Available online at www.sciencedirect.

com

Journal of the European Ceramic Society 33 (2013) 2767–2774

MgAl2O4 spinel as an effective ceramic bonding

in a MgO–CaZrO3 refractory
Edén A. Rodríguez a,∗ , G.-Alan Castillo a , Tushar K. Das a , R. Puente-Ornelas a ,
Yadira González a , Ana-María Arato a , J.A. Aguilar-Martínez b
a Universidad Autónoma de Nuevo León, Av. Universidad S/N, Cd. Universitaria, San Nicolás de los Garza, Nuevo León, C.P. 66451, Mexico
b Centro de Investigación en Materiales Avanzados, S.C. (CIMAV), Alianza Norte No. 202, Parque de Investigación e Innovación Tecnológica (PIIT), Nueva carr.
Aeropuerto Km. 10, Apodaca N.L. 66600, Mexico
Received 13 December 2012; received in revised form 22 April 2013; accepted 25 April 2013
Available online 7 June 2013

Abstract
The influence of MgAl2 O4 spinel addition as a ceramic bonding in the MgO–CaZrO3 refractory was established by the evaluation of physical and
microstructural characteristics in terms of density, porosity, crystalline phases, phase distribution and morphology. X-ray diffraction analyses and
scanning electron microscopy with microanalysis have been used. The mechanical behavior has been evaluated in terms of cold crushing strength
at room temperature and modulus of rupture at 25 and 1260 ◦ C. Static and dynamic resistances tested by chemical attack of clinker raw constituents
have been carried out at 1450 ◦ C. Results showed that thermo-mechanical properties significantly improved with increasing the content of spinel.
Microstructural analysis revealed that spinel phase aided to develop a strong bond between MgO and CaZrO3 refractory aggregates. Finally, the
refractory bodies exhibited a good thermal stability and an excellent chemical resistance against the clinker raw material.
© 2013 Elsevier Ltd. All rights reserved.

Keywords: MgO–CaZrO3 matrix; Spinel; Refractory; Bonding; Rotary cement kiln

1. Introduction enhance the corrosion process of refractory bricks. Thus, under

Magnesia based refractories have been widely used in cement
rotary kilns and steel ladles due to their high melting point, no
toxicity and good resistance to basic slags and clinker phases.
However, they have some disadvantages, such as high ther-
mal conductivity, poor thermal shock resistance and infiltration
resistance against chemical attack.1–3 In order to improve these
properties, some metal oxides and compounds such as SiO2 ,
CaO, Al2 O3 , MgAl2 O4 , Cr2 O3 , ZrO2 , CaZrO3 and FeAl2 O4
have been used as doping agent. These oxides may react with
MgO to form a second phase which improves the process of
sintering.
In recent years, the burning zone of rotary cement kilns has
been exposed to alkali salts and waste by-products such as rubber
or other organic products from animal sources; these materials
∗ Corresponding author. Tel.: +52 81 83294020x1619; fax: +52 8183294020.
E-mail addresses: earc22@hotmail.com, eden amaral rdz c@hotmail.com
(E.A. Rodríguez).
the severe service operating conditions present in the rotary
cement kilns, industry has demanded the development of new
refractory products for increasing the lining life.3–12 Many stud-
ies conducted by the refractory industry responds to the need for
the development and implementation of new chrome-free basic
bricks that meet the requirements presented in the kilns.
In other hand, calcium zirconate is one of the most promis-
ing high refractory compounds used in the steel and cement
industry.13 Particularly in the sintering zone, calcium zirconate
(CaZrO3 ), as a second phase of magnesia refractories works
effectively improving the resistance against both clinker phase
melting and alkali attack.13 Specially, CaZrO3 presents a high
melting point (∼2340 ◦ C) and it is compatible with the MgO
since does not react at any temperature and does not form
a liquid phase up to temperatures higher than 2060 ◦ C.14–16
Besides, it is compatible (does not react) with the main phases
of Portland cement clinker in the solid state (C2 S, C4 AF and
C3 S).17 Moreover, the solid–solid bonding between the com-
patible MgO and CaZrO3 phases is highly resistant to the
infiltration of fluxes from the clinker. Previously, Obregón

