Journal of Thermal Analysis and Calorimetry, Vol.

90 (2007) 1, 133–141

THERMAL EVOLUTION OF A SLATE

P. J. Sánchez-Soto1*, A. Ruiz-Conde1, R. Bono1, M. Raigón2 and E. Garzón3

1
Instituto de Ciencia de Materiales de Sevilla (ICMSE), Centro Mixto Consejo Superior de Investigaciones Científicas
(CSIC) – Universidad de Sevilla, c/Américo Vespucio 49, Isla de la Cartuja, 41092 Sevilla, Spain
2
Departamento de Ingeniería Mecánica y de los Materiales, Escuela Superior de Ingenieros, Universidad de Sevilla, 41092 Sevilla, Spain
3
Departamento de Ingeniería Rural, Universidad de Almería, La Cañada de San Urbano, 04120 Almería, Spain

The thermal evolution of a slate rock sample (Berja, Almería, Spain) has been studied. The phase minerals identified in this sample
were mica (illite), chlorite (clinochlore) and quartz as major components, with minor microcline, iron oxide and a mixed-layer or
interstratified phase (montmorillonite-chlorite). This slate is highly silico-aluminous (48.33 mass% silica, 22.04 mass% alumina),
and ca. 20 mass% of other elements, mainly Fe 2O3 (8.35 mass%), alkaline-earths and alkaline oxides.
Two main endothermic DTA effects, centered at 640 and 730°C, were observed. The more important contribution of total mass loss
(7.15 mass%) was found between 500–900°C, with two DTG peaks detected at 630 and 725°C. All these effects were associated to
the dehydroxylation of structural OH groups of 2:1 layered silicates mixed in the slate. The dehydroxylation of the layered silicates
evidenced by dilatometry, produced a rapid increase of expansion between 600–800°C. The thermal evolution of the slate upper
800°C indicated the first sintering effects associated to shrinkage, which is also favoured by its low particle size (average 23 mm)
and the presence of a liquid or vitreous phase as increasing the heating temperature. The application of thermal diffractometry to the
slate sample allowed to study the formation of dehydroxylated crystalline phases from the layered silicates after heating. At
1000°C, b-quartz, dehydroxylated illite, iron oxide, relicts of microcline and the vitreous phase were present in the sample. All
these results are interesting to know the thermal behaviour of a complex mineral mixture as identified in the slate.

Keywords: chlorite, dehydroxylation, dilatometry, DTA-TG, illite, interstratified phase, slate, thermal diffractometry

Introduction They can be also applied in paving roads, cement

fabrication, as fillers in plastics and paintings, and as
Slates are rocks constituted mainly by layered minerals, carrier in agriculture, being a raw material for the
such as illite, kaolinite and chlorite, besides quartz. preparation of lightweight aggregates for concrete
Other minerals can be present in low proportion, such after firing (~1200–1500°C) resulting a very porous
as feldspars, carbonates (calcite, dolomite), iron oxide material [2, 7–9, 13, 14]. The wastes of slate processing
and sulfides [1–9]. They are considered as compacted are interesting for new and interesting applications,
metamorphic rocks of fine grain characterized by for instance in the fabrication of materials such as the
slaty cleavage [3–5]. According to an industrial ceramic foams with low mass, low density and low
description [6], slate is a sedimentary derivation rock thermal conductivity [10–13, 15]. The ceramic foams
with fine granulation, weakly metamorphized, have useful applications, such as high-temperature
characterized by perfect cleavage, thus being able to filters, thermal barriers and lightweight structures [15].
form big sheets of just a few millimeters in thickness, There are several studies on thermal behaviour
in a continuous flat surface. Its physical properties, of slates with a particular mineralogy composition
such as preferential cleavage, average hardness, low [8, 11, 15–20]. These investigations attempt to describe
porosity, high mechanical resistance of its composing the complex thermal transformations of the phase
minerals to bad weather, etc., allow for its utilization as minerals that constitute the slate. For instance,
revestments in various areas of building construction Rodriguez et al. [8] studied a slate sample (Badajoz,
and architecture (floor tiles, slabs, roof tiles) with the Spain) containing mainly quartz and, in less
shape of plates because they are resistant, proportion, muscovite, chlorite and feldspar, with
unchanging, insulating and waterproof. High-quality secondary minerals like pyrite, hematite and goethite.
slates used as ornamental rocks show a low porosity These authors reported that chlorite disappears at
and water absorption capacity, very low thermal and 800°C by thermal treatment of the slate, muscovite is
electrical conductivity and relative high resistance at sligthly modified and quartz is undranged, being this
temperature changes [7–13]. The colour of the slates can process exothermic in accordance with their results
be violet, gray, beige, redish, green and black [6, 7, 13]. obtained by DTA [8]. Cambronero et al. [15] investigated

* Author for correspondence: [email protected]

1388–6150/$20.00 Akadémiai Kiadó, Budapest, Hungary

© 2007 Akadémiai Kiadó, Budapest Springer, Dordrecht, The Netherlands
 SÁNCHEZ-SOTO et al.

