee 1992; Lu, Jeffrey and Lawson 2000). Subramanian and Ferrajuolo (1976) found that Ni extraction was generally not affected by chloride addition during oxygen pressure leaching of nickeliferous sulfide concentrates at 110°C though chloride additions were lower than what was used in the studies by Lu and Lawson (1998), Lu et al. (2000), whilst Crossley et al. (1998) found the increase in Ni extraction was only slight during oxygen pressure leaching of a nickel sulfide concentrate at 110°C in the presence of chloride ion, increasing from 56.7% to 65.9%. Both studies, however, did find that copper extraction was significantly enhanced in the presence of chloride. The leaching of base metal sulfides in oxygenated mixed chloride-sulfate media has formed the basis for the CESL process (Jones 1996) and the BHAS process in use at Port Pirie (South Australia) for the leaching of copper-lead matte (Lawson, Cheng, and Lee 1992; Meadows, Ricketts, and Smith 1987).

Lu and Lawson (1998), Lu et al. (2000) found that higher oxygen partial pressures enhanced Ni dissolution and have suggested that oxygen likely serves a dual function, serving as the primary oxidant during the initial stages of leaching, with ferric iron taking over the role during the latter stages, with oxygen regenerating Fe(III) via oxidation of Fe(II). The ferrous ions are derived from the dissolution of pentlandite itself as well as iron sulfides in the concentrate. External addition of ferric iron (0.28 M) to an oxygenated mixed sulfate-chloride lixiviant at 85°C did not affect the overall extent of Ni leaching from pentlandite but did promote more rapid leaching kinetics during the initial stages of leaching. Cobalt dissolution was found to much slower than Ni dissolution; more than 90% of the Ni was dissolved from the concentrate after 10 hours of leaching whilst only 40% of the Co dissolved over the same period. The difference in dissolution kinetics of Ni and Co indicated that Co was not present solely as a replacement for Ni in pentlandite and was present as or in a separate phase. Lu, Jeffrey, and Lawson (2000), however, did not determine what the other co-bearing phase was though they performed X-ray powder diffraction (×RD) on the residues, which consisted mostly of elemental sulfur and pyrite and, given that SEM revealed the presence of unreacted particles of pyrite; it is likely that pyrite was a host for Co in the concentrate used in the study.

2.3. Chloride lixiviants

Chloride leaching involves the dissolution of base metal sulfide concentrates or ores exclusively in chloride media; the dissolution can be carried out in the presence of an oxidant such as air, oxygen, chlorine, ferric, or cupric chloride, with leaching being performed either under pressure or at atmospheric conditions. Alternatively, sulfide dissolution can also be carried out under non-oxidative conditions. Several laboratory investigations have been carried out on the leaching of nickel flotation concentrates in chloride lixiviants at atmospheric pressure using a variety of oxidants or under non-oxidative conditions; the key features of these laboratory studies are summarized in Table 3. Laboratory investigations are described under the relevant sub-headings, and distinction is made based on the type of oxidant that has been used.

Table 3. Summary of laboratory investigations carried out on chloride leaching of nickel sulfide concentrates at ambient pressure.

Reference	Nickel sulfide mineralogy	Nickel grade (wt% Ni)	Concentrate feed particle size	Mode of leaching	Leaching conditions
Ghali and Girard (1978)	Concentrate(s) A & B – pentlandite, awaruite, millerite & heazlewoodite Concentrate C – pentlandite Matte – heazlewoodite, metallic Ni	Concentrate A −20.1% Concentrate B−11.0% Concentrate C −5.14% Matte − 74.0%	Concentrate(s) A & B – P75 250 µm Concentrate C – P100 37 µm Matte – P99 250 µm	Batch stirred tank leaching, ambient pressure	1–5.7 M FeCl3, 80°C 100 rev/min, solids conc. 320 g/L, up to 66 hours leaching time
Ferron, Williamson, and Zunkel (1996)	Concentrate #1 & #2 - Pentlandite	Concentrate #1–3.58% Concentrate #2–4.00%	56 to 90% −37 µm	Batch stirred tank leaching, ambient pressure	95°C 7.5 h leaching time, 73–146 g/L Fe(III), 151–289 g/L Cl-, 10% w/v pulp density, impeller tip speed of 100 or 170 cm/s ORP controlled experiments − 95°C 7.5 h leaching time, 10% w/v pulp density 300–525 mV vs SCE via addition of either 300 g/L Fe(III) (as chloride) or Cl2 gas
Fang (1997a)	Pentlandite	7.74%	Not specified	Batch stirred tank leaching, ambient pressure	60 - 99°C 10–35 g/L Cu(II), 80–250 g/L NaCl, up to 12 h leaching time, 21% & 100% O2, 600 rev/min, pulp density not specified
Honey, Muir, and Hunt (1997)	Pentlandite	14.5%	P80 80 µm	Stirred tank leaching, ambient pressure	Leaching at 75, 110 & 115°C (reflux) over a 2-h period in 3 M chloride lixiviant. Oxygen-chlorine used as oxidant with cupric chloride as a catalyst (Cu conc. not specified)
Harris et al. (2004)	Pentlandite	18.65%	Not specified	Ambient pressure	2 or 4 N HCl plus MgCl2, up to 400 g/L total Cl-, 5% w/v solids, 95°C 4 h leaching time, ORP of 170–510 mV vs SHE, NaClO3, NaOCl, Cl2 and O2 used as oxidants, stirring rate not specified
Malkhuuz (2006)	Concentrate #1 – pentlandite Concentrate #2 – pentlandite, millerite	Concentrate #1(low MgO) −13.5% Ni,8.76% MgO Concentrate #2(high MgO) −25.75% Ni,17.4% MgO	Not specified	Batch stirred tank leaching, ambient pressure	60 or 100°C 2 molal MgCl2, 2–8 molal HCl, up to 24 h leaching time, pulp density not specified, 0.5 molal of either CuCl2 or FeCl3 added in some tests
Kshumaneva et al. (2009)	Pentlandite	39.4%	−50 + 20 µm	Batch stirred tank leaching, ambient pressure	0–4 M HCl, 0–4 M FeCl3, 0–4 M CuCl2, 40 - 98°C up to 300 min leaching time, solid:liquid ratio of 1:50

2.3.1. Ferric and cupric chloride leaching

The earliest study into leaching of nickel concentrates using ferric or cupric chloride as an oxidant was carried out by Ghali and Girard (1978) who investigated the leaching of three different nickel sulfide concentrates varying in grade and mineralogy in ferric chloride (FeCl₃) (refer to Table 3). Nickel extractions from the concentrates and sulfide oxidation to elemental sulfur were reported to increase with increasing ferric chloride concentration. The nickel dissolution kinetics were reported to slow down after approximately 70% of the Ni had dissolved. Overall, Ghali and Girard (1978) found ferric chloride to be an effective lixiviant for the extraction of nickel from sulfide concentrates and nickel matte.

Ferron, Williamson, and Zunkel (1996) studied the leaching of a bulk Cu-Ni sulfide concentrate from the Duluth Gabbro (Minnesota) grading 18.0% Cu, 3.58% Ni, and 0.14% Co using ferric chloride, the process being based on the Cuprex Metal Extraction Process (CMEP). Major sulfide minerals in the concentrate were chalcopyrite, cubanite, pentlandite, and pyrrhotite. Preliminary batch leaching experiments found that the extraction of base metals from the concentrate was affected by the agitation speed, feed grind size, and the ORP. Best leaching results were obtained when the concentrate was ground to 89% passing 37 µm, an impeller tip speed of 170 cm/s and a ferric ion concentration of 97 g/L (196 g/L Cl-), where greater than 93% of the base metals were extracted after leaching for 7.5 hours at 95°C Further leaching tests were performed with a stand-in concentrate (18.0% Cu, 4.00% Ni, no cubanite) to determine if selective leaching of Ni could be carried out through control of the ORP (vs saturated calomel electrode, SCE); the potential being maintained at either 300–370 mV through the addition of ferric chloride (300 g/L Fe³⁺) or 515–525 mV using chlorine gas. Ferron, Williamson, and Zunkel (1996) found that base metal extraction was dependent on the ORP with higher potentials favoring greater metal dissolution. There was little opportunity to selectively leach Ni whilst retaining Cu in the residue, with less than 20% of the Ni being leached at an ORP of 300 mV. Excess ferric ion in solution at the conclusion of leaching was stated to be of concern though no reason was given as to why, and the authors suggested that ferric ion could be controlled to low levels at the conclusion of leaching through counter current leaching or by control of the ORP via chlorine addition. Overall, Ferron, Williamson, and Zunkel (1996) demonstrated the suitability of ferric chloride for leaching bulk sulfide concentrate from the Duluth Gabbro.

Kshumaneva et al. (2009) investigated the leaching of a high-grade pentlandite concentrate in aqueous solutions of ferric chloride, as well as cupric chloride, and mixtures of both in hydrochloric acid. The leaching rate and extent of Ni dissolution from pentlandite was influenced by temperature and the composition of the chloride lixiviant. The highest extractions for Ni and Co from the pentlandite concentrate were observed when leaching using a 3:1 FeCl₃-HCl mixture and when using a 2:2 FeCl₃:CuCl₂ mixture as the lixiviant. The leach residues from pentlandite dissolution in FeCl₃-HCl and FeCl₃-CuCl₂ were characterized via XRD and electron microscopy and revealed that pentlandite dissolution did not form any intermediary phases during leaching and the residues consisted of un-leached pentlandite and elemental sulfur. An interesting observation made was that during ferric chloride leaching of pentlandite, un-leached pentlandite grains were occluded in elemental sulfur, and leach kinetics was therefore controlled by diffusion of the lixiviant through this layer of elemental sulfur. Characterization of the residues obtained from FeCl₃-CuCl₂ leaching of pentlandite via SEM revealed that un-leached pentlandite grains were essentially free of adhering elemental sulfur. The implications of these findings were that pentlandite dissolution kinetics is faster in mixtures of FeCl₃ and CuCl₂ than FeCl₃ alone and that there is synergism between FeCl₃ and CuCl₂ during leaching.