0955-2219/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jeurceramsoc.2013.04.027
2768 E.A. Rodríguez et al. / Journal of the European Ceramic Society 33 (2013) 2767–2774

et al. demonstrated through corrosion studies, the use of cement
clinker with MgO–CaZrO3 based fine grained materials for these
purposes.17 CaZrO3 can dissolve in the liquid clinker phase and
increases its viscosity. This process restricts infiltration of the
cement clinker phase into the refractory.18,19
MgO and CaZrO3 have been suggested as chrome-
free refractories for cement kilns20–24 to replace not only
magnesia–chromite, but also magnesia-oversaturated spinel
refractories currently in use. MgO spinel bricks are highly resis-
tant to thermal shock, non-sensitive against reducing/oxidizing
conditions, but vulnerable to thermal overload. Spinel also forms
low-melting phases with the result of premature wear.25–27 The
analysis of MgO–CaZrO3 composite as refractory materials in
rotary cement kilns carried out by Kozuka et al.11,12 points out
a good behavior on service of this material but peeled off easily
in high stressed areas. Recently, the behavior against corrosion
by clinker of MgO–CaZrO3 matrices at high temperatures was
established by Serena et al.28,29 Their results underline the good
corrosion behavior of the MgO–CaZrO3 materials, and support
their use as a matrix in magnesia chrome-free bricks for the
burning zone of rotary cement kilns.
Magnesium aluminate spinel is a technologically important
material with very attractive properties such as high melting
point (2135 ◦ C), high mechanical strength at elevated tempera-
ture, high thermal shock resistance and low thermal expansion
coefficient.
Taking into account the previously mentioned, the aim of
the present study is to investigate at laboratory and industrial
scale the physical, mechanical and chemical properties as well
as microstructural characteristics of new free-chrome brick com-
posites of a MgO–CaZrO3 matrix doped with low contents of
MgAl2 O4 spinel (2.5–6 wt.%) in order to meet the lining require-
ments present in the rotary cement kiln.30–34
2. Experimental procedure
2.1. Raw materials and refractory formulation
Different raw materials were used for the preparation of
the refractory mixtures at laboratory scale. Commercial high
purity magnesia (MgO at industrial grade purity), laboratory
synthesized calcium zirconate (CaZrO3 ), zirconia (ZrO2 ) and
MgAl2 O4 spinel with 98 wt% purity. The chemical analysis of
the raw materials obtained by X-ray fluorescence technique is
given in Table 1. It is important to mention that all the staring
raw materials used in this stage of the investigation have particle
size smaller than 45 ␮m. The addition of a small quantity of fine
zirconia particles in all formulations is mainly for avoiding the
free lime in the refractory matrix and promoting the formation
of an “in situ” calcium zirconate phase.
The phase analysis of raw materials at laboratory scale
was determined using a Siemens model D5000 diffractometer
equipment using, Cu K␣ radiation source (λ = 1.54056 Å) and
a setting of 40 kV and 30 mA. All XRD data were collected
under the same experimental conditions; in the angular range
5◦ ≤ 2θ ≤ 90◦ using θ/2θ configuration with a step scan of 0.05◦
and 1 s per step exposure time.
At the laboratory scale, the crystalline phases from the raw
materials were identified by the X-ray diffraction as: MgO (per-
iclase), CaZrO3 [calcium zirconate (CaZrO3 ) and a zirconium
dioxide stabilized by Ca2+ ions (Ca0.15 Zr0.85 O1.85 )], MgAl2 O4
[periclase (MgO) and magnesium aluminate spinel (MgAl2 O4 )]
and ZrO2 (baddeleyite).
In the first stage of the present work, the influence of
MgAl2 O4 spinel on MgO–CaZrO3 properties in a laboratory
scale was evaluated through the development of four refrac-
tory formulations. Variations carried out in the formulations are
reported in Table 2.
The second stage was conducted at an industrial scale. The
use of large amounts of high purity raw material was nec-
essary in order to obtain suitable final refractory properties.
The chemical composition of the raw material used for the
development of new refractory bricks is shown in Table 1. For
the sintered commercial magnesia a wide particle size distri-
bution composed by coarse, intermediate and fine grains was
used. The particle size distribution of sintered commercial mag-
nesia in mm was: −4.76 + 2.83 mm (7%), −2.83 + 1.19 mm
(29%), −1.19 + 0.595 mm (22%), −0.595 + 0.250 mm (12%),
−0.250 mm (6%) and −0.044 mm (24%). Particularly, in the
case of electrofused magnesia–calcium zirconate mixture, which