a slate sample (Orense, Spain) containing quartz, Differential thermal analysis (DTA) and thermo-
muscovite (sericite), chlorite, kaolinite, feldspars, gravimetric (TG) curves were obtained in air flow
calcite, pyrite and carbonaceous matter (graphite). (400 mL min–1) using a thermal analyzer system
In a study on Spanish roofing slate deposits, TG/DTA 6300 Seiko Instruments, model EXSTAR 6000.
Garcia-Guinea et al. [16] claimed that Spain is the The heating rate was 10°C min–1 up to 1000°C. Cal-
first country in world production of slate roofs. The cined alumina was used as the reference material and
same research group studied the thermal and optical Pt–Pt/Rh 13% thermocouples. Samples of 26 mg ex-
behaviour of a slate (Orense, Spain) used in building actly weighed were gently packed into cylindrical Pt
construction [17]. In a further study on this slate, crucibles.
Garcia-Guinea et al. [18] examined the dehydroxylation The total mass loss (average of three measurements)
and the Ostwald ripening effects [19]. was also determined using 1.0000 g of sample after
The aim of the present paper is to study the heating into a Hobersal ceramic furnace, model HD-230,
thermal behaviour of a slate sample. It was interesting up to 1000°C, with 1 h of holding time.
to investigate this slate using thermal methods because: The contraction-expansion behaviour of the slate
(a) at the present, there are scarce studies on this was investigated up to 1100°C using a Setaram Setsys
material [21] and (b) to relate the results with those dilatometer, model TMA-18, with alumina probe and
reported previously on other slate samples. This study cylindrical pressed samples of 4.05 mm height, at a
is a part of a research program on new deposits of heating rate of 6°C min–1.
slates in SW Spain (Granada and Almería) as raw X-ray powder diffraction diagrams (XRD) were
materials for new and improved applications. The obtained using a Bruker diffractometer, model D-501.
knowledge of mineralogy and physicochemical The patterns in disoriented preparations were obtained
characteristics will be worth to differentiate several with Ni-filtered CuKa radiation, graphite monochromator,
slate deposits. Large amounts of slates from Berja at 36 kV and 26 mA and scanning speed of 1° in
have been traditionally used as impermeabilization °2q min–1. The identification of crystalline phases
materials for roofs, beds and ponds. These slates have was performed using a software program, available
been even processed as compacted powders after with the equipment, and a SICOMP PC 16-20 system,
grinding the raw material. On the basis of the research taking into account the Joint Committee For Powder
results, new applications would be proposed. Diffraction Standards (JCPDS) files. Thermal
diffractometry from room temperature up to 1000°C
was also applied to study the thermal evolution of the
Experimental slate sample. The equipment was a Philips X’Pert
diffractometer using Ni-filtered CuKa radiation, at
Sample 40 kV and 40 mA, with a scanning speed of 0.05° in
The original compacted slate samples in fragments °2q s–1 from 5 to 80° 2q and q–q goniometer, with de-
(5–7 kg) were obtained at the deposit of Berja (Almería, tector X’Celerator. The XRD runs were performed
SW Spain). They show a gray colour, a greasy touch each 50°C of heating in static air. The high-temperature
and slaty cleavage. These samples were crushed, camera with temperature-controlled device was Anton
lightly ground and sieved to pass 63 mm to obtain a Paar HTK 1200.
representative and more homogeneous sample. This
powdered sample was dried at 110°C during 24 h.
Aliquots of dried sample (1–2 g) were gently ground Results and discussion
using an agate mortar for further analysis using
several techniques. Figure 1 shows the XRD pattern of the powdered slate
sample. The crystalline phase composition deduced
from X-ray study was mica (muscovite or illite), and
Analytical techniques chlorite, both layered silicates, quartz and minor
Chemical analysis was performed by X-ray fluorescence feldspars (microcline) and iron oxide. These minerals
(XRF) using a sequential Bruker SRS-3000 spectrometer correspond to a common slate sample [1–9]. The
with Rh X-ray tube. Cylindrical pressed samples were identification of the crystalline phases was carried out
prepared using a laboratory press. by their characteristic patterns [23]. A semiquantitative
The particle size analysis of the sample lightly assessment of the mineralogical composition (in mass%)
ground and sieved to pass 100 mm was performed by a of the slate sample, as a bulk, gave average figures of
wet aqueous method previously applied in layered 50–60% mica, 20–15% quartz, 10–15% chlorite,
silicates [22], using a Mastersizer X Laser equipment 5–10% feldspars (microcline) and ca. 5% iron oxide.
of Malver Instruments. The presence of chlorite was confirmed using the
treatments to differentiate it from kaolinite [24].

