METALS BY SELECTIVE DISSOLUTION

This disclosure concerns the recovery of metals from recycled products such as spent GTL catalysts. It concerns in particular a hydrometallurgical process that usefully can be carried out on the cobalt-based, PGM-bearing alloys that are typically obtained after smelting said catalysts.

Because of decreasing crude oil stocks, more and more diesel oil is produced using gas-to- liquid technology (GTL). Natural gas is hereby converted to syngas and then to high- quality liquid hydrocarbons that are virtually free of sulfur and aromatics.

The heart of the GTL process is the Fischer-Tropsch reaction, in which the syngas is converted into diesel. This reaction is performed in the presence of a catalyst. Most GTL catalysts consist of an alumina, silica or titanium dioxide substrate having a high specific surface area. Cobalt and traces of either PGM (platinum group metals, i.e. Pt, Os, Ir, Ru, Rh, Pd) or of Re form active catalytic sites at the surface of the carrier.

However, after time, due to contamination by toxins, catalysts loose their activity and have to be replaced. There are ecological and economical reasons why recycling of such spent catalysts is preferred to land-filling.

Typical GTL catalysts contain approximately 20% (by weight) of cobalt and 500 ppm of PGM. In view of the current and expected market value of these metals, it appears desirable to recover them with high yields. The carrier itself is less attractive from an economical point of view. It also has a negligible environmental fingerprint.

In view of the above, a preliminary separation of the desired metals from the carrier material is operated. This can be performed by smelting the catalyst according to known processes, yielding a metallic alloy comprising the cobalt and the PGM, and a slag phase comprising the carrier material. The object of this disclosure concerns the further processing of the cobalt-based alloy, i.e. alloys containing more than 50% Co by weight. They also contain significant amounts of precious metals, along with some residuals from the slag phase.

In a typical hydrometallurgical process, such a metallic alloy is leached in acidic conditions. This normally leads to the formation of H ₂ . Extreme caution is then necessary, as the contained PGM are well known to catalyze the reaction of H ₂ with any oxidizing agent such as O ₂ from e.g. air present in the reaction environment. The risk of spontaneous explosion is therefore considerable.

RU-2252270 discloses a flow sheet for treating mattes essentially composed of copper- nickel sulfides, and some cobalt and PGM. The process uses a first atmospheric leach followed by a two-step autoclave leach. The net result of the process is a copper sulfide residue, also containing the PGM, and a nickel sulfate solution. This solution also contains any cobalt present in the feed. The disclosure does not concern the treatment of metallic alloys. Moreover, neither the alloy leaching step using copper sulfate with cementation of metallic Cu, nor the oxidizing copper leaching step, are divulged.

CA-2235429 describes a process in which copper sulfate is used for the dissolution of a.o. cobalt from iron-bearing mattes and alloys. During this step, copper cement is produced. This cement is calcined to form a copper oxide, which is subsequently used for the neutralization of the acidic leaching solution for the precipitation of iron as goethite (FeOOH). This flow sheet is only applicable when no PGM are present; such precious metals would indeed invariably accompany the copper cement and end up in the goethite instead of in the copper bearing phase, which is not desirable.

US-6440194 describes a process for the dissolution of cobalt, nickel and iron from an alloy prepared by aluminothermy. This alloy also contains PGM and some sulfur. Cobalt, nickel and iron are dissolved using sulfuric acid, while copper and the PGM do not react, which is expected in the non-oxidizing conditions applied. The reaction products comprise H ₂ , H ₂ S and CuS, further demonstrating that the conditions are reducing. Neither the alloy leaching step using copper sulfate with cementation of metallic Cu, nor the oxidizing copper leaching step, are divulged.

US-3954448 describes a process for the recovery of cobalt from cobaltiferous materials. The process is mainly pyrometallurgic in nature, but the resulting alloy or matte is dissolved in the presence of copper sulfate. Copper cement precipitates, which is removed by means of filtration, together with the PGM. The selective leaching of copper from the obtained residue, and the recirculation of the copper sulfate to the leaching step, are not taught.

US-4468302 again concerns copper-nickel mattes, not alloys. The residue of a first acid leach is roasted to remove the sulfϊdic sulfur. The obtained copper and nickel oxides are then split in a first part that is smelted to form anodes, and in a second part that is leached using air to form an electrolyte. The anodes are then refined using said electrolyte. PGM are recovered from the anode slimes.

WO-2000/073520 describes a 2-stage leaching process for the recovery of cobalt from mattes and/or alloys. A mixture of copper sulfate and sulfuric acid is used for the dissolution of cobalt; selectivity towards copper and iron is established by choosing adequate pH levels. Although the examples feature good yields for cobalt, the process is not suitable for PGM-containing alloys, because PGM would report to the iron residue from the second leaching step, which is not desirable.

