Monday, 3 December 2018

Process Mineralogy '18: conference diary

Process Mineralogy '18 was MEI's 5th conference in this series, all of which have been held at Cape Town's Vineyard Hotel. The 3 day event was held between 19th and 21st November 2018.

Vineyard Hotel Conference Centre
Monday 19th November
A beautiful day in Cape Town, and this morning MEI's Jon Wills opened the conference, and welcomed the 85 delegates from 16 countries to the Vineyard Hotel,  thanking our sponsors Zeiss, ThermoFisher Scientific, Bruker, Petrolab and our media partners International Mining and Mining Review Journal Zambia.

Following Jon's introduction, our consultant, Megan Becker, of the University of Cape Town, spoke movingly of her great friend and mentor, Prof Dee Bradshaw, who was well known to all in the process mineralogy field, and who died in June after a long battle with cancer.

Our first keynote speaker was Alan Butcher, of the Geological Survey of Finland, who presented "When Scientists and Engineers Talk – Lessons from the Oil Industry and Applications to Mining".

The mineralogy of a rock (the individual constituents defined by their unique combinations of chemistry and crystallography) and the texture of a rock (the way in which the minerals are arranged geospatially, particularly with respect to each other, voids, pores and fractures), together largely control, or influence in some way, every major geological, petrophysical and engineering attribute relevant to the petroleum industry, whether it be density (sonic), hardness and brittleness (frackability), electrical conductivity (resistivity), radioactivity (gamma ray), rock type and lithotype (sedimentary, igneous or metamorphic components).

Can the same be said for mining? Almost certainly yes, according to Alan, who showed how the application of mineral and petrographic information is often quite different, as unlike the oil industry - where the commodities (oil and/or gas) are not measured directly, only the areas in the rock where they reside (mostly pores) - in mining, we investigate the nature of the commodity directly (ore-forming minerals).

This leads us to tackle similar challenges but from completely different perspectives. In mineral processing, we are typically concerned with predicting, measuring, monitoring and improving the behaviour of the (mostly valuable) components as they pass through the drilling, blasting, comminution and separation stages of the mining cycle. There is no question that an ore’s performance is strongly controlled and influenced by its original starting mineralogy and texture.

Alan reviewed which minerals are most likely to cause significant production problems in both worlds. He compared the analytical work flows used to characterize the so-called menace minerals, and suggested what learnings can be made. For example, knowing that one has the right type of pore-lining clay in a reservoir rock is critical to a petroleum engineer; likewise, being made aware that clay is present in an ore before it is processed (rather than after) is very helpful to a minerals engineer. The two industries typically approach characterizing clays from completely different angles, partly because the applications are different, but also the budgets and risk factors are not on the same scale.

But that aside, given that the minerals in each case may be similar (or virtually identical), and that it is applied mineralogists (scientists) who document and deliver the information to the engineers, it is important that all professionals involved in the greater natural resources community share their best practices across what are traditionally separate industrial segments, and that innovations are openly discussed so they can be adopted, if found to be useful and Alan's presentation was a major contribution to popularizing such a cross-disciplinary approach between the oil and gas and mining industries.

Alan Butcher (right), with Willem Swart of Advanced Laboratory Solutions,
a leading supplier of laboratory and scientific instrumentation in South Africa

After this fine introduction to the conference, Keith Taylor, of Oxford Instruments NanoAnalysis, UK, showed how, in order to understand the efficacy and efficiency of a mineral processing operation, it is of great importance that accurate, reliable data is acquired on the particle/grain morphologies and the compositions of the phases which are found.  This data must be acquired in a timely manner in order to enable quick feedback to be provided on the processing operation.

Advances in Energy Dispersive X-ray Spectrometry (EDS) Detectors used on Scanning Electron Microscopes (SEMs) mean that it is now possible to perform this analysis at the highest speeds whilst gathering a sufficiently high level of counts to confidently characterise both major and minor elements during the automated mineralogy process.  Matt reviewed the latest generation of hardware and how it can be applied to best meet the rigorous demands of automated mineralogy, using industrial and academic examples.

Keith Taylor and Matt Hiscock, of Oxford Instruments, talking to MEI's Jon Wills

Marcelene Voigt, of the University of Cape Town, South Africa presented an investigation into the the robustness of using the X-ray computed tomography (XCT) and the grey level co-occurrence matrices (GLCM) method for the quantification of mineral texture in 3D.