134 J. Therm. Anal. Cal., 90, 2007

 THERMAL EVOLUTION OF A SLATE

Fig. 1 XRD powder diagram of the slate sample. The main crystalline phase minerals identified are M=mica (muscovite, illite),
C=chlorite (clinochlore), Q=quartz, F=K-feldspar (microcline) and O=iron oxide. Mica muscovite, JCPDS card 7-0042
and chemical composition (K,Na)(Al,Mg,Fe) 2Si3Al0.9O10(OH)2; chlorite (clinochlore), JCPDS card 7-0078 and chemical
composition (Mg,Fe,Al)6(Si,Al)4O10(OH)8; quartz, JCPDS card 33-1161; feldspar (microcline), JCPDS card 2-0513 and
chemical composition Al2O3K2O6SiO2; iron oxide, JCPDS card 5-0637 and chemical composition Fe 2O3

However, it is worth to note that further X-ray treatments [30]. The high value of silica is associated
analysis revealed the presence of a mixed-layer or with the presence of free quartz, as found by XRD
interstratified phase in this slate sample (Fig. 2). It (Fig. 1). In general, the chemical data were in well
can be thus seen that this interstratified phase would agreement with the mineralogical composition
correspond to mixed-layer montmorillonite-chlorite. determined by XRD. The K2O content was associated
The presence of this kind of phase has not previously with the presence of potassium mica (muscovite or
been reported in the literature concerning slates. illite) and potassium feldspar (microcline). The
Information on the types of layer present in content of Fe2O3 is 8.35 mass%, which can be
mixed-layer minerals, their proportion, and the associated to free iron oxide, besides the contribution
pattern of interstratification can be obtained directly of iron in the crystalline silicate structures. The same
from mathematical analysis of the positions and can be said for MgO content, mainly associated to
intensities of XRD basal reflections [25–27]. This clinochlore, and even CaO and Na2O and the presence
complex analysis will be a matter of additional of exchangeable cations. In particular, part of CaO
work [28]. It should be emphasized that the presence content could be also associated to the presence of
of interstratified phases would influence the thermal interstratified montmorillonite-chlorite as a mineral
behaviour observed in mineral samples, as phase (Fig. 2). Finally, the total mass loss at 1000°C
demonstrated in a previous paper [29]. must be associated to the dehydroxylation of layered
On the other hand, the chemical composition of silicates, which are present in major proportion.
the slate sample is given in Table 1. Silica and alumina In a previous study of slate samples,
are the oxides with the highest percentages, which Cambronero et al. [15] have found similar results as
allows to qualify this sample as silico-aluminous. shown in Table 1 concerning the chemical analysis of
However, more than 20 mass% of the total chemical their sample. Their results (in mass%) have showed
composition of the slate is formed by impurity oxides, that Fe2O3 is 9.95% and the mass loss after heating at
in particular iron, alkaline-earths and alkaline oxides, 600°C ranges between 6.5–7.1%, although SiO2
which will form a vitreous or liquid phase by thermal content in that slate is slightly higher (54.68%) than

Fig. 2 XRD powder diagram in the zone 2q=5–30° to identify the presence of a mixed-layer or interstratified phase montmorillo -
nite-chlorite denoted by asterisks (*), showing the relative intensities of each peak. Symbols meaning as in Fig. 1. Mont -
morillonite-chlorite, JCPDS card 7-0027 and chemical composition AlCaNaSi 4O11

J. Therm. Anal. Cal., 90, 2007 135

 SÁNCHEZ-SOTO et al.

Table 1 Chemical analysis (mass%) by XRF of the slate sample

SiO2 Al2O3 Fe2O3 TiO2 CaO MgO Na2O K2O LOI* Total
48.33 22.04 8.35 1.15 4.43 3.43 1.84 3.32 7.02 99.91
*
LOI = loss on ignition, total mass loss after heating at 1000°C for 1 h

