A PROCESS FOR NICKEL AND COBALT RECOVERY FROM AN ELUATE BY USING SOLVENT EXTRACTION, AND NICKEL- OR COBALT-BEARING PRODUCT OBTAINED BY SAID PROCESS

The present invention relates to a process for nickel or cobalt recovery from nickel- or cobalt-bearing ores or solutions, preferably from nickel laterite ores, with application in the hydrometallurgical industry.

Hydrometallurgy is becoming increasingly important in the exploitation of nickel deposits. Abundant deposits of high-grade sulfide ore, simple to exploit are being exhausted, and environmental legislation is becoming ever more rigorous, restricting the implementation of once widely accepted pyrometallurgical processes. These two factors combined are increasing the importance of laterite deposits - including those with low nickel content - as the world's sources of nickel and cobalt, thus stimulating developments in hydrometallurgy. Leaching is a hydrometallurgical process, which uses an aqueous, usually acidic, solution to dissolve and separate target metals from the rest of the ore, for later recovery. Several leaching processes have been developed over the years, of which the most commonly used for nickel laterites is pressure acid leaching. It has been used for processing laterites with low magnesium content in Moa Bay since 1959. With a processing capacity of approximately 2 Mt of ore per annum, the plant was built by Freeport Sulphur, owner of the Nicaro plant (Caron process). Moa Bay was seized by the Cuban government during the communist revolution in 1960 and resumed operations in the following year. The pressure acid leach process has many advantages. It is a known process, which can be applied to low-grade laterites and, unlike smelting, and the Caron process, provides high recoveries of nickel and cobalt. The metals can be produced individually, thus adding more value, and there are no roasting

or drying steps, which are very energy intensive. Drawbacks include the fact that pressure leaching is only economically feasible when applied to limonites with low contents of acid-consuming gangue. In addition, high investment in autoclaves and flash tanks is required. The aforementioned processes for extracting nickel from laterites, be they hydrometallurgical or not, require high investment and/or have high energy costs. Therefore, a new, simpler and cheaper method has been developed: atmospheric leaching, which includes leaching in agitated tanks and heap leaching. Heap leaching involves placing run-of-mine ore in piles of different heights. The leaching solution is then distributed over the top of the pile, and the solution percolates through the pile down to the base, where the resulting liquor is recovered. The main advantage of heap leaching is its low cost, while its main drawback is the difficulty of treating ores with high clay content. This clay absorbs the leaching solution, swells, and blocks the pores in the pile, hence reducing permeability and increasing the consumption of acid. Atmospheric leaching processes involve the use of chelating agents to facilitate the dissolution of target metals. In this case, implementation costs are lower because autoclaves are not required. The solid-liquid separation stage poses big problems for this circuit. Due to poor sedimentation characteristics of the pulp in the counter- current decantation (CCD) stage, approximately 10 percent soluble nickel and cobalt are lost in the tailings. In order to minimize this significant loss, series of at least six large thickeners (over 50 meters in diameter) for each step are used in solid-liquid separation, to achieve proper settling of solids and obtain a clear overflow. Therefore, the capital cost of a single CCD stage with conventional thickeners can reach 30 percent of that of a titanium autoclave used for high- pressure acid leaching (HPAL). If chlorides are present, as in the case of some

laterite deposits in arid regions, even more expensive thickeners are required. In addition to capital costs, operational costs are also high because, they include not only the energy consumption of each rake, but also the flocculants required for the settling of fine-grained materials. Flocculant consumption varies from 200 to 800 grams per ton of solids, which increases by up to 10 percent the overall operational cost of the plant.

Another difficulty to be overcome in the recovery of target metals from the resulting solution, called pregnant leach solution (PLS), is posed by impurities present and low metal concentration. Ion exchange is an efficient method of overcoming these barriers. It is effective at low concentrations, and selectivity can be greatly improved by the correct choice of an ion-exchange resin. Resin-in-pulp technology, in which the resin is applied directly to the leach effluent, has become an interesting option because, owing to the high metal- recovery properties and high selectivity of ion-exchange resins, it replaces the costly solid-liquid separation stage following leaching and prior to metal recovery.