This can be achieved by testing the quality of the data that the GLCM outputs (statistics and heat maps) provides in response to changes in XCT conditions (e.g. artefacts, resolution, dual energy, calibration), as well as its response to different mineral texture types (e.g. anisotropic features). She showed how these findings will lead to a better understanding of how the GLCM captures the mineral texture information and assess the best conditions for the XCT analysis. It will further establish the versatility of the method to other mineral texture types and the factors affecting the underlying generic elements (e.g. grain size and mineralogy).

Marcelene's paper is part of the on-going research to develop a 3D mineral texture quantification method to be used in geometallurgy.

During the first poster session
After an extended coffee break, giving everyone time to view the posters and mix with the exhibitors, Lunga Bam, of Stellenbosch University, South Africa showed how X-ray computed tomography (XCT) is increasingly used to visualize and quantify ore characteristics in 3D for a variety of minerals processing applications.

The success of the technique relies on effective X-ray penetration which is most easily obtained for small, low density samples. However, it has been highlighted that sample size not only determines image quality but also whether the volume is an accurate representation of the ore characteristics.

Linked to this, is the challenge of differentiating minerals with insufficient density differences, e.g. chalcopyrite and pyrite. These issues are magnified in high density ore samples. Lunga described recent work that has developed protocols for XCT scanning of such samples.

Measurements of liberation are typically undertaken on materials at sizes where, in most instances, the inherent textural features of the ore have been destroyed. Understanding the contribution of textural characteristics to the evolution of mineral liberation can therefore be challenging.

Elaine Wightman, of Australia's JKMRC showed how, in previous work a method had been developed to measure and interpret the textural features of HQ core that contained both vein structures and disseminated grains.  Using random masking to simulate breakage at different sizes, it was shown that the presence of vein structures lead to liberation of sulphide minerals at coarser particle sizes.  An experimental program was then undertaken to break the core sections and measure the progeny particles created using a combination of areal (2D) and volumetric (3D) measurements.

Elaine described how data from these measurements show evidence of non-random breakage occurring and provides insights into which meso-scale features of the ore are the contributing factors.

Elaine Wightman (left) with Gaynor Yorath, Desh Chetty, Megan Becker, Solly Theron and Pieter Botha

Comminution indices are used to represent the grindability of a certain minerals combination in a specific process. An analogous behavior for the same comminution process is expected from two rocks that have similar index values. However, as shown by Leandro Voisin, of the University of Chile, indices are obtained assuming the breakage of homogeneous and constant materials without considering mineralogical and textural properties at any scale, so eventual differences in the process may not be explained by comminution indices.

Leando described how two different copper ore samples with similar Bond Work index were used to evaluate the influence of mineralogy and texture at meso and micro-scale at laboratory scale using the Magotteaux Ball Mill®, controlling physical and chemical grinding parameters. The effect on grinding was assessed through kinetic experiments and the results confirmed that rock properties at minor scales may affect grinding performance.

Humidity cell tests (HCTs) are a routine analytical tool for assessing the long-term variation in sulfide oxidation kinetics within potential mine waste. The need to understand the role that mineral texture plays in controlling HCT behaviour is integral, and analytical protocols for this are the subject of research by conference sponsor Petrolab Ltd, UK.

James Strongman presented a general protocol, from sample preparation, through to analysis methods and interpretation that has been developed from the analysis of HCT feed samples from Savage River, Australia and an IOCG deposit in Scandinavia.

From this study, and using process mineralogy tools, quantified textural parameters are calculated for the balance of acid-generating minerals and acid-consuming minerals along with the implications this carries for true acid-generation potential. Quantifying the textural controls in this manner are a powerful tool for understanding mineral textures and provide the overseeing geochemist with sufficient information and confidence to interpret current and likely future HCT behaviour.

James Strongman, of Petrolab, with Tata Steel's Sunil Tripathy and Gajanan Kapure

Image analysis aiming at process design or control might be based on several types of microscopy, each providing advantages and limitations. Automated mineralogy systems based on SEM are dominant for ore characterization, but, as discussed by Reiner Neumann, of CETEM, Brazil, in the final paper of the morning, there are relevant cases where such systems cannot perform the required characterization, such as with iron ores and complex multi-metallic rare-earth ores.