the present results (Table 1), Al2O3 is 23.52%, CaO is silicates, mainly chlorite and mica. The sharp endothermal
0.54% and MgO is 3.75%, and the sum (Na2O+K2O) DTA effect at 573°C corresponds to the quartz a®b
is 5.96%. These authors [15] identified muscovite phase transition. On the other hand, the total mass
(sericite), chlorite, kaolinite, quartz, calcite, feldspars, loss after dynamic heating of the slate sample is
pyrite and graphite (powder<45 mm). However, they 7.15%, in good agreement with that reported in
claimed that the chlorite mineral is chamosite, but no Table 1 (1000°C for 1 h). The first mass loss in the TG
additional results to confirm it were reported. In other diagram calculated from room temperature up to
studies on slate samples, Rodriguez et al. [8, 9, 20] ca. 200°C is very low (0.27%); the second one, from
identified quartz, muscovite, chlorite and small 200 up to 650°C, amounts 2.22%; the third mass loss,
amounts of feldspars, which correspond to a common from 650°C up to ca. 770°C, amounts 4.14%; finally,
slate. They reported similar chemical analysis results there is a fourth mass loss from 770 to 1000°C which
as compared to those presented in Table 1, with amounts 0.52%. The mass loss is almost constant
6.72% Fe2O3 and a mass loss after heating at 1000°C above 900°C. Associated to these thermal changes,
of 5.54%. The SiO2 content is higher (58.77%) than DTG curve shows two main peaks centered at ca. 630
the present results (Table 1). These authors found and 725°C, the last of larger intensity.
20.95% Al2O3, 0.27% CaO and 1.87% MgO; the sum Previous studies on thermal evolution of chlorites
(Na2O+K2O) was 2.49%. However, the contribution [31, 32] reported that the minerals of chlorite group
of CaO in the present slate sample is more important present a typical DTA curve with an endothermal effect
than that found in precedent papers concerning chemical between 575–610°C and other one centered at 740°C.
analysis data on slate samples. It is associated to the The first one is associated to dehydroxylation of
presence of the interstratified crystalline phase brucite-like layers and the second one to dehydroxylation of
(Fig. 2), and probably in part as exchangeable cations. mica or 2:1 talc-like sheet layers. However, in the pres-
On the other hand, Rodriguez et al. [8, 9, 20] ent case, this second endothermal DTA effect is
indicated that ‘the presence of magnesia suggest the superposed with the endothermal effect of dehydr-
mica nature of the slate sample’. This asseveration is oxylation of mica (muscovite or illite). This assumption
not consistent with the mica muscovite that they explains the larger mass loss observed in the TG dia-
identified by XRD in their slate sample. gram of the slate sample above 600°C and the increase
Concerning the particle size of the slate sample in intensity of DTG peak centered at 725°C. On the
studied in the present paper, it was found a high other hand, the dehydroxylation of mica (muscovite,
proportion of sample less than 100 mm (80 mass%). illite) has been previously studied by several authors
After sieving, the particle size analysis showed that [33–38]. Several factors, such as particle size, time, tem-
the largest contribution of particle size distribution is perature and structure of layer silicates, influence the
between 1–100 mm. It was calculated an average thermal behaviour. It should be noted that chlorite and
particle size of 23 mm. Taking into account this result,
the mica muscovite mineral identified by XRD
(Fig. 1) might be considered as sericite or illite (naturally
occurring clay-grade micas) [5]. The term ‘illite’ was
proposed as a general term given to a mixture of
minerals, including muscovite and feldspar, whose
physical properties resemble those of the mica family [5].
It has been generally accepted that the main
difference between mica and illite is that micas
contain more potassium as interlayer cations, whereas
illite contains more water and silica.
Figure 3 shows the DTA-TG curves for this slate
sample. It can be observed two main endothermal
DTA effects centered at 640 and 730°C (asymmetric)
Fig. 3 DTA-TG-DTG curves of the slate sample. The heating
accompanied by mass losses between 500–900°C. rate was 10°C min–1 up to 1000°C in air flow
They are associated to dehydroxylation and water (400 mL min–1), using calcined alumina as reference
elimination of structural OH groups of the layered material and sample mass of 26 mg

136 J. Therm. Anal. Cal., 90, 2007

 THERMAL EVOLUTION OF A SLATE

mica (muscovite) are 2:1 layer type silicates [39]. The

minerals of chlorite group are characterized by an addi-
tional brucite layer with subdivision in subgroups: di-
and trioctahedric chlorites. Thus, clinochlore is associ-
ated to this last subgroup [39], being identified as the
chlorite mineral present in the slate sample (Fig. 1). Illite
has a 2:1 layer structure having a plane of octahedrally
coordinated cations (gibbsite) sandwiched between 2 in-
ward pointings sheets of tetrahedra (silica). Further dis-
cussion on the structure of 1:1 and 2:1 layer silicates and
the interlayer material to understand their differences in
thermal stability and decomposition is included in [40]. Fig. 4 Dilatometric curves of the original slate sample up to
a – 1100°C and the b – derivative curve. Heating rate
In general, the thermal analysis by DTA-TG (Fig. 3) was 6°C min–1
confirmed the mineral phase composition of this slate
sample as found by XRD (chlorite, mica and quartz).
However, the presence of mineral interstratified phases although may be necessary to heat the samples for
influences the thermal behaviour, as found in a previous significantly long period to reach equilibrium, the
paper on layered silicates [29]. determination of eutectic temperature is very difficult.
Dilatometric analysis of this slate sample (Fig. 4, According to the SiO2–Al2O3–K2O ternary phase
curve a) shows a slow and continuous expansion by diagram at 1 atm pressure [38, 45, 46], the minimum
dynamic heating up to ca. 400°C, with constancy from temperature of liquid formation is 985°C. Then, it is
400–500°C, and a rapid increase between 600–800°C clear that a decrease of that temperature and a greater
as a result of mica and chlorite dehydroxylation. The amount of liquid will be produced if different oxides,
shape of the dilatometric curve is in accordance with such as MgO, CaO and Na2O (Table 1), are also
those previously reported in the literature which present. On the other hand, it should be noted that the
characterize illite samples [41, 42]. Concerning high amount of iron oxide and the mineralogy, for
dilatometric curves of slates, very scarce in the literature, instance illite, quartz and oligoclase, characterize
the dilatometric behaviour of the present slate sample some raw materials used for ceramics production [47].
up to 900°C is in good agreement with the results Figure 5 shows the dilatometric behaviour of the
reported by Cambronero et al. [15]. slate previously heated at 1100°C. It can be thus seen
After heating at 800°C, a first step of sintering that the only thermal effect recorded in the diagram is
seems to take place with a rapid shrinkage (Fig. 4, curve a). the quartz a®b phase transition, being more appraised
From ca. 930°C the dilatometric curve shows an in the derivative curve (Fig. 5, curve b). This dilato-
increase in expansion up to ca. 1020°C. This change metric behaviour is consistent with previous results
can be explained by a bloating effect associated to the concerning ceramic raw materials containing quartz
gas trapped in pores or the release of oxygen gas from after firing [41, 42, 47] and slate samples thermally
the reduction of Fe2O3 [43, 44]. These gases or trapped treated under vacuum, air and N2-5% H2 atmos-
atmospheric gases expanding in the closed pores with phere [15].
increase in temperature. Finally, the increase in liquid The XRD patterns at room temperature
phase by a second step of sintering produces a rapid (original) and after heating the slate sample at 600°C
shrinkage of the slate sample. These thermal changes (Fig. 6) show chlorite, mica (illite), quartz and
of expansion-shrinkage can be also observed in the feldspar (microcline), besides the interstratified phase
derivative curve (Fig. 4, curve b). The low particle montmorillonite-chlorite. Although there are some
size of the slate sample favours the sintering, but there little changes in the values of d-spacings as compared
is also an important factor: the chemical composition. with those of the original mineral phases, the layered
The amount of impurities in this slate (Table 1) silicate phases observed by XRD at this temperature
suggests the formation of a large proportion of must be the dehydroxylated phases. Chlorite X-ray
vitreous or liquid phase as increasing temperature. patterns disappear after further dynamic heating,
The content of oxides distinct of silica and alumina which is accounted for in the mass loss (TG-DTG curves,
will form the vitreous phase, which is a liquid after Fig. 3) and the associated endothermal DTA effect
progressive thermal treatments and, hence, it will (Fig. 3). First of all, there is a decrease in intensity of
favour the sintering process. Iron oxide would be the chlorite X-ray patterns from room temperature to
responsible of the bloating effect, as observed by 600–650°C, which coincides with the region of the
dilatometric analysis (Fig. 4 curve a) in the presence first endothermal DTA effect centered at 640°C and
of a liquid phase. In this kind of complex systems, the DTG peak at 630°C. This is in accordance with