Ion-exchange technology was first used in hydrometallurgy on a large scale for uranium extraction in the Soviet Union in the 1950s, and innovations in the application of ion exchange were developed as the uranium market increased. One of these early innovations with commercial application was the resin-in-pulp (RIP) process for uranium-bearing ores, which has the advantage of eliminating the solid-liquid separation stage. Also applicable to gold recovery, the resin-in-pulp process was developed from the carbon-in-pulp (CIP) process. Replacing activated carbon with ion-exchange resins has many advantages, because resins offer higher loading capacity and loading rate, are more abrasion-resistant, and are less likely to be contaminated by organic matter. The first commercial plant to use RIP for gold recovery was the Golden Jubilee Mine in South Africa. The industrial operation at this mine served as a basis for analyzing the advantages of RIP over CIP.

The mineral industry has shown deep interest in the utilization of solvent extraction for concentration, separation and purification of metals.

Application of this technology faces great challenges, because there is no extractant for nickel only, without co-extraction of other metals, and the development and commercialization of a new extractant to meet this need is very unlikely to occur. Therefore, application of the solvent extraction technique directly to an acid-leach effluent containing Ni, Co, Cu, Zn, Ca, and Mg is very restricted, with little chances of a selective recovery free of impurities. An interesting alternative is to apply solvent extraction directly to the eluate (i.e. the product from the ion-exchange stage with polymeric resins). This product is considered to contain high concentrations of nickel and cobalt and low concentration of impurities, but the metals still have to be separated. The expectation is that solvent extraction may be applied successfully to this type of product, as in the case of an eluate from resin-in-pulp operations for nickel recovery, in view of high nickel concentration and few impurities.

Industrial experiments on the application of solvent extraction technology directly to the leach effluent have not been successful, particularly because of the high levels of impurities and the non-existence of an extractant fully selective for nickel. This obstacle can be overcome by inserting, between the ore-leaching and solvent-extraction stages, a purification stage such as the ion- exchange option, with resins applied to the leach pulp, which has the advantage of eliminating the expensive solid-liquid separation stage.

In the conventional nickel-ore treatment process (Bulong, Murrin Murrin, Cawse); the solvent extraction operational units are fed with solutions, which contain impurities - mainly zinc and copper - to some extent.

Murrin Murrin

The process carried out in sulfuric medium uses Cyanex 272 as selective extractant for cobalt. This is a phosphinic acid-based organic solvent,

which extracts cobalt selectively, plus such impurities as manganese, magnesium, copper and zinc. The raffinate contains nickel.

Bulong

Bulong uses pressure leaching, and the effluent is conveyed to the iron removal stage, with lime addition, before entering the solvent extraction circuit. Bulong tried to avoid nickel precipitation and releaching processes by applying solvent extraction directly. One of the major disadvantages was the passing of the leach solution through two solvent extraction circuits, as well as gypsum formation. The process in sulfuric medium uses Cyanex 272 initially, as selective extractant for cobalt. This is a phosphinic acid-based organic solvent, which extracts cobalt selectively, plus such impurities as manganese, magnesium, copper and zinc. The raffinate contains nickel. Bulong also uses a second solvent extraction stage, with versatic acid selective for nickel, to minimize the concentration of impurities and increase nickel concentration. This stage was found to be the most critical for the commissioning of the plant, because of excessive gypsum precipitation and scaling.

Some critical points are worth mentioning:

•The effluent solution from the previous neutralization stage becomes saturated in calcium after limestone addition. This calcium is co- extracted with nickel in the versatic-acid circuit, particularly in the last steps. Gypsum precipitation is reduced by decreasing the organic feed rate and strict pH control. Anyway, gypsum build-up was detected in nickel solvent extraction.

• It was also found that the versatic-acid circuit was contaminated with Cyanex, which also carries calcium at higher pH levels into the circuit.

• In the first Cyanex circuit, removal of cobalt and copper is not quantitative, and this combined with low selectivity of versatic acid in the second

step may render production of LME specification nickel difficult.

