Hydrometallurgical recovery of sequential metal values from nickel saprolite ores to prepare magnesium chloride products

Abstract

Processes are provided in which successive steps of hydrometallurgical value extraction are provided, to extract iron, aluminum, nickel, cobalt, manganese and magnesium from nickel saprolite ores. Processes are provided that are substantially free of added sodium, to facilitate the production of high purity magnesium chloride products. Precipitations are carried out with an olivine slurry, so that additional metal values are extracted from the olivine materials while facilitating thorough precipitation of metal values.

Description

SODIUM FREE HYDROMETALLURGICAL RECOVERY OF SEQUENTIAL METAL VALUES FROM NICKEL SAPROLITE ORES TO PREPARE HIGH PURITY MAGNESIUM CHLORIDE PRODUCTSFIELD

[0001] The invention is in the field of inorganic chemistry, particularly sequential hydrometallurgical value extraction from nickel saprolite ores to produce high purity magnesium chloride products.BACKGROUND

[0002] Nickel is an important input for a variety of technologies that are required to facilitate a transition to a lower carbon economy. Nickel saprolite ores, an important source of nickel, are traditionally treated by smelting, which can be associated with relatively high carbon emissions. Technologies for reducing the emission of gaseous carbon dioxide associated with the production of nickel and other metal values from mineral feedstocks are accordingly an important tool for addressing anthropogenic climate change (as for example disclosed in PCT Publication WO2022 / 113025).SUMMARY

[0003] Processes are provided in which successive steps of hydrometallurgical value extraction are carried out on a mineral feedstock. In particular, processes are provided for producing magnesium chloride and segregated metal values from a nickel saprolite mineral feedstock. The processes may include comminuting the nickel saprolite mineral feedstock. The saprolite mineral feedstock is treated to leach iron, aluminum, nickel, cobalt, manganese and magnesium metal values from the comminuted mineral feedstock with an HCI acid leachant in a leach slurry substantially free of sodium, to produce a leach slurry comprising a solid siliceous residue and a loaded leach solution.

[0004] Iron and aluminum are precipitated from the loaded leach solution in the leach slurry with pH adjustment by addition of a ground olivine mineral slurry and a chlorine gas oxidant (the oxidant added so as to continuously oxidize ferrous iron in the loaded leach solution to ferric iron. The ground olivine mineral slurry may comprise an olivinemineral ground in the recycle MgCl2 brine solution. In select embodiments, precipitating iron and aluminum comprises addition of a stoichiometric excess of the ground olivine mineral slurry. The olivine mineral may comprise magnesium, nickel and cobalt, in which case the magnesium, nickel and cobalt values are leached from the olivine mineral in the leach slurry. The concentration of iron in the HCI acid leachant may be at least 15 g / L and the concentration of iron in the Fe / AI depleted solution may be less than 10 mg / L. The concentration of aluminum in the HCI acid leachant may for example be at least 100 mg / L and the concentration of aluminum in the Fe / AI depleted solution is less than 10 mg / L.

[0005] Precipitation of iron and aluminum produces an Fe / AI depleted solution and a solid amorphous silica residue (which as in all residues may be washed). The Fe / AI depleted solution is separated from the solid amorphous silica residue, and the solid amorphous silica residue may comprise silica and iron and aluminum and chromium- containing solids.

[0006] Nickel and cobalt are precipitated from the Fe / AI depleted solution, wherein the precipitating is with addition of a MgO slurry comprising Mg(OH)2, to produce a Ni / Co depleted solution and a first nickel and cobalt mixed hydroxide precipitate product. The first nickel and cobalt mixed hydroxide precipitate product may for example be recycled to the step of leaching.

[0007] Optionally, further precipitation of nickel and cobalt from the Ni / Co depleted solution may be carried out, wherein the precipitating is with addition of a further MgO slurry comprising Mg(OH)2, carried out so as to reduce the concentration of nickel and cobalt in the Ni / Co depleted solution and to produce a secondary nickel and cobalt mixed hydroxide precipitate product. The secondary nickel and cobalt mixed hydroxide precipitate product may be recycled to the step of precipitating nickel and cobalt from the Fe / AI depleted solution and / or to the step of leaching.

[0008] Manganese may be precipitated from the Ni / Co depleted solution with addition of a third MgO slurry comprising Mg(OH)2, in the presence of CL, so as to produce a MgCL product solution and a solid manganese dioxide residue precipitate product.

[0009] The MgCl2 product solution may be separated from the solid manganese dioxide residue precipitate product; and, a portion of the MgCl2 product solution may be recycled as a recycle MgCl2 brine solution to the step of comminuting the nickel saprolite mineral feedstock.

[0010] The MgCl2 product solution may be subjected to heat to produce a concentrated MgCl2 solution product, and the concentrated MgCl2 solution may be subjected to crystallization, for example by evaporation, flaking or prilling, so as to crystalize a MgCl2.6H2O crystal product. The MgCl2.6H2O crystal product may be subjected to pyrohydrolysis to produce an MgO product and hydrochloric acid for recycle.

[0011] Sodium hydroxide may be added to the concentrated MgCl2 solution product to produce a Mg(OH)2 precipitate and a sodium chloride solution product.BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 is a schematic illustration of an integrated continuous process for hydrometallurgical value extraction from a mineral feedstock.