J. Therm. Anal. Cal., 90, 2007 137

 SÁNCHEZ-SOTO et al.

Fig. 5 Dilatometric curves of the original slate sample after a

first thermal treatment at a – 1100° C and b – deriva-
tive curve. Heating rate was 6°C min –1

the results reported in the literature concerning

chlorite thermal transformations [32], where the
intensities of the basal reflections are changed
markedly by the first stage of reaction, but the spacing
is modified only slightly. In a second step, chlorite
patterns disappear from 650°C to higher temperatures, Fig. 6 XRD powder diagrams in the zone 2q=5–30°
which coincides also with the more intense corresponding to the original slate sample and after
endothermal DTA effect centered at ca. 730°C and dynamic thermal treatment at several temperatures up
the DTG peak at 725°C (Fig. 3). Chlorite X-ray to 1000°C. Symbols meaning as in Figs 1 and 2. Note:
From 600°C and upper, the layered silicates are
patterns are not present at 800°C. Mica (illite), as a
identified as crystalline dehydroxylated phases
dehydroxylated phase, is observed when the slate
sample is being heated (Fig. 6). This phase shows
iron oxide. The observation of dehydroxylated mica
more appraised changes in the range 700–800°C,
at 1000°C is consistent with previous results [33, 38].
when the second endothermal DTA effect takes place
The presence of a vitreous or liquid phase in the slate
and the last mass losses are produced (Fig. 3). The
sample after thermal treatment at 1000°C can be
interstratified phase montmorillonite-chlorite is not
inferred from the XRD background because it is
destroyed by heating and it can be observed even at
amorphous to X-rays (Fig. 8). The disappearance of
600°C as a dehydroxylated phase. The X-ray patterns
chlorite (dehydroxylated phase) at 800°C in this slate
of that phase disappear at 800°C. Quartz is present in
is in good agreement with the results reported by
the slate sample, but from 600°C the quartz phase
Rodriguez et al. [8] when studying a slate sample
must be the b-phase according to its characteristic thermal
containing mainly quartz and, in less proportion,
inversion, as observed previously (Figs 3 and 5). The
muscovite, chlorite and feldspar. These authors
X-ray patterns of microcline (alkaline feldspar) start
reported also that muscovite is slightly modified and
to disappear from 800°C (Fig. 6). This is an indication
quartz is undranged by thermal treatment, but they
of the formation of a vitreous or liquid phase as
considered that this process is exothermic according
increasing the thermal treatment, with progressive
to their DTA results. The present results on the
incorporation of alkaline ions, as suggested by
thermal transformation of a slate sample containing
dilatometry (Fig. 4).
mainly mica (illite), chlorite and quartz demonstrates
According to the above results, the most
that the thermal decomposition process of this sample
important thermal changes of the crystalline phases
is undoubtely endothermic (Fig. 3), with changes
which constitute the slate sample are evidenced
(expansion) in the dilatometric behaviour (Fig. 4).
between 600–800°C. Figure 7 shows a more detailed
The previous alteration of muscovite to sericite or illite
picture concerning these phase changes after dynamic
(considered naturally occurring clay-grade micas)
heating. It should be noted that XRD pattern of
with subsequent particle size reduction can be other
dehydroxylated mica (zone 25–26 °2q) does not
factor which would influence the thermal behaviour
disappear but changes progressively to coincide with
of the slate, in particular their dehydroxylation
the more intense quartz XRD peak.
process [18, 33, 34, 37, 47, 48].
Finally, the complete XRD diagram (5–80 °2q)
Returning with previous results on thermal
of Fig. 8 shows the state of the slate sample after
evolution of slates, Cambronero et al. [15] studied the
dynamic thermal treatment concerning the phase
same slate sample previously investigated by
changes up to 1000°C. Dehydroxylated mica can be
García-Guinea et al. [17, 18]. The thermal behaviour
observed, besides quartz, relicts of microcline and