The characterization of iron ores is typically performed using reflected light microscopy. Optical image analysis systems can recognize most common iron-bearing minerals, but the discrimination between quartz, other transparent gangue minerals and the embedding resin is very tough.

Reiner presented a correlative microscopy setup that combines reflected light and cathodoluminescence. The correlated use of these signals provides contrast to discriminate resin from quartz and allows the identification of several transparent minerals, besides classifying iron minerals by mineralogy and texture.

Reiner Neumann (left) with Zeiss's Shaun Graham, Allister McBride and Ben Tordoff,
and XPS-Glencore's Michelle Kelvin and Elizabeth Whiteman

A long break for lunch

The classification of phases measured by SEM-EDS automated mineralogy often relies on user-defined mineral chemical criteria. The design of such classification rules can be complicated and time-consuming in addition to being prone to errors due to lack of user experience or knowledge of complex mineral occurrences.

Pieter Botha, of Hippo Geoscience, Australia, described the development of a set of analysis tools that would augment the design of phase classification criteria for SEM-EDS automated mineralogy data. Mathematical methods were applied to a set of raw energy dispersive x-ray spectra to identify the main chemical clusters. These clusters were then used to define initial classification rules and input spectra for phase chemical quantification. Some challenging aspects in automated mineralogy, for example, mixed spectra deconvolution, polished surface imperfections, and grain size limitations were also addressed.

Micro-XRF technology in automated mineralogy has created new opportunities to analyse large samples of rock and drill core with minimal sample preparation, as well as gaining insights from traditional polished sections or briquettes.  The strength of micro-XRF above other techniques is in identifying trace elements.

However, micro-XRF has had shortcomings, for example detecting super light elements (SLE) with low atomic values below sodium (Z < Na).  Consequently, minerals such as calcite and fluorite that can be differentiated only by the fluorine (F) and oxygen (O) content cannot be discriminated from each other.

Samuel Scheller, of Bruker Nano GmbH, Germany, introduced the new M4 TORANDOPLUS, which uses SLE windows and x-ray tube, providing the capability to detect light elements and therefore distinguish carbonates from fluorites.  Samuel discussed these technological advancements and new capabilities using real-world applications in mineralogy in combination with the automated mineralogy software AMICS.

Samuel Scheller (2nd left) with Andrew Menzies, Mark Pownceby, Frances Wall and Joerg Blieffert

Automated mineralogy at its core is a computerised point-counting technique that removes operator bias when determining bulk mineralogy and derived information. However, vendor implementation details can introduce unintended bias unless operators spend significant time to detect and resolve identification artefacts.

Michael Owen, of Thermofisher Scientific, USA, proposed a new approach that directly models mineral complexity and SEM-EDS interactions. It automates mineral deconvolution quantification for boundary phases and tiny inclusions; allows full mineral solid solutions; and elemental impurity substitutions in mineral definitions. This reduces operator tuning for mineral ID significantly when compared with existing commercial products such as QEMSCAN or MLA.

Michael Owen (centre) with Gaynor Yorath and Philippe Giaro

According to Ivan Fernandes, of Helmholtz Institute Freiberg for Resource Technology, Germany, the minerals industry needs a quick, reliable and powerful predictive tool to perform techno-economic and sustainability evaluation of the raw materials value chain, being able to compare different process flowsheets and adapt to the ever increasing ore variability and complexity.

He described a study demonstrating how the integration of quantitative mineralogical data coupled with simulation tools (HSC Sim) and LCA software (GaBi), are used to truly develop a predictive geometallurgical tool. An MLA-based process simulation takes into account a range of geometallurgical variables, such as sized data, mineral locking, liberation, mineral chemistry and shape to predict metallurgical responses in energy/water/resource efficiency, energy consumption, mass and energy balance, besides the typical grade and recovery.

In addition, this approach creates a direct link from automated mineralogy data to the prediction of environmental footprint of a system, from mineral processing to metallurgy, giving sustainability thinking an important role in the decision-making process.

Ivan Fernandes with Anna Vanderbruggen

The polymetallic Cu-Pb-Zn-Ag Swartberg deposit in South Africa hosts three distinct ‘early stage’ geometallurgical domains across five lithostratigraphic ore types.  Variability within each domain is expressed as noticeable differences and similarities in bulk mineralogy, grain size distribution, gangue mineral association and textural orientation.

