Conference PaperPDF Available

Modern Applications of Electron Probe Microanalysis in Applied Mineralogy and Industrial Mineral Processing

Authors:

Abstract

Electron probe microanalysis (EPMA) is a well-known electron beam analytical technique capable of providing quantitative elemental compositional information on solid materials. Within the applied mineralogy and mineral processing sector, accurate chemical compositions of ore minerals and/or phases are important for downstream mineral processing and profitable metal recovery. This paper briefly examines current applications of the EPMA technique, which has shifted according to market trends within the mining and metal industry. Specific commodity types have seen an increase in economic relevance, requiring routine or non-routine analytical procedures. For example, the evaluation of polyphase, non-stoichiometric rare-earth oxides, heavy minerals, and kimberlite indicator minerals are commonplace. The exploitation of increasingly complex, lower grade ores has also become necessary, often requiring detection limits down to trace levels. Finally, the future outlook of the EPMA technique is briefly discussed with respect to recent innovations and their possible application in industry.
Celebrating nine decades of groundbreaking advancements in minerals and metallurgy, the Mintek@90
Conference reects on the past while focusing on the future of a sustainable mineral industry. Explore
the legacy and pioneering technologies that continue to shape the sector globally.
Mintek@90 Proceedings
11-12 November 2024 | Sandton Convention Centre Johannesburg, South Africa
1934 - 2024 | 90 Years of Excellence in Mineral Innovation
MintekMintek@@90 90
ConferenceConference
Gearing the Industry for a
Sustainable Mineral Future
THE SOUTHERN AFRICAN INSTITUTE OF MINING AND METALLURGY
JOHANNESBURG 2024
11-12 November 2024
Sandton Convention Centre, Johannesburg, South Afrca
MintekMintek@@90 Conference90 Conference
Gearing the Industry for a Sustainable Mineral Future
Published by The Southern African Institute of Mining and Metallurgy
Minerals Council South Africa, 7th Floor, Rosebank Towers, 19 Biermann Avenue, Rosebank, 2196
Republic of South Africa
© The Southern African Institute of Mining and Metallurgy, 2024
ISBN 978-1-7764673-7-2
The papers in this volume have been for the most part prepared from Word documents supplied by the authors,
with additional typesetting and formatting by The Southern African Institute of Mining and Metallurgy.
Desktop prepared by Camera Press, Johannesburg
The Organising Committee would like to thank the reviewers listed below for the eorts in the reviewing process
L. Auret
M. Bambo
O. Bazhko
L. Bbosa
T. Bungane
Q. Campbell
M. Chabalala
D. Chetty
A. Cherkaev
K. du Preez
P. de Vaal
M. Erwee
A. Garbers Craig
M. Gericke.
B. Joja
D. le Roux
R. Letsoalo
M. Manuel
G. Marape
P. Mdluli
K. Mudzanani
J. Moema
T. Mokhena
T. Moodley
G. Ndlovu
L. Nelson
S. Nkwanyana
J. Olivier
D. Phillpotts
H. Potgieter
Q.R. Reynolds
S. Robertson
M. Safari
A. Singh
K.C. Sole
S. Swanepoel
M. Tangstad
S. Tsebe
Contents
Page No
Improving energy eciency in tumbling mills through insights into slurry and media dynamics
T.L. Moodley and I. Govender ......................................................................................... 1
Establishment of fully-edged off-line fabrication capabilities for process and product development of rolled metal products
J.S. Moema, C.W. Siyasiya, N.D.E. Hadebe, K.V. Morudu, and M.J. Phasha .................................................. 9
Potential use of gold in improving the sulphur tolerance of palladium catalysts in diesel-engine exhaust systems
J. Aluha ........................................................................................................... 19
Assessing mining impact on water resources using full cost accounting: A comparative case study between two deep-level
gold and uranium mining sites in South Africa and Germany
Q. Roode-Gutzmer .................................................................................................. 37
Relating mineralogical data to ball-milling parameters of a multi-component ore: A UG2 case study towards limited modelling
T. Somo, S. Nkwanyana, V. Govender, D.H. Rose, D. Chetty, M. Manuel, and T. Nghipulile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
UX7 spiral: Ultrane processing made possible
R. Ceko ............................................................................................................ 65
Development and demonstration of t for purpose pyrometallurgical processes and technologies: Mintek’s perspectives
S. Tsebe and E. Matinde ............................................................................................. 81
Development of a binder for heap leaching of low permeability ores
S.W. Robertson and J. Petersen ...................................................................................... 95
Unlocking resources through sensor-based sorting
L. von Ketelhodt and A. Singh ........................................................................................ 105
Applications of data analytics, modelling and control in the mining, minerals and metals industries
K.S. Brooks and A. Higginson ......................................................................................... 115
Modern applications of electron probe microanalysis in applied mineralogy and industrial mineral processing
Y. Thakurdin ....................................................................................................... 125
Real-time optimisation of a base-metal grinding mill using grind curves and reinforcement
J. Olivier and W. Shipman ............................................................................................ 139
Exploring an ensemble models for predictive maintenance optimization at iThemba LABS
E Nkadimeng, V. Maluleke, T. Mokoena, and N, Stodart. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Reviewing weathering on kimberlites and its implications on mineral carbonation: A case study on the Cullinan/Premier
kimberlite mining operation
A. Mantshontshoa and Z. Nkosi ....................................................................................... 165
Leveraging machine learning for enhanced compliance in responsible gold sourcing: A Rand Renery case study
A. Hefer and S. Janse Van Rensburg ................................................................................... 179
Human rights issues related to unrehabilitated derelict mines – The signicance of a multidisciplinary approach towards
effective mine rehabilitation in South Africa
L.K. Moeletsi ....................................................................................................... 193
Carbon emission reduction through carbon dioxide sequestration
P. Ndlovu, S. Babaee, and P. Naidoo ................................................................................... 205
Mintek’s role in advancement of computational modelling tools in minerals processing and extractive metallurgy research
and operations
A.V. Cherkaev, W.J. Shipman, H.K. Mittermaier, M. Khama, T. Moodley, I. Govender, and L. Bbosa ............................. 215
A framework for technology selection in large scale systems using lessons from manufacturing and other large industrials in
the design of exible mining systems
C. Mukonoweshuro and A. Botha ...................................................................................... 229
The future of phytoremediation technologies in the treatment of Acid Mine Drainage: A review
C. Chigwede, P. Matshona ,and M. Safari ............................................................................... 247
Comparison of the effectiveness of various alkaline generating agents in neutralising acid mine water: An experimental and
simulation study
T.M. Mogashane, J.P. Maree, and L. Mokoena ........................................................................... 259
Governance and the mining industry in South Africa
L. Raputsoane. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Investigating the viability of gold tailings as a partial cement replacement in concrete
J.P. Kanjee and P.B. Nkambule ....................................................................................... 283
Mintek@90
Sandton Convention Centre, 11 – 12 November 2024
The Southern African Institute of Mining and Metallurgy
125
Modern applications of electron probe microanalysis in
applied mineralogy and industrial mineral processing
Y. Thakurdin
Mintek, South Africa
Electron probe microanalysis (EPMA) is a well-known electron beam analytical technique
capable of providing quantitative elemental compositional information on solid materials.
Within the applied mineralogy and mineral processing sector, accurate chemical
compositions of ore minerals and/or phases are important for downstream mineral
processing and profitable metal recovery. This paper briefly examines current applications
of the EPMA technique, which has shifted according to market trends within the mining
and metal industry. Specific commodity types have seen an increase in economic relevance,
requiring routine or non-routine analytical procedures. For example, the evaluation of
polyphase, non-stoichiometric rare-earth oxides, heavy minerals, and kimberlite indicator
minerals are commonplace. The exploitation of increasingly complex, lower grade ores has
also become necessary, often requiring detection limits down to trace levels. Finally, the
future outlook of the EPMA technique is briefly discussed with respect to recent
innovations and their possible application in industry.