138 J. Therm. Anal. Cal., 90, 2007

 THERMAL EVOLUTION OF A SLATE

Fig. 7 XRD powder diagrams in the zone 2q=5–30° corresponding to the slate sample after dynamic thermal treatment at 600,
700 and 800°C. Symbols meaning as in Fig. 1, but from 600°C and upper the layered silicates are identified as crystalline
dehydroxylated phases. The presence of a (dehydroxylated) mixed-layer or interstratified phase montmorillonite-chlorite is
denoted by asterisks (*)

Fig. 8 XRD powder diagram in the zone 2q=5–80° of a slate sample after thermal treatment at 1000°C using the temperature-con-
trolled device as in Fig. 6. DM is dehydroxylated mica; Q=quartz; F=K-feldspar (microcline); O=iron oxide.
Dehydroxylated mica (muscovite): JCPDS file 46-0741

raw clay material containing pyrite and carbon. On

of this slate sample in air, vacuum and N2–5% H2 the other hand, it should be noted that the presence of
atmosphere was investigated by dilatometry and XRD illite and/or chlorite as dehydroxylated phases, could
after thermal treatments of the sample. These authors be used to evaluate the possible firing temperature of
emphasized the oxidation or reduction conditions a ceramic sample [51].
which influence the thermal transformations [15]. The above results emphasize that the thermal
Unfortunately, their XRD diagrams were not behaviour of slate samples can be very complex according
recorded using dynamic heating conditions. They to the original mineral phases and the conditions of
observed mullite, hercinite and amorphous silica after thermal treatment. The problem will be more difficult
heating at 1100°C for 30 min [11, 15]. Heating at considering the particle size, chemical composition
1150°C in air produced mullite, hematite, spinel-like (especially impurities) and the thermodynamic
phases (hercinite), quartz and alumina, and above equilibria in this kind of systems [46, 49, 50]. In the
1200°C, the crystalline phases detected by XRD are present case, however, the thermal behaviour of a
quartz, mullite and hercinite [11, 15]. However, the particular slate sample in air up to 1000°C has been
same authors reported the formation of hieratite investigated using thermoanalytical techniques, in
(K2SiF6) with presence of quartz after sintering at particular thermal diffractometry under dynamic
1150°C for 30 min in N2–5% H2 atmosphere. These conditions of heating as an approach to study the
authors claimed that the formation of hieratite can be phase transformations.
produced from the thermal decomposition of
muscovite, with composition KAl2(Si3Al)O10(OH,F)2
which they identified in the sample slate. It should be Conclusions
remarked that Rodríguez et al. [8], studying other slate
sample, found the presence of secondary minerals like In this investigation, the thermal evolution of a slate
pyrite, hematite and goethite. These secondary sample has been studied. The main minerals identified
minerals could influence the thermal behaviour of the by XRD were mica (illite), chlorite (clinochlore) and
slate [49, 50]. In this sense, Decleer and Viaene [48] quartz, with minor interstratified phase (montmorillo-
have previously discussed the thermal reactions in a nite-chlorite), microcline (K-feldspar) and iron oxide. The

J. Therm. Anal. Cal., 90, 2007 139

 SÁNCHEZ-SOTO et al.