Henry Gordon, of Stellenbosch University, South Africa,  assessed the influence of such characteristics on the liberation, recovery, selectivity, throughput and grade of copper minerals within each geometallurgical domain when subjected to flotation in a size by size context. The results have implications for improving the classification of mineralogy-based domains as a function of distinct responses during flotation testing.

Pyrometallurgical processing of polymetallic ores from the Zeehan Field, Western Tasmania, occurred intermittently between 1896 and 1948, primarily focussed on recovering Pb, Ag and Cu. Zinc was not recovered and reported to the slag which was disposed of in two piles (North and South) located near Zeehan.

Anita Parbhakar-Fox, of the University of Tasmania, Australia, described how geometallurgical and geoenvironmental characterisation of slag samples was carried out using XRD, SEM, LA-ICPMS, XRF and static testing.

Samples from the North and South piles contain on average 12% and 18% Zn respectively by bulk analysis. The micro-texture of the slag is highly heterogeneous and includes distinctive dendritic wurtzite and sphalerite, Zn-rich glass phases (containing up to 34 wt.% Zn) and Zn-bearing olivine and pyroxene. Elevated levels of As, Cd, Co, Cu, Pb and Sb within the slag and geoenvironmental assessments indicate these materials are potentially acid forming in their current state. Anita discussed opportunities to reprocess the slag to economically extract Zn by leaching.

Anita Parbhakar-Fox (centre) with Nick Wilshaw, Frances Wall, Nathan Fox and Felicity Wilshaw

Knowledge of the distribution of minerals and elements within a deposit is key to determining the most efficient mineral processing procedure, for example Mn within Au-Ag epithermal deposits.  Cyanide is the primary reagent in gold recovery, and Mn reacts easily, oxidizing the cyanide and transforming it into cyanate, increasing the cyanide comsumption, affecting recovery, and increasing production costs.  In addition, the process solution generates a gel that obstructs plant pipe-lines affecting the process and maintenance.

Andrew Menzies, of Universidad Católica del Norte, Chile, showed how combined automated mineralogy and elemental scanning (µXRF) on two high grade samples from El Peñón in Chile quickly and efficiently identified key information regarding Mn and Au relationships.  Crushed samples identified the Au-Ag mineralogy (QEMSCAN), whilst elemental rock mapping (M4 Tornado) showed Mn does not have a spatial relationship with Au-Ag mineralisation at the micro-scale.  Thus, it was possible to apply geometallurgical processes to remove the Mn phases prior to the application of cyanide without influencing Au-Ag recovery. 

After a good first day, we made use of the warm sunshine for the first of our late afternoon sundowners.

More on day 1 on the posting of 19th November.

Tuesday November 20th
In bio- and acid heap leaching processes, a substantial fraction of the mineral grains are positioned below the surface of the ore particles. Leaching performance of the non-surface mineral grains has potential to be affected by factors including the presence of surfactant, the operating temperature and agglomeration pre-treatment effect.

Mhadi Ghadiri, of the University of Cape Town, showed how non-destructive 3D X-ray micro-computed tomography (µCT), an imaging technique, can be used to quantify the leach performance at the mineral grain scale within the ore particles.

The copper ore deposits found in the north western part of Botswana contain mainly chalcocite.  The plant processing this mineral shut down due to a global fall in copper prices, and decreased grade and recoveries, subsequently leading to high and unsustainable operational costs.

Poloko Nenguba, one of 5 representatives from the Botswana International University of Science and Technology, presented an investigation of how minerals are distributed and associated in the host rock that was processed, analysing the quantitative mineralogical analyses results carried out by XRF, XRD and automated Scanning Electron Microscope technique (QEMSCAN) by Bulk Modal Analysis (BMA) and Specific Mineral Search (SMS).

This analysis consequently identified better methods of processing the minerals of interest. The quantitative analysis of the mineral deportment further provided knowledge on copper speciation, elemental deportment, mineral liberation and their association, as well as the grain size distribution.

Poloko Nenguba with Megan Becker and Elizabeth Whiteman
Knowledge of mineral composition and phase distribution within ilmenite concentrates is important in understanding the inter-granular inclusions of impurities which might have direct effects on the market value and subsequent treatment process.