2.3.2. Gaseous oxidants (oxygen, chlorine gas)

Fang (1997a) investigated the oxygen leaching of a nickel sulfide concentrate containing Cu and Co at ambient pressure in an aqueous solution of sodium chloride and cupric chloride. The base metal content in the concentrate used in the study was 7.74% Ni, 2.01% Cu, and 0.69% Co. Nickel was present primarily as pentlandite and copper as chalcopyrite. Pyrrhotite constituted the main gangue sulfide mineral in the concentrate and contained 0.37% Ni. The leaching temperature, oxygen partial pressure, and initial Cu(II) concentration in solution exerted a significant effect on the rate of Ni and Co dissolution; dissolution was virtually complete in the presence of 35 g/L of Cu(II) after 10 hours at 85°C when 100% O₂ atmosphere was utilized. The effect of temperature on Ni and Co leaching was more complicated with the leaching rates and extents increasing from 60°C to 85°C and then dropping at 99°C though no explanation was given for this finding. The decrease in leaching rate and extent at 99°C could possibly be due to reduced solubility of oxygen in solution at this temperature. A similar finding was made by Meadows, Ricketts, and Smith (1987) where a decrease in Cu recovery above 85°C was observed during oxygen leaching of a copper-lead matte in a mixed sulfate-chloride lixiviant at ambient pressure, which the authors suggested was due to reduced oxygen solubility in solution. Characterization of the leach residue by SEM-EDX revealed the presence of hematite, goethite, ferrihydrite, elemental sulfur, covellite, and chalcocite. The presence of iron oxides and oxyhydroxides in the residue was suggested to be due to oxidation of ferrous iron in solution to ferric iron, with subsequent hydrolysis of the ferric iron (Fang 1997a). The presence of covellite and chalcocite in the leach residue was suggested by Fang (1997a) to be due to dissolved Cu(II) and Cu(I) undergoing cementation reactions with pentlandite and pyrrhotite, which resulted in Fe(II) and Ni being leached into solution at the expense of copper precipitating as its sulfides.

A follow-up study was carried out by Fang (1997b) to investigate the behavior of Fe, Cu, and S during oxidative leaching of a Ni-Cu-Co concentrate in an aqueous solution of NaCl (150 g/L). The copper concentration in solution was reported to decrease with time, particularly under non-oxidative conditions, which was attributed to the reaction of dissolved copper with sulfide minerals in the pulp. The influence of ORP and pH on the dissolution behavior of Cu and Fe was explored with copper dissolution favored by a low pH (pH ≤1) and high ORP (≥450 mV vs Ag/AgCl). Low iron concentrations (0.01 g/L) were favored by high pH (pH ≥2) and ORP (≥450 mV vs Ag/AgCl). Concentrate dissolution at a pH ≥2 to reject iron, however, resulted in a greater loss of copper to the leach residue. Sulfide oxidation to sulfate increased with increasing ORP for which a value of 410 mV vs Ag/AgCl was suggested to be suitable to maintain adequate formation of elemental sulfur rather than sulfate. The relationship between pH and ORP on Ni and Co dissolution from the concentrate and deportment to the residue was not reported.

Honey, Muir, and Hunt (1997) performed a limited number of experiments on chloride leaching (3 mol/L total chloride) of a Western Australian nickel sulfide concentrate (P₈₀ 80 µm) of grade 14.5% Ni, 0.4% Co, and 0.4% Cu under atmospheric pressure, utilizing a mixture of oxygen and chlorine gases as the oxidant and cupric chloride as redox mediator. From the experiments, it was found that at 75°C essentially no leaching occurred over the two-hour period, whilst high Ni extractions (above 90%) could be achieved within two hours when leaching at 110°C and 115°C. Key experimental details such as stirring rate, cupric chloride concentration, and concentrations of oxygen and chlorine gas were not specified. Honey, Muir, and Hunt (1997) assessed the various hydrometallurgical processes available at the time for direct production of nickel metal from sulfide flotation concentrates and found that chlorine-oxygen leaching at atmospheric pressure in the presence of cupric/ferric chloride, and sulfuric acid pressure oxidation were more economical than other competing technologies such as bioleaching and roast-leach type processes.

2.3.3. Non-oxidative dissolution and use of other oxidants

Non-oxidative leaching of nickel sulfides in hydrochloric acid media formed the basis for the original Falconbridge matte leach process for selective leaching of Ni from smelter matte (Thornhill, Wigstol, and Van Weert 1971) and the AMAX Chloride Refining Process for leaching of Ni and Co from mixed sulfide precipitates (Jha and Meyer 1983; Jha, Carlberg, and Meyer 1983). van Weert, Mah, and Piret (1974) studied non-oxidative hydrochloric acid leaching of nickeliferous pyrrhotite concentrates in the context of selectively dissolving pyrrhotite whilst leaving Ni enriched in the leach residue. Both Dyson and Scott (1976) and Filmer and Balestra (1981) showed that very little Ni dissolution from nickel sulfide flotation concentrates takes place during non-oxidative leaching in mineral acids without ‘thermal activation’ which is described in Section 2.6.2.2. The studies described here, however, were carried out in mixed hydrochloric acid-brine solutions, which have been suggested to have enhanced proton activity (Senanayake 2007).

Jaguar Nickel Inc. patented a process for the dissolution of base metal sulfide bearing material at ambient pressure in a hydrochloric acid-magnesium chloride lixiviant (30–150 g/L HCl, 80–350 g/L MgCl₂) at temperature from 75°C to the boiling point (Harris et al. 2004). The patent was not limited to the use of MgCl₂, and other chloride salts (Na, K, Ca) could be used though MgCl₂ was preferred due to its recyclability. A key aspect of the invention was the control of the ORP (250–600 mV vs SHE) such that sulfidic sulfur is converted to hydrogen sulfide (H₂S) rather than elemental sulfur or sulfate, and precious metals are retained in the leach residue. Higher redox potentials (>600 mV vs. SHE) resulted in dissolution of precious metals and formation of sulfate rather than H₂S, whilst lower redox potentials (<250 mV vs SHE) were too reducing for base metal sulfide dissolution. The oxidants covered in the patent included alkali perchlorates, chlorites, hypochlorites, and peroxides, as well as gaseous chlorine and oxygen. Examples were given in the patent showing the effectiveness of HCl-MgCl₂ as a lixiviant both in the presence and absence of an external oxidant. In the absence of an oxidant, leaching a concentrate containing 18.5% Ni, 1.38% Cu, and 0.19% Co in 2 N HCl plus MgCl₂ at a total Cl concentration of 300 g/L resulted in extraction of 66%, 79%, and 58% of the Ni, Cu, and Co, respectively, after leaching for 4 hours at 95°C Leaching under similar conditions but with FeCl₃ (230 g/L total Cl-) instead of MgCl₂ resulted in poorer Ni extraction (47%) and slightly lower Cu extraction (75%); cobalt recovery was higher in the presence of ferric chloride (70%). No explanation was given by the inventors for the observation though Malkhuuz (2006) reported a similar finding, which was suggested to be due to cupric and ferric chloride retarding metal extraction by reaction with H₂S to form elemental sulfur and copper sulfides on the particle surface. Further examples were provided by the inventors demonstrating the effectiveness of HCl-MgCl₂ for concentrate leaching in the presence of oxidants such as Cl₂, NaClO₃, and NaOCl. In one example, leaching the Ni concentrate in a HCl-MgCl₂ lixiviant using 5.25% w/v NaOCl as an oxidant at an ORP of 398 mV (vs SHE) resulted in Ni, Cu, and Co extractions of 98.2%, 85.8%, and 86.2%, respectively. The inventors suggested that HCl-MgCl₂ was an effective lixiviant due to enhanced proton activity in high ionic strength chloride solutions, with similar claims being made by Malkhuuz (2006) and Senanayake (2007).

Malkhuuz (2006) studied the dissolution of individual nickel sulfide minerals (millerite, violarite, and heazlewoodite) in HCl-MgCl₂ solutions. Among the nickel sulfides studied, heazlewoodite was found to be the most easily leached at all acid concentrations studied (1–10 molal). Millerite and violarite were found to be refractory over an acid concentration range of 1–6 molal though 56.7% of the contained Ni in millerite was dissolved in 10 molal HCl at 60°C. The leaching behavior of two different nickel sulfide concentrates in HCl-MgCl₂ was also studied by Malkhuuz (2006). The best leaching results with respect to Ni and Co extraction for both concentrates were achieved after leaching for 1 to 2 hours at 100°C in 6 to 8 molal HCl plus 2 molal MgCl₂ with 95% of the Ni and 75–76% Co being extracted. Characterization of leach residues via XRD under optimal leaching conditions for Ni showed that the residues consisted of pyrite, talc, and quartz, which were inert during leaching; pyrite inertness, however, could be an issue if it were a major host for cobalt, which might explain the lower Co extractions observed by Malkhuuz (2006). Malkhuuz (2006) found that CuCl₂ and FeCl₃ were not beneficial toward sulfide leaching due to reaction of Cu(II) and Fe(III) ions with H₂S to form passivating layers on the particle surfaces; copper enrichment on particle surfaces in leach residues was observed by SEM-EDX which could support the theory for Cu(II) inhibition due to sulfide precipitation. van Weert, Mah, and Piret (1974) found ferric ions had an inhibitory effect during non-oxidative dissolution of nickeliferous pyrrhotite concentrates but was cautious in ascribing this to the formation of elemental sulfur on the particle surface given that no elemental sulfur layer could be detected by EPMA. Filmer and Balestra (1981) whose work is described in more detail later have shown that non-oxidative dissolution of sulfides is dependent on the sulfur-to-metal stoichiometric ratio at the reaction interface and found that concentrates low in sulfur produce a more cathodic mixed potential in hydrochloric acid and are therefore more reactive. This would suggest that the inhibition in non-oxidative dissolution in the presence of ferric ions observed by van Weert, Mah, and Piret (1974), Harris et al. (2004), and Malkhuuz (2006) could possibly be due to ferric ions leaching metals from the sulfide particle surface thus producing a metal-deficient sulfide layer, resulting in a shift toward an anodic potential and reduced reactivity in hydrochloric acid.

2.4. Nitric acid leaching

Nitric acid functions both as a lixiviant and an oxidant and has been utilized in the leaching of a range of base metal sulfide concentrates. It is interesting to note that despite the stated benefits of using nitric acid as a lixiviant, there are few published studies into nitric acid leaching of nickel sulfide concentrates which are summarized in this series.

