Thermal decomposition of chromite spinel with chlorite admixture

Article abstract of DOI:10.1016/j.tca.2008.07.003

The behaviour of minerals in a South African chromite ore during the increasing of the temperature has been studied. Firstly, the changes produced during the ignition process have been examined by means of thermal and differential analysis (TGA-DTA) until 1200 °C. The characterization of the initial mineral and those obtained after heating at several temperatures in room atmosphere has been performed by X-ray diffraction (XRD). Moreover, voltammetric analyses have allowed to determine the variation of the iron oxidation degree in the studied materials. Light microscopy was applied to find more information about the different phases by their colour. During the heating, a wide range of complex exothermic and endothermic transformations take place. Decomposition compounds were identified, which were produced by heat decomposition, loss of structural water, element substitutions and oxygen absorptions and desorptions, caused mainly by the variation of the iron oxidation degree. The spinels of the chromite ore decompose in other spinels, with a partial change of the iron oxidation degree. From nearly 800 °C, chrome oxide (Cr₂O₃) comes off from the chromite forming another phase, and almost at 1000 °C, a slow decrease of weight was detected, caused among others to the formation of a magnetite phase. Simultaneously, the silicates undergo strong modifications, including decompositions and incorporation of iron (II) in their structure and producing other silicates stable at high temperatures, which modify the behaviour of the pure spinels. Moreover, at 1200 °C these silicates decompose to cristobalite (SiO₂).

Full text of DOI:10.1016/j.tca.2008.07.003

Thermochimica Acta 476 (2008) 11–19
Contents lists available at ScienceDirect
Thermochimica Acta
journal homepage: www.elsevier.com/locate/tca
Thermal decomposition of chromite spinel with chlorite admixture
S. Sánchez-Ramosa,b, A. Doménech-Carbóa, J.V. Gimeno-Adelantadoa,∗
,
J. Peris-Vicentea, F.M. Valle-Algarraa
a Department of Analytical Chemistry, Faculty of Chemistry, University of Valencia, C/Doctor Moliner 50,
46100-Burjassot, Valencia, Spain
b Escuela Superior de Cerámica, C/Ceramista A. Blat 22, 46940 Manises, Valencia, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
The behaviour of minerals in a South African chromite ore during the increasing of the temperature has
been studied. Firstly, the changes produced during the ignition process have been examined by means of
thermal and differential analysis (TGA–DTA) until 1200 ◦C. The characterization of the initial mineral and
those obtained after heating at several temperatures in room atmosphere has been performed by X-ray
diffraction (XRD). Moreover, voltammetric analyses have allowed to determine the variation of the iron
oxidation degreeinthestudied materials. Light microscopywasapplied tofindmore information aboutthe
different phases by their colour. During the heating, a wide range of complex exothermic and endothermic
transformations take place. Decomposition compounds were identified, which were produced by heat
decomposition, loss of structural water, element substitutions and oxygen absorptions and desorptions,
caused mainly by the variation of the iron oxidation degree. The spinels of the chromite ore decompose
in other spinels, with a partial change of the iron oxidation degree. From nearly 800 ◦C, chrome oxide
(Cr2O3) comes off from the chromite forming another phase, and almost at 1000 ◦C, a slow decrease of
weight was detected, caused among others to the formation of a magnetite phase. Simultaneously, the
silicates undergo strong modifications, including decompositions and incorporation of iron (II) in their
structure and producing other silicates stable at high temperatures, which modify the behaviour of the
pure spinels. Moreover, at 1200 ◦C these silicates decompose to cristobalite (SiO2).
Received 19 February 2008
Received in revised form 30 June 2008
Accepted 14 July 2008
Available online 25 July 2008
Keywords:
Chromite
X-ray analysis
Microscopy
Voltammetry
Ceramic industry
Thermal behaviour
© 2008 Elsevier B.V. All rights reserved.
1. Introduction
been made by the products from previously reduced refractory
chromites with the addition of reducing agents.
Pure chromite has
a
spinel crystalline structure and the
Iron oxidation state in the chromites has been studied by means
of solid-state electrochemistry. This methodology, which is based
on the mechanical transference of solid microparticles to graphite
electrodes in contact with a suitable electrolyte. was selected
because of its ability to directly characterize the oxidation state of
electroactive species [6–8]. This approach has been applied recently
to characterize iron oxides [9–12]. Chemical and structural changes
occurring as a result of thermal treatments were correlated with
the variations in the electrocatalytic effect exerted by the studied
materials on the reduction of dissolved oxygen and the oxidation of
water to oxygen in alkaline aqueous solutions. This is based in the
reported electrocatalytic ability of NiFe2−xCrxO4 spinel for promot-
ing O2 evolution in alkaline aqueous media [13] and the catalytic
effect exerted by Fe3O4 cathodes on the electrochemical reduction
of dissolved O2 [14].
theoretical chemical composition is FeCr2O4. Natural chromite
structure is described as a cubic spinel with a formula unit
(Mg,Fe2+)O·(Cr,Al,Fe3+)2O3, owing to the partial natural substitu-
tion of Fe(II) by Mg, Cr by Al and even by Fe(III) in lower proportion.
These occurrences depend on the type of mineral which is usually
found with the chromite, especially metha- and orthosilicates.
On the other hand, natural chromite minerals form extended
solid solutions with binary spinels of FeCr2O4, MgCr2O4, FeAl2O4
and MgAl2O4. The decomposition of natural spinels strongly
depends upon the chemical potential imposed in the forms of tem-
perature, pressure and pH differences in aqueous media.