Mercy Ramakokovhu, of the Tshwane University of Technology, South Africa, described how characterisation of the ilmenite concentrates from Richards Bay, South Africa was investigated along with mineral phase changes associated with the leaching process. A combination of characterisation techniques including the High-resolution Scanning Electron Microscopy and X- Ray diffraction techniques were used.

Mercy Ramakokovhu with her colleague Richard Mbaya
Manganese ores differ extensively in their mineralogical and bulk chemical composition. The result is that these ores follow different mineralogical morphology development paths as the reduction reaction proceeds in the production furnace.

For example the composition and distribution of initial liquid oxide formed at specific temperatures, as well as iron and manganese metallisation patterns due to their distribution in the ore minerals, are significant determining factors in the relative importance of different parallel reduction reaction mechanisms in each ore type. Applied mineralogy is a vital application field to better understand manganese ore reduction reaction mechanisms.

Theresa Coetsee, of the University of Pretoria, South Africa, showed how analysis of reduced ore phase chemistry and phase distribution was used to better explain differences in reduction rate variation with ore type. 

Theresa Coetsee with Gajanan Kapure and Sunil Tripathy
Phosphorus is one of the most deleterious elements in iron ore as it follows iron in the reduction process forming iron phosphides that make steel brittle. Excess P increases the cost of the steelmaking process and the steel industry has placed an upper limit of 0.07-0.08% P.  Goethite-containing P is abundant in many iron ores and it is difficult to remove without discarding valuable iron-containing units. The goethite forms during supergene metasomatic enrichment of BIF-derived ores with contained P typically associated with other impurity elements (e.g. Si and Al).

Mark Pownceby, of CSIRO Minerals Resources, Australia, discussed a study focussing on determining the distribution and association of P within goethite. Detailed characterisation of goethite-rich high-P iron ores was conducted using XRF, XRD and EPMA to measure their P and other impurity element contents and their distribution. Using this knowledge, the workers speculated on possible P-bearing species that may be present and also possible P substitution mechanisms in goethite.

Igor Tonžetic, of the University of Pretoria showed how the optimization of iron oxide discrimination on a QEMSCAN system was explored through the systematic testing of measurement and analysis parameters. Specifically the influence of backscatter electron (BSE) stabilizers (also known as BSE amplifiers), gun alignment, magnification field sizes and Mg/Ca solid solution in iron oxides were studied. Furthermore, the effect of stage height focus versus beam focus were compared and contrasted. Novel measurement parameters were presented that improve iron oxide distinctions including appropriate 3rd point BSE calibration standards and increased beam dwell times. Finally, a particle characterization mechanism that uses a de-convoluted iron oxide dilation erosion procedure was introduced.

Musarrat Safi, of Council for Geoscience, with Felicity and Nick Wilshaw

Lev Filippov
After the extended lunchbreak, due to a 'no-show' from the University of Johannesburg, Lev Filippov of the Université de Lorraine, France, discussed the Tabuaço tungsten project in Northern Portugal.

The deposit is composed of two skarns layers, namely “Main” and “Lower” skarns, which display significantly different mineralogical and geochemical features. Both skarns contain fine-grained disseminated scheelite but the Lower-skarn gangue is dominated by silicates whereas the Main-skarn gangue contains calcium-bearing minerals as fluorite, apatite and vesuvianite in close association with scheelite.

Preliminary feasibility studies show that direct separation of calcium-bearing minerals by flotation with fatty acids is very difficult due to their similar surface properties. Several routes for each skarn type have been proposed i.e. direct flotation for the Lower skarn and combined gravity/flotation route for the Main skarn. Moreover, the flotation of the Main Skarn has been deeply investigated in terms of depressants reaching a maximum WO3 enrichment ratio of 10 for the optimal conditions. Finally, new collector formulations have been developed to improve the separation contrast, leading to a WO3 enrichment ratio of 27 in the final concentrate.

Middle group (MG) chromitites of the Bushveld Complex are typically processed to recover chromite, with platinum group elements (PGE) derived as a by-product. Oxidised PGE ores, however, have proved problematic for the recovery of PGE by flotation. Desh Chetty, of Mintek, South Africa, showed how oxidised MG chromitites, subjected to flotation to recover PGE after gravity concentration to recover chromite, were characterised in order to understand PGE behaviour during processing. Characterisation was performed using automated scanning electron microscopy, laser ablation ICP-MS and bulk chemical assays.