[0013] Figure 2 is a simplified flow diagram for an exemplified pilot plant campaign (PP1 R).

[0014] Figure 3 includes four line graphs illustrating the results of PP1 R rush assay trends.

[0015] Figure 4 includes four line graphs illustrating the PP1 R L circuit profile solution deportments.

[0016] Figure 5 includes four line graphs illustrating PP1 R ON circuit profile solution deportments.

[0017] Figure 6 is a line graph illustrating PP1 R L / ON circuit metal deportments and reagent additions.

[0018] Figure 7 is a simplified flow diagram for an alternative exemplified pilot plant campaign (PP2R) for MHP and SP Circuits.

[0019] Figure 8 is simplified flow diagram for a PP2R MnR circuit.

[0020] Figure 9 includes four line graphs illustrating the results of PP2R rush assay trends.

[0021] Figure 10 includes two line graphs illustrating the results of PP2R MHP circuit profile solution assays.

[0022] Figure 11 includes two line graphs illustrating the results of PP2R MHP circuit profile precipitate assays.

[0023] Figure 12 includes two line graphs illustrating the results of PP2R SP circuit profile solution assays.

[0024] Figure 13 is a line graph illustrating the results of PP2R MHP circuit performance and reagent dose.

[0025] Figure 14 is a line graph illustrating the results of PP2R MHP grade and estimated seed recycle ratio.DETAILED DESCRIPTION

[0026] Figure 1 illustrates a process in which metal values are leached (“HCI leaching”) from a comminuted (“crushing and grinding”) mineral feedstock with an acid leachant (“HCI solution”). Crushing and grinding takes place in recycled brine solution containing magnesium chloride (“Brine Recycle”). Recycling of a brine solution in this way mitigates the addition of water, which would otherwise necessitate the energetically expensive increased use of evaporation to manage water levels in the circuit.

[0027] In the HCI Leaching step, HCI is used at high strength (typically 32-36% HCI by weight in water; for example a typical product from an HCI production facility attached to a chlor-alkali plant). The raw saprolite materials contain a variety of silicate minerals including magnesium, iron, aluminum, nickel, cobalt, manganese and minor impurity elements. The chemistry may therefore be represented as comprised of the following major reactions:Mg2SiO4+ 4HCI = 2MgCI2+ SiO2+ 2H2O (1)Ni2SiO4+ 4HCI = 2NiCI2+ SiO2+ 2H2O (2)Fe2SiO4+ 4HCI = 2FeCI2+ SiO2+ 2H2O (3)

[0028] Other minerals present such as iron oxides or aluminum oxides may also react with HCI to form additional salts in solution:FeO(OH) + 3HCI = FeCh + 2H2O (4)AIO(OH) + 3HCI = AICI3 + 2H2O (5)

[0029] The leaching temperature may be close to the boiling point, for example as to ensure rapid extraction. Acid addition may for example range from 500 to 1000 kg HCI per dry tonne of solid saprolite feed and will vary with the chemical composition of the feed. The brine recycle solution may be modulated so as to ensure that acid leaching is performed with a high total salt content as MgCl2.

[0030] The HCI Leaching time can for example vary from 1 hour to 8 hours. The leaching may be carried out in a series of reactors with acid added largely in the first reactor. In this way, the acid concentration in leaching solution drops from the first reactor to the last reactor as the extraction process approaches completion. The solids contain silica from the leaching process and other partially reacted minerals.

[0031] The entire HCI leaching slurry, comprising a solid siliceous residue and a loaded leach solution, may then be directed to an iron and aluminum removal process, for example a single stage process, using olivine addition for pH adjustment and oxidation by chlorine gas.

[0032] The iron and aluminum content in the loaded leach solution may be precipitated by the addition of ground olivine to the slurry. The olivine may for example be ground in a recycle brine solution containing magnesium chloride to minimize the addition of fresh water to the circuit. The pH adjustment may be conducted with a stoichiometric excess of olivine mineral, so as to ensure that the removal of iron, aluminum and chromium is maximized. The temperature for iron and aluminum precipitation may for example be 75 °C to the boiling point. Precipitation time for iron and aluminum precipitation may for example be 1 to 8 hours, and may be carried out in a series of continuous reactors, with the olivine slurry added to the first iron and aluminum removal tank in series.

[0033] As the olivine reacts, ferrous ion is leached into solution. Ferrous ions will not precipitate with the addition of olivine. Therefore chlorine gas is added via a sparger toone or more of the reactors to oxidize ferrous to ferric, so as to maximize removal of iron from solution. The combined leach and olivine neutralization slurry may then be directed to a solid / liquid separation step in order to separate the Fe / AI depleted nickel, cobalt and magnesium-containing solution away from the silica / iron / aluminum / chromium-containing solids (“Amorphous Silica Residue”). The solids may be washed to remove entrained chloride salts and may then, for example, be dried and used as a supplementary cementitious materials (SCM) for cement making.