Some interesting studies have been realized about the viabil-
ity of the use of chromites as chrome-diffusive covering materials
in steel [1–5]. The introduction of chrome in the surface has
In the case of pure iron chromite, the phase reaction in spinels
can be defined as the decomposition products at high temperatures
[15], which are FeO and Cr2O3. The chemical strongly depends on
the mineral composition and oxygen partial pressure, because the
∗
Corresponding author. Tel.: +34 963544533; fax: +34 963544436.
E-mail address: jose.v.gimeno@uv.es (J.V. Gimeno-Adelantado).
0040-6031/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.tca.2008.07.003

12
S. Sánchez-Ramos et al. / Thermochimica Acta 476 (2008) 11–19
chromite mineral is a solid solution of pure spinel end members
and the Fe(II) will oxidize to Fe(III) in the presence of an excess of
oxygen.
cia, Spain); Wolfram Carbide disk mill (Fritsch Pulverisette 9,
Idar-Oberstein, Germany); crucible (5% Au/Pt ZGS); glass discs con-
formator (Pt/Rh 30 mm diameter). XRF analyses were performed
in an X-ray fluorescence spectrometer (Phillips PW with rhodium
anticathode controlled by SuperQ/Quantitative software version
1.1, Eindhoven, Netherlands).
Studies by XRD were carried out with X-ray diffractometer
Siemens D500 (using CuK␣) controlled by DIFFRACT/AT version 3
software. Munich, Germany) with DIFFRACTplus Evaluation Pack-
age Release 2001 and Powder Diffraction File (PDF) Release 1999
Database, provided by International Centre for Diffraction Data
(ICCD).
Many investigations on chromite mineral phase equilibria have
been undertaken [16,17]. However, the main focus of the previous
research works was on the high temperature region (above 1300 ◦C)
related to the production of ferrochrome alloys. By contrast, the lit-
erature on the phase transformation for oxidation conditions below
1200 ◦C, used for the extraction of sodium chromate from chromite
espinels is rather limited [18,19]. Finally, Tathavadkar et al. used dif-
ferent techniques for the study of the decomposition of a refractory
chromite with an amount of free silica less than 1% [20].
Hellwege and Hellwege [21] have proposed the following
sequence for the decomposition of the spinel at low temperatures
(less than 600 ◦C):
Photographs were taken (144×) with a light microscope SZX12
Olimpus (Milan, Italy) coupled to a digital camera Olympus 4.5 M.p.
noise reduction system F 1.8.
Square wave voltammograms (SQWVs) were obtained with a
CH I420 equipment in a conventional three-electrode cell using
well-deaerated HCl (Panreac, Barcelona, Spain) solutions at 298 K.
A sample-modified composite electrode was used as a work-
ing electrode, a Pt-wire auxiliary electrode and a AgCl (3M
NaCl)/Ag reference electrode completed the three-electrode cells
after immersion of modified electrodes in suitable deaerated solu-
tions at 298 K. The potential scan was routinely initiated at +1.0 V
in the negative direction.
Spinelss → (T < 600 ◦C) SpinelI + SpinelII
→ SpinelIII + Sesquioxide (T > 600 ◦C)
(Mg0.45; Fe0.55)(Cr0.61; Al0.29)2O4 → (Mgx; Fe1−x)(Cr1−y; Aly)2O4
+(Fe)(Fe1−a; Ala)2O4 → (Mg)(Cr1−y˙ ; Aly˙ )2O4 + (Fe1−a˙ ; Ala˙ )2O3
It means that the oxidation of Fe(II)–Fe(III) produces firstly
maghemite (␥-Fe2O3), forming a metastable defective spinel,
although the reported transition temperature of maghemite to
hematite (␣-Fe2O3) is around 250 ◦C.
Moreover, some authors have studied the thermic behaviour of
previously purified chlorites, providing interesting data about the
temperature decomposition of the different analyzed specimens
[22–24].
The aim of this paper was for determining the behaviour of a
South African chromite ore during its heating. The study was carried
out from a chrome spinel mineral with an appreciable amount of sil-
icates. The mineralogical composition of the sample was as follows:
magnesiochromite ferroan (spinel), labradorite, chlorite chromian
and free silica. The chemical composition was: 42.88% Cr2O3, 6.53%
SiO2, 15.35% Al2O3, 21.24% Fe2O3 0.48% CaO, 12.76% MgO, 0.16%
Na2O, 0.15% MnO and 0.44% others (TiO2, K2O, etc.), where all the
iron was as Fe(II) [25]. Different materials were obtained by ignition
at several temperatures (450, 600, 800, 1000 and 1200 ◦C), from the
initial chromite ore, and the modification of the weight was con-
sidered. The obtained samples were studied by XRD in order to
characterize the formed compounds at each temperature. On the
other hand, a TGA-DTA of the mineral was carried to 1200 ◦C. The
study was completed by the observation by light microscopy of the
most significant changes detected in the powder. Finally, solid-state
voltammetry was used for determining the changes in the oxida-
tion state in the samples using Fe(III) oxide and Fe(II) oxalate as
reference compounds.
2.2. Electrochemical procedure
Experiments were performed in 1 M HCl with nanopure aqueous
solutions. Paraffin-impregnated graphite electrodes (PIGEs) con-
sist on cylindrical rods of 5 mm diameter of graphite impregnated
under vacuum by paraffin. The samples (ca. 10–20 ␮g) were com-
pletely powdered in an agate mortar and pestle, and then placed
on a glazed porcelain tile forming a spot of finely distributed mate-
rial. Then, the lower end of the PIGE was gently rubbed over that
spot of sample, and finally cleaned with a tissue paper to remove
ill-adhered particles [6]. Iron (III) oxide (Fe2O3, Aldrich) was used
as reference material for electrochemical experiments. Electro-
catalytic experiments were performed in O2-saturated aqueous
solutions of NaOH (Panreac) using uniform deposits of the different
chromites over PIGEs.