Table 1
Chemical composition of raw materials.
Raw materials Oxide compounds (% by weight)

MgO ZrO2 CaO Al2 O3 Fe2 O3 SiO2 K2 O Na2 O

Laboratory scale
Sintered MgO 97.90 – 1.64 – – 0.30
Sintered CaZrO3 – 72.80 26.57 – – – – –
MgAl2 O4 28.61 – 0.32 70.36 – 0.51 – –
ZrO2 98.0 – – – – 1.50
Clinker raw material 0.88 – 44.79 3.03 1.60 11.86 0.47 0.24
Industrial scale
Sintered MgO 98.91 – 0.85 0.08 0.05 0.11 – –
Electrofused MgO–CaZrO3 50.00 36.43 13.57 – – – – –
MgAl2 O4 34.00 – 0.20 64.0 1.0 0.40 – –
ZrO2 – 97.72 – – – 2.10 – –
Clinker raw material 0.88 – 44.79 3.03 1.60 11.86 0.47 0.24
 E.A. Rodríguez et al. / Journal of the European Ceramic Society 33 (2013) 2767–2774 2769

Table 2 examined by Scanning Electron Microscope (Leica, SEM S440)

Refractory composition of materials. equipped with an energy dispersive spectrometer (EDS) for qual-
Compositions (wt%) itative and semi-quantitative microanalysis.
Samples/Bricks MgO CaZrO3 MgAl2 O4 ZrO2
Physical properties (such as apparent porosity and bulk den-
sity) were evaluated by the Archimedes method. The cold
Laboratory scale crushing strength (CCS) was determined using an Instron
MZ 78.0 21.5 – 0.5
MA1 75.5 21.5 2.5 0.5
mechanical testing machine at a strain rate of 1 mm/s. For CCS
MA2 75.0 21.5 3.0 0.5 evaluation, the maximum load supported was registered as the
MA3 74.5 21.5 3.5 0.5 structural failure and the average value from ten specimens was
Industrial scale calculated for every formulation listed in Table 2.
MCZ 85.5 14.0 – 0.5 Laboratory scale specimens for chemical stability were tested
MAS1 83.0 14.0 2.5 0.5 using a static chemical attack method by clinker powders (see
MAS2 81.9 14.0 3.6 0.5 Table 1) at 1450 ◦ C for 4 h. In addition a coatability test by
MAS3 80.7 14.0 4.8 0.5
clinker on the refractory specimen is determined according to the
MAS4 79.5 14.0 6.0 0.5
following criteria: (i) Null, when the pellet of clinker is removed
by just inverting the specimen upside down, (ii) moderate, if the
pellet of clinker is removed by slightly hitting, and (iii) strong,
was synthesized in an electric arc furnace, only coarse and when the pellet of clinker is practically fused with the refractory
intermediate grains were used. The particle size distribu- specimen. The analysis was complemented by SEM.
tion of synthesized magnesia–calcium zirconate mixture in For the industrial scale, mixtures according to compositions
mm was: −4.76 + 2.83 mm (20%), −2.83 + 1.19 mm (44%), given in Table 2 were prepared. Prior to the drying and fir-
−1.19 + 0.595 mm (28%) and −0.595 + 0.250 mm (8%). Fine ing process, dextrine was added as an additive for binding
particle grains (<45 ␮m) were employed only for ZrO2 and mixtures from the green refractory bricks and thus enhancing
MgAl2 O4 . The addition of a fine particle size for spinel and the stability during subsequent handling. Powders were uni-
zirconia was done with the aim of enhancing the activity with axial pressed into a metallic mold obtaining refractory bricks of
the surrounding matrix. 228 mm × 114 mm × 76 mm. After pressing, green bricks were
Finally, from the XRD analysis carried out on the raw material dried at 120 ◦ C for 24 h and then sintered at 1650 ◦ C for 7 h in
used at the industrial scale (using the same conditions of XRD an industrial tunnel kiln. Physical properties such as apparent
described above), the crystalline phases were identified as: MgO porosity, bulk density, cold crushing strength, cold modulus of
(periclase-MgO), electrofused MgO–CaZrO3 (periclase-MgO, rupture and hot modulus of rupture (at 1260 ◦ C) were measured