• Metal concentrations in the feed to solvent extraction are low hence, long circuits are required.

Cawse Cawse produces a mixed Ni-Co precipitate, which is releached with ammonia. The resulting ammoniacal solution is then subjected to solvent extraction with LIX 84-1 , which is an oxime that selectively extracts nickel. The great majority of impurities remain in the raffinate, and the nickel-rich solution is subjected to electrowinning. The process also includes separate steps for copper reextraction (stripping), cobalt reextraction, and regeneration of the organic phase.

Some characteristics specific to this operation are worth noting:

• Mixed hydroxide precipitation (MHP) of nickel and cobalt followed by releaching with ammonia provide high selectivity for the target metals.

Only the base metals are directed to the solvent extraction circuit.

•The washing of both loaded and unloaded organic phases aids selectivity.

•Zinc, the only metal which does not meet LME specification, can be removed from the system by ion-exchange resins.

• Some ammonia transfer to the electrowinning system may occur.

•The circuit is robust, and the cathodes have good physical appearance. Goro

Goro uses pressure sulfuric acid leaching followed by removal of such impurities as Fe, Al, Cr and Cu, with lime addition, and ion exchange resins for copper recovery. The product from this purification is fed to two solvent

extraction steps, the first of which with Cyanex 301 for recovery of nickel and cobalt, and the second with Alamina 308 for nickel separation from cobalt.

Some typical characteristics of this approach are worth noting:

• Solvent extraction is applied directly to the purified sulfuric- leach effluent.

• Cyanex 301 can be contaminated with copper, requiring an ion-exchange stage for removal of the metal.

• The Cyanex 301 circuit concentrates both Ni and Co, reducing the size of downstream operations. • There is the potential of cross-contamination of the organic.

Solvent extraction is one of the most commonly used techniques for purification of nickel-bearing solutions, as well as for nickel separation from cobalt. However, when applied to nickel this technique still has some drawbacks, such as the following: • There is no ideal, commercially available extractant for nickel only as there is for copper.

• The investment cost of solvent extraction technology in nickel hydrometallurgical plants is often 3-6 times higher than that in copper plants.

• Nickel solvent extraction circuits need to be adapted to the rather limited performance of available reagents.

•Available extractants do not always match the chemistry of leach systems.

•Available extractants do not always produce a suitable electrolyte for the metal production route. • Solvent extraction circuits for nickel are more complex than those for copper and, as in the case of uranium circuits, require many stages including washing stages.

•The solutions often contain economic amounts of cobalt and

substantial amounts of copper and zinc.

The concentration of impurities in the leach solution increases with the type of ore feed leached, in the following order: LATERITES > SULFIDE CONCENTRATES > MATTE > HYDROXIDE PRECIPITATE > SULFIDE PRECIPITATE. Impurity removal and recovery of nickel and cobalt from this leach effluent generally requires a stepwise approach, in which hydrometallurgical operations - precipitation, selective leaching, and a multistage solvent extraction circuit - are used. The use of more than one type of extractant in an integrated hydrometallurgical circuit generally results in contamination of one extractant with another. Olympic Dam (oximes and amines) and Bulong (versatic acid and Cyanex 272) have had cross-contamination problems. What is observed is that the circuits are highly complex and of limited efficiencies, which creates several operational problems in the plants.

These problems arise because the industrially proven methods of purification of nickel-ore leach solutions - such as precipitation, selective releaching, and solvent extraction - still have operational limitations. Hence, the need to evaluate alternative methods of purification, such as resin-in-pulp ion exchange followed by a single solvent-extraction step, and to technically optimize their use so as to obtain not only economic advantages but also higher efficiencies in the overall recovery of nickel.

Since the reagents are not selective for nickel only, addition of an intermediate purification step is likely to overcome the deficiencies of commercially available extractants. However, the techniques developed to date, such as precipitation - followed, or not, by releaching - and solvent extraction, still have several operational limitations. Therefore, metallurgists should continue the search for flowsheets, which allow the use of existing conventional extractants, along with incorporating a number of unit operations that may minimize problems, as well as optimizing pre-existing unit operations in

conventional flowsheets.