3. Results and discussion
The samples were obtained by heating of the natural chromite
ore (3.5 g in a Pt crucible) at the following temperatures: 450, 600,
800, 1000, 1100 and 1200 ◦C in an oven, maintained during 30 min.
3.1. Voltammetric study
In order to determine the oxidation state of iron in the untreated
natural chromite minerals and thermally treated specimens, solid-
state voltammetric experiments were performed upon immersion
of sample-modified graphite electrodes into 1.0 M HCl. The results
for a) the untreated chromite ore and the specimens resulting
from its ignition at (b) 600, and (c) 1100 ◦C are shown in Fig. 1.
The voltammogram of chromite ore at room temperature (Fig. 1a)
shows a unique cathodic peak at ca. −650 mV.
2. Experimental
2.1. Instrumentation
This peak is attributable to the electrochemical reductive disso-
lution of chromite Fe (II) to Fe metal, as suggested by comparison
with the voltammetric response of iron (II) containing clays [9,10].
Following the literature data [6–12], the electrochemical process
can be represented as:
TGA and DTA experiments were performed using a Univer-
sal V 3.0G TA instrument, with ␣-alumina powder as reference
material. Throughout the experiment, the sample temperature and
weight-heat flow changes were continuously monitored using a
data logging device.
Glass discs for the XRF analyses were made from the sam-
ples using the following instrumentation: muffle furnace (GALLUR,
max T 1300 ◦C with temperature ascent speed regulator, Valen-
FeCr2O4(solid) + 8H+(aq) + 2e− → Fe(solid) + 2Cr3+(aq) + 4H2O
(1)

S. Sánchez-Ramos et al. / Thermochimica Acta 476 (2008) 11–19
13
ca. −150 mV, accompanied by a sharp peak at −650 mV (Fig. 1c).
This electrochemical response denotes that hematite turns to mag-
netite, whose presence is characterized by the reduction peak at
−150 mV [27], which can be described as:
Fe3O4(s) + 8H+(aq) + 2e− → 3Fe2+(aq) + 4H2O
(3)
The peak at −650 mV is caused by the electrochemical reduction
of the remaining iron (II) in the chromite ore heated at 1100 ◦C.
The remaining of Fe(II) in chromite ore sample for the whole
ignition process has not been ever detected in previous works about
the study of this kind of minerals and can be considered as a strong
novelty. The peak corresponding to the electrochemical reduction
of iron (II) appears in the voltammograms obtained by chromite
ore in all the samples ignited at the considered temperatures. It
may indicate the incorporation of the remaining iron (II) during the
heating in a stable compound after the decomposition of chlorite.
It might be also caused by a magnetite thermal reduction at the
higher considered temperature.
In order to obtain additional information on structural changes
accompanying thermal treatment of chromite ores, we introduce
here a novel methodology based on the different catalytic ability of
the studied systems, acting as electrode modifiers, with regard to
selected electrochemical processes. Here, it is studied the variation
of the catalytic effect exerted by chromite samples on the electro-
chemical reduction of dissolved oxygen and the oxidation of water
in O2-saturated aqueous NaOH solutions.
This can be seen in Fig. 2, where cyclic voltammogram recorded
at: (a) unmodified, and (b) chromite-modified PIGEs immersed into
O2-saturated 0.10 M NaOH. A reduction peak appears ca. −0.45 V,
corresponding to the reduction of dissolved O2. In the positive
region of potentials, a prominent rising current was obtained, cor-
responding to the oxidation of water to O2. As can be seen on
comparing Figs. 2a and b, both the current for O2 reduction and the
current for water oxidation recorded at chromite-modified elec-
trode are significantly enhanced with respect to the currents at
the bare graphite electrode. For our purposes, the relevant point to
emphasize is that the chromites studied here exert a significantly
different catalytic effect towards such electrochemical processes
which can be reflected in changes in the peak current at −0.45 V
and the maximum current at the extreme positive potential of the
voltammograms (+1.25 V). The variation of the catalytic current for
both processes, i(cat), with the temperature at which the chromite
sample was treated, is shown in Fig. 3, expressed as the quotient
between the actual current for a given sample, and the maximum
current in the series, i(cat)/imax(cat).
As recently studied by Singh et al. [13], octahedrally coordi-
nated surface Fe2+ ions are considered the active centres for the
catalysis of electrochemical reduction of O2 in alkaline aqueous
media. According with the proposed reaction scheme for chromite
ores, the catalytic effect on that process exerted by the studied
chromite samples should decrease from the parent chromite to
the sample treated at 450 ◦C, further increasing. This prediction
agrees well with the observed results, as can be seen in Fig. 3
(squares).
Fig. 1. Square wave voltammograms for: (a) pristine chromite specimen, (b)
specimen after ignition at 600 ◦C, and (c) after heating at 1100 ◦C, attached to
paraffin-impregnated graphite electrodes in contact with 1.0 M HCl. Potential scan
initiated at +0.85 V in the negative direction. Potential step increment 4 mV; square
wave amplitude 25 mV; frequency 5 Hz.
This was confirmed by the appearance of an oxidation peak at
−0.30 V, displaying typical features for oxidative dissolution (strip-
ping) processes, in voltammograms initiated at −850 mV in the
positive direction. This peak corresponds to the oxidation of Fe solid
to Fe2+ ions in solution phase [9,10].