calcium zirconate – CaZrO3 ), MgAl2 O4 (periclase-MgO and based on ASTM standards (830-93, 133-97, 583-80). Phase anal-
spinel-MgAl2 O4 ) and ZrO2 (baddeleyite-ZrO2 ). yses of the sintered bricks were carried out by X-ray diffraction
The effect of MgAl2 O4 spinel on a MgO–CaZrO3 refrac- study, Philips X’pert equipment using, Cu K␣ radiation source
tory was studied through the development of five refractory (λ = 1.54056 Å) and a setting of 40 kV and 30 mA. All XRD
formulations; the first one without addition of spinel (MCZ data were collected under the same experimental conditions; in
formulation) and the rest with variations in the spinel content. the angular range 10◦ ≤ 2θ ≤70◦ using θ/2θ configuration with a
Table 2 is shown the refractory formulations. It is important step scan of 0.05◦ and 1 s per step exposure time. Sintered bricks
to mention that the reduction of calcium zirconate content in morphology was examined with a scanning electron microscope
industrial formulations comparing to laboratory mixtures is due (Joel, JSM-6490) equipped with EDS.
to a better cost/benefit ratio. In the other hand, the use of an In order to determine the corrosion resistance and coata-
electrofused MgO–CaZrO3 mixture in industrial bricks as raw bility adherence in the refractory bricks, it was carried out
material leads to a higher resistance against chemical attack both static and dynamic chemical attack evaluation (ASTM
in comparison with the sintered CaZrO3 used in laboratory 768-85 and 874-85, respectively) by clinker raw powders (see
scale. Table 1 for clinker raw material chemical composition). The
static method intends to make raw material contact with the
2.2. Preparation and characterization of refractory refractory and evaluates clinker adherences by heating. Using
samples and bricks a crucible (114 mm × 114 mm × 76 mm) made from the refrac-
tory brick, a hole with 50 mm in diameter and 50 mm in depth
For laboratory scale, refractory mixtures with compositions was bored in the brick. Afterwards, the hole was filled with
according to the percentage listed in Table 2 were prepared. The the clinker raw material. Then crucible sample was heated in
powders were uniaxial pressed in a metallic mold for obtaining a gas furnace for 4 h at 1450 ◦ C. The adherence of the clinker
samples of 10.5 mm in height by 11.5 mm in diameter. After on the refractory was determined according to the same criteria
pressing, green compacts were sintered at 1650 ◦ C using natural established in the laboratory scale: null, moderate and strong
gas for 7 h in an industrial tunnel kiln. The phase analysis of adherences.
the sintered samples was determined using the same Siemens For the dynamic method a laboratory rotary kiln was lined
model D5000 diffractometer used for raw materials characteri- with the refractory to be tested. The temperature was increased
zation. Microstructure characteristics of sintered samples were up to 1450 ◦ C with an air/O2 burner. The raw material (the
2770 E.A. Rodríguez et al. / Journal of the European Ceramic Society 33 (2013) 2767–2774

Table 3
Physical properties of MgO–CaZrO3 and MgO–CaZrO3 –MgAl2 O4 refractory bricks.
Formulation Physical properties

Apparent Bulk density Cold crushing Modulus of Hot modulus of rupture

porosity (%) (g/cm3 ) strength (MPa) rupture (MPa) @ 1260 ◦ C (MPa)

MZ 3.2 3.51 130 – –

MA1 2.6 3.58 237 – –
MA2 2.2 3.59 252 – –
MA3 2.1 3.61 321 – –
MCZ 18.4 2.94 39.6 9.6 5.7
MAS1 17.7 3.01 54.2 12.7 14.1
MAS2 17.1 3.02 56.5 11.6 12.3
MAS3 18.1 2.98 47.2 8.7 11.3
MAS4 18.6 2.97 49.3 9.5 11.8

