The oxidized ores of the Main Sulfide Zone, Great Dyke, Zimbabwe: Turning resources into minable reserves – mineralogy is the key ! Thomas Oberthür Frank Melcher, Malte Junge, Marek Locmelis, Peter Buchholz, Herwig Marbler, Dennis Kraemer, Gila Merschel, and Michael Bau Outline of Talk The Great Dyke of Zimbabwe Main Sulfide Zone, Great Dyke Pristine ores - Geochemistry - Mineralogy Oxidized ores - Geochemistry - Mineralogy - Metallurgical options 4 Hartley Mine Ngezi Mine Unki 100 km Mimosa Mine Mafic Sequence Ultramafic Sequence 1 Stratigraphy of the Great Dyke upper Mafic Sequence, + 1150 m middle lower Gabbros, Norites, Gabbronorites MSZ Bronzitite Ultramafic Sequence, + 2300 m succession Dunites/Serpentinites, Harzburgites, Olivine-bronzitites, Pyroxenites, Dunite Websterite succession Chromitites 1 Great Dyke, Zimbabwe MSZ 1 Pt Sulfide, Pd metal subzone Base Ni, Cu PGE subzone PGE Vertical zonation pyroxenite or “offsets“ Main Sulfide Zone, Mimosa Mine, Great Dyke (Prendergast; 3 Wilson) cpy po moncheite sperrylite cpy Mimosa MIM-603 Hartley, SD-04 Hartley Mine GD-11-D MoS2 moncheite cooperite pn Platinum-group (contains up to 3000 ppm Pd !) minerals po (PGM) cpy 50 µm Pt-Fe alloy PtS Grains from concentrates (electric pulse, Univ. Leoben) PtAs2 (Pt,Pd)(Te,Bi)2 PGM - Main Sulfide Zone • Cooperite - Braggite • Moncheite • Maslovite • Sperrylite • Merenskyite • Michenerite • Rh-Ir-Ru-Pt-sulfarsenides • Kotulskite • Laurite • Pt-Fe-alloy • "PtSnS" 4 ~ 70 % are Pt-minerals ! PGM by dominant PGE (%) Pt 67.7 Ir Ru Rh Pd 2 3.7 19.8 6.6 n = 781 2 Regional variation – PGM proportions – Great Dyke Bi,Te > As Hartley (Pt,Pd)-bismuthotellurides Mhondoro Pt-Fe alloys Ngezi Unki others Mimosa PtAs2 (PGE)-sulfarsenides As > Bi,Te Pd in pentlandite (ppm) example : Unki 1 4 7 10 Pt peak = BMS-1 13 16 19 Pd peak 310 cm 22 25 28 Base of MSZ 31 0 500 1000 1500 2000 R-13 ~ 80 % of the Pd is hosted in pentlandite ! 1 Distribution of PGE PtFe Laurite [RuS2] PGE are bimodally distributed – Pt, Os, Ru: discrete PGM – Pd, Rh, Ir: in sulfides (mainly pentlandite) 13 PGM / PGE in the exogenic cycle MSZ MSZ PGM pristine oxidized (Pt,Pd)(Bi,Te)* 30-70% PGE oxides + PtAs2 ~ 25 % (Pt,Pd,Ni)S ? Pd in pentlandite ?? ~ 80% of Pd total Oxide ores • Great Dyke ~ 250 Mt of resources ! • Bushveld some 100 Mt (Platreef, MR, UG-2) • Characterization of the mineralogy of the PGE • Development of new processing technologies (“conventional“ PGE recovery only ≤ 30% ! ) 3 Oxide ores, Hartley Platinum Mine Resources: ~ 250 Mt of ore ! 1 Hartley Mine, Zimbabwe 1 Ngezi test pit Sampling “Old Wedza“ (Mimosa) MIMOSA (MIM 08-13) Martin Prendergast‘s outcrop ppm / ppb 0 1000 2000 3000 4000 5000 Cu Ni Au Pd m 0.75 Pt 1 Oxide Ores • Geology • Geochemistry • Mineralogy NGEZI (NRC146) ppm / ppb ppm / ppb 0 1000 2000 3000 4000 5000 1000 2000 3000 4000 5000 6000 1 Ni 2 Cu 3 Pt Pd Cu 4 Au Ni 5 1,45 m 1,45 Pt 8.00 m 6 Au Pd 7 8 9 Oxidized MSZ Oxidized MSZ Hartley Mine (HOP-20x) Ngezi (NRC 146) 1 Pt Pt/Pd = 2.43 (n=9) oxide MSZ Pt/Pd = 1.28 (n=12) sulfide Pd Au Geochemical trends: Sulfide ores oxide ores 1 234 100 gain Cu 50 Au V Ti Ni Fe Co Zn Ru Pt 0 Al LOI Mn Ir Si Rh Mg Ca Cr (Os) K Pd relative change [%]relative -50 Na -100 S loss MSZ: gains/losses