particle size distribution allowed to calculate an average evidenced. This crystalline phase showed some
particle size of 23 mm in samples sieved under 100 mm changes at 700–800°C, when the more intense DTA
(ca. 80 mass% of the total). The chemical composition and DTG peaks are observed. The interstratified phase
of the slate is highly silico-aluminous, with montmorillonite-chlorite, present as dehydroxylated
48.33 mass% silica and 22.04 mass% alumina. The con- phase at ca. 600°C, disappears at 800°C. The progressive
tent of other elements as oxides is ca. 20 mass% of the disappearance of microcline from 800°C is associated
total chemical composition, in particular iron, alka- to an increase of vitreous or liquid phase by heating,
line-earths and alkaline oxides. These results are in well which is in well agreement with the dilatometric
agreement with the mineralogical composition. Although behaviour. At 1000°C, b-quartz, dehydroxylated
interlayer cations must be also present in the structure of illite, iron oxide, relicts of microcline and the vitreous
layered silicates, the alkaline oxides are associated to phase are present.
illite and microcline. The present investigation has allowed to know the
Thermal analysis of the slate sample by DTA-TG main changes associated to the thermal transformations
and dilatometry confirmed the mineralogical composition. of a slate sample constituted by a mineral mixture of
The thermal behaviour was interpreted considering several layered silicates and a mixed layer or
the structure of 2.1 layered silicates assuming: (1) a interstratified phase. Although the mineral mixtures
first step of dehydroxylation of brucite-like layers of of slate samples can be very complex, and the
chlorite (clinochlore) at ca. 640°C, and (2) a second reported results on the thermal evolution of slate
step of dehydroxylation of 2:1 talc-like layers of this samples are very scarce, the results of this paper have
silicate at ca. 730°C, which is overlapped with the provided a better understanding of the thermal
dehydroxylation of illite, with octahedrally coordinated behaviour of slates with a previous characterization.
cations (gibbsite), besides the dehydroxylation of the
interstratified phase. In consequence, mass loss is
larger in the second step (4.14 mass%) and the DTA Acknowledgements
and DTG effects are more intense than those observed
at the first step. The dehydroxylation of the layered Financial support by Spanish Ministry of Education and
silicates has been also evidenced by dilatometry, with Science, Research Project CTQ2005-998 cofinancied with
Feder, is greatfully acknowledged. This paper is dedicated to
a rapid increase of expansion between 600–800°C,
the memory of Professor and friend Dr. Joaquín Rivas Florido,
and upper 800°C the first sintering effects associated recently deceased in Utrera (Sevilla, Spain).
to a shrinkage of the sample. The sintering process is
also favoured by the low particle size of the sample
and the presence of a liquid or vitreous phase which References
can easily be formed taking into account the chemical
composition of the sample (Table 1). An increase in 1 H. G. F. Winkler, Petrogenesis of Metamorphic Rocks,
expansion up to ca. 1020°C was interpreted by a bloating Springer Verlag, 1976.
effect. This effect is caused by gas trapped in the 2 P. Sousa Santos, Tecnologia de argilas aplicada as argilas
pores and the release of oxygen gas, the last being brasileiras, Vol. 2, p. 148, Blücher Ltd.,
characteristic of raw materials containing iron oxides. Sao Paulo (Brasil) 1975.
The presence of free quartz was detected by DTA 3 D. S. Woods, Current views of development of slaty
(original slate sample) and dilatometry (after a cleavage, An. Rev. Earth Sci., 2 (1974) 1.
4 R. L. Bates and J. A. Jackson, Glossary of Geology, Am.
previous heating of the sample up to 1100°C)
Geol. Inst., 1987.
according to its characteristic a®b thermal phase 5 W. A. Deer, R. A. Howie and J. Zussman, An Introduction
transition at 573°C. to the Rock-forming Minerals, Longman, Hong Kong 1992.
The application of thermal diffractometry has 6 Anon, What is slate?, Altivopedras Company, Minas
allowed the examination of the thermal behaviour of Gerais, Brazil 2002, in www.altivopedras.com.
slate under dynamic conditions of heating. It has been 7 M. Lombardero and J. M. Quereda, Recursos minerales de
thus confirmed by XRD the formation of EspaÔa, Textos Universitarios, J. García and J. Martínez,
dehydroxylated crystalline phases from layered silicates Eds., CSIC, Madrid, Spain 1992, pp. 1118–1150
after heating, with slight differences in their XRD 8 M. A. Rodríguez, F. Rubio, J. Rubio, M. J. Liso and
J. L. Oteo, Bol. Geol. Min., 106 (1995) 437.
patterns as compared with the original (hydroxylated)
9 M. A. Rodríguez, J. Rubio, F. Rubio, M. J. Liso and
layered silicates present in the mixture. Chlorite J. L. Oteo, Clays Clay Miner., 45 (1997) 670.
(clinochlore), as dehydroxylated phase previously 10 T. U. Pritchard, Development of Expanded Slate for
formed at ca. 600°C, disappear from 650°C to higher Horticultural and Aggregate Use, Eurothen 2000, Lisbon,
temperatures, being not present at ca. 800°C. The 19–21 January 2000.
formation of dehydroxylated illite was also 11 M. T. Vieira, L. Catarino, M. Oliveira, J. Sousa,
J. M. Torralba, L. E. G. Cambronero,