The following characteristics of current technology are worth mentioning:

• Development of new extractants is unlikely. *The market for nickel extractants is small, and development costs are high.

• The ideal extractant would be one that extracts at pH=1 and reextracts (or strips) at pH=4.

• If a new successful extractant is synthesized in laboratory, the costs of providing commercial production, pilot testing, as well as other development costs, may exceed US$ 5 million.

• For a market of 300 tpa, the reagent may be too expensive to be attractive.

• Phosphinic acid-based extractant Cyanex 272 renders the design and development of cobalt solvent extraction circuits simpler than those of nickel circuits.

The following are the main commercially available and commonly used solvents for cobalt and nickel:

^■ D2EHPA This is a non-selective nickel and cobalt extractant in sulfuric medium. It operates over a narrow 3.5-4.5 pH range, and its molecular formula is (C ₄ H ₉ CH(C ₂ H ₅ )CH ₂ O) ₂ POOH.

The solvent operates on a hydrogen cycle, as shown by the equation: 4RH + Ni ²⁺ → R ₂ Ni(RH) ₂ + 2H ⁺

Metal extraction by D2EHPA in sulfuric medium is shown in Table 1 attached hereto. This is a robust, low-cost extractant, but its selectivity is poor, as shown in Table 1. Careful pH control, as well as washing with solutions

of the target metal for removal of impurities, may be required.

^■ Versatic acid

This is another non-selective nickel and cobalt extractant in sulfuric medium, which operates over a higher, but also narrow, pH range from 5.5 to 6.5.

Its molecular formula is R ₃ COOH, and the extracted species are believed to be NiV2.2HV and CoV ₂ .2HV, where V represents a versatic acid cation.

Addition of pyridine carboxylate esters can increase the pH50(Ca-Ni) value by approximately 1 to 1.8 units while the addition of nonyl pyridine can reduce Ca extraction to very low levels. However, both these pyridine reagents are not commercially available. Table 2 shows nickel and calcium extractions by versatic acid in sulfuric medium, and how these extractions vary with the pH. ■ Phosphinic acids

The two main phosphinic acids used in solvent extraction are Cyanex 272 and Cyanex 301.

Cyanex 272 is a selective extractant for cobalt and nickel. It operates over a pH range of 3 to 6, in either sulfuric or hydrochloric medium. It also extracts manganese, magnesium, zinc and copper.

Its molecular form is shown in Formula 1 below. It operates selectively for nickel and cobalt at pH 3-5, and also extracts magnesium, copper, zinc and manganese at pH below 5; calcium at pH 5.5; and iron at pH 1-2. Iron stripping can be accomplished with a solution containing 100 g/L sulfuric acid.

R OH

FORMULA 1 - Molecular structure of Cyanex 272 extractant.

Cyanex 272 is a robust reagent, but cobalt can oxidize the diluent. This can be prevented by excluding air and using a sacrificial antioxidant.

Cyanex 301 , the sulfur-containing analogue of Cyanex 272, is illustrated in Formula 2 below. It selectively extracts nickel, cobalt and copper from leach solutions, being a strong extractant for these, as well as for zinc, in sulphate solutions. Nickel and cobalt stripping is accomplished with a 6N HCI solution. Iron and copper stripping, however, does not occur quickly and can lead to formation of bisulfides and loss of loading capacity. If the stripping stage is reductive, the reagent can be regenerated.

FORMULA 2 - Molecular structure of Cyanex 301 extractant. ■ LIX 84-1 ketoxime

This is a chelating extractant, selective for nickel, copper and cobalt (II) from sulfate solutions. It operates in the narrow 4-5 pH range and has slow kinetics for nickel and cobalt. Its molecular structure is illustrated below.

Where:

"R" represents either C ₉ H ₁₉ or C ₁₂ H ₂₅ .

"A" in the case of salyciladoximes represents one hydrogen (H), and in the case of oximes represents either C ₆ H ₅ or CH ₃ .

FORMULA 3 - Molecular structure of LIX 84 extractant.