amount of 3–5 kg) was fed continually during 4 h rotating the
kiln at 1 r.p.m. After the test, the degree of infiltration and the
microstructure damage of the refractory were evaluated by elec-
tronic microscopy on polished sections.
3. Results and discussion
3.1. Physical properties
Physical properties of each refractory formulations studied
during the first stage of this investigation are presented in Table 3.
According to the results, in general the apparent porosity was
established around 2.1–3.2% for the refractory specimens, these
values could be registered due to a very fine powder and the
high pressure of pressing used during the conformation stage
(200 MPa). From the table we can observe that the apparent
porosity tends to diminish with the addition of the spinel phase.
In fact, in order to avoid thermal overloads it is desirable that the
refractory bricks present a certain level of porosity which may
promote the adhesion of a given coating.
According to the density, the tendency is to increase with the
increment of spinel content. Since the spinel phase has a lower
melting point than the rest of the components in the refractory
matrix, it could be located in the high energy areas, i.e. the grain
boundary between MgO and CaZrO3 particles. Then, the spinel
located in the grain boundaries may act as a doping agent inhib-
iting the grain boundary movement. As the grain boundaries are
restrained to displace, the porosity will be consumed impacting
positively in the densification process. Beside, the spinel phase
could be developing a ceramic bonding that provides an easy
mass transport between MgO and CaZrO3 particles.
The mechanical analysis for the pressed specimens showed
an increment in the cold crushing strength as the spinel con-
tent was increased. The values of the crushing strength for
the specimens MZ, MA1, MA2 and MA3 were determined
as 130 MPa, 237 MPa, 252 MPa and 321 MPa, respectively. As
observed, there is an increment of approximately 245% between
the specimens MZ and MA3. This large increment in the crus-
hing strength allows us to scale this study at the industrial level.
In industrial bricks, the overall apparent porosity values
were estimated around 17–18%. The porosity level of the
bricks produced at the different conditions was suitable for
the purposes of the investigation, since it could be helping to
develop a good adherence of the coating.
The maximum crushing strength value in the refractory bricks
was found for the MAS2 specimen, which showed minimum
porosity value. There is maximum increment in resistance of
approximately 42.67% between the specimens MCZ and MAS2,
due to the addition of MgAl2 O4 .
For cold and hot modules of rupture the addition of MgAl2 O4
spinel resulted in higher values compared to MCZ specimen.
Moreover, the optimum content of spinel for both modulus of
rupture was observed for the MAS1 specimen. The cold modulus
increased 32.9% by the addition of MgAl2 O4 , while the hot
modulus increased 144%.
As was observed in the laboratory and the industrial scales
studies, the overall properties of the MgO–CaZrO3 refractory
matrix were improved through the addition of the spinel phase.
3.2. X-ray diffraction analysis
The X-ray diffraction (XRD) study obtained from each
refractory mixture studied in a laboratory scale revealed the
presence of the same phases prior to sintering. However, after
sintering new peaks are observed. In Fig. 1A, CaZrO3 and
Ca0.15 Zr0.85 O1.85 peaks can be appreciated. Clearly, the increase
in intensity of Ca0.15 Zr0.85 O1.85 phase, corresponds to the free
lime reacted with zirconia giving a zirconium dioxide stabi-
lized by Ca2+ ions during the sintering process. This phase is
advantageous, since it can possibly develop a microcracks sys-
tem and then strengthening the refractory matrix. On the other
hand, it is well known that microcraks are good stress absorbent
to prevent cracking propagation.14–16 According to diffraction
patterns of MA1, MA2 and MA3 mixture, it was not possible to
detect spinel phase peak intensity due to the small concentration
used.
In the industrial scale, the X-ray diffraction (XRD) study
obtained from each refractory brick revealed the presence of the
same phases prior to sintering. The phase analysis in the refrac-
tory matrix without spinel addition showed periclase (MgO), as
well as calcium zirconate (CaZrO3 ). This indicates that there is
no evidence of formation of new phases.
 E.A. Rodríguez et al. / Journal of the European Ceramic Society 33 (2013) 2767–2774
Fig. 1. Comparative XRD phase analysis of refractory mixtures based on
MgO–CaZrO3 with additions of spinel. (A) XRD at laboratory scale and (B)