Specimens prepared by heating chromite at 600 ◦C (Fig. 1b)
show a prominent peak at −420 mV, and accompanied by a weaker
peak at −650 mV. The peak at −420 mV can be described in terms
of the electrochemical reductive dissolution of hematite, in agree-
ment with extensive literature data [6–11,15], whereas the signal
at −650 mV is attributable to the electrochemical reduction of
remaining iron (II) in natural chromite minerals. This voltammetry
indicates that hematite is formed from chromite ore during ther-
mal turnovers. The electrochemical reduction of hematite can be
described as:
In turn, the variation of the catalytic current for the reduction of
dissolved oxygen in aqueous alkaline solution parallels the afore-
mentioned pass from Fe2+ to Fe3+ at 450 ◦C and its subsequent
re-formation of Fe2+ at temperatures above 800 ◦C. Eventually, this
catalytic effect can be influenced by structural changes in the spinel
structure and the formation of high-valence chromium species [14].
The variation of the catalytic current with temperature of thermal
treatment agrees well with that observed for thermal analysis of
samples (vide infra).
Fe2O3(s) + 6H+(aq) + 2e− → 2Fe2+(aq) + 3H2O
(2)
Specimens prepared from natural chromite minerals by heat-
ing at 1100 ◦C show voltammograms consisting of a broad peak at

14
S. Sánchez-Ramos et al. / Thermochimica Acta 476 (2008) 11–19
Table 1
Weight increment of the sample by heating at each temperature (compared with
the initial)
T (◦C)
Weight increment (%)
450
+0.89
600
+0.66
800
+0.76
1000
+1.10
1100
+1.08
1200
+0.99
Weight increment can be considered as negative LOI.
3.2. Ignition test
The variation of the weight at different temperature heating was
studied. The results are shown in Table 1.
From these obtained data, it can be deduced that the weight
of the sample get higher when heated at 450 ◦C, under the assay
conditions. This increasing is minor at 600 ◦C, and then the slope
returns to rise from 600 to 1000 ◦C, reaching the maximal value.
Above this temperature, a progressive loss of weight was observed
until 1200 ◦C.
In principle, it could be deduced from this assay that a global
weight rising happens while the temperature reached 1000 ◦C.
This can be caused by the increasing of the iron oxidation state
by oxygen absorption. Between 450 and 800 ◦C, the increasing of
weight would be partially hindered by mineral decomposition in
the ore, with loss of several compounds and/or desorption of oxy-
gen, caused by the lessening of the iron oxidation state at these
temperatures. From 1000 ◦C, a continual process with loss of weight
was observed.
3.3. Thermal analysis
Chromite ore sample was analyzed by thermogravimetric study
until 1200 ◦C, with the aim to obtain more precise informations
from the heating process of the mineral. Two temperature gradi-
ents, 10 and 20 ◦C min−1, were tested. The obtained results were
similar in both cases, so that the final assay was performed at
20 ◦C min−1. Experimental data in the ignition test (Section 3.2)
could differ from the thermogram (Fig. 4), because they were
obtained from a continuous heating at a fixed temperature gradient.
Therefore, specific results found by ignition at the several tempera-
tures can be only used as basic information to propose hypotheses
about the processes which can happen during the heating.
Fig. 2. Cyclic voltammograms for: (a) unmodified, (b) chromite-modified PIGEs
immersed into O2-saturated 0.10 M NaOH. Potential scan initiated at 0.0 V in the neg-
ative direction; extreme potentials +1.25 and −0.85 V. Potential scan rate 50 mV/s.
3.3.1. Thermogravimetric analysis
The following observations were extracted from the analysis of
the obtained thermogram curves (Fig. 4):
In the TGA curve, the first interval points to a continuous increas-
ing from the initial temperature to 180 ◦C (A), due to a low weight
gain cause of the heating of oven atmosphere. The density decrease
produces a rising of the sample weight with a similar effect to
“Archimedes’ Principle”.
The slope of the TGA curve diminished between 180 and 400 ◦C
(interval A-B) due to two simultaneous and opposed effects:
(a) a weight increasing by the initial oxidation Fe2+–Fe3+ (as
maghemite phase ␥-Fe2O3 [28], which turns to hematite ␣-
Fe2O3 from 400 ◦C) [20]. However, this transition normally
takes place at around 250 ◦C. This increase in transition temper-
ature can be caused by the presence of Al(III), as they occupy
holes in the maghemite structure [21].
(b) A low decrease of weight due to the loss of structural water
(dehydroxylation stage) of the chlorites, which maintains the
typical bilayer structure of chlorites and other similar minerals,
and is normally removed at this temperature range.
Fig. 3. Variation of the i(cat)/imax(cat) ratio with the temperature of thermal treat-
ment for chromite samples studied here corresponding to the catalytic oxidation
of water to oxygen (solid squares) and the catalytic reduction of dissolved oxygen
(squares). From cyclic voltammograms using the conditions described in Fig. 2.
In the interval B–C of TGA curve, the slope shows a higher
value from 400 ◦C to nearly 550 ◦C. This is caused because of the

S. Sánchez-Ramos et al. / Thermochimica Acta 476 (2008) 11–19
15
Fig. 4. DTA and TGA thermogram curves obtained by means of the ignition test applied to the South African chromite ore.
oxidation of from Fe(II) to Fe(III) happens without other effects, as
the previously discussed elimination of water.
In the interval 6–7 of the DTA curve, the slope becomes more
negative, which is attributed to the thermal reduction of the
sesquioxide phase to ferrous-magnetite phase, according to the
deductions taken from voltammetry analysis (Section 3.2). Further-
more, from 1200 ◦C (endothermic peak 7), the slope of the DTA
curve turns strongly positive. This point to the occurrence of an
exothermic process, owing possibly to decomposition of silicates
to free SiO2 and the incorporation of their elements to the spinels.