during weathering HOP-20x vs. CD-02+SD-04 After Gresens (1967). Al = constant 2 Oxide Ores • Geology • Geochemistry • Mineralogy Sample Depth Degree of PtAs2 (Pt,Pd) Pt-Fe (Pt,Pd) PGM Oxide minerals Sulfides metres oxidation S alloys (Bi,Te) MHR 4-4.5 1 1 - - 2 Goethite, chromite None 194 MHR 6.5-7 - - 5 - 5 Goethite, hematite, None 195 magnetite, chromite MHR 10-10.5 9 - 1 1 11 Goethite, hematite, Sparse relicts 196 magnetite, rutile in silicates MHR 15-15.5 4 12 - 28 44 Magnetite, Po, pn, cpy, py. 197 chromite Incipient oxidation at rims 2 Sperrylite SEM image Relict PGM Sperrylite BSE image polished section 50 µm smec HOP-102 HOP-103 goe 50 µm Cooperite SEM image opx cpx Relict sulfides talc pn Relict pentlandite: 127 ppm Pd up to 6500 ppm Pd up to 450 ppm Pt cpy 1 HOP-206 Replacements coop mich mer Replacement of (Pt,Pd)(Bi,Te)* HOP-206 (5910a) Ngezi Adit A (5711b) PGE - “Oxides“ Ngezi Mine Complete replacements (or neoformations ?) Ngezi Adit A (both 5711b) talc cpx opx mag Fe- serp mag Pd-S FeOOH Pt-S + 100 um 50 um PGE in FeOOH, Mn-Co-OH, smectites (Pt, Pd) Sn opx amph 3 ca. 200 ppm Pt Mn-Co-Ni-OH Pd-Pt smectite Pt Ni-Cu-cl 20 um 20 um Mimosa Mimosa Mn-Ni-Co-OHx = 200-400 ppm Pt ! FeOOH Mn-Ni-Co-OH Ni-chl 50-80 ppm Pd Opx (Pd,Pt) <1 µm Pt(Fe?) <1 µm FeOOH FeOOH = (<15 ppm Pd, Pt) 50-80 ppm Pd ! Pt(Fe?) <1 µm PGE in FeOOH and in Mn-Co-OH Ptx-neoformation ? 2 Pt and Pd in Mn-Co and Fe-hydroxides Hydroxides Pt-bearing 100000 Fe hydroxides Mimosa 10000 Mn-Co hydroxides MIM-11 Hartley MIM-12 HOP-206 HOP-207 1000 HOP-209 Pt Pt ppm MIM-11 Mn Mn-Co hydroxides Mimosa HOP-207 Mn HOP-209 Mn 100 MIM-10 Mimosa Hartley 10 10 100 1000 10000 Pd ppm Mimosa and Hartley Mineralogy of PGE/PGM in oxide ores 1. Relict PGM 2. Relict sulfides 3. PGE-oxides/hydroxides replacing PGM 4. Ptx (neoformation) 5. PGE in FeOOH 6. PGE in Mn-Co-OH 7. PGE in silicates (smectites) Mineralogical balance ? Pt distribution MSZ sulfide ores oxidized ores bimodal polymodal 1) discrete PGM ~ 25% PtAs2 & (Pt,Pd,Ni)S (Pt,Pd)(Bi,Te)* PGE - “oxides“ > 95% Ptx (neoformation) Pt > Pd in FeOOH Pt > Pd in Mn-Co-OH 2) Pt in pn and py Pd > Pt in smectites < 5% ~ 75% 3 Pd distribution MSZ sulfide ores oxidized ores bimodal polymodal 1) discrete PGM (Pd,Pt,Ni)S ~ 10% (Pd,Pt)(Bi,Te)* ~20% PGE - “oxides“ Pt > Pd in FeOOH Pt > Pd in Mn-Co-OH 2) Pd in pentlandite Pd > Pt in smectites ~80% ~ 40% ca. – 50% 4 PGM / PGE in the exogenic cycle MSZ MSZ PGM pristine oxidized (Pt,Pd)(Bi,Te)* 30-70% PGE oxides + PtAs2 ~ 25 % (Pt,Pd,Ni)S ? Pd in pentlandite ?? ~ 80% of Pd total 1 Metallurgical options for oxide ores Method Recovery • Grainger Bros. (1926) Bulk - Gravitation << 50% (?) • Hartley Mine (1997-99) Bulk - Flotation 30 % (?) • TML Process BHP – lab scale Bulk – Roasting (300-425°C) (1992/1994) Br° in NaBr in H2SO4 Pt = 85 % leach (70°C, 2h) Rh = 65 % CIP Au = 95 % problems: costs of chemicals; aggressive acids; environmental issues • Evans Bulk – various/hydrometallurgy ? (2002) – pyrometallurgy (ConRoast) ? 