The aim of the invention is to offer a high yield and safe process for the separation and recovery of metals, in particular of cobalt and of the PGM.

To this end, a process is divulged for the recovery of PGM from a cobalt-based, PGM- bearing alloy, comprising the steps of:

- powdering the alloy to a mean particle size of less than 250 μm; - providing a mixed tank reactor equipped with means for feeding H ₂ SO ₄ and a CuSO ₄ solution; - leaching in a mixed tank reactor the powdered alloy in a first acidic aqueous solution comprising H ₂ SO ₄ and CuSO ₄ , thereby obtaining a CoSO ₄ solution and a PGM-bearing metallic Cu residue;

- isolating the PGM-bearing metallic Cu residue from the first CoSO ₄ solution, by solid/liquid separation;

- leaching in a mixed tank reactor the PGM-bearing metallic Cu residue in a second acidic aqueous solution comprising H ₂ SO ₄ and an oxidizing agent, thereby obtaining a CuSO ₄ solution and a PGM concentrate;

- isolating the PGM concentrate from the CuSO ₄ solution, by solid/liquid separation; - returning the CuSO ₄ solution to the step of leaching the powdered alloy.

It is preferred to use powders with a D ₅₀ of less than 100 μm. The powder can be obtained e.g. by atomization of the alloy.

In the step of leaching the powdered alloy, it is useful to maintain the pH of the first acidic aqueous solution below 4, preferably between 1 and 2, by providing an adequate quantity H ₂ SO ₄ .

It is also useful to maintain the oxygen reduction potential of the first acidic aqueous solution above O mV vs. Ag/ AgCl by addition of Cu ²⁺ in the step of leaching the powdered alloy. Typically, a stoichiometric amount of Cu ²⁺ (vs. Co) is needed. The step is thus best performed without addition of any other oxidizing agent such as air, that is, using Cu ²⁺ as the only such agent

During this first leaching step, the temperature of between 60 and 90 °C is particularly suitable.

In the step of leaching the PGM-bearing metallic Cu residue, it is useful to maintain the pH of the second acidic aqueous solution below 4, preferably between 1 and 2, by addition of an adequate quantity of H ₂ SO ₄ . Typically, a stoichiometric amount of acid (vs. Cu) is needed. During this second leaching step, the temperature of between 60 and 90 ⁰ C is particularly suitable.

The proposed process relies on the oxidative characteristics of aqueous copper solutions, whereby Cu ²⁺ ions oxidize not only any less noble metal such as Co, but also any free hydrogen, according to the respective reaction (1) and (2):

Cu ²⁺ + Co° -> Co ²⁺ + Cu° ( 1 )

Cu ²⁺ + H ₂ ^ Cu° + 2H ⁺ (2) The above-mentioned equations show that no hydrogen will be formed, as long as Cu ²⁺ ions are in solution. It thus appears that the competing reaction (3) of acid with cobalt is strongly inhibited, at least if the acidity remains moderate:

2H ⁺ + Co° -> Co ²⁺ + H ₂ (3)

The net result is a total absence of free H ₂ , and thus a highly safe process.

A schematic representation of the process for the treatment of alloys of cobalt and PGM is given in Figure 1. The reference numerals and letters correspond to the following process steps and streams:

(1) First leaching; (2) First solid/liquid separation; (3) Second leaching; (4) Second solid/liquid separation;

(A) Powdered Co-PGM alloy; (B) CuSO ₄ solution; (C) CoSO ₄ solution; (D) PGM-bearing Cu residue; (E) PGM concentrate.

In the first leaching step (1), the Co-PGM alloy (A) should be finely powdered in order to ascertain sufficient dissolution kinetics. In acidic conditions, the CuSO ₄ will selectively dissolve the cobalt and other less noble metals according to equation (1), leaving the PGM, together with metallic copper, in the residue (D).

During leaching, at least some CuSO ₄ should remain present. Even a minimal amount (such a 1 mg/L) of Cu ²⁺ indeed ensures that no H ₂ will be formed. While the Cu ²⁺ could be measured directly, it is more practical to monitor the ORP (oxidation-reduction potential) as an indicator of the presence of Cu . If the ORP decreases to below 0 mV vs Ag/AgCl, no more Cu ²⁺ is present in solution and the risk of H ₂ formation arises.

It is useful to monitor the pH as some acid can be consumed in the leaching step, e.g. if oxidized metal compounds are present in the alloy. While a low pH increases the reaction kinetics, it also tends to increase the risk of H ₂ formation if the Cu ²⁺ would at any time during leaching become exhausted. A high reaction temperature also favors rapid reactions; there is however a tradeoff, as evaporation becomes excessive when approaching the boiling point of the solution.