[0034] Olivine with a high magnesium content is desired for iron and aluminum precipitation. Olivine is typically represented as magnesium silicate but olivine typically also contains ferrous silicate as well. The reactions for neutralization may be represented shown below:4HCI + Mg2SiO4 = 2MgCI2+ 2H2O (6)4HCI + Fe2SiO4= 2FeCI2+ 2H2O (7)2FeCI2+ Cl2(g) = 2FeCI3(8)4FeCI3+ 3Mg2SiO4 + 2H2O = 4FeO(OH) + 6MgCI2+ 3SiO2(9) 4AI Cl3+ 3Mg2SiO4+ 2H2O = 4AIO(OH) + 6MgCI2+ 3SiO2(10)4CrCI3+ 3Mg2SiO4+ 2H2O = 4CrO(OH) + 6MgCI2+ 3SiO2(11 )

[0035] Olivine often contains nickel and cobalt. These elements will also react with the acidic leachate. Nickel and cobalt extraction from olivine is accordingly beneficial to the production of nickel and cobalt as mixed hydroxide from the overall process. The extraction of magnesium from olivine may for example be carried out so as to lead to a higher overall production of magnesium in all forms.

[0036] As illustrated in Figure 1, nickel and cobalt recovery takes place in a Ni / Co MHP Production step. Nickel and cobalt are present in solution as NiCI2and CoCI2salts. The recovery of Ni / Co may be orchestrated in many ways, including the direct precipitation of mixed hydroxide precipitate. This may for example be done directly from the solution coming from the Iron and Aluminum Precipitation step. For example, magnesium oxide may be slurried in water to form magnesium hydroxide and then dosed into the precipitation circuit, with the attendant reactions illustrated as follows:MgO + H2O = Mg(OH)2(12)NiCI2+ Mg(OH)2= Ni(OH)2+ MgCI2(13)C0CI2 + Mg(OH)2= Co(OH)2+ MgCI2(14)

[0037] Other metals may also precipitate with the Ni / Co in minor amounts, such as Cu, Zn, Mn and Fe (remaining iron in solution). Some magnesium oxide / hydroxide may also remain in the mixed hydroxide precipitate (MHP). This can be minimized by recycling of the MHP solids in the precipitation circuit to provide repeated opportunity for the remaining MgO / Mg(OH)2 to react with soluble nickel and cobalt. The selectivity of MHP precipitation can be enhanced by using a two stage MHP precipitation, in which the second stage precipitate is recovered and recycled to the first stage MHP precipitation process or to the discharge from the main leaching step (where acid is present to redissolve the Ni / Co and other metals from the second stage leach). Figure 1 only indicates the recycle of this solid to the leach process.

[0038] The MHP is recovered by solid / liquid (S / L) separation and washing. A pressure filter may for example be used with a “squeeze” cycle to minimize the entrained moisture in the washed MHP cake prior to shipping.

[0039] The Ni / Co MHP Production precipitation step may for example be carried out between 25- 90 °C and terminal pH may for example be in the range of 5-8. In some embodiments, pH measurement may be difficult in a strong salt solution, in which case magnesium oxide addition may be controlled by stoichiometry rather than, or in addition to, pH. The Ni / Co MHP Production precipitation time may for example be 1-8 hours. Seed recycling may be used to maximize particle size and minimize contamination. The process (as in all steps) may be conducted continuously.

[0040] Manganese removal may then be carried out on the Ni / Co depleted solution. Manganese is generally an undesirable impurity in magnesium hydroxide, and typically cannot be selectively precipitated as a hydroxide in the presence of magnesium hydroxide. Accordingly, oxidation and precipitation are used to remove manganese from solution, as represented by the following reaction:MnCI2+ Cl2(g) + 2MgO = MnO2+ 2MgCI2(15)

[0041] The manganese dioxide precipitate may be filtered and washed (“Washing of Mn Residue” in Figure 1). This precipitate may optionally be added to the cement residue from leaching or may be sold as a source of manganese dioxide.

[0042] The product of the process at this point is a high purity magnesium chloride (MgCl2) product solution, free of any added sodium ion from NaOH or NaOCI. This solution can for example be concentrated by evaporation to prepare a range of products. In one embodiments, a strong MgCl2 solution (containing e.g. 28-33% MgCl2) may be prepared as a final product. In an alternative embodiment, a crystal product of MgCl2.6H2O may be prepared, for example by evaporation and crystallization or prilling (in which magnesium chloride is placed in a fluid bed dryer as droplts or in a spray tower as droplest to be cooled and crystalized) or flaking (by placing hot concentrated solution on a cold surface). In a further alternative embodiment, an MgO product can be prepared, along with HCI for recycle, for example using fluid bed or spray pyrohydrolysis, as represented by the following reaction:MgCI2+ H2O = MgO + 2HCI(g) (16)

[0043] In a fourth alternative embodiment, the MgCl2.6H2O salt may be prepared by evaporation and crystallization and then directed to a magnesium metal making process such as the Norsk Hydro technology (see Pinfold, Primary magnesium production at Becancour, In Proceedings of Metallurgical Society of Canadian Institute of Mining and Metallurgy, Extraction, Refining, and Fabrication of Light Metals, Pergamon, 1991 , Pages 43-53). This technology uses multi-stage dehydration and finally molten salt electrolysis to make magnesium metal and chlorine gas. The chlorine gas may be reacted with hydrogen to make HCI for recycle to leaching.