Keywords: electron microprobe, rare earth elements, indicator minerals, X-rays,
microscopy
INTRODUCTION
As essential component of materials characterisation is chemical composition. Without chemical data,
it is not possible to fully understand the properties of an ore, mineral, metal or slag, and the processes
through which these solids were formed or produced. For this reason, bulk chemical tests (e.g. wet
chemistry or powder X-ray fluorescence spectrometry) are in high demand, and often the first step in
sample characterisation. Electron-beam instruments such as the electron probe microanalyser (EPMA)
differ from bulk chemical methods in that they provide spatially resolved data - this means that
chemical compositions can be obtained at the micrometre (micron - µm) scale, allowing for materials
classification at the particle level. This spatially resolved information is critical for mineral processing
and metal extraction.
Basic principle of EPMA
An electron microprobe is primarily used to measure the phase chemical composition (in weight
percent, wt. %) of solid materials. Currently, almost every element can be measured, with the exception
of a few light elements (H, He, Li). As mentioned above, EPMA is not a bulk chemical method but
measures the chemistry of a substance (mineral/metal/phase) at the micron scale. The material to be
analysed is mounted on a glass slide or placed within epoxy resin prior to grinding and polishing to
produce a flat surface. Afterwards, the specimen is placed within the instrument and exposed to a high
energy electron beam. Chemical signals are collected from the surface, and a small volume beneath the
surface (penetrated by electrons) which varies in size according to the instrument operating conditions
and the composition of the material being analysed.
The electron microprobe is similar to the standard scanning electron microscope (SEM) in its
instrumental configuration.
126
It consists of an electron gun that generates electrons (usually emitted from a tungsten filament) that
move through a series of electromagnetic lenses to produce a focused analysis spot on the sample
surface. Electron bombardment results in the displacement of inner shell electrons (ionisation) from the
target material atoms, creating a vacancy that is subsequently filled by an outer shell electron. This
orbital transition releases energy in the form of a package of radiation called X-rays. The emitted X-rays
have a characteristic wavelength depending on the element being analysed. These characteristic X-rays
are filtered/reflected through wavelength dispersive crystals that only allow certain frequencies to pass
through (wavelength dispersive spectrometryWDS) prior to measurement at an X-ray counter. The
number of measured counts for a specific element is compared to that of a standard of known
composition. This count ratio (‘k-ratio’) is then used to calculate the concentration of a specific element
within the phase being analysed. The k-ratio value is further refined using a number of corrections
(called matrix effect corrections) that account for the complex particle interactions involved in X-ray
generation and measurement.
Comparison with EDS
Energy dispersive spectrometry (EDS) involves the detection and measurement of characteristic X-rays
produced through the electron beam-sample interactions described above. However, the produced X-
rays are differentiated by their energy only, rather than wavelength, and are not filtered through WDS
crystals. This means a greater proportion of X-rays are permitted to reach the EDS detector. The
simultaneous processing of multiple X-ray signals results in broader elemental peaks and hence a lower
spectral resolution. A reduced spectral resolution is especially problematic when quantifying elements
that lie within a similar region of the X-ray spectrum (e.g., rare earth elements (REE) and platinum
group elements (PGE)). SEMs or automated SEMs (AutoSEM) are usually equipped with EDS
spectrometers for elemental acquisition, although most EPMA instruments contain an EDS
spectrometer as well.
EDS allows for capturing of elemental data far more rapidly than measurement with EPMA, since all
elements can be measured simultaneously in a ‘standardless’ procedure. However, the low spectral
resolution and the standardless approach of EDS limit its application to qualitative or semi-quantitative
applications (in most cases – see Newbury and Ritchie, 2013). Even though quantitative EDS is made
possible by utilising appropriate reference materials (and by adhering to a rigorous analytical protocol)
certain elemental suites (REE, PGE, light elements) cannot be fully resolved with this system, especially
at low concentrations (Ritchie and Lowers, 2018).
Spatial resolution is also an important factor to consider. The resolution of EPMA is dependent on the
accelerating voltage, probe current, spot size and the nature of the analysed substrate (Buse and Kearns,
2020). These parameters define the analytical volume, which is the area from which X-rays are generated
within the sample, and is the critical limit on spatial resolution (McSwiggen, 2014). Importantly, the
analytical volume often exceeds the analytical spot size, as demonstrated by Monte Carlo simulations
of electron interactions in certain materials (see Berger and Nissen, 2014). For practical purposes, the
spatial resolution of EPMA using regular thermionic filaments (e.g., tungsten) is no less than 1 micron
at standard operating conditions. Resolution is improved by utilising finer lanthanum hexaboride
(LaB6) filaments or by using a field emission electron source (FE-EPMA, see below). Non-field emission
SEM-EDS instruments utilise lower probe currents than EPMA (ideal for electron imaging), however,
the interaction volume may not be significantly reduced due to the application of high acceleration
voltages (15-20 kV) required for excitation of all elements during EDS measurement.
Due to the similarities between instruments, SEM-EDS and EPMA are inevitably compared. Put simply,
the SEM system is primarily designed for electron imaging (e.g., back-scattered or secondary electron
imaging) complemented by semi-quantitative chemical acquisition through EDS. EPMA provides fully
quantitative, spatially resolved chemical compositional information imaging and mapping is also
possible but with some restrictions on sample size, working distance and field of view. Of course, some
exceptions do exist such as WDS spectrometers mounted on SEMs, however, these configurations are
less common.
127
MATERIALS AND METHODS
A note on sample preparation
As mentioned above, a flat surface is required for the best results with EPMA. The second requirement
is a conductive surface to prevent the accumulation of charged particles on the specimen. Conductivity
is achieved by coating the sample surface with a thin film (~20 nm) of carbon or metal (e.g., aluminium
or gold).
While carbon/metal coating is common practice (achieved using evaporative or sputter coating
machines), the thickness of the coating is not always viewed as significant. However, overly thick
coatings may result in the loss of counts and/or spurious results, especially if the relative coating
thickness between reference materials and analysed samples differs significantly (Nash, 1992). Coating
thickness is often overlooked because the nanometre-sized thickness of carbon or metal films is not easy
to measure precisely. Some manufacturers include quartz thickness control monitors on coating
machines, but this apparatus is not present on all instruments. Another frequently applied method of
carbon film thickness estimation is by observation of interference colours formed on polished brass,
glass or other interfaces placed alongside samples during coating (Goldstein et al., 2017). Colour-based
estimations, however, are somewhat subjective and need to be observed at all locations within the
coating chamber (Zhang and Yang, 2016).
Instrumental setup
All electron microprobe analyses discussed in this paper were performed using a JEOL JXA-8230
Superprobe situated in Mintek (Randburg, South Africa). The instrument possesses four WDS
spectrometers and a single EDS analyser. The system is additionally equipped with high contrast back-
scattered electron and secondary electron detectors for imaging purposes.
Typical spot analyses are performed at an accelerating voltage of 20 kV with a beam current of 30 nA,
and a 1-5 µm spot size. Counting times are 10 seconds on peak, and 5 seconds on each of the two
background positions adjacent to the peak. Matrix corrections are carried out using the ‘phi-rho-z’ or
ϕ(ρz)’ method (Pouchou and Pichoir, 1991). Mass absorption corrections are calculated using the
‘FFAST’ tables published by NIST (Chantler et al., 2005). Specific reference materials are analysed as
procedure depending on the elemental suite required. Mintek possesses an extensive standard library
consisting of over 500 oxides/alloys, allowing for measurement of almost all solid elements on the
periodic table. For regular oxide analyses, measured standards include jadeite (Na Kα, Al Kα),
wollastonite (Si Kα, Ca Kα), orthoclase (K Kα), periclase (Mg Kα) rhodonite (Mn Kα), hematite (Fe Kα),
eskolaite (Cr Kα), bunsenite (Ni Kα) and rutile (Ti Kα). Elements are measured using various WDS
crystals distributed across four separate channels, these include: TAP (Na, Al, Mg, Si), LIF (Mn, Cr),
LIFL (Ti, Fe, Ni) and PETL (Ca, K). LDE crystals are used for light element detection if required (e.g., F,
C, N, and B). The order of elements is configured such that elements prone to beam migration effects
are included in the first series of measurements (e.g., Na and F). For overlapping element peaks (e.g., Ti
V, Cr, Mn, REE, PGE), overlap correction procedures may be applied. The above instrumental
parameters and selection of reference materials may differ based on the desired elemental suite and
required detection limit.