Glencore's Kidd concentrator and Kidd mine are located in Timmins Ontario Canada and is one of the world’s largest massive sulphide deposits.  Elizabeth Whiteman and co-workers from XPS – Expert Process Solutions, Canada, have worked with Kidd concentrator monitoring performance based on monthly composites during 2015-2017. 

Using automated mineralogy (QEMSCAN, LA-ICP-MS) on sized samples of flotation feed, Cu concentrate, Cu tailings/Zn feed, Zn concentrate and Zn tailings, mineralogical drivers of flotation performance were examined and reported in the presentation.

Björn Lewandowski
Since 2014, fluorite (calcium fluoride) has been one of the 20 most critical materials indexed in the EU Raw Materials Initiative. Currently it is the only fluoride containing raw material to be used for production of fluorine. Barium sulphate, a common gangue of fluorite, behaves in a similar way with collectors as calcium fluoride, making the separation via flotation more complicated. It is of great interest to optimize the productivity of these flotation plants. For quantification of calcium fluoride, X-ray spectroscopy is commonly applied. The main disadvantage of this method is the necessity of a high-energy lamp for detection of fluorine.

Björn Lewandowski, of University of Duisburg-Essen and Niederrhein University of Applied Sciences, Germany), described the application of Raman spectroscopy to a model flotation system containing calcium fluoride and typical gangue particles. Calibration with pure substances showed very good quantification of calcium fluoride and barium sulphate. Application to an industrial flotation system also showed very promising results.

Uwe König, of Malvern Panalytical B.V., The Netherlands, discussed how fast and frequent monitoring of the mineral content and related process parameters brings value to bauxite mining, refining and aluminium production.

X-ray diffraction (XRD) is a critical process tool to efficiently use energy during aluminium production. Traditionally quality control of electrolytic baths, alumina and bauxites have relied on calibration based single peak methods or more advanced full pattern techniques. Recent tests showed that the same XRD measurement used to determinate composition and bath parameters can be used to predict the bath temperature and to calculate liquidus temperature and superheat. This technique saves valuable time and costs for additional temperature monitoring and allows fast counteractions on changing conditions to prevent bath solidification.

Even though today was dull and wet, Kirstenbosch Botanical Gardens was a great venue for the very informal conference dinner, a relaxing respite from the conference atmosphere.

More on the conference dinner on the posting of November 21st.

Wednesday November 21st
The final day of the conference specialised in high-tech metals, as a prelude to the Hi-Tech Metals '18 conference which begins tomorrow.

The day began with a keynote lecture from Frances Wall, of the Camborne School of Mines, UK, who reviewed common problems, and progress towards solutions, in the process mineralogy of rare earths.

The geochemistry and mineralogy of rare earth element (REE) deposits is diverse and ranges from carbonatite-related deposits and alkaline rocks to mineral sands, ion adsorption clays, marine crusts, nodules and clays, by-products of phosphate and bauxite, and re-use of waste materials. Despite the large number of recent exploration projects, very little additional REE production has started. An in-depth understanding of the mineralogy is essential for process design and all of the deposit types have mineralogical advantages and challenges, which Frances reviewed and explained.

Most critical element ore deposits are complex and display a high degree of variability, arising from their inherent geological and mineralogical characteristics, which impact their beneficiation. Process mineralogy for rare and critical elements, including REE, Li, and Nb aim to help with the flowsheet development to avoid extensive and time-consuming bench testing. 

Process mineralogy can provide critical quantitative data i.e., for Li-bearing minerals that may not be recoverable or marketable, the deportment of REE, the type of Nb phases and their liberation.

Tomas Hrstka, of SGS Minerals, Canada, discussed examples of automated mineralogy applied in a geometallurgical framework at an exploration to feasibility level.

Jochen Petersen, of the University of Cape Town, discussed how in-situ leaching of REE from low grade clay deposits formed through subtropical weathering of granites presents a potential alternative due to its low cost and limited infrastructure.

Jochen showed how a REE enriched ion-adsorption clay sample from Madagascar was characterised using a combination of process mineralogy techniques from simple optical microscopy to advanced synchrotron analysis to understand the deportment of the different REE within these clay ores.