XRD at industrial scale.
Small peaks related to the magnesium aluminate spinel could
be identified in XRD patterns at 2θ = 43◦ , 53◦ and 18◦ for the
bricks with different spinel contents (MAS1 to MAS4). The
intensity of MgAl2 O4 peaks almost does not increase at higher
spinel addition (Fig. 1B).
3.3. Microstructural analysis of refractory mixtures
Fig. 2A shows a SEM image of the sintered microstructure
corresponding to the MgO–CaZrO3 matrix (MZ formulation).
A homogeneous microstructure matrix composed by two well-
distributed phases and a low porosity can be observed. In
Fig. 2B, several 120◦ dihedral angles appear through the entire
microstructure which indicates an excellent sintering process. It
is well known that 120◦ dihedral angles represent the equilib-
rium between grains boundaries, since there is no more energy
available, i.e. no more curvature to be used for the sintering
process. Magnesia (gray grains) and calcium zirconate (white
grains) were identified by EDS analysis. At higher magnifica-
tion an excellent ceramic bonding between MgO and CaZrO3
grains was detected (see Fig. 2C). Finally, it is important to
notice the cuasi-spherical shape of calcium zirconate particles
2771
indicating once again a good tendency to full densification in
other words a good sintering process (see Fig. 2D).
The result obtained by microstructural analysis of the
MgO–CaZrO3 –MgAl2 O4 matrix (MA1, MA2 and MA3)
revealed, as in the MgO–CaZrO3 matrix, the presence of
well-distributed magnesia and calcium zirconate grains in the
microstructure as mainly phases, a low porosity (see Table 2)
and dihedral angles. The MgAl2 O4 spinel phase was not found at
low magnifications, but at higher magnifications, it was revealed
that spinel was located at the boundaries between MgO and
CaZrO3 particles. It seems this phase act as bonding between
these particles (see Fig. 3A).
It is important to point out that the addition of CaZrO3 grains
in all the refractory mixtures produced a grain refinement in
the MgO matrix. Calcium zirconate avoided abnormal grain
growth in all the cases providing to the matrix a mechanical
reinforcement (see Fig. 3B).
Fig. 4A shows a SEM image of the sintered microstructure of
the brick without the addition of spinel (MCZ brick). It can be
observed a homogeneous microstructure matrix formed by two
well-distributed phases. Using EDS, the dark gray grains were
identified as magnesia, while the light gray grains were identi-
fied as calcium zirconate. This microstructure showed calcium
zirconate aggregates (fused grains) surrounded by the periclase
groundmass. At higher magnification an excellent bond linkage
between MgO and CaZrO3 grains was observed.
In Fig. 4B and C, it is shown the microstructure corre-
sponding to refractory bricks with addition of MgAl2 O4 spinel
(MAS1, MAS2, MAS3 and MAS4). This microstructural anal-
ysis revealed, the presence of well-distributed magnesia (dark
gray grains) and calcium zirconate grains (light gray grains) in
the microstructure as the main phases. In these images, calcium
zirconate aggregates (fused grains) are surrounded by the peri-
clase groundmass. The spinel phase was also located at the grain
boundaries, as for the MCZ specimens.
Thus, it is proposed that spinel (MgAl2 O4 ) phase was diffused
into the grain boundaries due to its low melting temperature.
Besides, the fine particle size of spinel phase contributed to the
arrangement in those specific zones. A high magnification SEM
image for MAS4 brick is shown in Fig. 4D, in which spinel
phase was identified in the boundaries between magnesia and
calcium zirconate particles. The detailed EDS analysis (labeled
MgAl2 O4 in Fig. 4D) revealed the presence of Mg and Al on
the boundary between MgO and CaZrO3 .
Small particles of spinel (MgAl2 O4 ) located at the grain
boundaries and triple points of the coarse aggregate of electro-
fused MgO–CaZrO3 , are not compatible with the CaZrO3 and
MgO. Because of this, the sintering process of these materials
takes place in the presence of liquid phase (the liquid phase in
the current system, represented by MgAl2 O4 spinels could be
acting as a ceramic bonding). As it was predicted by Liddle and
Brett,35 the liquid phase could appear between MgAl2 O4 and
CaZrO3 at above 1650 ◦ C. The appearance of the liquid phase
below this temperature is related to the presence of transition
phases as the calcium aluminates. For CaO–MgO–Al2 O3 sys-
tem, in MgO–C12 A7 –C3 A co-stability triangle, the liquid phase
appeared at 1321 ◦ C as the eutectic point.36
2772 E.A. Rodríguez et al. / Journal of the European Ceramic Society 33 (2013) 2767–2774