In the interval C–D (550–700 ◦C) of TGA curve, the thermogram
shows a decrease of weight, owing to the loss of hydroxy ions (OH−)
from the structure of the chlorite (silicate phase of the mineral) dur-
ing its decomposition. This process overcomes the sample weight
by oxidation of Fe(II).
In the interval D–F of TGA curve, the increasing of the weight due
to the exothermic oxidation of Fe(II), continues until nearly 1100 ◦C.
The slope was smaller than in interval B–C, and this decrease was
more marked for the interval E–F (850–1100 ◦C). This was caused by
the minor amount of Fe(II) in the sample, probably together to iron
(II) migration from the chromite spinel to the silicate, after chlo-
rite decomposition. The formed iron (II) silicate is stable at these
temperatures and do not undergo oxidization.
3.3.3. Extracted deductions from thermal analysis
As summary it can be deduced that the behaviour of the
chromite spinel ore during the heating can be considered highly
complex. Indeed, the following processes have been established by
a series of phenomena: (i) overlapped changes of phases due to
oxygen absorptions and desorptions, (ii) substitutions in the net of
the cationic compounds, (iii) decompositions of the mineral and the
siliceous gangue (which natureandamountis determinant inallthe
studied process), between others. The process can be considered as
a global procedure, including a first oxidation of the iron in the min-
eral and followed by a thermic reduction of the iron oxidation state,
process interfered by the chlorite behaviour.
Finally from 1100 ◦C (point F) to higher temperatures, a strong
fall of the weight was clearly observed in TGA curve. This effect
is caused by the thermoreduction of iron (III) to lower oxidation
degree, probably magnetite phase, among others. These deductions
are in agreement to the results obtained by voltammetric studies
(Section 3.1).
Therefore, TGA and DTA curves are consequence of the
behaviour and interaction between the three compounds present
in the mineral sample. Although the main compound, the chromite
spinel (Mg,Fe2+)(Cr,Al)2O4 is responsible of the general behaviour
of the curves. Also they are affected by the silicates in the sample,
as labradorite (Na2.84Ca4.16Al12Si20O64) and, specially, chromium
chlorite (Mg5.1Al1.2 Si3Cr0.7O10(OH)8).
On this account, Tathavadkar et al. [20] have carried out a study
about TGA and DTA curves from 130 ◦C. They used a chromite
mineral spinel with less than 1% in silicates as impurities. The
most important results were the apparent stability of the sample
weight until 300 ◦C with a fast increasing of the weight, and an
exothermic peak between 300 and 400 ◦C. That corresponds to
the peak 3 of the DTA curve of the present study (Fig. 4), although
slightly switched. The exothermic interval is even followed by a
gain step until 637 ◦C, which is smaller than in the previous range.
This was followed by an interval with a higher positive slope until
3.3.2. Differential thermal analysis
At the beginning of the analysis, the DTA curve indicates an
endothermic process caused by the heating, which provokes an
endothermic peak at 200 ◦C (peak 1).
The DTA curve shows an exothermic step (interval 1–3) with
a slight and irregular slope to the peak 2, and an increasing of the
slope to nearly 420 ◦C (peak 3), according to the proposed processes
deducted from TGA for these temperature intervals (Section 3.3.1).
From 420 ◦C, an endothermic process occurs with a peak at 650 ◦C
(peak 4) and followed by an increasing of the heat-flow with an
exothermic peak (5), related to the process described for this inter-
val in TGA curve (Section 3.3.1), together with the decomposition
of original pure chromite spinel in two other spinels [21]. From
800 ◦C (peak 5 of the TDA curve), the heat-flow showed a strong
fall, due probably to the chromite segregation process (endothermic
process), which ends nearly at 1100 ◦C (peak 6).

16
S. Sánchez-Ramos et al. / Thermochimica Acta 476 (2008) 11–19
900 ◦C, exclusively corresponding to an endothermic process. The
weight decrease due to the loss of dehydroxylation stage of the
chlorite (explained by endothermic peak 4 and exothermic peak 5
of the present work) has not been observed by Tathavadkar et al.
Moreover, the DTA curve shows stability from 750 to 930 ◦C. Those
authors neither explain the evolution of the curve to 1100 ◦C [20].
It is important to indicate that in this work the thermal reduction
of the iron (III) in the last step has been detected for the first time
in the mineralogical study of this kind of materials.
On the other hand, Damodaran and Somasekar obtained pure
chlorite from a chromite mineral. The chromium chlorite is sep-
arated from the remaining minerals by grinding to −200 mesh
(B.S.S.) by hand picking using heavy liquids and a Frantz,s Iso-
dinamic Separator [22]. The purified fraction of the chromium
chlorite was subjected to Differential Thermal Analysis. The curve
has two endothermic peaks, one at 720 ◦C and the other at 850 ◦C,
and another exothermic peak at 825 ◦C.
Other authors deal with different kinds of chromium chlorite,
reducing the first endothermic peak even to 600 ◦C, and they pro-
posed the presence of another exothermic point in the majority of
the studied chromium chlorite, at 350 ◦C. This peak indicates a weak
oxidation, suggesting the absence of oxidation during heating in
chlorites which does not show this peak [24,26]. Those previously
exposed are according to the TDA curve (Fig. 4). The mentioned
peaks would correspond to the exothermic peaks 3, 5 and the
endothermic peak 4 in the Fig. 4. The second exothermic peak was
not visible in DTA.
On the other hand, the fact that the endothermic points found
in this study appear at minor temperatures would have a justi-
fication. Laphman has correlated the percentage of Cr2O3 of the
chlorite content with the temperature of the first endothermic peak
and obtained a curve which shows a decrease on the decomposi-
tion temperature when the amount of Cr2O3 increases [24]. The
oxide content in the chromium chlorite studied by these authors
was 3–4%, whereas the chlorite of the mineral described was 9.3%
in this work. Another interesting difference was the absence of iron
in the composition of the chromium chlorite of the studied sam-
ple. However, the referred sample shows a little amount (total iron
0.4–1.8%).
and 1200 ◦C have been observed by light microscopy in order to
identify the kind of chlorite.