Ouellet, Torma, and Bolduc (1975) studied the leaching of a copper-nickel sulfide concentrate and ore (refer to Table 4) containing pentlandite, pyrrhotite, and chalcopyrite, in a mixed sulfuric acid-nitric acid lixiviant. The Ni and Cu grade in the concentrate used by Ouellet, Torma, and Bolduc (1975) was 5.14% Ni and 11.50% Cu. A similar approach was taken by Fossi et al. (1977) in nitric acid leaching of nickel matte and MSP. The use of sulfuric acid reduces the amount of nitric acid required for leaching since nitric acid only serves as an oxidant and is also not being used to complex the dissolved metals in solution. Additionally, leaching in a mixed nitric acid-sulfuric acid solution would be preferable as Fossi et al. (1977) state that the presence of nitrate ions in solution is undesirable for subsequent nickel electrowinning and by taking this approach nitrate formation in solution should theoretically be eliminated; this was not found to be the case in practice, and elaborate methods were required to separate Ni from a mixed sulfate-nitrate solution. The leaching experiments by Ouellet, Torma, and Bolduc (1975) were performed under a range of conditions at atmospheric pressure. Nickel extractions from the concentrate were very high with almost complete extraction occurring at 85°C (97.6%), however Cu and Co extractions were lower, being only 79.9% and 81.4%, respectively, at 95°C Optimal leaching conditions for metal dissolution were determined to be a temperature of 85°C nitric acid concentration of 6.3 mol/L, sulfuric acid concentration of 0.8 mol/L, leaching time of 90 min, and a pulp density of 25% (Ouellet, Torma, and Bolduc 1975). The extent of Fe extraction during leaching was not reported by Ouellet, Torma, and Bolduc (1975); however, jarosite was identified in leach residues via XRD suggesting that at least some Fe must undergo hydrolysis and precipitation during leaching.

Table 4. Summary of laboratory investigations carried out on nitric acid leaching of nickel-bearing sulfide concentrates at ambient pressure.

Reference	Nickel sulfide mineralogy	Nickel grade (wt% Ni)	Concentrate feed particle size	Mode of leaching	Leaching conditions
Ouellet, Torma, and Bolduc (1975)	Ore & concentrate -pentlandite	Ore − 1.83% Concentrate − 5.14%	P100 32 µm	Ambient pressure leaching	50 - 95°C 0–7.5 M HNO3, 0–0.95 M H2SO4, 10–45% pulp density, up to 120 min leaching time
Shukla, Mukherjee, and Gupta (1978)	Pentlandite, millerite	6%	P100 149 µm	Ambient pressure leaching	40 - 100°C 10% pulp density, up to 10-hours leaching time, agitation rate of 400 rpm, 10–16% HNO3 Roast-leach – concentrate roasting at 450 - 700°C calcine leached in 15% HNO3 at 100°C for 4 hours

Shukla, Mukherjee, and Gupta (1978) conducted nitric acid leaching on a copper-nickel flotation concentrate grading 6% Ni and 9% Cu, with nickel being present as pentlandite and millerite and copper as chalcopyrite. Direct nitric acid leaching as well as roasting followed by nitric acid leaching of the calcine were explored in the study. Close to 100% recovery of Ni from the concentrate was obtained by direct leaching in 15% HNO₃ at 100°C for 8 hours at a pulp density of 10%. Copper extraction from direct leaching was reported to be only 82%, whilst 82% of the iron present in the concentrate was also dissolved during leaching. Shukla, Mukherjee, and Gupta (1978) reported that 60% of the HNO₃ used for leaching could be regenerated simply by oxidizing the evolved NO gas using oxygen and scrubbing the NO₂ using water. Roasting prior to nitric acid leaching was also found to be effective for leaching Ni and Cu from the concentrate, provided roasting temperatures were between 450°C and 500°C The benefits of roasting prior to leaching were reported to be reduced dissolution of iron in subsequent leaching, thereby improving the purity of the leach liquor, and reducing nitric acid consumption during leaching.

2.5. Bacterial leaching

Bacterial leaching (also termed bioleaching, bio-oxidation, or bacterial oxidation) takes place in an acidic sulfate medium where ferric ions act as an oxidant, the ferric ions oxidizing the mineral sulfides to metal ions and elemental sulfur (Watling 2006). The role of the microorganisms in bioleaching is to regenerate the ferric ions and oxidize elemental sulfur to sulfate, generating acid in the process (Watling 2006). Bacterial leaching processes are characterized by the class of microorganisms involved in leaching which impose limitations on operating temperatures; most bioleaching processes utilize mesophilic microorganisms and operate below 40°C (Norris 1997; Watling 2006). Bacterial leaching of nickel sulfide ores and concentrates has been extensively researched and is the subject of an excellent review by Watling (2008). Discussion of research carried out to date on the bioleaching of nickel flotation concentrates is limited to studies after 2008, and readers are referred to the review by Watling (2008) for a more extensive review of nickel concentrate bioleaching and bioleaching of whole nickel sulfide ores. Since the review by Watling (2008), there have been further studies carried out on the bioleaching of nickel concentrates which are summarized in Table 5. Bioleaching has been carried out using a variety of microorganisms (mesophiles, moderate thermophiles, and thermophiles) on nickel and copper-nickel sulfide concentrates. A survey of the research carried out since 2008 on the bioleaching of nickel concentrates (refer to Table 5) shows that the majority of the research has focused on the use of moderate thermophiles and thermophilic archaea due to the much faster leaching kinetics these enable.

Table 5. Summary of laboratory and pilot plant investigations carried out since 2008 on bioleaching of nickel sulfide concentrates using various microorganisms.

Reference	Nickel sulfide mineralogy	Nickel grade (wt% Ni)	Concentrate feed particle size	Microorganisms	Mode of leaching	Leaching conditions
Neale et al. (2009)	Pentlandite, trace gersdorffite	5.24%	D90 varied between 10 & 35 µm	Thermophiles	Continuous stirred tank leaching	70°C 10% w/v solids, pH 1.1–1.3, 6-day residence time, Eh 600 mV vs Ag/AgCl
Cruz et al. (2010)	Pentlandite	5.9%	P80 47 µm	Mesophiles & moderate thermophiles	Shaker flasks	34°C (mesophiles), 50°C (thermophiles), 5% w/v solids, pH 1.2–2.4, 0–10 g/L Fe(II), up to 25 days leaching time
Gericke and Govender (2011)	Pentlandite, trace gersdorffite	5.6%	D90 varied between 10 & 40 µm	Moderate thermophiles & thermophiles	Batch & continuous stirred tank leaching	45°C (moderate thermophiles), 65 – 75°C (thermophiles), 3–6 days residence time, 10% w/v solids, pH 1.25–1.7, Eh 430–580 mV vs Ag/AgCl (thermophiles), 420–670 mV vs Ag/AgCl (moderate thermophiles)
Wakeman, Honkavirta, and Puhakka (2011)	Pentlandite, gersdorffite, niccolite	Concentrate #1–10.5% Concentrate #2–14.3%	Not specified	Mesophiles	Shaker flasks & continuous stirred tank leaching	6 – 27°C 2, 5 & 10% w/v solids, up to 56 days leaching time, pH 3
Mwase, Petersen, and Eksteen (2012)	Pentlandite	3.4%	83% passing 38 µm coated onto granite pebbles	Thermophiles	Column	65 – 80°C 10 g/L H2SO4, 2 g/L Fe, 88 days leaching time, lixiviant flow rate of 1000 mL/day, aeration rate of 130 mL/min
Norris (2017)	Concentrate #1 – Pentlandite, violarite Concentrate #2 & #3 – pentlandite	Concentrate #1–10.5% Concentrate #2–8.8% Concentrate #3–7.6%	Concentrate #1 – D80 75 µm Concentrate #2 – D80 88 µm Concentrate #3 – D80 9 µm	Moderate thermophiles & thermophiles	Continuous stirred tank leaching	Moderate thermophiles − 49°C pH 2.5 (concentrate #1), pH 1 (concentrate #2, #3), 5% w/v solids Thermophiles − 77°C pH 1.3, 5% w/v solids
Sun et al. (2020)	Pentlandite	27.4%	P100 74 µm	Mesophiles	Chemical leaching – shaker flasks Bioleaching – batch stirred tank leaching	Chemical leaching: pH 1, 33°C 3% w/v solids, 0–9 g/L Fe^3+ Bioleaching: pH 2–3, 33°C 2% w/v solids

2.5.1. Bioleaching using mesophilic microorganisms

The use of mesophiles operating at temperatures below 45°C for the bacterial leaching of nickel sulfide concentrates has been the subject of investigation from a limited number of researchers (Cruz et al. 2010; Sun et al. 2020; Wakeman, Honkavirta, and Puhakka 2011). Sun et al. (2020) investigated pentlandite bioleaching using mesophiles. The pentlandite used in the study was a high-quality sample prepared by flotation and magnetic separation that graded 27.43% Ni. This study aimed to elucidate the mechanism of pentlandite bioleaching and maximizing Ni extraction was not the focus. Nickel release from pentlandite was reported to take place via transformation of pentlandite, to millerite and finally to Ni²⁺ which is released into solution. Cruz et al. (2010) evaluated the use of mesophilic and moderate thermophilic microorganisms for bioleaching of a nickel concentrate grading 5.9% Ni (as pentlandite). Bioleaching the nickel concentrate using mesophiles at 34°C in the pH range of 2.0–2.2, solid concentration of 5% w/v, and in the presence of 2.5 g/L Fe(II) resulted in Ni extraction of 50% after 5 days of leaching. Nickel dissolution reached close to 80% after 5 days of leaching at 50°C using moderate thermophiles, confirming the much slower leach kinetics when using mesophilic microorganisms.

Wakeman, Honkavirta, and Puhakka (2011) investigated the bioleaching of nickel concentrates produced as a by-product of talc mining using mesophilic microorganisms. Two different flotation concentrates were trialed for bioleaching that differed in terms of their nickel and arsenic contents, the first concentrate graded 10.5% Ni and 1.3% As and was obtained as a by-product of talc processing, the second concentrate was produced from reverse flotation of the first concentrate and graded 14.3% Ni and 15.4% As; arsenic was present in both concentrates as gersdorffite and niccolite. Bioleaching of both concentrates in shaker flasks at 27°C pH of 3 and a pulp density of 2% resulted in extraction of 95% of the Ni after 56 days of leaching; cobalt extraction was 85% and 47% for concentrate #1 and concentrate #2, respectively. Nickel and cobalt extractions worsened with increasing pulp density, decreasing temperature and during leaching using stirred tank reactors, though the latter was attributed to poor mixing. Arsenic extractions were reported to be very low during bioleaching of concentrate #1, the leach liquor tenor being only 13 mg/L but this reached as high as 2.3 g/L during leaching of the arsenic rich concentrate (#2). The latter was attributed to the lower total iron content in the concentrate (25.3% vs 35.0%) with an iron: arsenic ratio of 4 in the concentrate being crucial to ensure arsenic removal from solution by co-precipitation with ferric iron (Wakeman, Honkavirta, and Puhakka 2011).