RESULTS AND DISCUSSION
Routine analyses: chromite, heavy minerals and indicator minerals
The requirement for quantitative elemental data is often questioned by clients, especially if the ideal
compositions of most minerals and alloys are known. Numerous books and data repositories exist to
provide such information such as the publication by Deer et al., (2013), the ‘Handbook of Mineralogy’
(Anthony et al., 2024) and websites such as ‘Webmineral’ (Barthelmy, 2014). Organisations likely also
possess in-house databases containing catalogued historical data for various minerals and phases. One
basic consideration is the compositional variation of naturally formed minerals (especially oxide
minerals). The ranges in elemental values of certain minerals vary considerably and unpredictably,
which has implications for ore processing operations. For example, the ideal formula of chromite is
128
FeCr2O4. It is well known, however, that natural chromites incorporate significant (and variable)
quantities of Al and Mg, which substitute for Cr and Fe atoms respectively (Table I). The concentrations
of Cr and Fe may differ according to lithology – for South African chromites sourced from different
chromitite layers of the Bushveld Complex (Lower Group, Middle Group, Upper Group), these
variations may be significant (~1-10 %; Sciarone, 1998). It is important to note that lithological grades
are not always consistent, especially for non-UG2 chromite. The ratio of Cr to Fe is an important
indicator of grade in the chrome industry (Kleynhans et al., 2023), and this ratio is primarily controlled
by the Cr content in chromite. The mineral chemistry of chromite may also affect the smelting behaviour
of chrome ore (e.g., high-temperature reduction, Geldenhuys, 2013).
Table I. Representative EPMA elemental compositions for various oxide minerals
‘Heavy minerals’ generally refer to dense minerals (>2.96 g/cm3) prone to sinking in liquid separation
media. Of particular commercial importance are the minerals ilmenite (FeTiO3) and rutile (TiO2), which
are primary sources of titanium. Heavy mineral deposit types include mineral sands accumulated along
coastal beaches or collected within sedimentary basins.
Ilmenite may incorporate Mg, Al, Cr and Mn within its structure through substitution with Fe2+. The Cr
content in ilmenite is critical, as variations of up to ~1-2 % may be deemed unsuitable for processing
(e.g., Australian mineral sands Pownceby et al., 2020). Additionally, the alteration of ilmenite results
in the formation new solid-solution phases such as pseudorutile and leucoxene. Representative EPMA
analyses of ilmenite, rutile and leucoxene obtained from a South African beach deposit are shown in
Table I. Leucoxene is defined as Fe-Ti oxide phases with TiO2 greater than 70% and (Ti/Ti+Fe) between
0.7 and 0.9 (Frost et al., 1983). Notice the significant variation in Fe and Ti between rutile, ilmenite and
leucoxene these chemical variations are the primary control on the magnetic susceptibility of such
minerals (Cavanough et al., 2006; Thakurdin and Ashwal, 2016). Since the concentration method for
heavy mineral sands is magnetic separation, this chemical information is critical for optimal calibration
of separation equipment (Contreras et al., 2018). Apart from physical separation processes, the chemistry
of heavy minerals can also influence hydrometallurgical separation procedures. For example, Schirmer
et al., (2020) demonstrated that the Mn content in heavy minerals affected the crystal stability of such
phases during leaching, resulting in lower overall Ti yields. Finally, the integration of EPMA data in
refining AutoSEM databases has become common in industry (discussed below) improvement in
particle characterisation and the identification of altered phases (pseudorutile) in a Senegalese heavy
mineral deposit was recently demonstrated using combined AutoSEM and EPMA data (Kanzari and
Graul, 2024).
Diamonds continue to possess considerable value in both the commercial and industrial sectors. Natural
diamonds are sourced from packages of mantle rocks brought to the surface during localised volcanism
(creating narrow volcanic tubes called kimberlite pipes). Diamond occurrences are rare, however, so
other mineral tracers are utilised to assess the diamond bearing potential of kimberlite-associated rocks,
prior to attempting time-consuming (and expensive) separation and isolation of diamonds from a large
rock mass. Such tracers include minerals crystallised at high temperatures and pressures (mantle
conditions similar those present during diamond-formation) which are entrained during kimberlite
Chromite Ilmenite Rutile Leucoxene Hematite Grossula r 1 Grossular 2 Almandine-pyrope
Element (wt %) n = 25 n = 25 n = 54 n = 5 n = 3 n = 47 n = 5 n = 5
MgO 8.39 (0.97) 0.89 (0.72) -0.08 (0.07) 0.04 (0.03) 0.05 (0.02) 0.12 (0.07) 9.39 (1.14)
Al
2
O
3
16.97 (1.58) 0.11 (0.11) 0.07 (0.07) 1.27 (0.95) 0.30 (0.15) 26.97 (1.93) 19.5 (2.11) 21.53 (0.42)
SiO
2
-0.11 (0.17) 0.05 (0.09) 1.40 (1.54) 0.34 (0.15) 38.41 (0.34) 38.72 (0.56) 38.59 (0.78)
Cr
2
O
3
44.63 (1.30) - - - - - - -
FeO 28.05 (1.01) 45.13 (2.04) 0.25 (0.25) 13.65 (5.41) 90.16 (0.95) 10.01 (2.36) 5.40 (2.02) 22.83 (0.90)
TiO
2
1.05 (0.3 5) 52.81 (1.89) 99.46 (0.51) 78.92 (6.46) 0.21 (0.06) 0.13 (0.07) 0.83 (0.81) 0.08 (0.07)
V
2
O
5
0.06 (0.0 5) - - - - - - -
MnO 0.27 (0 .02) 0.73 (0.62) 0.01 (0.01) 0.37 (0.63) 0.05 (0.01) 0.20 (0.10) 2.08 (1.87) 0.80 (0.86)
P
2
O
5
- - - - - 0.06 (0.04) - -
CaO -0.04 (0.03) 0.01 (0.01) 0.78 (1.45) -23.48 (0.42) 34.08 (1.27) 5.94 (2.15)
Total 99.43 (0.44) 99.66 (0.37) 99.86 (0.31) 96.69 (3.17) 91.79 (0.73) 99.27 (0.40) 100.72 (0.40) 99.16 (0.77)
standard deviations shown in parentheses
129
magma ascension through the crust. These are called kimberlite indicator minerals (KIMs) and include
phases such as garnet, clinopyroxene, ilmenite, chromite and olivine, among others (McClenaghan,
2005). Garnet analyses show considerable variation in chemistry and are categorised according to
chemical end-members (e.g., grossular, pyrope, spessartine), however, intermediate compositions are
frequently observed (e.g., almandine-pyrope, Table I). The precise chemistry (obtained using EPMA) of
KIMs is used firstly to constrain the pressures and temperatures of crystallisation – this establishes that
the minerals are indeed mantle-derived. Thereafter, calculated pressures and temperatures are plotted
to construct the local geothermal gradient and constrain the ‘diamond window’ (Kjarsgaard et al., 2019)
which is used to assess the likelihood of diamonds forming in the local geological region. The indicator
mineral chemistry itself serves as a classification tool (using various chemical discrimination diagrams)
which can qualitatively establish the diamond-bearing potential of a suite of kimberlite-associated
rocks.