From the observations of the samples, the following deductions
were extracted:
The photography from the natural chromite ore shows particles
with brown and especially orange colour, typical in chrome spinels
with large amount of Fe(II), in several grey tones. Besides, other
white and white-yellow particles, characteristic in silica and sili-
cates, have been as well detected. However, the most interesting
observation was the presence of numerous inclusions with violet
colour, together marine blue interference colour (depending on the
luminosity). These points to the presence of chromium chlorite, dis-
carding the occurrence of antigorites and Fe–Mg chlorites, which
can be characterized by their green colour [25].
In the 450 ◦C-heated mineral, an appreciable diminishing of
the orange particles and the appearance of brown-red ones was
observed. This can be explained by the partial oxidation of Fe(II) in
the spinel and the formation of other spinels with Fe(II) and Fe(III).
In the 600 ◦C-heated sample, a small amount of green particles
appear and the violet colour disappears. This indicates the decom-
position of chlorite and formation of olivine (Fe and/or Mg silicate).
Moreover, a gradual increasing of the green colour is observed in
the 800 ◦C-heated sample.
In the 1000 ◦C–1200 ◦C-heated sample, the green colour
strongly enlarge, owing to the separation of the formation of a Cr2O3
phase.
In all cases, white particles may be observed, due to the presence
of other minority compounds, as silicates or several spinels, among
others.
3.5. Study of the mineralogical composition of the materials
obtained by ignition of the studied chromite ore by X-ray
diffraction
In order to study the behaviour of the unknown sample during
the heating process, the diffractograms of the samples obtained by
ignition at the referred temperatures were measured, comparing
them with that obtained for the original sample. The Fig. 5 displays
the superposed diffractograms obtained from (below to above):
original South African chromite ore sample [25], and those heated
at 450, 600, 800, 1000, 1100 and 1200 ◦C.
3.4. Study by light microscopy
The accurate chemical composition of the obtained minerals by
the interpretation of the diffractograms shows a high difficulty,
owing to, among others, to the similar values of the interplanar
Powdered and sieved samples from the South African chromite
ore, natural and obtained after heating to 450, 600, 800, 1000, 1100
Fig. 5. Overlapped diffractograms obtained from the chromite ore and the materials obtained by the ignition at several temperatures. From below to above: original chromite
ore, 450, 600, 800, 1000, 1100, and 1200 ◦C. The identified phases were: (᭹) spinels, (ꢀ) cristobalite, (×) Cr2O3 and (ꢁ) chlorite.

S. Sánchez-Ramos et al. / Thermochimica Acta 476 (2008) 11–19
17
distance characteristic of the different spinels, and even to sev-
3.5.2. Mineral identification
eral olivines, because the ignition at several temperatures produces
specific minerals with different accurate elements ratio (Al/Cr,
Fe(II)/Mg, Fe(III)/Al, Fe(II)/Fe(III), etc.), almost all not included in
a database.
In this work, the diffractograms, together with the results
obtained by voltammetry (Section 3.1), thermal analysis (Section
3.2) and light microscopy (Section 3.3), allow the approximate
determination of the mineralogical and chemical composition of
the sample to establish the process which happens during the heat-
ing.
The identification of the mineralogical composition extracted
from the X-ray diffraction study was summarized as follows:
As previously explained, the initial mineral is clearly composed
by a chrome spinel, magnesiochromite ferroan. (PDF Card 09–0353,
together with two silicates, chromium chlorite (magnesium alu-
minium chromium silicate hydrate PDF Card 72–1385), and, in
lesser amount, labradorite (PDF Card 85–0878). The presence of
a slow amount of free silica can be also envisaged [25].
In the diffractograms corresponding to the samples from 450
to 800 ◦C, the peaks of the initial chromite stabilize, matching to
compounds which already have the generic formula of the natural
chromium chromite (Mg,Fe2+) (Cr,Al,Fe3+)2O4. This spinel is formed
by the partial substitution of Al(III) by Fe(III), which is obtained
by the thermal oxidation of Fe(II) induced at these temperatures.
This can also explain, in this range of temperatures, the presence
of other spinels, as the standard magnesium aluminium iron oxide
(PDF Card 71–1235) in the sample heated at 450 ◦C and magnesium
iron aluminium oxide (PDF-Card 73–2159) in that ignited at 800 ◦C.
In the samples heated at 1000, 1100 and 1200 ◦C, a large amount
of peaks, corresponding to chromium oxide PDF-card 85–0730,
appear. This oxide comes from the segregation of two sources. The
first one is chromium chlorite (magnesium aluminium chromium
silicate hydrate, Mg5.1Al1.2 Si3Cr0.7O10(OH)8), which shows higher
content of chrome than other chlorites and has decomposed at
minor temperatures. The other source is the chromite spinel. The
intensity of the peaks of Cr2O3 remains constant until 1200 ◦C with-
out a significant increasing of the peak intensity, so probably the
segregation of this last one is almost total about 1100 ◦C in both
cases. As consequence, the chrome spinel would disappear, form-
ing a unique phase, in which the standard which better fit the peaks
is magnesium aluminium iron oxide, PDF Card. However, probably
Fe(III) would lessen the oxidation degree, by transformation of the
hematite phase into magnetite.
3.5.1. General observations
From the analysis of the diffractograms of the Fig. 5, the follow-
ing important generic observations have been stood out.