[0044] In a fifth alternative embodiment, there is an option to add sodium hydroxide to the concentrated MgCL solution to prepare Mg(OH)2 precipitates. This would result in a solution of NaCI which could either be recycled via a chlor-alkali process or crystallized and sold as NaCI or disposed, as summarized in the following reaction: MgCI2+ 2NaOH = Mg(OH)2+ 2NaCI (17)

[0045] Although various embodiments of the invention are disclosed herein, many adaptations and modifications may be made within the scope of the invention inaccordance with the common general knowledge of those skilled in this art. Such modifications include the substitution of known equivalents for any aspect of the invention in order to achieve the same result in substantially the same way. Terms such as “exemplary” or “exemplified” are used herein to mean “serving as an example, instance, or illustration.” Any implementation described herein as “exemplary” or “exemplified” is accordingly not to be construed as necessarily preferred or advantageous over other implementations, all such implementations being independent embodiments. Unless otherwise stated, numeric ranges are inclusive of the numbers defining the range, and numbers are necessarily approximations to the given decimal. The word "comprising" is used herein as an open-ended term, substantially equivalent to the phrase "including, but not limited to", and the word "comprises" has a corresponding meaning. As used herein, the singular forms "a", "an" and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a thing" includes more than one such thing. Citation of references herein is not an admission that such references are prior art to the present invention. Any priority document(s) and all publications, including but not limited to patents and patent applications, cited in this specification, and all documents cited in such documents and publications, are hereby incorporated herein by reference as if each individual publication were specifically and individually indicated to be incorporated by reference herein and as though fully set forth herein. The invention includes all embodiments and variations substantially as hereinbefore described and with reference to the following examples and drawings.

[0046] Example 1. Sodium Free Leaching and Olivine Neutralization

[0047] A continuous pilot plant operation was developed to demonstrate sodium free leaching and olivine neutralization with chlorine addition for iron oxidation and removal.

[0048] The pilot plant Leach (L) and Olivine Neutralization (ON) circuits consisted of four leaching reactors and five neutralization reactors. Chlorine gas was used for oxidation of ferrous iron. Ferrous came from the olivine that was used to neutralize free acidity and precipitate iron and aluminum from the leach solution. Chlorine was addedin the third neutralization reactor (ON3). A diagram of the two circuits is in Figure 2, showing the mixing of ore slurry (at -34% solids nominally) with 32% w / w hydrochloric acid in the first leach reactor (L1) followed by a gravity fed cascade through the rest of the reactors. Olivine slurry (60% w / w and finely ground) was added in the ON1 reactor, and no further additions were made in subsequent reactors. The final pulp discharging from the ON circuit was collected in batches and processed through a plate and frame filter press, followed by a partial washing of the cake within (adding 2.45 kg of wash water for every 1 kg of ore processed).

[0049] The leaching circuit had a total residence time of 4.58 hours while the olivine neutralization had a total residence time of 5.62 hours. The pilot plant operated over a 5 day period. Chlorine was added as a pure gas via a flow controller into the 3rdolivine neutralization tank.

[0050] Throughout the campaign, the pilot plant operators took hourly measurements of key parameters, specifically focusing on flowrates and temperatures. These hourly measurements were averaged over twelve-hour time periods between midnight and noon for each day (AM and PM) and are summarized in Table 1 and Table 2. In general, the controlled flowrates (ore slurry, acid, and olivine) were relatively stable with only minor deviations between periods. The flow of chlorine gas was modulated by the ORP in ON3; originally a probe (measuring in mV relative to a Ag / AgCI reference electrode) was inserted in the reactor pulp and measured either as a live reading of the pulp or by collecting a small sample from the reactor and using an off-line ORP probe. The ORP target for control was -600-650 mV. This was a minimum to achieve oxidation of ferrous iron to ferric iron for effective precipitation.Table 1 : PP1R Summarized Physical L Circuit DataTable 2: PP1R Summarized Physical ON Circuit Data1rotameter that replaced digital flowmeter did not have a calibration curve for chlorine gas available* flow measured with digital flowmeter before equipment failure

[0051] Samples from each circuit were collected at regularly scheduled intervals, consisting of:• Rush / control samples - circuit discharge liquor samples collected every four hours and analyzed for a select number of elements.• Profile samples - liquor and solids samples collected from each reactor within a circuit. Each circuit was profiled once per day.• Composite samples - key feed and discharge streams (liquor and solids) were subsampled every four hours over a twelve-hour period, combining the subsamples to generate an overall composite covering the entire period. These samples are used to generate metallurgical balances for the circuits.

[0052] Rush / control samples were collected from the discharge of both the L and ON circuits every four hours, and analyzed for nickel, iron, magnesium, and aluminum. The trends observed through those control assays are shown in Figure 3. There was little change in the nickel concentrations between the L and ON circuits, while there was a marked change in the other three. The iron content fell from ~20 g / L to very low levels for the majority of the campaign, with 0.7 mg / L as the analytical reporting limit. Initially the iron content in solution was higher, caused by the issues with securing a reliable flow of chlorine gas during and after the digital flow controller’s failure, and given the retention time of the ON circuit, several hours required to flush out the impacted pulp. Conversely, the magnesium tenor rose between L and ON, owing to the leaching and exchange of magnesium in the olivine added in ON1 with the free acidity and Fe / AI ions in the L solution. Much like iron, the aluminum tenors fell sharply, from -200 mg / L to <0.2 mg / L for most of PP1 R. These results show that chlorine addition to the olivine circuit is very effective in removing iron from solution to very low values.