It extracts nickel at pH above 4, with stripping being accomplished with a solution containing either 25 g/L sulfuric acid or more than 250 g/L ammonia, as shown by the equations below: R ₂ Ni + 2H+ + SO ₄ ^2" → 2RH + Nl ²⁺ + SO ₄ ^2"

R ₂ Ni + 2NH ₃ + 2NH ⁴⁺ → Ni(NH ₃ ) ₄ ^2" + 2RH

R ₂ Ni + 4NH ₃ + 2NH ⁴⁺ → Ni(NH ₃ ) ⁶²⁺ + 2RH

A modified version of this extractant, named LIX 87QN, is more used in ammonia strip circuits. On average, extraction time is 3-4 minutes, while strip time is

6-10 minutes.

This extractant has a number of operational drawbacks. Cobalt

(II) can be easily oxidized to cobalt (III) in ammoniacal solutions, but only cobalt

(II) can be extracted. Furthermore, extracted cobalt (II) can still oxidize to cobalt (III), which renders the already slow stripping stage with sulfuric acid much more difficult.

Despite this, cobalt (III) can be stripped with a reductive stripping solution using iron or zinc and acid. In any case, cobalt extraction and stripping cause reagent degradation, which reduces nickel-over-zinc selectivity, increases the number of organic phase separation stages and reduces nickel transfer capacity.

•Alamine 308 (tri-iso-octylamine)

This tertiary amine is used as an ion-exchange extractant, for

extracting CoCL/ ^' from hydrochloric-acid leach solutions. Nickel does not form chloro complexes and is not extracted, whereas copper and zinc are. Zinc is extracted at lower chlorine concentrations than cobalt.

Since it contains one basic hydrogen atom, it readily reacts with a variety of organic and inorganic acids, forming amine salts, which are capable of undergoing ion-exchange reactions, as shown by the equations below.

■ Protonation

[R ₃ N] _O rg + [HA] _aq «→ [R ₃ NH ⁺ A-] _O rg

^■ Ion exchange [R ₃ NH ⁺ A-] _org + [B ^" ] _aq → [R ₃ NH ⁺ BIr ₉ + [Alaq

The extent to which B will exchange for A is a function of the relative affinity of both with the organic cation and of their respective solvation energies.

The regeneration stage of the target species can be accomplished with a wide variety of solutions of such inorganic salts as NaCI,

Na2CO ₃ and (NhU) ₂ SO ₄ . The choice of the reagent for the stripping stage depends on the overall recovery of the process, but in general, basic agents which reverse the protonation reaction provide better results in a smaller number of steps. The equation below gives the recovery by Na ₂ COa. 2[R ₃ NH ⁺ B ] _O rg +[2Na ⁺ +CO ₃ ² ] _aq 2[R ₃ N] _org + H ₂ O + CO ₂ +

2Na _aq ⁺ + 2B ^" _aq

In some cases, the formation of anionic complexes and their subsequent extraction depend on the concentration of this anion, as in the case of cobalt extracted as a chloro complex. In view of the limitations of currently available techniques which have now become conventional, a flowsheet is proposed, in which the introduction and rearrangement of unit operations offer several advantages.

Flowsheet 1 attached hereto illustrates the unit operations involved according to

the invention. The advantages are as follows:

• Minimization of the operational problems commonly found in commercial plants.

• Increased process efficiency and higher overall nickel and cobalt recoveries.

• Economic gains resulting from reduced capital and operational costs.

For nickel laterite ores, it is being proposed that the beneficiated ore be subjected to leaching, which can be either acidic or basic, such as sulfuric acid, hydrochloric acid, or ammonia leaching. This operation can be performed in pressure vessels; in agitated tanks under atmospheric conditions; in leaching heaps; or a combination of two or more types of leaching.

The leach effluent, as obtained in pulp form without any type of solid-liquid separation, is conveyed to the ion-exchange stage with polymeric resins, which are applied directly to the pulp. Basically, this process, known as resin in pulp, is an alternative of purification of the solution for effective, selective recovery of nickel over all the impurities present in the effluent. This process makes it possible to both eliminate the onerous solid-liquid separation stage and achieve highly efficient purification, and hence substantial gains in terms of process economics are expected. At the same time, it is being proposed a simplification of the subsequent stages in the process flowsheet.