2.5.2. Bioleaching using thermophilic microorganisms

Gericke and Govender (2011) evaluated different bioleaching strategies for the extraction of Ni and Cu from a low-grade Ni-Cu sulfide concentrate from the Aguablanca mine, Spain, containing pentlandite and chalcopyrite. Leaching strategies included concentrate regrinding to various levels of fineness and use of either moderate thermophilic or thermophilic microorganisms with redox control (oxidizing conditions) during leaching (Gericke and Govender 2011). Bioleaching using thermophilic microorganisms with no redox control (oxidizing conditions) at 70°C 10% solids, feed particle size of D₉₀ 10 µm and 6-day residence time resulted in extraction of 95% and 99% extractions of Cu and Ni, respectively. Implementing redox control during thermophilic leaching in the first leach stage resulted in an increase in chalcopyrite leaching rate and the stated benefits of redox control were a reduced residence time and use of a coarser grind size during leaching. Staged bioleaching with redox control using moderate thermophiles at 45°C grind size of D₉₀ 10 µm, 20% solids, and 7 days overall residence time resulted in extractions of 91% and 94% for Cu and Ni, respectively. Thermophilic bioleaching of a Cu-Ni sulfide concentrate from the Aguablanca mine at 70°C was also tested at scale in an integrated pilot plant producing copper cathode and nickel hydroxide precipitate (Neale et al. 2009). Thermophilic bioleaching under optimal conditions resulted in extractions of 95.1% and 99.4% for Cu and Ni, respectively.

Mwase, Petersen, and Eksteen (2012) investigated the column leaching of base metal sulfides using sulfuric acid-ferric sulfate solution and thermophilic microorganisms as part of a two-stage leaching process for extracting base metals (Ni, Cu, Co) and subsequently PGMs from a PGM-bearing flotation concentrate derived from Platreef ore (South Africa). The effect of temperature on column leaching using thermophiles was evaluated with the temperature being varied between 65°C and 80°C. Base metal extractions worsened with increasing temperature with the best leaching results being achieved at 65°C over an 88-day period, with Ni, Cu, and Co extractions being 99.5%, 91.1%, and 83.5%, respectively. Norris (2017) investigated the bioleaching of nickel and copper-nickel sulfide concentrates using moderate thermophiles and high-temperature thermophiles (thermophilic archaea) in continuous stirred tank reactors (refer to Table 5). Leaching of the nickel concentrate using a mixed culture of moderate thermophiles at 49°C and pH 1.4 resulted in Ni extraction of 85% after a residence time of 2.7 days. Leaching of a finely ground copper-nickel concentrate grading 7.6% Ni and 4.3% Cu under identical conditions resulted in Ni extraction of 89%, however, Cu extraction was only 50%. Leaching of the same finely ground copper-nickel concentrate using thermophilic archaea at 77°C and pH 1.3–1.4 resulted in extraction of 92% of the Cu after a residence time of 2.5 days. The work of Gericke and Govender (2011) and Norris (2017) demonstrated that use of thermophilic archaea operating at or above 70°C for bioleaching of nickel concentrates results in high extractions of Ni and Cu relative to moderate thermophiles operating at 45–49°C. The main drawback to the use of thermophilic archaea is that the microorganisms can only tolerate lower pulp densities (up to 12.5% w/v) relative to moderate thermophiles (20% w/v), the consequence being lower metal tenors in solution and larger process equipment requirements/reduced throughput.

2.6. Concentrate pre-treatment strategies

Several pre-treatment strategies for nickel concentrates have been evaluated by researchers to render the concentrates more amenable to hydrometallurgical processing. Concentrate pre-treatment options generally fall into two strategies: mechanical activation and pyrometallurgical pre-treatment. Research carried out to date on the use of these strategies to enhance the dissolution of nickel from sulfide concentrates is reviewed and summarized. There has only been one published study on the use of sonication to enhance leaching, which has also been briefly described.

2.6.1. Mechanical activation

The use of ultra-fine grinding prior to sulfide leaching forms the basis for several piloted and commercialized processes, which were presented in Part I of this series. The application of mechanical activation to the pressure leaching of nickel sulfide concentrates was reviewed in the second part of this series. Here, laboratory investigations into the use of mechanical activation prior to atmospheric leaching of nickel sulfide concentrates are reviewed, and a summary of these investigations is presented in Table 6.

Table 6. Summary of laboratory investigations carried out to date on mechanochemical treatment of nickel sulfide concentrates prior to atmospheric leaching using a range of lixiviant systems.

Reference	Nickel sulfide mineralogy	Nickel grade (wt% Ni)	Milling equipment	Milling conditions	Leaching media	Leaching conditions
Kulebakin (1977)	Pentlandite	25%	Planetary mill, 500 g of beads (wet milling), 1200 g (dry milling)	Wet milling − 25 g concentrate, 125 mL water, 7 min milling time Dry milling − 3g	Thiobacillus Ferrooxidans in 9K nutrient medium^a + 9 g/L Fe(III)	Ambient pressure, 8 hours, solids conc. 20 g/L, temperature not specified
Maurice and Hawk (1999)	Pentlandite	8.10–11.76% (depends on size fraction)	Shaker mill, 4.8 mm stainless steel beads, 30% filling factor	5:1 charge ratio^b, 1–30 min, dry conditions	1 M FeCl3 +1 M NaCl +0.25 M HCl	70 - 90°C 60–150 min, solids conc. 10 g/L, ambient pressure
Baláž et al. (1998, 2000)	Pentlandite	23.19%	Attrition mill, 2000 g of 2 mm beads, 86% fill factor	100g concentrate, 200 mL water, 15–60 min milling time at 25°C	Water, 0.25 M Fe2(SO4)3 +0.25 M H2SO4	45 - 90°C 120 min, solids conc. 4 g/L, ambient pressure

^aAqueous sulfuric acid solution containing dissolved salts [(NH₄)₂SO₄, KCl, K₂HPO₄, MgSO₄·7 H₂O, Ca(NO₃)₂, FeSO₄.7 H₂O].

^bMass of grinding media to mass of concentrate.

Kulebakin (1977) studied the bioleaching of a pentlandite concentrate that had been pre-treated by mechanical activation in a planetary mill under wet and dry conditions. Mechanical activation under both wet and dry conditions was reported to result in amorphization of the crystal structure and an increase in surface area from 0.5 m²/g (untreated) to 3.7–3.8 m²/g for milled samples. The increase in surface area of pentlandite after mechanical treatment did not necessarily translate to an increase in bacterial leaching rate; dry milled samples were found to be coated in a layer of magnetite, which was believed to inhibit bacterial leaching. Wet milling was reported to result in the formation of nickel sulfate (NiSO₄) on the surface, probably because of surface oxidation; the leaching rate of pentlandite was found to be five times faster after wet milling.

Baláž et al. (1998, 2000) studied Fe₂(SO₄)₃ leaching of a pentlandite concentrate grading 23.19% Ni, 0.75% Cu, and 0.22% Co subjected to mechanical activation in an attrition mill for varying times (0–60 min). In all cases, wet milling of the concentrate resulted in an increase in base metal dissolution relative to the as-received concentrate during leaching using a mixture of 0.25 M Fe₂(SO₄)₃ and 0.25 M H₂SO₄ for 120 min at 90°C Leaching of the as-received concentrate only resulted in Ni, Cu, and Co extractions of 12.74%, 19.31%, and 36.36%, respectively. The best results were achieved by wet milling for 60 min followed by leaching, where Ni, Cu, and Co extractions were 72.90%, 99.78%, and 89.73%, respectively. The specific surface area of the concentrate was measured and was found to increase with time up to 30 min; further increase in the grinding time did not result in an increase in surface area, and this was suggested to be due to recombination of fine grains. The degree of amorphization was reported to increase with increasing milling time, and based on these findings, Baláž et al. (2000) suggested that amorphization during grinding must be responsible for the increased dissolution observed. Leaching of the mechanically activated concentrate was also carried out in water at 90°C for 120 min where it was found that Ni, Cu, and Co extractions of 30.76%, 8.05%, and 6.62%, respectively, indicating that oxidation of pentlandite must be taking place during wet grinding in the attrition mill (Baláž et al. 1998). The difference in dissolution between Ni and Co suggests that cobalt is not exclusively hosted in pentlandite and Baláž et al. (2000) suggested that pyrite in the concentrate may be a host for Co.

Maurice and Hawk (1999) investigated the ferric chloride (FeCl₃) leaching of a mechanically activated pentlandite-chalcopyrite concentrate, activation being achieved either in a shaker mill (stainless steel balls or autogenously) or in a horizontal ball mill (autogenously only). Mechanical pre-treatment was reported to result in an increase in Cu and Ni dissolution from the concentrate. The best results for Ni and Cu extraction were achieved by grinding concentrate in a shaker mill with stainless steel grinding media for 20–30 min followed by leaching in 1 M FeCl₃ at 90°C. Autogenous milling was less effective than grinding in stainless steel media; however, improvements in Ni and Cu extraction were still achieved. Mechanical pre-treatment was reported to result in an increase in specific surface area and formation of crystalline defects in the concentrate which was responsible for increased dissolution in the lixiviant (Maurice and Hawk 1999). While the reported studies have demonstrated increased Ni leaching rates from mechanically activated nickel concentrates, very little has been done to elucidate the structural and chemical changes occurring in the nickel concentrate(s) as a result of mechanical activation.