[0053] Once per day, profile samples were collected from the L and ON reactors, analyzing the solution and solids in order to track performance kinetics within the circuits. The L reactors were sampled at 20:00, while the ON circuit was sampled at 08:00. The analyses were used to calculate the metal deportment to the solution phase in each reactor, with the results for L1 through L4 shown in Figure 4. The indicated nickel extraction was generally high with the exception of April 23, with iron and magnesium following suit. Aluminum tended to show decreasing liquor deportments through the reactors, with a larger degree of spread from day to day. The exception on April 23 is attributed to low acidity in the L circuit within that time period, with free acidities titrated between 2.5 g / L HCI in L1 and 1.1 g / L HCI in L4. The preceding and subsequent days were higher (April 22 was 17.5 g / L HCI in L4, while April 26 was 10.1 g / L HCI in L4).

[0054] The calculated liquor deportment for the ON reactors is summarized in Figure 5. The lower acidity and extractions in the L circuit on April 23 were seen to have an impact on the April 24 ON profile set, with lower overall deportments of all four main elements. Aside from that sample set, overall nickel deportment remained high, whileiron fell to near zero and magnesium slowly rose from ON1 to ON5 as the olivine reacted. Aluminum was effectively removed in ON2, while iron persisted until ON3 where the oxidation took place.

[0055] Composite Samples and Metallurgical Balances

[0056] The composite samples of the key circuit feeds and discharge streams were collected twice daily, corresponding with the same AM / PM time periods used to average the physical readings and measurements taken by the plant operators. For the L circuit, key streams included the ore slurry and L4 discharge pulp; the technical grade acid was assumed to be 32% w / w as procured. The analyses of the SMT ore slurry are in Table 3, and show a very consistent feed composition with 1.7-1.8% Ni, 8.5-8.9% Fe, -16% Mg, and -0.16% Mn in the ore itself and little in the way of dissolved metals in the solution phase of the slurry (-20 mg / L Mg, -10 mg / L Si, and -8 mg / L Ca). The discharge pulp, summarized in Table 4, shows much lower levels of main elements in the residue, averaging 0.11 % Ni, 1.8% Fe, 2% Mg, and -0.01 % Mn. The leach liquor was enriched in these elements, with approximately 4.6 g / L Ni, 21.5 g / L Fe, -42 g / L Mg, and 424 mg / L Mn. Additionally, up to -200 mg / L Co was also present, along with -120 mg / L Cr and -220 mg / L Al.Table 3: PP1R L Circuit Ore Feed Pulp AnalysesTable 4: PP1R L Circuit Discharge Pulp Analyses

[0057] For the ON circuit, the olivine slurry and the ON5 discharge pulp were composited, with analyses shown in Table 5 and Table 6, respectively. Much like the ore, the olivine slurry showed little variation from period to period, averaging 0.28% Ni, 5.23% Fe, 28.5% Mg, and 19.2% Si as the major elements in the olivine itself. The liquor phase showed trace elements present, with ~32 mg / L Na and 26.5 mg / L Mg as the most prominent dissolved metals. The average olivine addition was 0.31 g / g relative to SMT ore.Table 5: PP1R ON Circuit Olivine Pulp Analyses

[0058] The composition of the ON5 discharge pulp is analogous to the L4 pulp, with inverted iron and aluminum; the nickel grade in the residue was 0.16% Ni, 11.4% Fe, 4.77% Mg, and 0.45% Al on average, with 4.4 g / L Ni, 101 mg / L Fe, 56.5 g / L Mg, and <1 mg / L Al in the solution. Of note is that while the average iron tenor was 101 mg / L, that value is inflated by the high iron in the start-up periods (161 and 683 mg / L) compared tothe rest of the plant, which was <50 mg / L and as low as <0.2 mg / L. The silicon grade of the residue was just under 30% overall in PP1 R.Table 6: PP1R ON Circuit Discharge Pulp Analyses

[0059] The overall deportment trends are shown graphically in Figure 6, along with the acid and olivine additions relative to SMT ore.

[0060] Overall, PP1 R demonstrated that oxidation of the ferrous iron from the olivine can be accomplished through the use of chlorine gas under continuous conditions, effectively removing sodium addition in the form of bleach from the flowsheet.

[0061] Example 2. Sodium Free Production of Mixed Hydroxide Precipitate and Manganese Removal

[0062] The second campaign under sodium-free conditions aimed to confirm the use of magnesium oxide slurry as the alkaline reagent instead of sodium hydroxide as part of the testing of the sodium free process. As shown in Figure 7, the MHP circuit consisted of four reactors in series, with the filtrate / wash from the ON circuit (producedin PP1 R from Example 1) fed to MHP1 where MgO slurry was added to raise the pH and precipitate nickel. The discharge from the fourth reactor was directed to a thickener, where flocculant was added (Magnafloc 338 before changing to Magnafloc 611 after it showed better performance) to settle the solids. A high seed recycle of underflow solids was started as soon as possible, and any extra underflow was directed to vacuum filtration as required. Overflow solution from the MHP thickener was directed to the SP circuit (a surge vessel was used to break the circuits, not shown in Figure 7), which consisted of three reactors and a thickener in a similar manner as the MHP circuit. Magnesium oxide slurry was once again added to the first reactor and the same flocculants were used in SP. The overflow solution from the SP thickener was directed to the manganese removal (MnR) circuit, while the underflow was collected in batches (if required) as very little in the way of solids was generated.