140 J. Therm. Anal. Cal., 90, 2007

 THERMAL EVOLUTION OF A SLATE

F. L. González-Mesones and A. Victoria, J. Mater. 30 K. H. Schüller, Process Mineralogy of Ceramic Materials,

Process. Technol., 92–93 (1999) 97. Ed. W. Baumgart et al., pp. 1–27, Ferdinand Enke,
12 R. Caligaris, N. Quaranta, M. Caligaris and E. Benavides, Sttutgart, Germany 1984.
Bol. Soc. Esp. Ceram. V., 39 (2000) 623. 31 G. W. Brindley and T. S. Chang, Am. Miner., 59 (1974) 152.
13 M. Regueiro and M. Lombardero, Innovaciones y avances 32 G. W. Brindley and J. Lemaitre, Chemistry of Clays and
en el sector de las rocas y minerales industriales, Summa Clay Minerals A. C. D. Newman Ed., Monograph No. 6,
S.A., Madrid 1997. The Mineralogical Society, London 1987, p. 319.
14 T. S. Valera, A. P. Ribeiro, F. R. Valenzuela-Díaz, 33 H. Takeda and B. Morosin, Acta Crystall., B31 (1975) 2444.
A. Yoshiga, W. Ormanji and S. M. Toffoli, Annual 34 S. Guggenheim, Y. H. Chang and A. F. Koster van Groos,
Technical Conference-Society of Plastics Engineers, Am. Miner., 72 (1987) 537.
3 (2002) 3949. 35 K. J. D. MacKenzie, I. W. M. Brown, C. M. Cardile and
15 L. E. G. Cambronero, J. M. Ruiz-Román and J. M. Ruiz, R. H. Meinhold, J. Mater. Sci., 22 (1987) 2645.
Bol. Soc. Esp. Ceram. V., 44 (2005) 368. 36 E. Murad and U. Wagner, Clay Miner., 31 (1996) 45.
16 J. Garcia-Guinea, M. Lombardero, B. Roberts and 37 G. E. Roch, M. E. Smith and S. R. Drachman, Clays Clay
J. Taboada, Trans. Inst. Min. Metall. (Sect. B: Appl. Earth Miner., 46 (1998) 694.
Sci.), 106 (1997) 205. 38 S. G. Barlow and D. A. C. Manning, Br. Ceram. Trans. J.,
17 J. Garcia-Guinea, M. Lombardero, B. Roberts, J. Taboada 98 (1999) 122.
and A. Peto, Mater. Construcc., 48 (1998) 37. 39 S. W. Bailey, Structures of Layer Silicates, in [23], p. 1.
18 J. Garcia-Guinea, V. Cardenas, V. Correcher, A. Delgado, 40 A. F. Koster van Groos and S. Guggenheim, in CMS
M. Lombardero and J. C. Barros, Bol. Soc. Esp. Ceram. Workshop Lectures, Vol. 3, Thermal Analysis in Clay
V., 39 (2000) 589. Science, J. W. Stucki, D. L. Bish and F. A. Mumpton eds.,
19 B. A. van der Pluijm, N.-C. Ho, D. R. Peacor and The Clay Minerals Society, Boulder, Colorado, 1990, p. 49.
R. J. Merriman, Nature, 392 (1998) 348. 41 P. Munier and J. Meneret, Bull. Soc. Fr. Ceram., 7 (1950) 6.
20 M. A. Rodríguez, F. Rubio, J. Rubio, M. J. Liso and 42 G. García Ramos, F. González García, P. J. Sánchez-Soto
J. L. Oteo, Bol. Soc. Esp. Ceram. V., 40 (2001) 101. and M. T. Ruiz, Bol. Soc. Esp. Ceram. V., 24 (1985) 67.
21 E. Garzon, A. Ruiz-Conde and P. J. Sanchez-Soto, 43 M. S. Tites and Y. Maniatis, Br. Ceram. Trans. J.,
Communication presented at VIII Congreso Nacional de 74 (1975) 19.
Materiales, Valencia 2004. 44 A. W. Norris, D. Taylor and I. Thorpe, Br. Ceram. Trans. J.,
22 P. J. Sánchez-Soto, M. C. Jiménez de Haro, 78 (1979) 102.
L. A. Pérez-Maqueda, I. Varona and 45 J. F. Schairer, J. Am. Ceram. Soc., 40 (1957) 215.
J. L. Pérez-Rodríguez, J. Am. Ceram. Soc., 83 (2000) 1649. 46 E. F. Osborn and A. Muan Eds., Phase diagrams for
23 G. W. Brindley and G. Brown, Crystal structures of clay ceramists, Plate 407, The American Ceramic Society,
minerals and their X-ray identification, Mineralogical Columbus, Ohio 1960.
Society, London 1980. 47 V. Nastro, D. Vuono, M. Guzzo, G. Nicéforo, I. Bruno
24 S. W. Bailey and J. S. Lister, Clays Clay Miner., and P. De Luca, J. Therm. Anal. Cal., 84 (2006) 181.
37 (1989) 193. 48 D. D. Eberl and J. Shrodon, Am. Miner., 73 (1988) 1335.
25 D. M. C. MacEwan and A. Ruiz-Amil, Interstratified Clay 49 J. Decleer and W. Viaene, Appl. Clay Sci., 8 (1993) 111.
Minerals, Soil Components, Vol. 2, Inorganic 50 A. Muan, J. Am. Ceram. Soc., 40 (1957) 121.
Components, J. E. Gieseking Ed., Springer-Verlag, 51 D. N. Papadopoulou, M. Lalia-Kantouri, N. Kantiranis and
New York, 1975, pp. 265-334. J. A. Stratis, J. Therm. Anal. Cal., 84 (2006) 39.
26 A. Ruiz-Amil, F. Aragón, E. Vila and A. Ruiz-Conde,
Clay Miner., 27 (1992) 257.
27 A. Ruiz-Conde, A. Ruiz-Amil, J. L. Pérez-Rodríguez and Received: June 26, 2006
P. J. Sánchez-Soto, J. Mater. Chem., 6 (1996) 1557. Accepted: April 12, 2007
28 A. Ruiz-Conde, E. Garzón and P. J. Sánchez-Soto, OnlineFirst: July 11, 2007
in preparation.
29 A. Justo, J. L. Pérez-Rodríguez and P. J. Sánchez-Soto,
DOI: 10.1007/s10973-007-7751-2
J. Thermal. Anal., 40 (1993) 59.

J. Therm. Anal. Cal., 90, 2007 141