2.6.2. Pyrometallurgical pre-treatment (roast-leach processes)

Several researchers have investigated pyrometallurgical pre-treatment of nickel sulfide concentrates at temperatures well below that utilized in smelting (<1300°C) to render the concentrates more amenable to leaching. Pyrometallurgical pre-treatment processes that have been studied include oxidation roasting and two-stage oxidation roast-reduction roasting (Caron-type processes), reduction roasting, chloridizing roasting (also termed salt roasting), and sulfation roasting. The choice of roasting process affects subsequent chemical changes in the original feed material and therefore the choice of lixiviant. In the case of chloridizing or sulfation roasting, the aim is to convert the base metals to water-soluble salts, and hence, leaching of the calcines is typically performed in water or dilute acids. In the case of reduction roasting, the sulfides are generally converted to metal-rich and conversely sulfur-deficient sulfides that are more reactive and readily leached under non-oxidative conditions in mineral acids. Whilst oxidative leaching could also be performed, the purpose of the reducing roast treatment is to avoid the formation of passivating layers during subsequent leaching. A summary of research carried out on pyrometallurgical pre-treatment of nickel sulfide concentrates is presented in Table 7 where it can be seen that chloridizing and sulfation roasting have received the most attention. Many of these studies have not proceeded beyond laboratory and pilot plant investigations. The only commercial application of roast-leach processes has been in the treatment of base metal (Ni, Cu, Co) bearing pyrrhotite concentrates in Canada and pyrite-pyrrhotite concentrates in Finland which are described in detail in Part I of this review; these operations are now closed. A potential issue with pyrometallurgical pre-treatment is the generation of SO₂ emissions, which by today’s environmental standards would necessitate the use of emission abatement equipment or a sulfuric acid plant (or Claus plant for H₂S emissions) for any commercial operation, resulting in an increase in capital cost.

Table 7. Summary of research carried out to date on pyrometallurgical pre-treatments of nickel concentrates to render them more amenable to subsequent hydrometallurgical processing.

Reference	Nickel sulfide mineralogy	Nickel grade (wt% Ni)	Additive	Roasting conditions	Leaching conditions
Caron-type processes & oxidation roasting
Forward, Kudryk, and Samis (1948)	Not specified	Concentrate − 16.7% Middling − 2.86% Matte − 55.6%	Oxidizing roast stage – air Reduction roast stage – H2	Oxidizing roast − 760 to 815.6°C Reduction roast − 350 to 600°C	8% NH3 +8% (NH4)2CO3, 21.1–37.8°C 5% pulp density, 2 hour leaching time
Queneau et al. (1949)	Pyrrhotite, pentlandite	<1%	Oxidizing roast stage – air Reduction roast stage – CO + CO2 + H2 + H2O	Oxidizing roast − 704.4 to 1148.9°C Reduction roast − 537.8 to 926.7°C	6–10% NH3 +3–7% CO2, leach temperature, pulp density & time not specified
Leo and Georg (1954)	Not specified	5.6%	None	Not specified	Calcine leached in aqueous ammonia with aeration, however specific conditions not provided
Zhu et al. (2012)	Pentlandite, vaesite, violarite	6.20%	None	350–910°C	2 M HCl, 70 - 100°C 5 hour leaching time, L:S ratio varied up to 200:1
Reduction roasting
Dyson and Scott (1976)	Pentlandite, violarite, millerite	12.4–31.1%	None	650–850°C under a stream of CH4 gas for 1 hour	5 N HCl or 5 N H2SO4 at 100°C for up to 4 hours, solids conc. 250 g/L
Filmer and Balestra (1981)	Pentlandite	7.5%	None	850°C under a stream of H2 gas for various unspecified times	3–10 N HCl at 60°C for up to 4 hours, solids conc. 1.1 or 1.4%, 450 rpm agitation speed
Chloridizing roasting
Kershner and Hoertel (1961)	Siegenite in both concentrates	Cu concentrate – 0.7–0.9% Fe concentrate – 2–3%	NaCl	450–600°C 2 or 3 hour period	Calcine leached in water acidified to pH 0–1.5 (using HCl)
Smith and Iwasaki (1976, 1985)	Pentlandite	2.02%	Cl2	1^st stage – S2Cl2 conversion @ 300°C 2^nd stage – chlorination @ 100 - 300°C 3^rd stage – ferrous chloride decomposition (oxidation) @ 300 - 500°C	Distilled water for 10 min at a pulp density of 20% w/v; temperature of leach not specified
Mukherjee et al. (1985)	Pentlandite, millerite, violarite, bravoite^a	8%	NaCl	280–450°C up to 8 hours	Acidified water at pH 2
Imideev et al. (2014)	Pentlandite	6.88%	NaCl	300–500°C up to 3 hours	Distilled water at 90°C for 1.5 hours, S:L ratio 1:7
Xu et al. (2017)	Pentlandite	9.37%	NH4Cl	200–325°C up to 5 hours, N2 atmosphere	Distilled water at 90°C for 1 hour, solids conc. 200 g/L
Aleksandrov et al. (2019a, 2019b)	Pentlandite	5.78%	NaCl	300–500°C up to 3 hours	Distilled water at 90°C for 1.5 hours, S:L ratio 1:7
Li et al. (2020)	Pentlandite	9.35%	NH4Cl	1^st stage − 250°C 2 hours 2^nd stage − 650°C 1 hour	Distilled water at 80°C for 1 hour, solids conc. 50 g/L
Sulfation roasting
Fletcher and Shelef (1963, 1964)	Pentlandite	7.92%	Alkali metal sulfates (Li, Na, K, Rb, Cs)	680°C under a 90/10 air/SO2 mixture, no residence time stated	Calcine leached in boiling water
Fletcher and Hester (1964)	Pentlandite	4.69%	Na2SO4	670–690°C under air, 6-hour residence time	Calcine leached in 50 g/L H2SO4 at 60% solids and 80°C for up to 6 hours
Weir et al. (1983)	Ni & Co-bearing pyrite concentrates	-	Na2SO4	500–800°C under air, no residence time stated	Calcine is repulped in sulfate solution and then oxygen pressure leached at 100 - 180°C and 50–1500 kPa O2 overpressure
Rao, Natarajan, and Padmanabhan (2001)	Pentlandite & other unspecified sulfides	7%	Li, Na & K sulfate	500–650°C 4 to 6 hour holding time	Calcine water-leached twice at 60°C for 30 min at 30–40% solids
Yu, Utigard, and Barati (2014a)	Pentlandite	17.9%	None	650–775°C under air for up to 13 minutes	Water leach at 90°C for 30 min, solids conc. 4 g/L
Yu, Utigard, and Barati (2014b)	Pentlandite	17.9%	Na2SO4	660–740°C under a 95/5 air/SO2 mixture for up to 350 min	Water leach at 90°C for 30 min, solids conc. 4 g/L
Cui et al. (2018)	Pentlandite	5.7%	Mixture of (NH4)2SO4 & Na2SO4	300–550°C for up to 4 hours under a 80/20 N2/O2 mixture	Water leach at 95°C for 2 hours, S:L ratio not specified
Li et al. (2018)	Pentlandite	9.7%	Na2SO4	400–800°C up to 180 min holding time	Water leach at 80°C for 2 hours, S:L ratio 1:20
Li et al. (2019)	Pentlandite	8.93%	(NH4)2SO4	300–750°C up to 180 min holding time	Water leach at 80°C for 30 min, S:L ratio 1:4

^aNickel-rich variety of pyrite [(Fe, Ni, Co)S₂.

2.6.2.1. Caron-type processes and oxidation roasting

Roasting of sulfide concentrates under an oxidizing atmosphere to decompose sulfides into their respective oxides prior to leaching has been researched based on adapting the Caron process to treat nickel sulfide ores and concentrates (Forward, Kudryk, and Samis 1948; Queneau, Sproule, and Illis 1949). The patent filed by Queneau, Sproule, and Illis (1949) culminated in Inco deploying a modified Caron process for recovering Fe, Ni, and Cu from pyrrhotite concentrates between 1956 and 1982. A modified Caron process was originally pursued by Forward, Kudryk, and Samis (1948) in developing a process to extract Ni and Co from Lynn Lake flotation concentrate prior to discovering that sulfides could be dissolved in ammonia-ammonium carbonate in the presence of sufficient oxygen without any need for roasting, leading to the development of the Sherritt-Gordon ammonia leach process. Oxidation roasting has also been evaluated based on improving Ni extractability during hydrochloric acid leaching (Zhu et al. 2012) or selective oxidation of iron sulfides prior to ammoniacal leaching (Leo and Georg 1954).

The Caron process for extraction of Ni and Co from oxidized ores was patented in 1924 by M. H. Caron (Caron 1950) and involves reduction roasting to convert Ni and Co oxides to their respective metals followed by leaching of the reduced calcine in oxygenated ammonia-ammonium carbonate to dissolve Ni and Co (Forward, Kudryk, and Samis 1948). Based on the success of the Caron process in treating oxidized Ni ores at the Nicaro nickel plant, Cuba, the process was adapted by Forward, Kudryk, and Samis (1948) and Queneau, Sproule, and Illis (1949) to process sulfide ores and concentrates. In both cases, an oxidizing roast stage was required to drive off the sulfur prior to reduction. Queneau, Sproule, and Illis (1949) reported that the oxidizing roast temperature was important as Ni leaching suffered at higher roasting temperatures, below 1700°F (926.7°C) was ideal. Conard (2013) has stated that the oxidizing roast temperature is important to limit magnetite formation since substitution of Ni into magnetite, which is structurally similar to nickel ferrite, renders the Ni difficult to reduce to the metallic state during reduction roasting. The temperature and reducing gas composition employed during reduction of the calcines was also critical for Ni recovery during leaching, the goal being to maximize reduction of Ni, Co, and Cu to the metallic state with minimum formation of metallic iron, which adversely affects leaching (Forward, Kudryk, and Samis 1948; Queneau, Sproule, and Illis 1949). Forward, Kudryk, and Samis (1948) have suggested that metallic iron alloys with Ni, thereby inhibiting oxidation and leaching of Ni from the reduced product. To overcome this, Forward, Kudryk, and Samis (1948) devised a selective iron oxidation step, termed ‘digestion’ whereby the reduced calcine is slurried in ammonia-ammonium carbonate and heated to 100°F (37.8°C) with low agitation under an air or oxygen atmosphere above the slurry to selectively oxidize the metallic iron to ferric hydroxide. Leaching data presented by Forward, Kudryk, and Samis (1948) have shown that in the absence of this digestion step, Ni recoveries from a high-grade concentrate in subsequent leaching are only 37% whilst carrying out digestion either under an air or oxygen atmosphere increases Ni recoveries to 60% and 95%, respectively. Forward, Kudryk, and Samis (1948) found that this digestion step was not necessary in treatment of nickel matte, presumably due to the much lower Fe content (0.5% Fe) relative to the flotation concentrate (24.3% Fe) and middlings (51.3% Fe) which would not have inhibited leaching due to passivation. The patent of Queneau, Sproule, and Illis (1949) does not mention a digestion step prior to ammoniacal leaching. However, staff at Inco Ltd. (1956) reported that the reduced calcine after quenching was in a passive state and required ‘activation’ which was achieved by leaching in ammoniacal ammonium carbonate at high pulp density out of contact with air to leach ferrous iron, a different approach to what was taken by Forward, Kudryk, and Samis (1948). With respect to the ammoniacal leaching stage, both Forward, Kudryk, and Samis (1948) and Queneau, Sproule, and Illis (1949) found that the use of oxygen rather than air significantly increased dissolution kinetics; in practice, however, air was used at the Inco iron ore recovery plant (Conard 2013) likely due to the lower cost.