Results show the presence of varying morphologies of kaolinite and halloysite, and contribute to further understanding of the movements of Fe, Mn, La and Ce within the weathering cycle. Based on the understanding of the REE deportment, suitable processing routes can be designed and optimised to effectively recover the REEs.

Reiner Neumann, of CETEM, Brazil, showed how recovering REE as by-products of existing mineral operations is a low-investment option to increase supply in the short term. REE-minerals have been reported for the São João del Rei pegmatitic province (southern Minas Gerais State, Brazil) since the late 1940’s, but data is scarce and dubious.

The largest pegmatites in the Province have been mined since 1945 at the Volta Grande Mine, producing Sn (cassiterite) and Nb-Ta (microlite and columbite group minerals) concentrates. The detailed mineralogical and technological characterisation of streams from the beneficiation plants identified apatite with monazite and xenotime microinclusions as the most important REE-bearing mineral. Apatite is well liberated, and reports to the tailings, which contains up to 6.8 (wt.%) that could be reprocessed to generate a product adequate for phosphoric acid production with REE recovery as by-products and would also allow secondary recovery of cassiterite, fluorcalciomicrolite and columbite-(Mn).

Tailings and their disposal have been described as the single most important source of environmental impact from mining operations. Tailings may also provide a potential source for further beneficiation of valuable materials. The fundamentals for any processing or treatment of ore and tailings are initially dictated by the mineralogical and chemical characteristics.

Rita Kallio, of the University of Oulu, Finland, presented two case studies for tailings characterisation. The main objective was to increase mineralogical knowledge of the tailings with an emphasis on liberation analysis. Liberation analyses were conducted with INCAMineral in combination with post-processing the results with GrainAlyzer.

The first case, spodumene flotation tailings from a greenfield lithium mine project, focused on determining the liberation of feldspar and quartz. In the second case, tailings from a closed Fe-Ti-V-mine, the main interest was in the properties of ilmenite in various parts of the old tailings storage facility.

Following the lunch break Pekka Tanskanen, of University of Oulu discussed the on-line monitoring of the spodume heat treatment process using Timegated Raman spectroscopy. Spodumene is the most important hard rock source for lithium. Prior to leaching of lithium from spodumene the natural monoclinic crystal structure must be transformed into tetragonal ß-form by heat treatment. The heat treatment at about 1050 °C is highly energy consuming and can lead to lowered lithium recovery in the leaching stage if the concentrate is over-heated or under-heated. On-line monitoring of the spodumene phase transformation grade gives data for controlling the heat treatment process. 

Timegated® technology is built on the research carried out at VTT Technical Research Centre of Finland and Oulu University since 2009. Timegate Instruments Ltd is based in Oulu, and is exhibiting for the first time at an MEI Conference. The basis of the innovation is novel Timegated®  Raman spectroscopy technology that enables continuous, non-contact, quantitative and qualitative measurements of solids, liquids and slurries.

The sample is illuminated with a laser and the scattered photons provide mineralogical and polymorphic information while minimizing photoluminescence interference which has been a major challenge for conventional Raman spectroscopy. Timegated® Raman technology also minimizes thermal emission interference making high temperature measurements feasible.  

Jussi Soukkamaki, of Timegate Instruments with Rita Kallio and Pekka Tanskanen of University of Oulu

Michelle Kelvin, of XPS Expert Process Solution, Canada, showed how, in the field of process mineralogy, LA-ICP-MS is a valuable tool that can complement other mineralogical techniques used to characterise orebodies and concentrator performance.

Many economic metals are present at ultra-low levels as solid solution within minerals and require complex recovery techniques to separate them from the bulk ore. Understanding metal occurrence is necessary for implementing the proper recovery strategies.

LA-ICP-MS is a cost-effective and versatile method of measuring precious-metals and other economic trace-elements in solid solution in various types of mineral phases while achieving low detection limits. To demonstrate the practical uses of LA-ICP-MS to process mineralogy, Michelle showed how XPS has collected data from a sulphide-rich heavy liquid separation and shaking table concentrate from the Kipushi Cu-Zn deposit, Katanga, DRC.