Since the product, known as eluate, from the ion exchange stage is highly pure, what is expected is not only a simplification of the subsequent stages, but also a facilitation of the operation of these units, thus minimizing the problems commonly detected in commercial plants. Therefore, a solvent extraction stage with Cyanex 272 reagent is being proposed for separating nickel from cobalt. As the eluate is a high-purity product with low concentrations of impurities, the solvent extraction stage for purification of the

solution may, or may not, be eliminated, depending on each specific ore type.

Henceforth, this simplified flowsheet prevents the occurrence of operational problems commonly found when dealing with highly impure solutions. Once separated, nickel and cobalt are individually recovered.

The nickel is sent to the electrowinning unit where it is recovered in metallic form, and cobalt may be recovered in various ways, preferably as an intermediate product, such as cobalt sulfide, hydroxide, carbonate, etc. In the event the nickel- bearing raffinate has a substantial concentration of impurities, a second solvent extraction stage using versatic acid becomes fundamental.

A number of positive results achieved with the invention are described below, as well as a number of advantages:

• Increased process efficiency and high nickel and cobalt recoveries; • Simplification of the process with the reduction of the number of unit operations;

• Minimization of operational problems commonly detected in commercial plants of this type;

• Lower environmental impact - smaller water consumption and possibility of water recycle;

• Lower operational and capital costs;

• Higher quality of the process end products - high selectivity for target metals, high separation capacity, and high metal recovery efficiency;

• Lower technical and operational risk, owing to optimization of unit operations with respect to operational conditions, reagents, and purity of the feed solution;

With these objectives in mind, a process was developed for recovery of nickel or cobalt from ores or solutions containing them, comprising

the following stages:

(a) beneficiation or preparation of ores containing nickel and/or cobalt and other base metals;

(b) ore leaching: can be performed in various ways and combinations of these;

(c) resin in pulp: the leach product is subjected to treatment with resin in a counter-current flow, and the solid-liquid separation stage is eliminated;

(d) solvent extraction: the eluate obtained after elution of the resin is subjected to one or more solvent extraction steps for nickel separation from cobalt, and if necessary, one or more steps for purification of the nickel- bearing raffinate;

(e) nickel electrowinning: either the raffinate from stage (d) or the nickel-bearing solution from the previous stage is fed to the electrowinning stage for obtaining the end product in the form of cathode nickel;

(f) the solvent reextraction solution, which contains cobalt together with some impurities, shall be fed to a precipitation unit for cobalt recovery.

Detailed description: Ore: the nickel ore which is being proposed is of the laterite type, which can be treated conventionally by hydrometallurgical techniques.

Beneficiation/preparation: this stage includes such operations as crushing, scrubbing, attrition, magnetic separation, gravity separation, and ore size classification. In this stage, nickel upgrading is simply accomplished by separating fine-grained ore particles from coarser material. These fine particles, while constituting a small fraction of the overall ore mass, contain a large portion of the nickel content of the ore.

Charts 3, 4, 5 and 6 illustrate the nickel upgrade that can be

reached by reducing the mass recovery and concentrating the nickel recovery on this finer-grained fraction, usually at 100# mesh. Table 3 shows the nickel upgrading factor as a function of the silicon content at <200# mesh. Table 4 shows mass recovery as a function of silicon content at <200# mesh. Table 5 shows metallurgical nickel recovery of as a function of silicon content at <200# mesh. In addition, Table 6 shows the nickel content of the fraction at <200# mesh.

Leaching: ore leaching may be performed in various ways, such as, for example, in pressure vessels, or in agitated tanks under atmospheric conditions, or in the form of leach piles or heaps. A combination of two or more types of leaching is also possible. The most commonly used reagents are either acidic, such as hydrochloric acid or sulfuric acid, or basic, such as ammonium salts.