Using EPMA to create reliable datasets
Another basic application of EPMA in industry is quality control. This is especially relevant for bulk
modal mineralogical data collected by AutoSEM, X-ray diffraction (XRD) or micro-XRF spectrometry
(µ-XRF). A common practice when assessing modal mineralogical information is to cross-reference the
mineral chemistry (recalculated as elemental contributions to the bulk sample) with independently
performed bulk chemical assays. The proportion of elements measured using the two techniques should
coincide within an acceptable error margin. This process is referred to as data reconciliation or mineral-
chemical reconciliation, and is regularly applied using ideal mineral chemistries. The reconciliation
procedure is important in establishing data quality and reliability, and should be performed as routine
for quantitative datasets.
The quantitative measurement of mineral chemistries when performing mineral-chemical reconciliation
is less common (Subramanian et al., 2016), partly due to time and cost considerations. In most cases,
either ideal/literature mineral chemistries or semi-quantitative estimations of mineral chemistry (EDS)
are used for reconciliation. The assumption of ideal chemistry may be valid – for example, non-oxide
species such as sulphides (e.g., pyrite) often display chemistries that conform well with their structural
formulae. For complex oxides (e.g., REE-oxides), however, assuming the mineral chemistry may become
problematic since the actual mineral compositions deviate significantly from ideal chemistry or semi-
quantitative chemistry acquired using EDS.
The effect of using assumed ideal/assumed literature chemistry is demonstrated using REE
deportmental data obtained from an African carbonatite deposit. Deportment refers to how are
distributed within the minerals/phases present in the sample. REE-bearing minerals in this sample
included monazite, synchysite and bastnäsite, the relative bulk modal proportions of which were
obtained using AutoSEM (Table II). The REE deportment (Ce, La and Nd) was calculated using the bulk
proportion of the minerals (Table II) and the mineral chemistry obtained using EPMA (Table III
discussed in detail in later sections).
Table II. AutoSEM modal proportions of REE phases from a carbonatite ore
Mineral
Ideal Chemical Formula
Relative proportion (mass %)
Synchysite
Ca(Ce,La)(CO3)2F
91.0
Bastnäsite
(Ce, La)(CO3)F
8.5
Monazite
(La,Ce)PO4
0.5
Figure 1 compares the variation in calculated elemental deportment when using mineral chemistries
obtained from different sources these sources included measured EPMA data (Table III), literature
data (Handbook of Mineralogy) and ideal mineral data (Webmineral). The cerium-rich varieties of each
mineral were chosen from the Handbook of Mineralogy and Webmineral (bastnäsite-Ce, synchysite-Ce
and monazite-Ce) since they conformed best with measured EPMA data.
130
Figure 1. Elemental deportment of REE (Ce, La, Nd) within rare earth minerals calculated using mineral
chemistry from three different sources (EPMA, Handbook of Mineralogy, Webmineral).
The calculated elemental deportment changes significantly when different mineral chemistry sources
are applied (Figure 1). A better correlation is observed between measured EPMA values and the
Handbook of Mineralogy which makes sense since this database contains historical electron
microprobe data. REE deportment calculated using assumed chemistry from Webmineral provides
completely different information, particularly for La and Nd. The assumption of ideal (and literature)
mineral compositions is likely to become less reliable, as increasingly low grade deposits with multiple
enrichment mechanisms are exploited. This is because lower abundances/concentrations are
intrinsically more difficult to measure at the bulk level, and require more precise mineral chemistry for
mineral-chemical reconciliation and accurate calculation of elemental deportment.
Non-routine analyses: Rare earth-bearing minerals and phases
The demand for REE is well established and will maintain market relevance due to its requirement in
renewable technologies (Haque et al., 2014). A large number of REE ore deposits in southern Africa (and
worldwide) are carbonatite or carbonatite-laterite hosted (Harmer and Nex, 2016). REE enrichment
processes within such deposits often involve multiple episodes of hydrothermal alteration, resulting in
significant deviations from ideal mineral chemistry of REE-minerals. The precise elemental composition
of such REE phases therefore needs to be obtained in order to firstly identify the REE minerals present
(if possible), and secondly determine if the REE grade is suitable for mining and beneficiation.
The REE have similar atomic masses, sizes and electronic configurations, resulting in closely spaced or
overlapping energy signals on the X-ray spectrum. Wavelength dispersive spectrometry is ideally
suited to resolve such peak energies; however correction procedures are required to account for
unavoidable peak overlaps. The simplest correction protocol involves the measurement of peak heights
(in counts per second), in which the counts of the desired elemental peak and overlapping peak are
obtained to create a ratio or correction factor that can be applied to each measurement. Other more
rigorous correction procedures such as peak deconvolution may also be used (Takahashi et al., 1991).
While such correction procedures are well-known, some practical challenges are frequently
encountered when analysing REE-minerals from real world deposits. Table III shows EPMA rare earth
131
analyses from several carbonatite REE deposits in southern Africa and North America. Analysed
minerals include monazite, bastnäsite, synchysite and rhabdophane. The textures of monazite and
rhabdophane are shown in Figure 2.
Monazite elemental compositions are relatively consistent (compared to other REE phases), with
reasonably low standard deviations and near 100 weight percent totals. Textures are generally non-
euhedral, discounting a primary magmatic origin for such grains (Figure 2). Such textures may be
explained by alteration of primary monazite by iron-rich fluids which subsequently precipitated along
cracks within the original grains, creating a fine-grained secondary monazite enriched in Fe (Chetty et
al., 2017). The measured concentration of thorium is low, and was likely mobilised by alteration fluids,
further supporting a secondary monazite origin. A notable aspect of the monazite chemistry is the
fluorine content (~ 1 wt. %). The presence of F is not an analytical artefact, since similar proportions of
fluorine in monazite have been reported from the Huanglongpu (China), Hongcheon (Korea), Purulia
(India), Transbaikalia, Tomtor and Khaluta (Russia), Abyan (Yemen), Kizilcaoren (Turkey), Wincheeda
(Canada) and Mount Weld (Australia) carbonatite/phoscorite or carbonatite-laterite deposits (Chen et
al., 2017). The presence of F is not well understood some authors have speculated that F is incorporated
from apatite, since monazite and apatite are often texturally associated (Catlos and Miller, 2017). The
observation of multiple impurities (Fe, Si, Ca, Sr) in the monazite dataset (Table III) point toward minor
incorporation of elements from multiple gangue and ore species (iron oxide, quartz, apatite, REE-
fluorocarbonates and strontianite all observed within the mineralogical suite) during fluid-assisted
alteration of primary monazite.
Mineral compositions that are similar to monazite but produce sub-100 weight percent totals are
generally assigned as rhabdophane (Table III, Figure 2), and essentially represent a hydrated phosphate
species formed through monazite alteration (Berger et al., 2008). Rhabdophane is also commonly
observed in carbonatite deposits, sometimes in greater quantities than monazite. Cook et al., (2023)
suggests that this is because monazite is partially replaced by rhabdophane during alteration.
Within synchysite, Ca, La and Ce proportions vary significantly (standard deviation>2), indicating a
range in compositions between synchysite and bastnäsite. Bastnäsite compositions are better
constrained; however a number of intermediate compositions are not shown since they could not be
assigned to a mineral phase name based on their elemental compositions alone. This apparent variation
in chemistry between bastnäsite and synchysite has been previously noted in the magmatic rocks of
Madagascar (Berger, 2008) and within the Mount Weld carbonatite deposit (Cook et al., 2023). It is a
considerable challenge to ascertain whether certain analyses represent real mineral compositions (since
so few are reported), or are mixtures of previously and newly formed/partially recrystallised REE-
phases. One approach involves assessing the level of observed impurities (e.g. Fe – Cook et al., 2023) or
simply through textural observation. Such parameters are not always obvious, however, since REE-
mineral textures may vary subtly (usually only slight changes in back-scattered electron image contrast)
and REE-minerals may be hosted in a variety of gangue phases that contribute differing elemental
components as impurities in the analysis.