[0063] The MnR circuit is shown in Figure 8, including a feed solution preheater followed by three reactors in series (MnR1 through MnR3). Chlorine gas was added to MnR1 to oxidize manganese and make it more amenable to precipitation, achieved through additional pH adjustment in MnR2 where magnesium oxide slurry was added. The discharge pulp was directed to a vacuum filter, and the filtrate to drums for storage.

[0064] A total of 3.06 hours retention time was used for the MHP circuit (4 stages) with 2.49 hours for the SP circuit and finally 1 .66 hours for the MnR circuit.

[0065] Beginning on the morning of May 13, PP2R ran until the afternoon of May 17, for approximately 115 hours of operation including start-up / fill operations.

[0066] Hourly readings and measurements by the plant operators were recorded and averaged over two twelve-hour periods per day (AM / PM); the summarized data for the MHP circuit is in Table 7. Originally the circuit was to use 20% w / w magnesium oxide slurry as the reagent, but due to pumping issues (plugging, settling within the lines) at the low flowrates of reagent, this was reduced to 10% w / w to increase the overall flow / speed within the lines. Additionally, while PP2R began using dose control of magnesium oxide relative to the nickel concentration in the combined PP1 R ON filtrate and wash solution (to be adjusted based on the analytical results during operation) midway through the campaign the circuit pivoted to pH control to help smooth outoscillations in the control. An added benefit was that pH control helped to absorb the changes in the circuit as underflow seed recycle began (bringing with it potentially unreacted magnesium units that would reduce the amount of fresh magnesium oxide required).Table 7: PP2R Summarized MHP Circuit Physical Data

[0067] The pH in the MHP circuit was measured twice, once using the circuit probes (pH probes immersed in the reactor pulps continuously for a live reading) and a secondtime using an offline bench probe, collecting a small volume of pulp from each reactor to measure on the bench. As such, when pH control was enacted, the bench pH readings were used to ensure that the circuit probes were not drifting / fai li ng such that they could not be relied on to control the magnesium oxide addition. The pH value used in MHP1 (typically between 6.1-6.3) was purely based on the circuit analysis feedback and amounts to an internal control value for this specific campaign / conditions.

[0068] The physical data for the SP circuit in PP1 R is summarized in Table 8. As with the MHP circuit, issues were encountered with pumping 20% w / w magnesium oxide slurry, and due to the even lower flowrate required in SP, the slurry was thinned to 5% w / w for the first two days of the campaign and switched to 10% w / w in the afternoon of May 14. Also, like the MHP circuit, SP switched from reagent dose control to pH control, the day after MHP made the change. The same pH measurement system was used for SP, utilizing bench and circuit probes to confirm the operability of the probes in the reactors. Seed recycle from the SP thickener was started on May 15, at a low overall flowrate to avoid “rat-holing” the small underflow bed that had formed up to that point.Table 8: PP2R Summarized SP Circuit Physical Data

[0069] Table 9 summarizes the physical data from the PP2R MnR circuit. This circuit diluted the magnesium oxide stream to 5% w / w prior to starting based on the performance of the SP circuit and was left at that slurry density for the entire campaign. The MnR circuit was also fully pH controlled from the beginning, adjusting the pH setpoint as required based on the nickel analyses in the control samples. While chlorine gas was added to MnR1 , a stainless steel rotameter meant for air was procured at the request of Atlas for use in this campaign (instead of the oversized PTFE unit used in PP1 R) as new PTFE rotameters had not arrived yet prior to PP2R beginning. This stainless rotameter began showing signs of corrosion on the float ball almost immediately, such that the readings were not trustworthy. The float would not fall to the bottom of the tube when the rotameter valve was closed, and eventually stoppedresponding to any valve changes at all as time progressed. This is attributed to corrosion of the float, with rust causing enough swelling to prevent its proper movement in the tube. As such, the chlorine gas flows in PP2R were not flow controlled, and instead the valve was opened just enough to maintain a high (-1000 mV) ORP in the circuit.Table 9: PP2R Summarized MnR Circuit Physical Data

[0070] All three of the circuits in PP2R were sampled every four hours during the campaign, analyzing for nickel, manganese, and magnesium. The results are shown in Figure 9 [was 12], split by circuit for Ni / Mn and all three magnesium trends together for comparison purposes. In MHP, the nickel was initially high (>600 mg / L) and then was gradually lowered to -500 mg / L before sharply dropping to 100 mg / L just before the pH control was implemented. After switching to pH control (May 15 PM) the nickel still exhibited oscillations, but they were not as pronounced, as small changes in the pH controller setpoint were less drastic than changes to the rate of magnesium oxide pumping.

[0071] Both nickel and manganese were effectively removed in the SP circuit, aside from some higher values during the first few samples. Typically, the nickel tenor was 4- 8 mg / L, while manganese was even lower, often <1 mg / L. In MnR, both elements were low, with <5 mg / L Ni after the initial start-up, and <1 mg / L Mn (and many samples at <0.04 mg / L). The magnesium from circuit to circuit did not change substantially, as theamount of magnesium oxide added relative to the magnesium already present in the ON filtrate / wash solution was very small.