In addition to Cu and Zn, the Kipushi operation is currently investigating strategies for concentrating Ge and Ga. A combination of gravity separation and recovery of Ge and Ga through flotation of Zn and Cu sulphides is being considered. The objective of the LA-ICP-MS analysis was to determine the proportion of Ge and Ga present in solid solution in sulphides and in discrete mineral phases.

Sphalerite is the most important Zn-bearing ore mineral, however, its flotation response during mineral processing can be complicated by the presence of trace impurities. Previous work has constrained the effects of Fe substitution on sphalerite mineral structure and flotation response, yet minimal work has focussed on the effects of other transition metals such as Cd and Co which readily substitute into the sphalerite mineral structure.

Using electro-kinetic techniques, Lebogang Babedi, of Stellenbosch University, South Africa, discussed the variations in surface charge, Cu-activation and collector adsorption arising from cation substitution in trace-element doped synthetic sphalerite.

These variations include differences in the onset pH during flotation and differences in the reaction products arising during collector adsorption. These differences can be linked to the nucleophilic and electrophilic characteristics of the trace metal substituent. The study highlights the need for fundamental scientific investigation which will enhance predictive capabilities and efficiencies during beneficiation of complex natural ores.

As noted by Andrea Guhl, of Freiberg University of Mining and Technology, Germany, while FEI’s MLA system is a staple in ore characterization and ore process assessment, automated mineralogy systems have been used for a multitude of other materials as well.

When looking at ashes, the high amount of X-Ray amorphous components have prevented a thorough characterization using XRD. Wet chemistry testing has been employed to assess the elemental content of this material, but fails to reflect the nature of ashes. Previous studies of ashes have occasionally highlighted interesting elemental compositions, which encouraged recycling efforts. However, turning waste into a resource requires an intimate knowledge of these materials.

Andrea stressed how evaluating particles – their morphologies and compositions – of ashes is essential in devising treatments for a targeted approach.

Andrea Guhl with Olga Guseva

MEI Consultant Megan Becker briefly summarised the conference, after which MEI's Amanda thanked the sponsors, chairpersons, authors and delegates and invited everyone back to the Vineyard for Process Mineralogy '20, which will be held in October 2020, prior to the IMPC, also in Cape Town.

Farewell drinks in the Vineyard gardens

The Proceedings of the conference are available on USB, and all authors have been invited to submit their final revised papers to Minerals Engineering for peer-review with a view to publication in a virtual special issue of the conference.

All the photos from the event can be viewed here.
Twitter @barrywills


  1. Just a quick word to thank you for the great organization of Process Mineralogy and thumbs up for your new Hi-Tech metal initiative. It was a double first for me : first time attending your conferences and first time in South Africa. Thank you for making this such a great experience on all levels.

    Philippe Giaro, University Liege, Belgium

  2. It was great to meet you Philippe. I hope that you will now be part of the MEI conference scene

  3. I am extremely happy to read your excellent summary. We, mineral engineers and practicing plant operators generally did not bother about mineralogy and were happy with chemical analyses. People talk about assay of tailings but do not talk whether any liberated and recoverable values reported in tailings due to operational conditions.
    Mineralogy has to be the backbone(first information) to plan mining(blasting) to beneficiation.

    I hope more focus and attention is given to mineralogy and properties of composite particles.

  4. I would like to thank the organizers as well as the participants for this absolutely fantastic conference. It was my first participation in a MEI conference. The atmosphere during the entire conference was very friendly. As the conference was relatively small compared to other conferences I attended in Germany, the networking with other researchers was much easier. Furthermore, the topics discussed during the conference were a nice insight into the current problems in mineral processing. The academical background of the participants showed that mineral processing is a topic, which affects many different researchers (geologists, material engineers, chemical engineers, etc.). I can highly recommend this conference and would be happy to attend other MEI conferences in the future.
    Many thanks to everyone!


    1. Many thanks for your kind comments Bjorn. We look forward to seeing you at the next one!

  5. Nice to see the summary again and our Tata Steel guys too.
    Hope to visit and meet in upcoming conferences.
    Congratulations for another successful conference...
    Rama Murthy

    1. Many thanks Rama. Yes, you must get to one of our conferences. It was good to meet Sunil and Gajanan


If you have difficulty posting a comment, please email the comment, and any photos that you might like to add, to and I will submit on your behalf