Until nearly 600 ◦C, a gradual displacement to the right side
of the most representative peaks of the original chrome spinel
ore. These effects may be caused (especially from 600 to 800 ◦C)
by the formation of other chrome spinels, due to the progressive
increasing of the oxidation state of the iron, or as consequence
of partial or total substitution of the oxides of different cationic
compounds. Indeed, the net parameters modify, providing dif-
ferent spinels stable at each temperature. In addition to these
other spinels, anhydride silicates (essentially of iron) can be found,
formed from the decomposition of the chromium chlorite, which
occurs at that temperature. Iron (II) cations incorporate to the struc-
ture of the silicates and their peaks would superpose to those from
spinels.
From 800 ◦C, new peaks corresponding to Cr2O3 appear, clearly
visible in the Fig. 5, in the 1000 ◦C-heated sample diffractogram.
This material partially segregates from the chrome spinels and the
chromium chlorite during their decomposition.
Other effect clearly appreciable in the sample heated at 1200 ◦C
is the appearance of free silica peaks as cristobalite, due to the
decomposition of a part of the present silicates (especially the
labradorite).
On the other hand, iron silicate (PDF-CARD 29–0722), a com-
pound highly thermostable, was detected in the mineral ignited
Scheme 1. Schematic summary of main processes produced during the heating process of natural chromite spinel ore.

18
S. Sánchez-Ramos et al. / Thermochimica Acta 476 (2008) 11–19
at 800 ◦C. This indicates that the iron incorporates as Fe2+ in a sili-
cate after the decomposition of the chromium chlorite, and partially
remains with this oxidation state in the whole process, including in
the sample at 1200 ◦C. However, it can also oxidize since 1000 ◦C.
From this temperature, the wollastonite (PDF-Card 76–1948) was
clearly detected. This phase probably comes from the gehlenite,
SiO2.Al2O3.2CaO, which has been formed after the decomposition
of the silicates by a loss of aluminium (maybe incorporated to the
spinel).
Thermal analysis showed a gradual and unsteady increasing of
the weight, mainly caused by the oxidation of FeO, originally found
in the chromite spinel. However, near 950–1000 ◦C, a gradual and
abrupt loss of weight due to the formation of a ferrous-magnetite
phase was observed. This process is stopped by the decomposition
of the chromium chlorite, which takes place in two steps: firstly the
dehydroxylation stage between 200 and 400 ◦C, and the second by
the decomposition of the silicates provoked by the loss of hydroxy
groups near 600 ◦C.
Analyses by X-ray diffraction, light microscopy and voltamme-
try indicate that the primary chromite spinel ((Mg,Fe2+)(Cr,Al)2O4)
decomposes at low temperatures in other two spinels: another
chromite spinel with a progressive introduction of iron (II) in the
structure (Mg,Fe2+)O·(Cr,Al,Fe3+)2O3 and Magnesium aluminium
Iron Oxide, MgAl0.8 Fe1.2 O4, which relative amount in each ele-
ment can vary depending on the oxidation degree of the iron by
the increasing of the temperature. On the other hand, after the
chromium chlorite decomposition, part of the Fe(II) would incor-
porate to the silicates forming a stable Iron Silicate (Fe2SiO4).
From 800 to 900 ◦C, the Cr2O3 segregates as a differentiable com-
pound, being removed from the chromite spinel and remaining
only in spinels as magnesium, aluminium and iron oxides. Finally,
a reduction process of the iron takes place, provoking the forma-
tion of a ferrous–magnetite phase, and the silicates decompose
near 1100–1200 ◦C, producing as cristobalite low and tridymite low
(SiO2), which normally turns into the other one at high tempera-
ture. The occurrence of Fe(II) in the chromite ore sample for the
whole ignition process is considered as a strong novelty, as it has
been never detected in previous studies about chromite ores. The
influence of the chlore chromite over the heating process have been
found very important in the entire heating process.
Finally, new peaks, already described and corresponding to
the standard cristobalite low (PDF-Card 76–0941), a silicon oxide
obtained by partial decomposition of the silicates, appear at
1200 ◦C. This material would be found together with tridymite low,
(PDF-CARD 76–0894), which also appears in the materials obtained
at 1000 and 1100 ◦C, and is considered as a source of cristobalite. In
the case of the labradorite and the iron silicate, segregated Fe(II)
would move during the thermal reduction to the spinel, conse-
quently increase the amount in this cation.
3.6. Global processes during heating
The initial compounds in the natural chromite spinel ore
are magnesiochromite ferroan [(Mg,Fe2+)(Cr,Al)2O4], chlorite
chromian (magnesium aluminium chromium silicate hidrate
Mg5.1Al1.2 Si3Cr.7O10(OH)8), labradorite (Na2.84Ca4.16Al12Si20O64
)
and quartz (SiO2). During the heating process, they undergo several
decomposition and sintering processes, together with elements
exchanges and substitutions, forming new compounds stable at
each temperature (Scheme 1). From the results found in this work,
these processes can be summarized as follows:
(a) Until 450 ◦C, the magnesiocheromite ferroan decomposes in
other two spinels, a chromite spinel (Mg,Fe2+)O·(Cr,Al,Fe3+)2·O3
and magnesium aluminium iron oxide (MgAl0.8Fe 1.2 O4). Iron
(II) gradually oxidizes to iron(III). The other minerals do not
undergo modifications.
Acknowledgments
Financial support from the Spanish Government “Ministerio de
Educación y Ciencia” and the European Union (R + D + I Projects
CTQ2005-09339-CO3-02/BQU and CTQ2008-06727-C03-02/BQU
and E.R.D.F.) is gratefully acknowledged.