The complex nature of REE oxides and phosphates also creates some practical analytical concerns. The
application of matrix corrections to mixtures of minerals containing one or more phases/impurities is
problematic, as the equations used to calculate elemental concentrations assume a single matrix (Llovet
et al., 2021).
132
Table III. EPMA elemental compositions for various rare earth-bearing minerals
Figure 2. Back-scattered electron images showing the textural setting of rare earth minerals (monazite,
rhabdophane) hosted within iron-manganese oxyhydroxide particles from a ferruginised carbonatite deposit.
mean mi n max s.d mean min ma x s.d mean mi n ma x s.d mean min max s.d
F1.00 0.73 1.31 0.11 5.74 3.99 7.76 0.99 6.44 4.77 8.73 0.84 - - - -
Cl - - - - - - - - - - - - 0.09 0.03 0.20 0.05
Na
2
O - - - - - - - - - - - - 0.05 0.02 0.12 0.03
Al
2
O
3
0.02 0.02 0.03 0.01 - - - - 0.10 0.03 0.29 0.12 0.50 0.02 4.29 1.17
SiO
2
1.32 0.41 3.48 1.34 0.88 0.03 2.04 1.00 0.08 0.02 0.27 0.07 0.32 0.19 0.61 0.14
P
2
O
5
28.93 27.52 29.62 0.52 - - - - 0.13 0.04 0.89 0.14 25.99 15.02 28.20 3.21
CaO 0.32 0.12 0.48 0.08 0.09 0.03 0.23 0.08 16.21 7.54 24.63 2.42 2.98 1.85 4.09 0.53
TiO
2
- - - - - - - - 0.63 0.63 0.63 -0.52 0.07 0.98 0.64
MnO - - - - - - - - 0.18 0.10 0.29 0.10 0.93 0.06 2.51 1.09
FeO 0.58 0.04 1.73 0.66 - - - - 0.60 0.04 5.96 1.21 1.33 0.05 5.52 1.46
SrO 0.33 0.12 0.67 0.18 0.14 0.11 0.17 0.02 0.80 0.36 4.24 0.56 1.09 0.31 1.56 0.32
Y
2
O
3
- - - - 0.07 0.00 0.10 0.04 0.70 0.07 2.98 0.69 1.38 0.44 2.05 0.45
Nb
2
O
5
0.23 0.23 0.23 - - - - - 0.28 0.20 0.31 0.05 0.13 0.12 0.14 0.01
La
2
O
3
21.02 20.14 22.71 0.62 22.98 22.01 24.84 0.65 13.70 8.77 21.30 3.05 16.38 5.87 23.66 4.21
Ce
2
O
3
35.11 32.89 36.09 0.71 36.66 34.90 37.69 0.63 25.43 20.55 31.17 2.03 28.06 12.42 48.30 7.75
Pr
2
O
3
3.29 2.89 3.64 0.21 3.65 3.14 4.08 0.25 2.76 1.83 3.75 0.43 2.24 1.03 3.43 0.60
Nd
2
O
3
8.76 8.28 9.28 0.27 10.29 8.78 11.69 0.69 7.87 4.45 11.71 1.85 8.27 3.77 13.45 2.61
Sm
2
O
3
0.45 0.19 0.60 0.08 0.49 0.19 0.75 0.13 0.71 0.08 1.49 0.39 0.76 0.11 1.36 0.42
Eu
2
O
3
0.18 0.10 0.27 0.04 0.09 0.09 0.10 0.01 0.19 0.09 0.30 0.06 0.21 0.07 0.37 0.10
Gd
2
O
3
0.20 0.13 0.26 0.05 - - - - 0.26 0.17 0.43 0.10 0.33 0.15 0.69 0.16
Tb
2
O
3
0.12 0.10 0.13 0.02 0.09 0.00 0.12 0.06 0.13 0.09 0.17 0.03 0.08 0.08 0.08 -
Dy
2
O
3
- - - - 0.05 0.00 0.10 0.07 0.28 0.10 0.64 0.16 0.24 0.11 0.41 0.09
Ho
2
O
3
- - - - - - - - 0.28 0.28 0.28 -0.23 0.14 0.39 0.12
Er
2
O
3
0.13 0.13 0.13 - - - - - 0.15 0.12 0.18 0.03 0.13 0.08 0.21 0.04
Lu
2
O
3
0.15 0.15 0.15 - - - - - - - - - - - - -
ThO
2
0.65 0.17 2.15 0.64 0.18 0.16 0.21 0.04 0.65 0.15 1.74 0.37 0.42 0.13 1.31 0.37
UO
2
0.17 0.15 0.20 0.04 0.16 0.16 0.16 -0.16 0.16 0.16 - - - - -
-O=F 0.42 0.31 0.55 0.05 2.42 1.68 3.27 0.42 2.71 2.01 3.68 0.35
Total 100.50 99.32 100.97 0.45 77.64 75.34 80.71 1.21 73.15 69.46 75.83 1.43 90.21 82.01 93.20 2.76
rhabdophane (n =19)
synchysite (n = 52)
bästnasite (n=29)
Element (wt %)
s.d = standard deviation
133
Moreover, the compositions of such phases deviate from common REE standards (e.g., REE
phosphates). Standardisation ideally involves matrix-matched material (i.e., the standards are as close
as possible to the measured, unknown minerals). However, reliable standards for complex REE-species
[(e.g., REE-fluorocarbonates such as bastnäsite and synchysite) are not widely available (if at all)]this
problem is exacerbated when lower concentrations of REE are measured.
Future applications of EPMA in applied mineralogy and mineral processing
As the costs for processing and beneficiation increase across the value chain, one may expect a decrease
in budget available to companies to source complex mineralogical data (including EPMA). Compared
to other chemical methods (e.g., EDS), EPMA does present a more costly option with longer turn-around
times and the requirement of specialised operators. However, economic forecasts do not predict a
reduction in the demand for metals. Additionally, the surge in demand for technology metals (e.g., REE)
requires the assessment of complex, challenging deposits containing non-ideal mineral occurrences
with lower overall metal grades. Geometallurgical programmes therefore should not avoid the usage
of quantitative analytical techniques, and in some cases may need these methods for accurate
characterisation.
Aspects of the EPMA technique may shift, however, with the analysis of increasingly complex minerals
with lower required limits of detection (trace level detection in the ppm range). Methods for accurate
trace element detection have been developed using EPMA (e.g. Batanova et al., 2015; Donovan et al.,
2016) and has seen applications in petrochronology (Williams and Jercinovic, 2017) and economic
geology (e.g. Agangi et al., 2015; Sciuba and Beaudoin, 2021). Within the minerals processing industry,
the utilisation of trace element analysis by EPMA is not common currently. Some applications of this
method include the analysis of impurities affecting the flotation of copper ore minerals (Benzaazoua et
al., 2002) and the analysis of diamond micro-inclusions (Weiss et al., 2008).
Apart from lower detection limits, the spatial resolution of EPMA has increased to the sub-micron scale
through the incorporation of field emission guns (FE-EPMA). However, element quantification at this
scale requires the utilisation of lowered accelerating voltages (<10 kV). Standard EPMA operating
conditions include an accelerating voltage of 15-20 kV and probe current between 20-30 nA these
values are sufficient for electron ionisation of all elements, especially for the detection of high energy
Kα signals. Reduction in voltages and currents can complicate element quantification as lower energy
radiation signals (e.g. L-lines) need to be selected as a result, studies on low voltage element
quantification are currently limited (Rinaldi and Llovet, 2015). The high spatial resolution of FE-EPMA
is still a significant advantage as fully quantitative, non-destructive measurement of micrometric
features is not easily achieved with other techniques (e.g. measurement of sub-micron scale magnetite
and ilmenite in carbonatites - Milani et al., 2017).
