Automated mineral mapping
in optical ore microscopy
Eric
PIRARD
Université de Liège GeMMe - Georesources & GeoImaging
Accuracy and limitations
EMC2012 Frankfurt
Automated Mineral Mapping
– Sampling
– Imaging
– Mineral Identification
EMC2012 FrankfurtAutomated Mineral Mapping
• Imaging
– Spatial Sampling
• A random point sampling gives an unbiased estimator
of the volume proportion of a phase (α) in a solid
1st Principle of Stereology
(Delesse,1848)
α
α
α
V
A
P
A
V
P
=
=
“Do more less well”
Systematic sampling on a “random” section
EMC2012 Frankfurt
Automated Mineral Mapping
• Sensing the mineral target
• Sensing principle
» Ex. X-ray fluorescence; light absorption; atomic force; …
• Signal processing and interpretation
» Database of spectra; Characteristic rays (Kα,…)
Goethite
EMC2012 Frankfurt
Automated Mineral Mapping
• Whiskbroom imaging mode
– Scanning beam
• or moving sample
Result of EDX mapping© QEM SCAN
Ex. Scanning Electron Microscopy
EMC2012 Frankfurt
Automated Mineral Mapping
• Pushbroom imaging mode
– Scanning linear sensor
Diffuse reflectance imaging
Specular reflectance imaging
Industrial vision of « marble » tiles
EMC2012 Frankfurt
Automated Mineral Mapping
• Array imaging mode
– CCD/CMOS camera
Reflected Light Microscopy
Typical Quantum Efficiency for an Si-detector
EMC2012 Frankfurt
Photonic Ore Microscopy
– Photons
– Filters
– Reflectance database
EMC2012 FrankfurtPhotonic Ore Microscopy
• MultiSpectral Imaging
• Conventional Ore Microscope
» Objective transmittance >50% @ 1100nm
• Scientific grade CCD camera
» Spectral sensitivity 350nm-1000nm
• Filter wheel
» Interference filters @ 50 nm spacing
Typical Quantum Efficiency Curve for an Si-detector
360 nm 438 nm 489 nm 591 nm 692 nm 870 nm Multispectral Image EMC2012 Frankfurt
Photonic Ore Microscopy
• MultiSpectral Imaging
• Calibration
» Correction for uneven illumination » Measure of reflectance standard
• Correlation with Specular Reflectance Database
» Quantitative Data File - QDFIII (Criddle & Stanley, 1993) » Extension to 1000 nm (Brea et al., IMA, 2010)
BORNITE 0 10 20 30 40 50 60 400 500 600 700 800 900 1000 Wavelength (nm) R e fl e c ta n c e ( % ) BN tfpe9 001_1 BN tfpe9 001_2 BN tfpe9 001_3 BN tfpe9 001_4 BORNITE QDFIII PYRITE 2000 2500 3000 3500 4000 4500 5000 5500 6000 400 500 600 700 800 900 1000 Wavelength R e fl e c ta n c e ( % ) Pyrite R tfpe9 011 - Pyrite PYRITE 1500 2000 2500 3000 3500 4000 400 500 600 700 800 900 1000 tfpe9 001 - Chalcocite Chalcocite COM CHALCOCITECOVELITE 0 1000 2000 3000 4000 5000 6000 7000 400 500 600 700 800 900 1000 Wavelength (nm) R e fl e c ta n c e ( % x 1 0 0 ) COV Ro (116) COV Re (116) Covelite COVELITE CR 286 15,00 20,00 25,00 30,00 35,00 40,00 45,00 50,00 55,00 400 500 600 700 800 900 1000 Wavelength (nm) R e fl e c ta n c e ( % ) PO (CR286_012) PO (CR286_014) PO (CR286_020) CUB - (CR286_019) CUB - (CR286_019) CP (CR286_001) CP (CR286_007)
Enhanced Pyrrhotite-Cubanite discrimination
EMC2012 Frankfurt
Photonic Ore Microscopy
• MultiRadial Imaging