Resin in pulp: the resin-in-pulp operation is performed in a series of several compressed air-agitated pachuca tanks, and counter-current contacting devices. Resin separation from the pulp is performed by screening. This type of ion-exchange process directly applied to the pulp not only eliminates the onerous solid-liquid separation, but also increases the efficiency of the recovery of the nickel contained in the leach effluent as a result of the sorption- leaching phenomenon, which enables nickel recoveries from both the solid and liquid phases. Elution of the resin can be performed in one or more steps, with such acidic reagents as sulfuric acid, hydrochloric acid, or ammonia, preferably with sulfuric acid, with concentrations that may vary from 2% to about 15% in weight. The eluate obtained from this stage may have nickel concentrations of more than 45 g/L, with low concentrations of impurities. This product, as obtained, with a high purity level, shall be directed to a single solvent extraction step with the purpose of separating Ni from Co.

Table 7 attached hereto shows an example of recovery of

nickel from the liquid phase along 10 resin-in-pulp contact reactors.

Table 8 attached hereto shows an example of elution of the loaded resin using sulfuric acid 10%.

Solvent extraction: the only objective of this stage of the flowsheet is Ni separation from Co. The feed to this stage is already a highly pure solution and therefore a second solvent extraction step for removal of impurities - as found in Bulong and Goro flowsheets - is unnecessary. This solution, which is basically composed of metal sulfates, mostly nickel sulfate in concentrations of more than 45 g/L, is contacted with a phosphinic acid-based extractant such as Cyanex 272, which is selective for cobalt along with a number of impurities, thereby providing Ni separation from Co.

Nickel electrowinning: the solvent extraction raffinate (i.e. the nickel-bearing effluent solution from the previous stage) is the feed to the electrowinning stage for obtaining the end product in the form of cathode nickel. Cobalt recovery: the solvent reextraction solution, containing cobalt and some impurities, is the feed to a precipitation unit where cobalt is preferably recovered as an intermediate product in the form of sulfide, hydroxide, oxide, or carbonate.

The scope of the present invention can be better understood from the examples presented below. It bears pointing out that said examples are illustrative only, and shall not be construed as limiting the scope of the invention.

Examples of high pressure sulfuric acid leaching and atmospheric leaching, and the combination of both:

EXAMPLE 1 : The particle size fraction between 32# and 200# (-32# +200#) of the nickel laterite sample is composed of 0.81% Ni, 0.15% Co, 20.16% Fe, 2.14% Mg and 3.31 Al. This fraction, together with a 96% sulfuric acid solution, was fed to the atmospheric leaching (AL) stage at 950 C temperature, with 385

rpm agitation and 33% solids for 6 hours.

An extraction of 90.4% Ni and 48.0% Fe occurred in AL, originating an effluent with concentrations of 4.5 g/L Ni, 51.3 g/L Fe, 5.4 g/L Al and 96.3 g/L residual free acidity, as well as a residue containing 0.10% Ni, 13.50% and 3.06% Al.

A portion of this liquor, together with a 96% sulfuric acid solution and another fraction (passing 200# mesh) of the sample (composed of 1.69% Ni, 0.10% Co, 19.70% Fe, 2.97% Mg and 3.06% Al), formed the feed to an autoclave, where High Pressure Acid Leaching (HPAL) was carried out at 2500 C temperature and 650 psi pressure, with 500 rpm agitation and 30% solids for 75 minutes.

On being fed to the HPAL stage, the liquor suffered dilution, and its concentrations became 0.03 g/L Ni, 0.31 g/L Fe and 0.03 g/L Al. After HPAL, the concentrations in the autoclave effluent were 12.6 g/L Ni, 14.5 g/L Fe and 4.6 g/L Al, while the residue contained 0.13% Ni, 23.40% Fe and 2.61% Al.

Precipitated Fe and Al regenerated about 116.1 kg/t of acid, which corresponds to 30% of gross consumption, and 65.5 kg/t of acid (corresponding to 17% of overall consumption) from the residual free acidity present in the AL effluent were reused. Therefore, only 53% (205.4 kg/t) of acid had to be added. In AL followed by HPAL Ni extraction was 93.7% and it bears highlighting that in HPAL the extraction was 95.9% Ni for a gross consumption of 387.1 kg/t of acid.