The development of the soft X-ray emission spectrometer (SXES Takahashi et al., 2010) has the
potential for future application in industry due to its ability to detect lithium Kα signals. Lithium is a
critical element for current and future technology development, and is therefore highly relevant within
the exploration geology and mining sectors. Recent instrumental modifications and the development of
specific detectors have allowed for further characterisation of lithium signals using EPMA/SXES, with
significant progress being made towards lithium quantification (Hassebi et al., 2024; Schweizer et al.,
2024 – and references therein).
Finally, the combined utilisation of fully quantitative EPMA data with emerging techniques has seen
some development, and is likely to continue. For example, combined EPMA and laser-induced
breakdown spectroscopy (LIBS) was used to examine chromite and silicates from the UG-2 layer of the
Bushveld Complex (Meima et al., 2022), while both techniques were utilised for the characterisation of
uranium ore from the Czech Republic (Krempl et al., 2023). In a similar manner, the large-area mapping
capability of µ-XRF may be combined with EPMA for quantitative textural analysis. The combination
of a plethora of modern techniques (LIBS, µ-XRF, Raman and FTIR spectroscopy) in conjunction with
established techniques (EPMA, AutoSEM, EDS) was recently demonstrated by Sardisco et al., (2024) for
the investigation of battery minerals. Thus, EPMA remains a useful technique for accurate phase
134
chemical analysis, and, together with other techniques, assures characterisation commensurate with the
increasing complexity of ores and their associated processing/beneficiation products.
ACKNOWLEDGEMENTS
The author extends gratitude to Mr. Archie Corfield for contributing and assisting with microprobe
analyses and sample preparation at Mintek. Dr Deshenthree Chetty and Ms. Marian Manuel are also
acknowledged for checking and refining the original draft of the paper.
REFERENCES
Agangi, A., Przybyłowicz, W. and Hofmann, A. (2015). Trace element mapping of pyrite from Archean
gold deposits–A comparison between PIXE and EPMA. Nuclear Instruments and Methods in Physics
Research Section B: Beam Interactions with Materials and Atoms, 348, pp.302-306.
Anthony, J.W., Bideaux, R.A., Bladh, K.W. and Nichols, M.C, Eds. (2024). Handbook of Mineralogy,
Mineralogical Society of America, Chantilly, VA 20151-1110,
USA. http://www.handbookofmineralogy.org/ [accessed 30 July 2024].
Barthelmy, D., Webmineral Mineralogy Database. https://www.webmineral.com/ [accessed 30 July
2024].
Batanova, V.G., Sobolev, A.V. and Kuzmin, D.V. (2015). Trace element analysis of olivine: High
precision analytical method for JEOL JXA-8230 electron probe microanalyser. Chemical
Geology, 419, pp.149-157.
Benzaazoua, M., Belem, T. and Bussiere, B. (2002). Chemical factors that influence the performance of
mine sulphidic paste backfill. Cement and concrete research, 32(7), pp.1133-1144.
Berger, A., Gnos, E., Janots, E., Fernandez, A. and Giese, J. (2008). Formation and composition of
rhabdophane, bastnäsite and hydrated thorium minerals during alteration: Implications for
geochronology and low-temperature processes. Chemical Geology, 254(3-4), pp.238-248.
Berger, D. and Nissen, J. (2014). March. Measurement and Monte Carlo simulation of the spatial
resolution in element analysis with the FEG-EPMA JEOL JXA-8530F. In IOP Conference Series:
Materials Science and Engineering (Vol. 55, No. 1, p. 012002). IOP Publishing.
Buse, B. and Kearns, S.L. (2020), July. Spatial resolution limits of EPMA. In IOP Conference Series:
Materials Science and Engineering (Vol. 891, No. 1, p. 012005). IOP Publishing.
Catlos, E. J., & Miller, N. R. (2017). Speculations linking monazite compositions to origin: Llallagua tin
ore deposit (Bolivia). Resources, 6(3), pp 36.
Cavanough, G.L., Holtham, P.N. and Powell, T.M. (2006). Magnetic susceptibility measurement applied
to the minerals industry. Minerals Engineering, 19(15), pp.1588-1595.
Chantler, C.T., Olsen, K.J., Dragoset, R.A., Kishore, A.R., Kotochigova, S.A. and Zucker, D.S. (2003). X-
ray form factor, attenuation and scattering tables (version 2.0). http://physics.nist.gov/ffast
[accessed 30 July 2024]
Chen, W., Honghui, H., Bai, T. and Jiang, S. (2017). Geochemistry of monazite within carbonatite related
REE deposits. Resources, 6(4), pp.51.
Chetty, D., Clark, W., Bisaka, K. and Thobadi, I. (2017). Mineralogical studies in developing a novel
approach for the extraction of rare earth elements from ferruginised rare earth element ores,
Process Mineralogy 2017, MEI Conferences. https://www.min-
eng.com/processmineralogy17/paps.html [accessed 30 July 2024]
135
Contreras, M., Gázquez, M.J. and Bolívar, J.P. (2018). Mineral Separation and Characterization of the
Ilmenite Ore Phases: Optimization of the TiO2 Pigment Process. Int. J. of Thermal & Environmental
Engineering, 16(2), pp.105-112.
Cook, N.J., Ciobanu, C.L., Wade, B.P., Gilbert, S.E. and Alford, R. (2023). Mineralogy and Distribution
of REE in Oxidised Ores of the Mount Weld Laterite Deposit, Western Australia. Minerals, 13(5),
pp.656.
Deer, W.A., Howie, R.A. and Zussman, J. (2013). An introduction to the rock-forming minerals.
Mineralogical Society of Great Britain and Ireland.
Donovan, J.J., Singer, J.W. and Armstrong, J.T. (2016). A new EPMA method for fast trace element
analysis in simple matrices. American Mineralogist, 101(8), pp.1839-1853.
Frost, M.T., Grey, I.E., Harrowfield, I.R. and Mason, K. (1983). The dependence of alumina and silica
contents on the extent of alteration of weathered ilmenites from Western Australia. Mineralogical
Magazine, 47(343), pp.201-208.
Geldenhuys, I.J. (2013). June. Aspects of DC chromite smelting at MINTEK—an overview. 13th
International Ferro-Alloys Congress (INFACON XIII), Kazakhstan, Almaty, pp. 9-12.
Goldstein, J.I., Newbury, D.E., Michael, J.R., Ritchie, N.W., Scott, J.H.J. and Joy, D.C. (2017). Scanning
electron microscopy and X-ray microanalysis, 4th edn. Springer.
Haque, N., Hughes, A., Lim, S. and Vernon, C. (2014). Rare earth elements: Overview of mining,
mineralogy, uses, sustainability and environmental impact. Resources, 3(4), pp.614-635.
Harmer, R.E. and Nex, P.A.M. (2016). Rare earth deposits of Africa. Episodes Journal of International
Geoscience, 39(2), pp.381-406.
Hassebi, K., Rividi, N., Fialin, M., Verlaguet, A., Godard, G., Probst, J., Löchel, H., Krist, T., Braig, C.,
Seifert, C. and Benbalagh, R. (2024). High-resolution x-ray emission spectrometry in the lithium
K range with a reflection zone plate spectrometer. X
-
Ray Spectrometry.
Kanzari, A., Graul, S., Delaporte, A., Hints, R. and Blancher, S. (2024). Mineral Exploration of the
Senegalese Grande Côte heavy minerals placer: QEMSCAN® characterisation of Fe-Ti oxides for
ore modelling. EGU General Assembly 2024, Copernicus Meetings, No. EGU24-19227.
Kjarsgaard, B.A., Januszczak, N. and Stiefenhofer, J. (2019). Diamond exploration and resource
evaluation of kimberlites. Elements: An International Magazine of Mineralogy, Geochemistry, and
Petrology, 15(6), pp.411-416.
