Application of principal component analysisand tof-sims to mineral recognition, surface

chemistry and separation by flotation

ROGER ST. C. SMART1, BRIAN HART

2, MARK BIESINGER2,

JAMES FRANCIS2, TESFAYE NEGERI

3

1ACeSSS, University of South Australia, Mawson Lakes, SA

5095, Australia (Roger.Smart@unisa.edu.au)2Surface Science Western, University of Western Ontario,

London, ON, Canada N6A 5B73Natural Resources Canada, 555 Booth Street, Ottawa, ON,

Canada K1A 0G1

In surface analysis of mineral particles from ores, process andwaste streams, different mineral phases are normally identified byimaging for a major element of their composition (e.g., Cu, Fe,and Zn) but this selection can be difficult with many multi-metalminerals (e.g., chalcopyrite, pyrite, and sphalerite) and with pre-cipitated, adsorbed, reacted and contaminant species in the outer-most molecular layers. Hence, a challenge in studying surfacechemistry and reactions of specific minerals is to find more reli-able methods of mineral phase recognition in these complex sur-face chemistries. This paper will show that principal componentanalysis (PCA) provides improvement in phase recognition, parti-cle selection, surface chemical changes and associated surfacespecies.

Statistical methods based on the monolayer-sensitive time offlight secondary ion mass spectrometry (ToF-SIMS) techniquehave been utilized to differentiate surface chemical factors pro-moting (hydrophobic species) or inhibiting (hydrophilic species)flotation Species recognition based on the PCA method clearlyidentified a statistical difference in copper intensities between thesphalerite and pyrite phases (Hart et al., 2006).The method hasalso been applied to concentrate and tails samples collected fromthe Inco Matte Concentrator demonstrating extensive Cu and Nitransfer between chalcocite and heazlewoodite minerals.

PCA of ToF-SIMS imaging data has elucidated surface chem-ical factors that differ between samples which have undergonehigh intensity (shear) conditioning (HIC) and those which havenot in parallel laboratory floats where sphalerite was separatedfrom a complex sulphide ore which includes pyrite other gangueminerals. A 5% increase in the overall Zn recovery was reportedin the HIC test. The statistical analysis has elucidated differencesin surface chemistries which illustrate the discriminating depres-sant action of adherent aluminosilicate (gangue) fine particlesand adsorbed ions on the surface of sphalerite grains as well assignificant transfer of Cu and Zn to pyrite surfaces in the non-HIC samples resulting in collector adsorption and inadvertentflotation. In the flotation test concentrates, HIC conditioningresulted in the removal of these gangue fines improving surfacecollector attachment and sphalerite recovery.

This method can be applied to core and crushed samples forinformation on relative reactivity and adsorption.

ReferenceHart, B., Biesinger, M.C., Smart, R.St.C., 2006. Min. Eng. 19, 790–798.

doi:10.1016/j.gca.2006.06.1108

The reactivity of uranyl phosphate mineralphases via microbial weathering: Short vs.

long term kinetics

C. SMEATON, C. WEISENER, D. FOWLE

GLIER, University of Windsor, Windsor, Ontario, Canada

(smeato3@uwindsor.ca; weisener@uwindsor.ca)

HypothesisThe mobility of uranium will depend strongly on its oxidation

state, with U(IV) species being less soluble than U(VI) speciesunder circumneutral to alkaline abiotic conditions. Under theseconditions, Uranyl species are often complexed by available car-bonate and phosphate leading to the precipitation of uranyl-phos-phate minerals. The formation insoluble phosphates is veryfavorable in terms of the natural attenuation of uranium. Howev-er, the reactivity and hence stability of uranyl phosphates in thepresence of a natural microbial consortia has yet to be deter-mined. By measuring the rates of microbial corrosion of naturaland synthetic uranyl-phosphate minerals and comparing quantifi-able mineral properties (e.g., solubility and composition) to twomicroorganisms (i.e., aerobic vs. anaerobic metabolism), it willbe possible to assess the corrosion potential of Uranyl U(VI)phosphate phases in natural settings.

ResultsPreliminary data for the corrosion of the autunite mineral

phase suggests that incongruent dissolution is occurring in thepresence of Shewanella putrefaciens. EPS and bacteria are shownsurrounding the autunite with clear evidence of etch features asso-ciated with the uranyl phosphate lathes. Furthermore, solutiondata reveals that uranium and phosphorous is released over timein the presence of S. putrefaciens(see Figure).

ConclusionsA comparison of short term kinetics will be presented for syn-

thetic meta-autunite (Ca[(UO2)(PO4)]2(H2O)6) and natural tor-bernite (Cu[(UO2)2(PO4)2] (H2O)10).

doi:10.1016/j.gca.2006.06.1109

Goldschmidt Conference Abstracts 2006 A597