TABLE 1 - Sam le com osition and AL roduct data

TABLE 3 - Comparison between results from HPAL and from AL followed b HPAL

EXAMPLE 2:

AL feed was formed by a fraction (-32# +200#) containing 0.36% Ni, 0.09% Co, 22.92% Fe and 0.57% Mg, and a 96% sulfuric acid solution. The operational variables used were 950C temperature, 385 rpm agitation, and 33% solids for 6 hours. AL produced a liquor with concentrations of 1.7 g/L Ni, 66.0 g/L Fe, and 90.4 g/L residual free acidity. The residue contained 0.06% Ni and 16.80% Fe, and extraction values were 86.4% Ni and 40.4% Fe. The subsequent stage (HPAL) was fed with a portion of this liquor, a 96% sulfuric acid solution, and an amount of ore corresponding to the fraction passing 200# mesh (0.68% Ni, 0.11 % Co, 23.70% Fe and 0.58% Mg). Addition of the solution to the liquor caused dilution, after which the feed concentrations became 0.02 g/L Ni and 0.58 g/L Fe. HPAL was performed at 2500C and 650 psi, with 500 rpm agitation and 30% solids, for 75 minutes. The resulting liquor contained 4.5 g/L Ni and 8.0 g/L Fe, and the residue contained 0.05% Ni and 26.50% Fe. Sulfuric acid regeneration of 88.8 kg/t, which occurred in the autoclave, corresponded to 30% of gross consumption. For this reason, the amount of acid added into the autoclave was 51 % less, that is, 144.0 kg/t were fed and 63.3 kg/t from AL (corresponding to 21% of gross consumption) were reused. Ni extraction values were 94.5% for a single HPAL stage, and 93.1% for AL followed by HPAL. TABLE 4 - Sample composition and AL product data

Atmospheric Leaching - AL

Composition (%)

Ni Co Fe Mg (-32# +200#)

TABLE 6 - Comparison between results from HPAL and from AL followed by HPAL

HPAL AL followed by HPAL

EXAMPLE 3:

The composition of the fraction (-32# +200#) used as feed to AL was 0.56% Ni, 0.03% Co, 27.65% Fe and 1.65% Mg and 5.10% Al. This fraction was added to a 96% sulfuric acid solution and leached at 950C, with 385 rpm agitation and 33% solids, for 6 hours. The AL effluent presented extraction values of 91.7% Ni and 56.6% Fe, with concentrations of 2.6 g/L Ni, 74.7 g/L Fe, 14.2 g/L Al and 65.9 g/L free acidity. The residue contained 0.07% Ni, 18.10% Fe and 3.39% Al. A portion of this liquor, together with a 96% sulfuric acid solution and the fine-grained fraction (-200#) of this sample, was used to feed the autoclave for HPAL. The fine-grained fraction composition was 1.13% Ni, 0.03% Co, 24.90% Fe, 3.10% Mg and 6.77% Al. After dilution caused by the sulfuric acid solution, the concentration values for the liquor that was fed to HPAL were 0.01 g/L Ni, 0.30 g/L Fe and 0.06 g/L Al. The operational parameters used for HPAL were as follows: 2500C temperature; 500 rpm agitation; 650 psi pressure; and 30% solids, for 75 minutes. HPAL produced a liquor containing 7.3 g/L Ni, 12.1 g/L Fe and 13.0 g/L Al, and a residue containing 0.06% Ni, 27.80% Fe and 5.55% Al. During HPAL, there was sulfuric acid regeneration of 30% (144.5 kg/t), and 43.1 kg/t (9%) from free acidity were reused, so that only 61% new acid, corresponding to 294.3 kg/t, had to be added. This liquor from HPAL presented an extraction value of 94.8% Ni, while for HPAL not preceded by AL the extraction value was 92.6%. It was also noted that 54.6% Fe and 35.3% Al was precipitated.

TABLE 9 - Comparison between results from HPAL and from AL followed by HPAL