Kleynhans, E.L.J., Beukes, J.P., van Zyl, P.G. and du Preez, S.P. (2023). Chemical beneficiation of
chromite ore to improve the chromium-to-iron ratio for ferrochrome production. Minerals
Engineering, 201, pp.108196.
Krempl, I., Novotný, K., Wertich, V., Škoda, R., Kanický, V. and Leichmann, J. (2023). Distinguishing
secondary uranium mineralizations in uranium ore using LIBS imaging. Spectrochimica Acta Part
B: Atomic Spectroscopy, 206, pp.106734.
Llovet, X., Moy, A., Pinard, P.T. and Fournelle, J.H. (2021). Reprint of: Electron probe microanalysis: A
review of recent developments and applications in materials science and engineering. Progress in
Materials Science, 120, pp.100818.
McClenaghan, M.B. (2005). Indicator mineral methods in mineral exploration. Geochemistry: Exploration,
Environment, Analysis, 5(3), pp.233-245.
McSwiggen, P. (2014). March. Characterisation of sub-micrometre features with the FE-EPMA. In IOP
Conference Series: Materials Science and Engineering (Vol. 55, No. 1, p. 012009). IOP Publishing.
136
Meima, J.A., Rammlmair, D. and Junge, M. (2022). The use of Laser Induced Breakdown Spectroscopy
for the mineral chemistry of chromite, orthopyroxene and plagioclase from Merensky Reef and
UG-2 chromitite, Bushveld Complex, South Africa. Chemical Geology, 589, p.120686.
Milani, L., Bolhar, R., Cawthorn, R.G. and Frei, D. (2017). In situ LA–ICP-MS and EPMA trace element
characterization of Fe–Ti oxides from the phoscorite–carbonatite association at Phalaborwa,
South Africa. Mineralium Deposita, 52, pp.747-768.
Nash, W.P. (1992). Analysis of oxygen with the electron microprobe: Applications to hydrated glass and
minerals. American Mineralogist, 77(3-4), pp.453-457.
Pouchou, J.L. and Pichoir, F. (1991). Quantitative analysis of homogeneous or stratified microvolumes
applying the model “PAP”, Electron probe quantitation, pp. 31-75
Pownceby, M.I., Bruckard, W.J. and Sparrow, G.J. (2020). Technological developments in processing
Australian mineral sand deposits. Journal of the Southern African Institute of Mining and
Metallurgy, 120(2), pp.105-112.
Rinaldi, R. and Llovet, X. (2015). Electron probe microanalysis: A review of the past, present, and
future. Microscopy and Microanalysis, 21(5), pp.1053-1069.
Ritchie, N.W. and Lowers, H.A. (2018). Exploring the Limits of EDS Microanalysis for Rare Earth
Element Analyses. Microscopy and Microanalysis, 24(S1), pp.750-751.
Sardisco, L., Apeiranthitis, N., Hirani, J., Franzel, M., Jolis, E.M., Lukkari, S., Michallik, R.M., Tepsell, J.,
Pearce, T.J. and Butcher, A.R. (2024). Battery Mineral Characterization—A Case Study of a Nickel
Reference Material. Materials Proceedings, 15(1), pp.83.
Schirmer, T., Achimovičová, M. and Goldmann, D. (2020). Influence of chemical and phase composition
in the hydrometallurgical processing of FeTi oxide phases. Hydrometallurgy, 191, pp.105250.
Schweizer, P., Brackx, E. and Jonnard, P. (2024). A new device for high-resolution Li K X-ray
spectroscopy using an electron microprobe. Spectrochimica Acta Part B: Atomic Spectroscopy, 218,
p.106994.
Sciarone, M. (1998). South African chrome ore for the production of charge chrome. Geology, 2(10), pp.1.
Sciuba, M. and Beaudoin, G. (2021). Texture and trace element composition of rutile in orogenic gold
deposits. Economic geology, 116(8), pp.1865-1892.
Subramanian, S., Tammishetti, V. and Rai, B. (2016). Concurrent reconciliation of chemical and mineral
assays for mineral processing circuits. International Journal of Mineral Processing, 146, pp.1-9.
Takahashi, H. (1991). Application for Spectrum Deconvolution with EPMA. Document for the JEOL User’s
Meeting in 1991, Application Note XM62.
Takahashi, H., Murano, T., Takakura, M., Asahina, S., Terauchi, M., Koike, M., Imazono, T., Koeda, M.
and Nagano, T. (2016). Development of soft X-ray emission spectrometer for EPMA/SEM and its
application. In IOP Conference Series: Materials Science and Engineering (Vol. 109, No. 1, p. 012017).
IOP Publishing.
Thakurdin, Y. and Ashwal, L. (2016). Relation of magnetite composition to magnetic susceptibility in
Mauritian Basalts. Abstract 35th International Geological Congress, Cape Town, South Africa American
Geosciences Institute. Paper number 4747.
Weiss, Y., Griffin, W.L., Elhlou, S. and Navon, O. (2008). Comparison between LA-ICP-MS and EPMA
analysis of trace elements in diamonds. Chemical Geology, 252(3-4), pp.158-168.
Williams, M.L., Jercinovic, M.J., Mahan, K.H. and Dumond, G. (2017). Electron microprobe
petrochronology. Reviews in Mineralogy and Geochemistry, 83(1), pp.153-182.
137
Zhang, R.X. and Yang, S.Y. (2016). A mathematical model for determining carbon coating thickness and
its application in electron probe microanalysis. Microscopy and Microanalysis, 22(6), pp.1374-1380.
Dr Yash Thakurdin
Senior Scientist
Mineralogy Division at Mintek
Dr Yash Thakurdin is a Senior Scientist within the Mineralogy Division at Mintek
(YashT@mintek.co.za). He holds an Honours degree in Geochemistry and a PhD in Geology (University
of the Witwatersrand). His current focus is ore mineralogy and materials analysis using electron
microscopy and spectrometry (EPMA, LA-ICP-MS, SEM, µ-XRF, XRD).
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Implementing a newly developed spectrometer for the soft x‐ray range of (35–130) eV based on reflection zone plates was successfully accomplished on an electron probe microanalyzer. In this context, we present the first spectra acquired using this setup, including those of elements such as Be (Kα), C (Kα), Mg (L 2,3 ), Al (L 2,3 ), and Si (L 2,3 ). We have also conducted an analysis of several lithium compounds and measured the emission of Li Kα from metallic Li, LiF, and LiNbO 3 . Some of the results were compared with density functional theory calculations. The spectrum obtained for the lithium‐bearing mineral amblygonite Li 0.75 Na 0.25 Al(PO 4 )F 0.75 (OH) 0.25 is chosen to discuss some of the challenges faced.