• Rotating incident polarizer
» “Information” about crystal anisotropy
{ }
θ{ }
θ y x y xmin
P
P
Max
B
=
,−
, Multiradial Image “Bireflectance” Image EMC2012 FrankfurtApplications
– Epithermal Cu-Au
– Stratiform Cu
– Carbonate rocks
EMC2012 FrankfurtWavelength Selection
• Phalaborwa : Cu-Ni Sulphides
» @ 437nm; 489nm; 591nm; 692nm (10nm FWHM)
Criddle & Stanley QDFIII, 1993
Pentlandite Chalcopyrite Violarite Pyrrhotite Cubanite Chalcocite Covellite Bornite Magnetite Califice A., 2008 EMC2012 Frankfurt
Post Processing
• Chelopech : Epithermal Cu-Au Paragenesis
Pyrite
Resin
False Colour Image Maximum Likelihood Classification
After Conditional Geodesic Propagation
EMC2012 Frankfurt
Modal Analysis
• Kansanshi : Stratiform Cu
• Supergene (mixed) zone
» Secondary Cu sulphides, » Malachite,… EMC2012 Frankfurt Siebels K., 2012 Chalcopyrite (A) Chalcopyrite (B) Copper Cuprite Digenite Malachite Chalcocite Rutile Molybdenite Pyrite 489nm, 590nm, 690nmModal Analysis
• Kansanshi : Stratiform Cu
• Supergene (mixed) zone
» Secondary Cu sulphides, » Malachite,…Curvas de Liberación Ccp RoCo
0 10 20 30 40 50 60 70 80 90 100 0 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-99 >99 % Liberación % C c p
Lib Vol Ccp MIXTAS Lib Sup Ccp MIXTAS Lib Vol Ccp TOTAL Lib Sup Ccp TOTAL
Cp1 Cp2 Cc Mal Py Cp 54,2% Ss-Cu 11,1% Ox-Cu 3,0% Hem 1,7% S-Fe 25,1% Gg 4.6% Other 0,3% EMC2012 Frankfurt Dufrasne Fl., 2010; Perez-Barnuevo L., 2011
Microstructure Analysis
• Carbonate Rock
EMC2012 Frankfurt Jaimes Contreras R., 2011 Multiradial Image PorosityMicrostructure Analysis
• Carbonate Rock
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 L o n g u e u r (µ m ) P90 P50 P10Percentiles of Intercept Length (Crystal Size Distribution)
EMC2012 Frankfurt
Accuracy & Limitations
– Visual check
– Chemistry, XRD
– Comparison with SEM
– Round Robin
EMC2012 Frankfurt
Accuracy & Limitations
• Validation
– Visual
» Point counting; Time Consuming; Subjective;…
– Chemistry
» Balance; Limited mineralogy; …
– XRD
» Major minerals;…
– Round Robin Test
» Interlaboratory test on « similar samples » » Detailed statistics of inter/intra variability » Initiative of IMA-CAM
EMC2012 Frankfurt
Accuracy & Limitations
• Comparison with Automated EM-based Mineralogy
– QEM-SCAN/MLA
• High resolution (PGE, Au)
• Non-stoechiometric minerals (Lcx)
• Trace / Precious elements partitioning
• Gangue mineralogy
• Process mineralogy oriented software
– Optics
• Cheap technology
• Fast imaging
• Large samples (
do more less well!)
• Good discrimination in some critical ores
» Iron oxide; Ni-Cu sulfides; …• Multiradial imaging
» Grain size analysis» Crystal orientation (EBSD for the poorest)
EMC2012 Frankfurt
EMERALD Erasmus Mundus
• European Master Degree in Georesources Engineering
• Innovative education in « Geometallurgy »
» Mineral Resources Characterization – Processing – Modelling - Management
• Worldwide network of associated universities
» Moscow, Queensland, Capetown, Hacceteppe, Minas Gerais, UChile, Kazakhstan,…