3 Metallurgical options for oxide ores BGR – (1) Establish mineralogical balance of PGE in the ores – (2) Metall. tests → mechanical pre-concentration (?) Step-wise crushing & milling Electric pulse disintegration Sieving Gravity separation Magnetic separation Preconcentration Selective leach 1 Metallurgical options for oxide ores BGR – previous work Preconcentration Selective leach NO Current work: Total leach experiments (JU Bremen) - hydrometallurgy - (1) one-step leach using siderophores: < 1% Pt recovery (2) two-step leach, inorganic acids + siderophores, unbuffered: 20 – 40 % Pt recovery (3) Bio-leaching (ongoing work, BGR labs) (4) Inorganic acids – ongoing work – BGR/DERA + JU Bremen Conventional processing of oxidized PGE ores problematic due to: missing BM sulfide association (Oberthür et al., 2013) Occurrence of naturally-floating Concentrate dilution; gangue (Becker et al. 2014) 15-30% Recovery Polymodal distribution of PGE (Oberthür et al., 2013) Therefore Hydrometallurgical leaching experiments with organic and inorganic lixiviants (Jacobs University Bremen, Germany) Leaching experiments with siderophores (biologically produced organic ligands) Siderophores / Metallophores: • low molecular weight organic molecules Structure of desferrioxamine B (DFO-B) • exuded by plants (phytosiderophores) and Yoshida et al. (2004) bacteria (microbial siderophores) to cope with nutrient deficiency or toxicity in natural systems • In modern environments mainly produced to bind and mobilize Fe(III), which is needed as a micronutrient But: High affinity for platinum and palladium Holmes et al. (2005) Leaching experiments with siderophores (biologically produced organic ligands) Oxidized MSZ Oxidized Platreef Great Dyke Bushveld 100 Oxidized MSZ (Great Dyke): 80 • Pt recovery up to ~80% • Selective leaching possible 60 • Low Pd and Ni mobilization 40 Oxidized Platreef (Bushveld): leached[%] • Very low recoveries 20 irrespective of leaching time, reagent concentrations, preleach, etc. 0 Pd Pt Ni Pd Pt Ni [%] leached by hydrochloric acid pretreatment (24h) [%] leached by siderophores (48h) [%] remaining in solid residue Siderophore experiments demonstrate principal viability of bioleaching ! (leaching with biogenic compounds) ! for the treatment of oxidized PGE ores Leaching experiments with inorganic lixiviants Oxidized MSZ Oxidized Platreef Great Dyke Bushveld 2B-Platinum Leach: 100 • Leach based on inexpensive inorganic reagents 80 • Initial lab-scale tests indicated very 60 promising leaching efficiencies 40 [%] leached [%] 20 0 Pd Pt Ni Pd Pt Ni [%] remaining in solid residue [%] leached by 2B-platinum leach (60°C; 1h) Test of applicability in optimization and upscaling experiments Samples totalling 150 kg of oxidized PGE ore Heap ore from the Mogalakwena Mine, South Africa Heap of oxidized PGE http://southafrica.angloamerican.com ore: 6 mio tonnes Nickel, Cr, Au, Pt and Pd grades of the oxidized PGE ore samples used for Samples were only optimization and upscaling experiments crushed, Pt Pd Ni Cu Cr [µg/kg] [µg/kg] [mg/kg] [mg/kg] [mg/kg] one sample was milled Average (n=6) 1184 924 2050 2088 798 First results – upscaling experiments 2B-Platinum leach: Upscaling experiments conducted on oxidized heap ore • Experiments at 20°C, 40°C and 80°C • Time series: samples taken after 1h, 2h, 4h, 8h • Upscaling experiments: Up to 78% Pt and 55% Pd into solution Extraction of PGE from oxide ores by hydrometallugical methods Advantages: 1.