BASF filed a patent in 1954 for an oxidizing roast-ammoniacal leach process, which is in essence a modified Sherritt-Gordon ammonia leach processes, whereby the sulfide material to be leached is subjected to a controlled oxidizing roast to oxidize the iron sulfides to iron oxide and reduce the sulfur content in the concentrate, though enough sulfur is retained in the roasted material to ensure the base metals are present as their sulfides (Leo and Georg 1954). Selective oxidation of the iron sulfides is possible as the kinetics of oxidation of pyrite and pyrrhotite are faster than the sulfides of Cu, Ni, and Co (Fletcher and Shelef 1963). By this approach, the inventors stated that an iron oxide residue with better settling characteristics could be produced and with low sulfur content, which would make it suitable for iron making. Additionally, by oxidizing the iron sulfides before leaching, the amount of ammonia consumed during leaching is reduced as well as a reduction in ammonium sulfate yield. Shukla, Mukherjee, and Gupta (1978) adopted a similar approach though nitric acid was used as the lixiviant, and they reported a reduction in nitric acid consumption due to reduced dissolution of iron after roasting. Only one example was provided by Leo and Georg (1954) involving roasting-ammoniacal leaching of a Cu-Ni sulfide ore grading 5.6% Ni and 2.9% Cu though no roasting or leaching conditions were specified, let alone extraction efficiencies during leaching.

Zhu et al. (2012) studied the dissolution of Ni from a Nu-Cu concentrate grading 6.20% Ni and 2.35% Cu, as part of a hydrometallurgical process for preparation of nickel ferrite. Nickel was present as a variety of different minerals (refer to Table 7) whilst Cu was present as chalcopyrite. Pressure leaching the concentrate in hydrochloric acid under an air atmosphere resulted in very little Ni dissolution with maximum Ni extraction being 35.89% after leaching for 4 hours in 2 M HCl at 160°C. It is not clear if the oxidant (air) was continuously replenished, which would explain the poor recoveries. The authors attributed the poor recovery to the formation of a sulfur film on the particle surface though this is doubtful given that chloride has been demonstrated to be beneficial during pressure leaching with respect to sulfur dispersion (Brown and Papangelakis 2005; Tong 2009). Zhu et al. (2012) roasted the concentrate under an air atmosphere at various temperatures to study the mineralogical transformations. A roasting temperature of 700°C was chosen for subsequent leaching work since roasting at that temperature decomposed the sulfides in the concentrate and Ni and Fe were converted to nickel (II) oxide (NiO) and hematite. X-ray diffraction carried out by Zhu et al. (2012) found that above 800°C nickel ferrite (NiFe₂O₄, mineral name trevorite) formed in addition to NiO. Presumably, nickel ferrite is refractory to leaching in HCl given roasting temperatures above 700°C were not chosen for further work. Complete Ni recovery from the 700°C roasted calcine was achieved by leaching in 2 M HCl at 95°C for 5 hours at a liquid:solid ratio of 200. The use of higher acid concentrations would possibly have accelerated leaching and allowed leaching at higher pulp densities though this was not explored by the authors. The authors also did not evaluate the effects of roasting temperature on subsequent Ni leaching. Only Ni dissolutions were reported, and it does not appear that there was an attempt to separate Ni, Cu, and Co prior to precipitation and calcination to produce the ferrite.

2.6.2.2. Reduction roasting

Dyson and Scott (1976) performed reduction roasting-acid leaching experiments on nickel sulfide concentrates of diverse mineralogy (refer to Table 7). Concentrates were roasted under a methane atmosphere in the temperature range of 650–850°C followed by non-oxidative dissolution in boiling 5 N mineral acid (HCl or H ₂SO₄). Hydrogen and carbon monoxide were reported to be as equally effective as methane for the reductive decomposition of the concentrates. The purpose of the reduction roasting stage was to decompose pyrrhotite and pyrite to an acid-soluble iron sulfide phase approaching the composition of troilite (FeS, the iron-rich end-member of the pyrrhotite group) which was readily dissolvable in acid with liberation of H₂S, instead of elemental sulfur that could coat the nickel sulfides and inhibit nickel sulfide dissolution. Other mineralogical changes reported during reduction roasting were decomposition of violarite to pentlandite and heazlewoodite and inversion of millerite (β-NiS) to αNiS, rendering the nickel sulfides more soluble in subsequent leaching. Optimal roasting conditions were determined to be 750°C under an atmosphere of CH₄ for 1 hour and boiling the roasted concentrates in hydrochloric or sulfuric acid extracted 97–99% of the nickel. In the absence of any thermal pre-treatment, Ni extractions were only 10–11%. Chalcopyrite was reported to be unreactive during non-oxidative dissolution unless an excess of HCl was added, which resulted in extractions of up to 40%. Filmer and Balestra also found Cu dissolution was impacted by HCl concentration with very little Cu dissolution occurring over a 6-hour period except in the presence of 10 M HCl.

Filmer and Balestra (1981) studied the non-oxidative dissolution of a copper-nickel sulfide concentrate grading 7.5% Ni and 11.6% Cu, as pentlandite and chalcopyrite. In the absence of any thermal pre-treatment, the Ni dissolution was very low with only 17% Ni (and 9.5% Fe) extraction after 6 hours of leaching at 60°C in 3 M HCl, consistent with the findings of Dyson and Scott (1976). Thermal activation under a hydrogen atmosphere at 850°C for an unspecified time converted pyrrhotite to troilite and chalcopyrite to a mixture of bornite and chalcocite whilst pentlandite and pyrite were not transformed. Additionally, the sulfur content in the concentrate decreased relative to the unactivated material (22.9% vs 26.9%) and the metal-to-sulfur-ratio increased (1.15 vs 0.91). The result being that leaching of this material under identical conditions resulted in an increase in Ni and Fe extraction of 70% and 91%, respectively. Filmer and Balestra (1981) found that reducing the sulfur content further through prolonged thermal treatment under a hydrogen atmosphere improved reactivity of the concentrate in acid with the highest Ni and Fe extractions being attained for a concentrate with a sulfur content of 18.2%. Jha, Carlberg, and Meyer (1983) reported similar findings in studying the non-oxidative leaching of MSP in hydrochloric acid, with prolonged hydrogen reduction improving leachability, which was attributed to greater sulfur removal from the MSP.

Dyson and Scott (1976) attributed the increase in concentrate reactivity after a reduction roast treatment to the liberation of sulfur as H₂S during leaching rather than elemental sulfur, which passivates the particle surface during dissolution. However, Filmer and Balestra (1981) have argued that this is not the likely mechanism as leaching is performed below the melting point of elemental sulfur, which would produce a uniform inhibiting layer, and instead, a porous sulfur layer is formed which inhibits the reaction rate only due to a reduction in surface area. Instead, Filmer and Balestra (1981) have suggested that the reactivity is due to the metal-to-sulfur stoichiometric ratio and the mixed potential and have shown that low sulfur concentrates (<20% S) have a mixed potential around −0.3 V (vs SCE) and result in complete Ni dissolution in mineral acid whilst high-sulfur concentrates produce an anodic potential and have low reactivity. Filmer and Balestra (1981) further confirmed the importance of mixed potential on Ni dissolution by performing experiments in which elemental sulfur was added at the start of leaching of an activated concentrate, which resulted in the mixed potential shifting to a more anodic potential (−300 mV to +50 mV vs SCE) and a cessation of Ni and Fe dissolution after 90 min. A similar observation was made by the authors during non-oxidative dissolution of Ni matte in the presence of sulfur addition. Jha, Carlberg, and Meyer (1983) also found that the addition of elemental sulfur during non-oxidative leaching of MSP dramatically reduced Ni extraction and suggested that the presence of oxidizing conditions adversely affects leaching but that this was due to H₂S oxidation to elemental sulfur. Finally, Dyson and Scott (1976) also found that pentlandite dissolution in mineral acid decreased to below 5% after grinding with elemental sulfur or pyrrhotite but attributed this to the formation of a film of elemental sulfur during grinding rather than to changes in mixed potential during leaching. Whatever the reason for reduced dissolution, control of the redox potential appears important in non-oxidative leaching of sulfides, and van Weert, Mah, and Piret (1974) have suggested that the non-oxidative dissolution of pyrrhotite could be extinguished by simply raising the redox potential, which was considered advantageous from a plant safety perspective.

2.6.2.3. Chloridizing roasting

The objective of chloridizing roasting is to convert Ni, Co, and Cu present as their sulfides into water-soluble chlorides, which are subsequently leached from the roasted material. This can be achieved using chlorine gas (Smith and Iwasaki 1976, 1985), ammonium chloride (Li et al. 2020; Xu et al. 2017) or sodium chloride (Aleksandrov et al. 2019a, 2019b; Imideev et al. 2014; Kershner and Hoertel 1961; Mukherjee et al. 1985). Kershner and Hoertel (1961) have suggested that the advantages of using sodium chloride relative to chlorine are a lower reagent cost, less corrosion problems, and that chloridizing roasting does not need to be carried out in a closed system unlike chlorine. Temperatures greater than 250°C are necessary when roasting pentlandite and chalcopyrite with alkali chlorides to convert Ni, Co, and Cu into water-soluble chlorides (Aleksandrov et al. 2019b; Mukherjee et al. 1985; Xu et al. 2017). Sulfur is converted to SO₂ and sulfate salts when roasting with alkali chlorides, whilst iron is converted to ferrous chloride, which undergoes decomposition to iron (III) oxide in the presence of oxygen via formation of ferric chloride as an intermediate (Aleksandrov et al. 2019a; Mukherjee et al. 1985). A minimum temperature of 300°C was reported by Smith and Iwasaki (1976) to be necessary to oxidize ferrous chloride to ferric chloride and iron (III) oxide. Higher temperatures were required (above 400°C) to decompose any iron (III) oxychloride formed though this could lead to ‘fritting’ (material fuses into a porous solid mass), while higher temperatures negatively affected copper recoveries. Decomposition of iron chlorides is advantageous for subsequent water leaching as this minimizes iron dissolution and contamination; however, at the same time, it is wasteful in that alkali chloride is consumed in the process. The use of chlorine gas rather than alkali chlorides is preferable as sulfur in the concentrate can be converted to elemental sulfur rather than SO₂ gas and metal/alkali sulfates (Smith and Iwasaki 1976). A disadvantage to using chlorine is the formation of gaseous sulfur chlorides though this was addressed by Smith and Iwasaki (1976, 1985) by contacting a portion of the concentrate with gas at 300°C the S₂Cl₂ reacting with sulfides in the concentrate to form metal chlorides and elemental sulfur.