After this first leaching step, the residue and the filtrate are separated by decantation or filtration (2). The products are a solution (C) containing most of the cobalt together with some residual CuSO ₄ , and a residue (D) containing the PGM along with metallic copper and possibly traces of undissolved alloy (A).

This residue (D) from this first leaching (1) and separation (2) steps can be further treated according to known ways, such as by copper electrowinning, whereby the PGM are further concentrated in so-called anode slimes. However, according to said preferred embodiment, a second leaching step is first performed on this residue.

This optional second leaching step (3) involves the selective dissolution of copper using an acid, preferably H ₂ SO ₄ , and an oxidizing agent. For the latter, air and ozone are suitable, but pure O ₂ and H ₂ O ₂ are usually the most interesting candidates from an economical point of view. If maximal dissolution of the copper is desired, stoichiometric amounts of acid and of oxidizing agents are to be used. Again, no H ₂ is formed in this step.

The next step (4) separates the solid PGM concentrate (E) from the leach solution (B) that now mainly consists of aqueous CuSO ₄ . A small amount of dissolved CoSO ₄ will also be present in this CuSO ₄ solution: some of it originates from the solution that impregnates the residue (D); the remainder comes from the dissolution of any undissolved alloy in that residue.

The filtrate of the separation step (4) is returned to the first leaching step (1), where it is used for leaching of a new quantity of Co-PGM alloy. Any cobalt that is present in this recycled solution (B) will finally report to the filtrate (C). This preferred embodiment thus loosens the constraints on the first solid/liquid separation in step (2), allowing to envisage a simple decantation in that step.

Here again, the pH and the temperature during leaching are chosen so as to favor fast reaction kinetics. A too high acidity is however to be avoided, in particular towards the end of the leaching step, as the resulting CuSO ₄ solution (B) has to be compatible with the first leaching step (1), to which it is recycled. After the separation step, the PGM concentrate can further be refined according to known techniques.

Because copper cementation is easily performed in a sulfate solution, this medium is definitely favored in the proposed process. Somewhat faster kinetics may however be obtained when a small amount or HCl is added to the first leaching operation, although a negative effect on the cementation of copper is to be expected. This means that some residual Cu ²⁺ will be present in the CoSO ₄ solution (C). Downstream removal of this Cu ²⁺ may be performed using a suitable reducing agent or a small amount of fresh alloy.

The amount of alloy that can be treated depends on the amount of CuSO ₄ made available. To limit the volume of the circulated solutions, it seems appropriate to keep the concentration of CuSO ₄ in the recycled stream (B) as high as possible. In practice, a certain maximum concentration should not be exceeded, so as to avoid the precipitation of CuSO ₄ crystals when the solutions cool down, such as during storage.

It is further noteworthy that the complete process, including the regeneration of the CuSO ₄ , can be performed using just one reactor, a filter and a storage tank. Example

About 41.2 g of atomized (100 μm D ₅₀ ) Co-Pt alloy is added to 500 mL of acidified CuSO ₄ solution (63.5 g/L Cu, 10 g/L H ₂ SO ₄ ). The alloy contains, by weight, 83% Co, 3.4% Cu and 0.2001% Pt. An overhead mixer is used for suspending and mixing the alloy particles. The solution is heated to 90 °C and the pH kept between 1.5 and 1.7 using diluted sulfuric acid (250 g/L H ₂ SO ₄ ). The ORP (vs. Ag/AgCl) is also continuously monitored. The color of the alloy slowly changes from grey to red, which indicates the formation of metallic copper. After 4 h, the leaching comes to an end, as indicated by the ORP of the solution stabilizing at about 90 mV. The obtained mixture is filtered. The filtrate (515 mL) contains 50 g/L Co, 8 g/L Cu and <1 mg/L Pt. The residue (38.5 g) contains 20% Co, 75% Cu and 0.2% Pt. The cobalt yield is calculated to be 77%, while no Pt had dissolved.

Then, the residue is repulped in fresh water (400 mL) and heated to 90 °C. Fresh H ₂ SO ₄ (1000 g/L) is used to control the pH between 1.5 and 1.8. Oxygen is sparged into the solution at a rate of 15 L/h. After 3.5 h, no more H ₂ SO ₄ is consumed, indicating that the reaction has come to an end. Solids and liquid are separated by filtration. The residue (2.7 grams) contains 10% Co, 8% Cu and 3.2% Pt, while the filtrate (450 mL) contains 65 g/L Cu, 17 g/L Co and <1 mg/L Pt.

The analytical results in Table 1 show that less than 1% of the cobalt and more than 99.4% of the Pt report to the second leach residue. Excellent separations and yields are thus achieved for these metals. The PGM concentrate (E) also contains silicates or any other inert fractions that were present in the starting Co-PGM alloy (A). Table 1 : Analysis of feed and products according to Example