Article
Full-text available
The Mount Weld rare earth element (REE) deposit, Western Australia, is one of the largest of its type on Earth. Current mining exploits the high-grade weathered goethite-bearing resource that lies above, and which represents the weathering product of a subjacent carbonatite. The mineralogy, petrography, deportment of lanthanides among the different components, and variation in mineral speciation, textures, and chemistry are examined. Microanalysis, involving scanning electron microscope (SEM) imaging, electron probe microanalysis (EPMA) and laser ablation inductively coupled-plasma mass spectrometry (LA-ICP-MS), was conducted on sized fractions of three crushed and ground laterite ore samples from current and planned production, and a representative sample from the underlying carbonatite. High-magnification imaging of particles in laterite samples show that individual REE-bearing phases are fine-grained and extend in size well below the micron-scale. Nanoscale inclusions of REE-phosphates are observed in apatite, Fe-(Mn)-(hydr)oxides, and quartz, among others. These have the appearance, particularly in fluorapatite, of pervasive, ultrafine dusty domains. Apart from the discrete REE minerals and abundant nano- to micron-scale inclusions in gangue, all ore components analysed by LA-ICP-MS contain trace to minor levels of REEs within their structures. This includes apatite, where low levels of REE are confirmed in preserved igneous apatite, but also Fe- and Mn-(hydr)oxides in which concentrations of hundreds, even thousands of ppm are measured. This is significant given that Fe-(Mn)-(hydr)oxides are the most abundant component of the laterite and points to extensive mobility and redistribution of REEs, and especially HREE, during progressive lateritisation. Late-formed minerals, notably tiny grains of cerianite, reflect a shift to oxidising conditions. REE-fluorocarbonates are the main host for REEs in carbonatite and are systematically replaced by hydrated, Ca-bearing REE-phosphates (largely rhabdophane). The latter displays varied compositions but is characteristically enriched in HREE relative to monazite in the same sample. Fine-grained, compositionally heterogeneous rhabdophane is accompanied by minor amounts of other paragenetically late, hydrated phosphates with enhanced MREE/HREE relative to LREE (although still LREE-dominant). Minor, relict xenotime and zircon are significant HREE carriers. Ilmenite and pyrochlore group members contain REE but contribute only negligibly to the overall REE budget. Although the proportions of individual mineral species differ, the chemistry of key ore components are similar in different laterite samples from the current resource. Mineral signatures are, however, subtly different in the lower grade southeastern part of the deposit, including higher concentrations of HREE relative to LREE in monazite, rhabdophane, florencite and Fe-(Mn)-(hydr)oxides.
Article
Full-text available
Rutile from a wide range of orogenic gold deposits and districts, including representative world-class deposits, was investigated for its texture and trace element composition using scanning electron microscopy, electron probe microanalysis, and laser ablation-inductively coupled plasma-mass spectrometry. Deposits are hosted in various country rocks including felsic to ultramafic igneous rocks and sedimentary rocks, which were metamorphosed from lower greenschist to middle amphibolite facies and with ages of mineralization that range from Archean to Phanerozoic. Rutile presents a wide range of size, texture, and chemical zoning. Rutile is the dominant TiO2 polymorph in orogenic gold mineralization. Elemental plots and partial least square-discriminant analysis suggest that the composition of the country rocks exerts a strong control on concentrations of V, Nb, Ta, and Cr in rutile, whereas the metamorphic facies of the country rocks controls concentrations of V, Zr, Sc, U, rare earth elements, Y, Ca, Th, and Ba in rutile. The trace element composition of rutile in orogenic gold deposits can be distinguished from rutile in other deposit types and geologic settings. Elemental ratios Nb/V, Nb/Sb, and Sn/V differentiate the rutile trace element composition of orogenic gold deposits compared with those from other geologic settings and environments. A binary plot of Nb/V vs. W enables distinction of rutile in metamorphic-hydrothermal and hydrothermal deposits from rutile in magmatic-hydrothermal deposits and magmatic environments. The binary plot Nb/Sb vs. Sn/V distinguishes rutile in orogenic gold deposits from other geologic settings and environments. Results are used to establish geochemical criteria to constrain the source of rutile for indicator mineral surveys and potentially guide mineral exploration.
Article
Full-text available
Electron probe microanalysis (EPMA) is a microanalytical technique widely used for the characterization of materials. Since its development in the 1950s, different instrumental and analytical developments have been made with the aim of improving the capabilities of the technique. EPMA has utilized crystal diffractors with gas detectors (wavelength-dispersive spectrometers, WDS) and/or solid-state detectors (energy-dispersive spectrometers, EDS) to measure characteristic X-rays produced by an electron beam. In this review, we give an overview of the most significant methodological developments of EPMA that have occurred in the last three decades, including the incorporation of large area diffractors, field-emission guns, high-spectral resolution X-ray grating spectrometers, silicon drift detectors, as well as more powerful Monte Carlo simulations, which have opened a wide range of new possibilities for the characterization of materials using EPMA. The capabilities of the technique are illustrated by a selection of representative applications of EPMA to materials science and engineering, chosen to show the current merits and limitations of the technique. Given the lack of coverage in previous reviews of the excellent capabilities of EPMA for measurements of thin films and coatings, that topic is covered in detail. We finally provide ideas for new research opportunities using EPMA.
Conference Paper
Full-text available
Electron microprobe data on titanomagnetites reveal varying abundances of Fe, Ti, Al, Mn, Mg, Ni, Zn and V. The components that displayed the greatest control on magnetic susceptibility are Fe2+/3+ and Ti4+. Magnetic susceptibility increases with total Fe (Fe2+ and Fe3+) abundance in titanomagnetite and decreases with increasing Ti4+ abundance. These observations are expected due to the general paramagnetism of Fe and diamagnetism of Ti components in magnetites, respectively. From the observed data, Ti4+ abundance in titanomagnetite (TiO2 = 3.25 – 26.17 wt. %) is likely the strongest controlling chemical element on magnetic susceptibility. It is a general first order assumption that greater abundances of magnetite within a sample results in higher magnetic susceptibilities. The abundance of magnetite in our samples shows an unexpectedly weak correlation to magnetic susceptibility; increasing amounts of magnetite resulted in only slightly higher magnetic susceptibilities.
Article
The main aim of this work is to demonstrate the potential of LIBS as a complementary technique to electron probe microanalysis (EPMA) for distinguishing and characterizing uranium mineralizations. Combining both methods can help estimate uranium oxidation states and monitor the possible mobilization of uranium in the environment by detecting oxygen and hydrogen using LIBS. It was confirmed that the LIBS signal of oxygen is proportional to oxygen content and that the strength of the oxygen signal is closely related to the oxidation state of uranium. The second assumption that the hydrogen signal is closely related to water (or hydroxyl group) content was also confirmed by detecting a stronger hydrogen signal from the presumed secondary mineralization. In contrast, hydrogen was not found in uraninite and quartz. When superimposed, images obtained with LIBS and EPMA show a clearly visible contrast between primary and secondary uranium mineralizations. Images of uranium obtained with the two techniques match perfectly, while the LIBS image of oxygen confirms the presence of an oxidized form of secondary uranium minerals (uranophane). The LIBS image of hydrogen clearly shows mineral phases containing water or a hydroxyl group, confirming that uranophane and other associated minerals contain greater amounts of hydrogen (water).
Article
A new approach using scanning Laser Induced Breakdown Spectroscopy (LIBS) was developed to unravel cryptic variations in the mineral chemistry of chromite, orthopyroxene, and plagioclase in drill core from Merensky Reef and UG-2. The LIBS method provides continuous spatially resolved chemical data, elucidating variations in mineral chemistry and in element partitioning, taking into account textural features of the rock itself, of individual minerals, and of immediate neighbourhoods. Since the method is relatively fast and easy to apply to numerous core sections, it is of special interest when comparing the lateral variability of diagnostic horizons such as Merensky Reef and UG-2. This approach for measuring Mg/Fe, Cr/Al and Ca/Na ratios is based on the strength of LIBS in detecting chemical variations using intensity ratios within a single matrix, which in this application is one single particular type of mineral phase. Supervised classification followed by k-means clustering was applied to identify those spectra that consist of pure chromite, orthopyroxene, and plagioclase. Measurements and data processing strategies were focused on minimizing the standard deviation in the LIBS-based ratios to be able to detect relatively small variations in mineral chemistry. The LIBS-based Cr/Al and Mg/Fe ratios were successfully validated with respect to electron probe microanalysis on multiple polished sections. Distinct trends could indeed be extracted from a continuous LIBS measurement of 6 m drill core through the Merensky Reef. The data correlate with whole-rock geochemistry and hydrous phases such as phlogopite. The LIBS-based approach opens up unique insights into chemical and mineralogical expressions of large- and local-scale processes. This allows unique and detailed interpretations relevant for the genesis of mineralizations and metal distributions, based on a comprehensive dataset.