(b) From 600 ◦C to nearly 800 ◦C, the decomposition of the chlorite
begins to affect the process, to form olivines, iron (II) and/or
magnesium silicates (green and red), and iron silicate (very
stable at these temperatures), where the Fe (II) incorporated
from chromite. Therefore, in the chromite the Fe(III)/Fe(II) ratio
grows up and the magnesium aluminium iron oxide increases
its proportion in the spinels. Moreover, olivines stable at these
temperatures are formed.
References
[1] A. Alujas Díaz, R. Quintana Puchol, J.E. Gonzalez Ruiz, H. Yee Madeira, Rev. Latin
Am. Met. Mater. 23 (2003) 2.
[2] M. Cugat, Deformación Metálica 13 (1998) 197.
[3] Y.Y.L. Ding, N. A: Warner, Thermochim. Acta 292 (1997) 85.
[4] Y.Y.L. Ding, Experimental study on the carbothermic reduction of composite
chromite pellets, in: Proceedings of the 1996 IChemE Research Event/Second
European Conference for Young Research, University of Leeds, Leeds, 1995, pp.
506–509.
[5] Y.Y.L. Ding, N.A. Warner, A.J. Merchant, Scand. J. Metall. 26 (1997) 55.
[6] F. Scholz, B. Meyer, in: A.I. Bard, Rubinstein (Eds.), Electroanalytical Chemistry,
Marcel Dekker, New York, 1998, pp. 1–87.
[7] T. Grygar, F. Marken, U. Schröder, F. Scholz, Electrochemical analysis of solids:
a review, Collect. Czech. Chem. Comm. 67 (2002) 163.
[8] T. Grygar, J. Electroanal. Chem. 405 (1996) 117.
[9] A. Doménech-Carbó, M.T. Doménech-Carbó, J.V. Gimeno-Adelantado, F. Bosch-
Reig, M.C. Saurí-Peris, S. Sánchez-Ramos, Analyst 126 (2001) 1764.
[10] A. Doménech, S. Sánchez, M.T. Doménech, J.V. Gimeno, F. Bosch, D.J. Yusa, M.C.
Saurí, Electroanalysis 14 (2002) 685.
(c) From 800 ◦C to nearly 1000 ◦C, these processes continue, and
the segregation of the chromium from chromium oxide (green
Cr2O3) is produced, which ends at this temperatures. Therefore,
the original chromite spinel disappears, whereas the amount of
the other sesquioxide phase rises. The Al2O3 and MgO incorpo-
rated from the silicates to the sesquioxide phase, whereas the
iron (II) remains in the silicates.
(d) From 1000 to 1200 ◦C, a magnetite phase is formed by thermal
reduction. The silicates decompose to cristobalite, and part of
the iron (II) incorporate to the spinel.
[11] S. Sánchez-Ramos, F. Bosch-Reig, J.V. Gimeno-Adelantado, D.J. Yusá-Marco, A.
Doménech-Carbó, Anal. Bioanal. Chem. 373 (2002) 893.
[12] P. Encinas, L. Lorenzo, M.L. Tascón, M.D. Vázquez, P. Sánchez-Batanero, J. Elec-
troanal Chem. 371 (1994) 161.
4. Conclusions
The ignition of the chromite spinel mineral provokes a complex
process consisting of several different transformations, as changes
in the oxidation degree in the iron, cations substitutions, decompo-
sitions, chemical reactions between the compounds, etc., producing
other spinels and more stable at each temperature. These pro-
cesses become more complex if the mineral contain an appreciable
amount of silicates, because their own transformation interferes
and strongly modifies the generic behaviour in the spinel.
[13] R.N. Sigh, J.P. Singh, B. Lal, M.J.K. Thomas, S. Bera, Electrochim. Acta 51 (2006)
5515.
[14] E.R. Vago, E.J. Calvo, J. Electroanal. Chem. 339 (1992) 41.
[15] A.M. Alper, High Temperature Oxides, Academic Press, New York, USA, 1970.
[16] A. Petric, K.T. Jacob, J. Am. Ceram. Soc. 65 (1982) 117.
[17] R.G. Richards, J. White, Trans. Br. Ceram. Soc. 53 (1954) 233.
[18] N.P. Lyakishev, M.I. Gasik, Metallurgy of Chromium, Allerton Press, New York,
1998, pp. 240–265.
[19] V.D. Tathavadkar, M.P. Antony, A. Jha, Scand. J. Metall. 33 (2004) 65.
[20] V.D. Tathavadkar, M.P. Antony, A. Jha, Metall. Mater. Trans. B 36B (2005) 75.

S. Sánchez-Ramos et al. / Thermochimica Acta 476 (2008) 11–19
19
[21] X.H. Hellwege, A.H. Hellwege, Landolt-Bornsted Numerical Data and Functional
Relation in Science and Rechnology, Group III 4. Magnetic and other Properties
of Oxides, Springer-Verlag, Berlin, Germany, 1970.
[22] K.T. Damodaran, B. Somasekar, Clays Clay Miner. 24 (1976) 31.
[23] M.H. Hey, Miner. Mag. 30 (1954) 227.
[25] S. Sánchez-Ramos, A. Doménech-Carbó, J.V. Gimeno-Adelantado, J. Peris-
Vicente, Talanta 74 (2008) 1592.
[26] M.N. Viswanathan, J.A.K. Tareen, P.N. Satish, Kurr. Sci. 43 (1974) 202.
[27] A. Doménech-Carbó, M.T. Doménech-Carbó, H.G.M. Edwards, Electroanalysis
19 (2007) 1890.
[24] D.M. Lapham, Am. Miner. 43 (1958) 921.
[28] Y. Yariv, E. Mendelovici, Mater. Chem. 5 (1980) 37.