Non-ferrous metal recoveries during water leaching of the calcines produced from chloridizing roasting have been reported to be greater than 95% for Ni and Cu, and 88% for Co. Iron extraction during leaching is variable and is largely dependent on the amount of chloridizing agent added and is less influenced by temperature above 350°C higher alkali chloride additions resulting in higher extractions of iron into solution (Aleksandrov et al. 2019b; Xu et al. 2017). There appears to be a trade-off between the amount of chloridizing agent required to obtain acceptable extraction of non-ferrous metals whilst minimizing iron dissolution during water leaching and there have been several attempts to minimize iron dissolution through various process configurations. Kershner and Hoertel (1961) found that iron extractions during subsequent leaching could be reduced by applying a steam treatment at 300°C to the salt-roasted material to hydrolyze iron chlorides; iron dissolution could be reduced from 30% to as low as 10% with steam treatment. Kershner and Hoertel (1961) did evaluate the increase in the temperature at which salt roasting is carried out (400 – 450°C) to minimize iron dissolution and found that iron dissolution decreased to 4%; however Ni, Co, and Cu recoveries slightly decreased but were still above 95%. Smith and Iwasaki (1976, 1985) tried to minimize iron dissolution during leaching by carrying out a selective oxidation of the chlorinated concentrate to oxidize iron chlorides to iron oxides though temperature needed to be limited to below 450°C to prevent decomposition of cupric chloride to insoluble cuprous chloride. Li et al. (2020) was able to reduce iron extraction to below 1% during water leaching whilst non-ferrous metal extractions were above 95% during ammonium chloride roasting of a nickel concentrate grading 9.35% Ni, by carrying out roasting in two stages prior to water leaching. The first stage involved roasting with ammonium chloride at 250°C to convert metal sulfides to chlorides followed by a second roast at higher temperature (650°C) which resulted in conversion of non-ferrous metal chlorides to sulfates and decomposition of iron salts to iron (III) oxide. Chloridizing roasting of Ni-Cu sulfide concentrates was piloted by Mukherjee et al. (1985) where a few tons of concentrate was roasted and water-leached, using NaCl in a slight stoichiometric excess as the chloridizing reagent. Non-ferrous metal recoveries (Ni, Cu) were reported to be 90%, and the roasting process was found to be suitable up to 350°C beyond which the material would undergo ‘fritting’ (Mukherjee et al. 1985). Kershner and Hoertel (1961) also encountered fritting issues in attempting to carry out continuous salt roasting of sulfide concentrates in a rotary kiln.

2.6.2.4. Sulfation roasting

The sulfation roasting process has been researched in the past for the treatment of whole nickel sulfide ores and low-grade concentrates between the 1960s and 1990s (Yu, Utigard, and Barati 2014a). The sulfation roasting of base metal bearing iron sulfide concentrates was in commercial operation at the Falconbridge iron ore plant (Canada) and the Kokkola works (Finland) which were described in Part I of this review. Sulfation roasting is a process whereby sulfide concentrates are roasted under an oxidizing atmosphere with the objective of converting iron sulfides to oxides and non-ferrous metal sulfides to water-soluble sulfate salts. Different strategies have been proposed for achieving this such as (a) roasting under an atmosphere of SO₂/SO₃, (b) roasting to oxidize sulfides followed by sulfation of the calcine through addition of an alkali sulfate, (c) direct addition of an alkali sulfate to the concentrate during roasting and combinations thereof (Fletcher and Shelef 1963). Yu, Utigard, and Barati (2014a) initially investigated the sulfation roasting of a pentlandite concentrate from Sudbury (Canada) under an oxidizing atmosphere (air) in the absence of an alkali sulfate. Formation of Ni and Co sulfates occurred once the iron sulfides in the concentrate had been oxidized; however, temperatures above 650°C resulted in decomposition of the sulfates to their respective oxides. Extractabilities of Ni and Co from roasted calcines via water leaching were poor, the best results obtained being 20% and 40%, respectively, from the calcine after roasting at 650°C.

Addition of alkali metal sulfates (Na, K, and Li) are reported to enhance the recovery of Ni during sulfation roasting with lithium sulfate being the most effective and potassium the least (Fletcher and Shelef 1963; Rao, Natarajan, and Padmanabhan 2001). Fletcher and Shelef (1964) conducted sulfation roasting experiments on a Ni-containing flotation concentrate in a fluidized bed reactor (refer to Table 7). After roasting for 1 hour in the absence of sodium sulfate, Ni extraction during water leaching was only 10.5–12.5%, whilst the addition of Na₂SO₄ increased Ni extraction to 55.5–59.0%. Thornhill (1954) in the original patent for the process in use at the Falconbridge iron ore plant proposed that Na₂SO₄ improved nickel recovery during sulfation roasting by reacting with nickel ferrite releasing NiSO₄, and by reaction with SO₃ to form sodium pyrosulfate (Na₂S₂O₇) which is an effective sulfation agent. Rao, Natarajan, and Padmanabhan (2001) also investigated sulfation roasting of a bulk sulfide flotation concentrate grading 7% Ni and found that the addition of an alkali metal sulfate (6 or 12 wt%) to roasting greatly improved Ni extraction during water leaching. Copper recoveries were not found to be influenced by alkali sulfate addition (92%) whilst cobalt recovery increased from 72% to 88% when an alkali sulfate was added regardless of the type of alkali metal sulfate used. Nickel extraction, however, was reported to be affected by the type of alkali sulfate used with the highest Ni extractions occurring with the use of lithium sulfate (Li₂SO₄); the addition of 12 wt% Li₂SO₄ increased Ni recoveries to above 60% when roasting concentrate at 500°C for 4 hours versus only 18.4% Ni recovery in the absence of any additive under the same conditions.

In a follow-up to their earlier study, Yu, Utigard, and Barati (2014b) explored methods to improve the extractability of Ni and Co during sulfation roasting of concentrate. Roasting in a 95/5 air/SO₂ mixture for 30 min at 720°C and subsequent water leaching resulted in an improvement in extraction to 40% of the Ni and 80% of the Co. Under the optimal roasting conditions (700°C 95/5 air/SO₂ mixture, 10 wt% Na₂SO₄, 150 min residence time), the nickel and cobalt recoveries increased to 79% and 95%, respectively, with only 4% iron dissolution. Nickel recoveries were lower relative to cobalt due to the formation of nickel ferrite (NiFe₂O₄) during roasting, which is not amenable to sulfation and therefore reported to the leach residue. A few authors have evaluated other measures to overcome the limitations of sulfation roasting regarding low extraction of Ni relative to Co and Cu. Fletcher and Hester (1964) in piloting a sulfation roasting process for treatment of a Cu-Ni sulfide concentrate grading 3.63% Cu and 4.69% Ni found that Ni extraction could be improved from 76% to 85% by regrinding the concentrate from 64% minus 74 µm to 90% minus 44 µm; the improvement in Cu recovery due to regrinding was only slight (95% vs 92%). Weir, Kerfoot, and Chalkley (1983) patented a process involving sulfation roasting of Ni and co-bearing pyrite concentrates with Na₂SO₄ as an additive followed by pressure leaching of the calcines and an oxygen overpressure of 50–1500 kPa to dissolve nickel ferrites and unreacted sulfides. Using this approach, Ni/Co recoveries and Fe rejection during calcine leaching were improved whilst Cu recoveries were generally unaffected. In one example, pressure leaching of a calcine produced from sulfation roasting a pyrite concentrate grading 2.2% Ni and 1.6% Co at 150°C and O₂ pressure of 140 kPa for 2 hours in 10 g/L H₂SO₄ resulted in Ni and Co recoveries of 89% and 93%. Leaching the same calcine at ambient pressure and 70°C in a 50 g/L H₂SO₄ solution for 2 hours resulted in Ni and Co recoveries of 68% Ni and 89% Co. In both cases, Cu recoveries were unaffected and were 94–95%.

Li et al. (2019) evaluated the use of ammonium sulfate [(NH₄)₂SO₄] for the sulfation roasting of a nickel concentrate grading 8.93% Ni but which contained Ni as silicate minerals such as willemseite and nepouite in addition to pentlandite. Ammonium sulfate was found to be suitable as a sulfation agent though large amounts were required with 200 wt% being optimal. Sulfation of the Ni-bearing silicate phases was suggested by Li et al. (2019) to be occurring based on their disappearance from the XRD patterns of roasted products; for nepouite this occurred at 450°C whilst in the case of willemsite, this occurred at 650°C. The sulfation of Ni-silicate phases would presumably enable the extraction of the contained Ni during subsequent leaching. Cui et al. (2018) also evaluated the use of (NH₄)₂SO₄ for sulfation roasting of a Cu-Ni sulfide concentrate and found that high additions (3 to 4 times the concentrate mass) were required to achieve adequate extraction of base metals in subsequent water leaching and even then, Ni and Co extractions were still low (approximately 52.5% and 60%, respectively) though Cu recoveries were high (approximately 90%) at an ammonium sulfate to ore mass ratio of 4 and roasting at 400°C for 2 hours. Increasing the roasting temperature to 500°C dramatically improved Co recovery to above 90% but did little to improve Ni recoveries. The addition of 0.4 g sodium sulfate per gram concentrate dramatically increased Ni recovery to approximately 90% with near-complete extraction of Cu and Co. By comparison, other studies utilizing alkali metal sulfates typically only used small additions to obtain an improvement in Ni extraction, of the order of 10 wt% (Li et al. 2018; Rao, Natarajan, and Padmanabhan 2001; Yu, Utigard, and Barati 2014b), showing that ammonium sulfate is not an effective sulfation roasting additive on a mass basis. The two commercial operations that employed sulfation roasting (Falconbridge, Kokkola) utilized sodium sulfate as a sulfation agent.