[0072] Profile sample analyses from the MHP circuit are shown in Figure 10 [was 13] and Figure 11 [was 14] for the solutions and solids, respectively. While all of the magnesium oxide was added in MHP1 , there is a slow and steady decline in both the nickel and manganese tenors in solution from MHP1 to MHP4, attributed to the slow reaction kinetics of magnesium oxide. The actual Ni / Mn concentrations were very dependent on the pH applied during the sampling time (20:00), accounting for the day- to-day differences for nickel to a degree, while the manganese appears to be more related to time (May 13 and 14 offering higher manganese in solution, while May 15 and 16 are both lower) and the level of oxidation in the batch of PP1 R solution in use at the time (and hours preceding). In terms of the precipitate grades, the nickel showed a gain in grade across the reactors on the first two sample sets, from 30% to 40% Ni, and once the seed recycle began the solids took on the composition of the original discharge solids from MHP4, hiding any gains due to the high seed recycle. The inverse is seen for magnesium, which showed a decline in precipitate grade with each reactor, and once the seed recycle started the trend flattened somewhat, but still showed some decrease between MHP1 and MHP4.

[0073] Figure 12 shows the solution assays from the SP circuit profile samples. While all showed a decreasing solution tenor with each reactor, the first sample set (May 14 08:00) showed the most drastic decline in falling from 242 mg / L Ni to 46.6 mg / L. The following profile sets only showed decreases of 5-10 mg / L across the circuit for both nickel and manganese.

[0074] All three circuits in PP2R used magnesium oxide from the same source; its analysis is shown in Table 10.Table 10: Analysis of Magnesium Oxide used in PP2R

[0075] The composition of the MHP feed solution throughout PP2R is shown inTable 11. The tenors of nickel, magnesium, and manganese were lower than that ofthe ON filtrate from PP1 R due to the wash water dilution (see Example 1), such that an average 2.87 g / L Ni was fed to MHP, along with 35.4 g / L Mg and 0.21 g / L Mn. The campaign began with the highest iron content solution to process it first during the startup and commissioning.Table'll : PP2R MHP Feed Solution Analyses

[0076] In addition to the feed solution, the seed recycle stream also represented an “in” to the MHP circuit, with analyses summarized below in Table 12. Operating between 20-28% w / w solids, the seed recycle precipitate contained an average of 43% Ni, 4.2% Mg, and 1 .8% Mn. Iron content was low at -0.22%, attributed to both the initial higher iron level in the MHP feed solution as well as the iron content in the magnesium oxide (0.29% Fe). The liquor portion of the seed recycle stream was rich in magnesium (38.2 g / L) and depleted in both nickel and manganese.Table12: PP2R MHP Underflow Recycle Pulp Analyses

[0077] The analyses of the discharge pulp from MHP4 are shown in Table 13, and were very similar to that of the seed recycle stream, at a high seed recycle ratio. Initially <1 % solids, the discharge slurry rose to 5.8% solids at the end of the campaign owing to the seed recycle. Nickel in solution averaged 506 mg / L, and 41.4% Ni in the precipitate, while magnesium averaged 36.8 g / L in solution and 4.62% in the solids. Manganese ranged from 1.5-2.1 % in the solids, and 30-160 mg / L in the solution.Table13: PP2R MHP Discharge Pulp Analyses

[0078] The calculated precipitation efficiency and magnesium oxide dose for each balancing period are shown in Figure 9 [was 16], Even though the reagent dose was decreased between May 14 AM and May 15 AM, the precipitation efficiency increased for both nickel and manganese, assumed to be related to the interplay between the first interval of MHP underflow recycling on May 14 AM. Figure 10 [was 17] shows the magnesium grades in the precipitate for each period along with the estimated seed recycle ratio. As the seed recycle increased, the magnesium content appeared to plateau and even potentially show signs of increasing again, though it is likely that not enough time was provided in PP2R to see the full effect (including regular removal of surplus MHP solids from the thickener for off-line filtration).

[0079] The analyses of the SP feed solution composite samples are in Table 14, showing similar levels of nickel, manganese and magnesium as the MHP discharge solution. The seed recycle stream for SP (beginning May 15 AM) ranged from 20-30%(averaging 22%) solids, which contained -25% Mg, 17% Ni, and 3% Mn, along with -3 mg / L Ni, 0.11 mg / L Mn, and -36 g / L Mg in the solution.Table14: PP2R SP Feed Solution AnalysesTable15: PP2R SP Underflow Recycle Pulp Analyses

[0080] The composition of the discharge from the SP circuit was similar to that of the recycle underflow, at an average of -0.9% solids as produced in the circuit. Even with most of the nickel precipitated in MHP, the SP solids still contained 16-30% Ni, as wellas 2-5% Mn and 10-25% Mg (Tables 15, 16). The solution had <10 mg / L Ni and <1 mg / L Mn for most of PP2R.Table16: PP2R SP Circuit Discharge Pulp Analyses

[0081] The composite feed solution analyses for the MnR circuit are shown in Table 17, and are very similar to that of the SP circuit discharge solution (averaging 6 mg / L Ni, 3 mg / L Mn, and ~34 g / L Mg). The level of manganese fell throughout the campaign, reflecting the changing level of oxidation in the MHP feed solution.Table17: PP2R MnR Feed Solution Analyses

[0082] The analyses of the discharge solution from the MnR circuit are summarized in Table 18; the circuit did not generate sufficient solids for analysis in the composite samples given the low manganese content in the feed. Similar to the rush / control assays, the MnR solution was low in both manganese and nickel, with an average of <0.2 mg / L Mn and 6.2 mg / L Ni, inflated by the higher values at the beginning of the campaign (later solutions were closer to 1-3 mg / L Ni and <0.04 mg / L Mn).Table18: PP2R MnR Circuit Discharge Solution Analyses

[0083] Overall, the PP2R campaign demonstrated under continuous conditions that magnesium oxide could replace the previously tested caustic for the precipitation of nickel and manganese and chlorine gas addition could replace the previously tested sodium hypochlorite (bleach) addition for manganese oxidation effectively removing sodium addition.