2.6.3. Ultrasonic assisted leaching

The use of sonication to enhance the leaching of nickel sulfide concentrates has barely received any attention with only one published study to date. Juanqin et al. (2010) studied the effect of ultrasonics on leaching of a Ni-Cu concentrate grading 9.67% Ni and 4.83% Cu, using 0.8 M sodium persulfate (Na₂S₂O₈) as an oxidant with a small quantity (0.001 M) of silver nitrate (AgNO₃). Sonication for 30 min prior to leaching was found by Juanqin et al. (2010) to result in an improvement in leaching rate by 6.79% relative to no pre-treatment. The authors also explored the effects of Na₂S₂O₈ concentration (0.6–1 M) and AgNO₃ concentration (0–0.002 M) on leaching and found that Ni recovery increased with increasing oxidant concentration. It is interesting to note that both Juanqin et al. (2010) and Mulak (1987) found that the addition of Ag+ accelerated Ni dissolution from pentlandite and heazlewoodite, respectively. Ahonen and Tuovinen (1990) obtained mixed results with respect to the effect of Ag+ ion addition during the column leaching of pentlandite-containing copper-zinc ores whilst Nakazawa, Hashizume, and Sato (1993) found Ag+ ions hindered pentlandite dissolution during bacterial leaching of a Cu-Ni sulfide flotation concentrate and Pawlek (1976) found that Ni extraction from a Cu-Ni concentrate worsened during oxygen pressure leaching in the presence of Ag+ ions. The effect of Ag+ ions on Ni dissolution from Ni-sulfide minerals is not clear and requires further attention.

3. Summary and research opportunities

The anticipated demand for class 1 nickel products, particularly nickel sulfate, for the EV industry creates a unique opportunity to revisit the direct hydrometallurgical processing of nickel sulfide concentrates/ores as well as nickel laterites. New nickel mines will likely be required to come online to meet the demand for class 1 nickel in the EV sector given the costs of refining class 2 nickel products such as ferronickel and nickel pig iron is high (Campagnol et al. 2017). Additionally, Campagnol et al. (2017) have stated that there is limited scope to switching over class 1 nickel output (cathodes and briquettes) destined for stainless steel to nickel sulfate production as this will cannibalize existing demand. Hydrometallurgy will play an important role in meeting class 1 nickel demand, and direct hydrometallurgical processing of base metal sulfide concentrates and ores offers several advantages over sulfide smelting-matte refining:

1. Concentrates unsuitable for smelting without blending, such as those high in arsenic and/or MgO can be processed

2. Potentially lower capital costs relative to smelting and matte refining, economically viable at small and large scale with ease of brown field site expansion via incremental plant addition

3. The adaptability of hydrometallurgical processes to treatment lower grade concentrates relative to smelting means that during flotation, less emphasis can be placed on gangue rejection (within reason) thereby improving Ni recoveries at the concentrator

4. Potentially higher overall recoveries of cobalt during direct hydrometallurgical processing relative to concentrate smelting and matte refining, where on average, 30–50% of the cobalt can be lost to the slag during smelting and converting

5. Production of marketable metals or compounds directly from concentrates, thereby eliminating the need for smelting and dedicated matte refining facilities that are capital intensive

6. Sulfide in the concentrate is converted to elemental sulfur or sulfate during hydrometallurgical processing rather than SO₂ (smelting and roasting), thus avoiding the need for emission abatement equipment or coupling sulfuric acid production to metal production

7. Potentially higher degree of plant automation

Against these advantages however, there are several limitations, such as:

1. Recovery of precious metals from flotation concentrates is more challenging as the species present in the original concentrate report to the leach residue (depending on the leaching conditions), and the concentrations are much lower relative to that obtained from matte leaching, being further diluted by un-leached gangue, elemental sulfur (depending on the specific leach conditions), and iron hydrolysis products

2. Leach liquor purification is more complicated relative to matte refining due to the presence of iron in the concentrate, though this can be partially addressed through choice of lixiviant for leaching and rejection of iron sulfides during flotation

3. Issues relating to potential water pollution, effluent disposal, and environmental stability of leach residues, i.e. elemental sulfur (if it is not recovered), jarosite decomposition

4. For acid-based leaching processes, a low-cost neutralizing agent is required to neutralize acid post-leaching. Limestone is the lowest cost option; however, for sulfate-based leaching systems, this produces gypsum and increases the volume of residue requiring disposal

Several research opportunities have been identified from this review that are deemed valuable for the hydrometallurgical processing of nickel sulfide concentrates and whole ores. Firstly, a technoeconomic assessment of the various hydrometallurgical processes currently available for processing nickel sulfide concentrates and their adaptability to meet changing demands for class 1 nickel, i.e. stainless steels vs lithium-ion batteries, is required, given the large number of lixiviants and processes that are potentially available. From the summary of the advantages and disadvantages of the various lixiviant systems that were presented in the first part of this review, sulfate-based leaching systems appear to be most adaptable and additionally, many of the purported benefits of other lixiviant systems (chloride media, nitric acid) can be realized in sulfate systems by adding, for example, chloride salts or nitrates/nitrites. On this basis, an understanding of the leaching characteristics and electrochemistry of economically important nickel sulfide and arsenide minerals and gangue species in different sulfate-based leaching systems is generally lacking in the literature and is required. Secondly, an understanding of the processing characteristics of nickel sulfide concentrates in relation to hydrometallurgical processing in sulfate media is warranted, specifically, how different ore types, i.e. massive, disseminated, high clay, etc., and the effect of ore alteration impact leaching. Finally, an investigation into the mineralogical transformations taking place during the leaching of nickel sulfide concentrates and the interactions taking place between sulfide minerals and non-sulfide gangue during concentrate leaching is lacking. The hydrometallurgical processing of nickel concentrates arising from the flotation of disseminated nickel sulfide ores and arsenic-rich ores represents another research opportunity given that concentrates derived from these ores tend to be unsuitable for smelting without blending to reduce the magnesium and/or arsenic concentration. The co-processing of nickel laterite and nickel sulfide resources represents yet another potential opportunity in Western Australia given the proximity of nickel laterite and sulfide resources to each other and research into the effects of laterite mineralogy (limonitic vs clay silicate) and sulfide source (concentrate, whole ore, and high MgO/As) and their interactions warrant attention. Better knowledge of leaching characteristics of nickel sulfide minerals and gangue and processing characteristics of sulfide concentrates can improve recoveries and kinetics during leaching, particularly in the development of lower cost leaching options possibly carried out at atmospheric pressure or without the need for ultra-fine grinding as an example. This is important because, as pointed out by Crundwell et al. (2011), the lower grade of nickel concentrates relative to mattes (5–20% Ni vs up to 70%) puts pressure on direct concentrate leaching processes to be more efficient relative to matte leaching operations to minimize equipment sizing and match an equivalent Ni production rate.

Finally, consideration of upstream impacts of hydrometallurgical processing on nickel sulfide flotation practice should be evaluated. As a result of the potentially less stringent requirements on concentrate quality for direct hydrometallurgical processing, this could result in potentially lower capital and operating cost for concentrators and increased base metal recoveries at the concentrator (Peters 1976). Two issues pose challenges for nickel concentrates destined for smelting, the magnesium and pyrrhotite contents in the concentrates to be smelted. The iron-to-magnesium ratio in the nickel concentrates destined for smelting is a key parameter that influences smelting temperature and slag viscosity (Mayhew et al. 2013; Senior and Thomas 2005), hence emphasis is placed on the rejection of Mg-bearing gangue during flotation of nickel sulfide ores (Lotter et al. 2008; Pietrobon et al. 1997) though this may be accompanied by the significant loss of Ni values whilst producing a Ni flotation concentrate of acceptable quality for smelting. Pyrrhotite is an issue for smelting as it increases the amount of SO₂ and slag produced during smelting and yields little Ni (Bruce and Orr 1986; Senior, Shannon, and Trahar 1994); however, rejection of pyrrhotite at the concentrator results in Ni losses due to the Ni content of pyrrhotite as well as fine pentlandite associated with pyrrhotite that is too fine to liberate (Bulatovic 2007; Kerr 2002; Senior, Shannon, and Trahar 1994). Direct hydrometallurgical processing could potentially address these issues as requirements on concentrate quality for leaching are expected to be less stringent and therefore more emphasis can be placed on Ni recovery during flotation (Jones et al. 2010; McDonald et al. 2012; Senior and Thomas 2005). Norton, Coetzee, and Barnett (1998) however have suggested that the sulfur-to-nickel ratio, rather than the iron-to-magnesium ratio, is of greater importance for bioleaching processes with emphasis being placed on pyrrhotite rejection during nickel flotation. The benefits of greater pyrrhotite rejection on nickel concentrate bioleaching were stated to be lower cooling and aeration requirements during leaching, and lower acid consumption (Norton, Coetzee, and Barnett 1998). Conversely however, in the context of co-processing of nickel laterites with sulfide feeds, pyrrhotite (and pyrite) in the concentrate would be desirable as this is a source of sulfuric acid and heat for leaching the laterites (McDonald and Li 2020) and other nickel oxide materials such as metallurgical slags (Perederiy et al. 2011). What this could theoretically entail for a nickel sulfide concentrator in a Western Australian setting is that a high-grade sulfide concentrate is produced for smelting, whilst a separate lower grade concentrate rich in pyrrhotite could be processed by hydrometallurgy, possibly via co-processing with nickel laterite ores. A similar concept is already being carried out at the Nadezhdin mill (Russia) where the pyrrhotite concentrate recovered from flotation operations is pressure leached and the sulfide concentrate recovered from the leaching operation is combined with high-grade Ni concentrate for smelting. Additionally, the Falconbridge and Inco mills in Canada processed nickeliferous pyrrhotite concentrates between 1955 and 1982 via roast-leach-type processes. These scenarios therefore highlight the need for upstream considerations in the hydrometallurgical processing of nickel sulfide concentrates.

Acknowledgements

The authors are grateful for the support provided by CSIRO Mineral Resources to Nebeal Faris through the ResearchPlus CERC Postdoctoral Fellowship program. The authors would like to thank Dr Robbie McDonald and Mr David McCallum of CSIRO Mineral Resources for taking time to review and provide valuable feedback during the preparation of this manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the CSIRO [Research Plus CERC Postdoctoral Fellowship program].