Claims

CLAIMS1 . A process for producing magnesium chloride and segregated metal values from a nickel saprolite mineral feedstock, comprising: a) comminuting the nickel saprolite mineral feedstock; b) leaching iron, aluminum, nickel, cobalt, manganese and magnesium metal values from the comminuted mineral feedstock with an HCI acid leachant in a leach slurry substantially free of sodium, to produce a leach slurry comprising a solid siliceous residue and a loaded leach solution; c) precipitating iron and aluminum from the loaded leach solution in the leach slurry with pH adjustment by addition of: a ground olivine mineral slurry, and a chlorine gas oxidant so as to continuously oxidize ferrous iron in the loaded leach solution to ferric iron, to produce an Fe / AI depleted solution and a solid amorphous silica residue; d) separating the Fe / AI depleted solution from the solid amorphous silica residue; e) precipitating nickel and cobalt from the Fe / AI depleted solution, wherein the precipitating is with addition of: a MgO slurry comprising Mg(OH)2, to produce a Ni / Co depleted solution and a first nickel and cobalt mixed hydroxide precipitate product; e’) optionally, further precipitating nickel and cobalt from the Ni / Co depleted solution, wherein the precipitating is with addition of: a further MgO slurry comprising Mg(OH)2, to reduce the concentration of nickel and cobalt in the Ni / Co depleted solution and to produce a secondary nickel and cobalt mixed hydroxide precipitate product;f) precipitating manganese from the Ni / Co depleted solution with addition of: a third MgO slurry comprising Mg(OH)2, in the presence ofCI2, to produce a MgCl2 product solution and a solid manganese dioxide residue precipitate product; g) separating the MgCl2 product solution from the solid manganese dioxide residue precipitate product; and, h) recycling a portion of the MgCl2 product solution as a recycle MgCl2 brine solution to step (a) of comminuting the nickel saprolite mineral feedstock.

2. The process of claim 1 , further comprising subjecting at least a portion of the MgCl2 product solution to heat to produce a concentrated MgCl2 solution product.

3. The process of claim 2, further comprising subjecting at least a portion of the concentrated MgCl2 solution to crystallization to crystalize a MgCl2.6H2O crystal product.

4. The process of claim 3, wherein crystallization comprises flaking or prilling and / or evaporation.

5. The process of claim 3 or 4, further comprising subject at least a portion of the MgCl2.6H2O crystal product to pyrohydrolysis to produce an MgO product.

6. The process of any one of claim 3-5, wherein at least a portion of the MgCl2.6H2O crystal product is subject to dehydration and molten salt electrolysis to produce magnesium metal.

7. The process of any one of claims 1 -6, wherein a portion of the MgCl2 product solution is subject to pyrohydrolysis to produce MgO and HCI.

8. The process of any one of claims 1-7, further comprising adding sodium hydroxide to at least a portion of the concentrated MgCl2 solution product to produce a Mg(OH)2 precipitate.

9. The process of any one of claims 1-8, wherein the ground olivine mineral slurry comprises an olivine mineral ground in the recycle MgCl2 brine solution.

10. The process of any one of claims 1-9, wherein precipitating iron and aluminum comprises addition of a stoichiometric excess of the ground olivine mineral slurry.

11. The process of any one of claims 1-10, wherein the olivine mineral comprises magnesium, nickel and cobalt, and magnesium, nickel and cobalt values are leached from the olivine mineral in the leach slurry.

12. The process of any one of claims 1-11 , wherein the concentration of iron in the HCI acid leachant is at least 15 g / L and the concentration of iron in the Fe / AI depleted solution is less than 10 mg / L.

13. The process of any one of claims 1-12, wherein the concentration of aluminum in the HCI acid leachant is at least 100 mg / L and the concentration of aluminum in the Fe / AI depleted solution is less than 10 mg / L.

14. The process of any one of claims 1-13, further comprising washing the amorphous silica residue.

15. The process of any one of claims 1 -14, wherein the solid amorphous silica residue comprises silica and iron and aluminum and chromium-containing solids.

16. The process of any one of claims 1-15, further comprising recycling a portion of the first and / or secondary nickel and cobalt mixed hydroxide precipitate product as a seed feed into step (e) and / or (e’) of precipitating nickel and cobalt from the Fe / AI solution.

17. The process of any one of claims 1-16, further comprising recycling the secondary nickel and cobalt mixed hydroxide precipitate product to step (e) of precipitating nickel and cobalt from the Fe / AI depleted solution and / or to step (b) leaching.

18. The process of any one of claims 1 -17, wherein a portion of the MgCI2 product solution is recycled to a step of milling of olivine to produce the ground olivine mineral slurry.

Images