Fig. 2. SEM images of the MZ sample: (A) Dark grains-MgO and bright grains-CaZrO3 . (B) Dihedral angles. (C) Ceramic bonding between MgO and CaZrO3 . (D)
Spherical grains of CaZrO3 .

Fig. 3. SEM images of the MZ sample: (A) Spinel location at high magnifications and its corresponding EDS. (B) CaZrO3 avoiding abnormal grain growth
phenomenon.

3.4. Coatability and corrosion resistance refractory bricks for industrial scale (static and dynamic meth-
ods). In general, the hot face in every refractory was no corroded
Table 4 shows the results obtained for the coating adhe- by clinker liquid phases. Anyhow, it was registered a brown
sion in the samples at laboratory scale (static method) and molten phase entered up into the matrix. According to this
test, for the laboratory scale there was not a good compatibil-
Table 4 ity between the clinker raw material and the refractory matrix
Adherence test with clinker raw material. due to a null adherence in every formulation tested. Fig. 5A
Brick type Static test Dynamic test shows a SEM image from the MA3 specimen, revealing nei-
ther a microstructure without trace of clinke infiltration, nor
Infiltration distance (mm) Adherence Adherence
the presence of any detriments new phase. Thus, taking into
MZ – Null – account that the MA3 specimen represents the highest spinel
MA1 – Null –
content; the microstructure of these specimens suggests an
MA2 – Null –
MA3 – Null – excellent chemical stability in contact with the clinker raw mate-
MCZ 9.0 Strong Moderate rial.
MAS1 6.4 Strong Moderate For the industrial scale, MCZ refractory brick showed the
MAS2 5.6 Strong Moderate maximum coating adherence. In addition, CaZrO3 is very com-
MAS3 5.7 Moderate Moderate
patible with clinker phases, permitting a good coating adherence.
MAS4 5.7 Moderate Moderate
It is clear that there is a strong tendency to coating adherence
 E.A. Rodríguez et al. / Journal of the European Ceramic Society 33 (2013) 2767–2774 2773

Fig. 4. SEM images of refractory sample at industrial scale: (A) Without spinel content. (B and C) With addition of MgAl2 O4 spinel. (D) EDS corresponding
toMgAl2 O4 spinel.

at low spinel contents. However, at higher contents the coating The examination of the infiltrated region by cement clinker
adherence becomes moderate in spite of a porosity value similar in accordance to the SEM analysis corresponding to static and
to that registered by MCZ brick, since the spinel (MgAl2 O4 ) has dynamic tests resulted in a slightly densification without corro-
a low coating adherence. sion evidence; i.e. areas close to clinker did not suffer any kind

Fig. 5. SEM images of sample attacked by clinker raw meal (A) at laboratory scale without trace of clinker infiltration, (B–D) at industrial scale revealing a slightly
dense microstructure due to the clinker infiltration labeled as CK.
2774 E.A. Rodríguez et al. / Journal of the European Ceramic Society 33 (2013) 2767–2774
of alteration. The bonding strength still remained in all refrac-
tory bricks. Fig. 5B–D revealed a slight dense microstructure
in all bricks due to the infiltration of clinker phases (labeled as
CK “clinker phase” in the figures) located as interstitial parti-
cles between periclase grains (dark gray) and calcium zirconate
grains (light gray).
4. Conclusions
Through our study we can conclude:
There is a good correlation between laboratory and industrial
results.
In the refractory mixtures (MgO–CaZrO3 ), an excellent bond
linkage between MgO and CaZrO3 grains was detected. In addi-
tion, calcium zirconate avoided MgO abnormal grain growth in
all the formulation studied here, thus this fact lead to a mechan-
ical reinforcement in the entire matrix.
Spinel phase was located at the boundaries between MgO and
CaZrO3 particles acting as a bond linkage among these phases
according to high magnification microstructural analyses and
EDS results.
The mechanical analyses indicate a cold crushing strength
tendency to increase with spinel addition.
References
1. Peng C, Li N, Han B. Effect of zircon on sintering composition and
microstructure of magnesia powder. Sci Sinter 2009;41:11–7.
2. Han B, Li Y, Guo C, Li N, Chen F. Sintering of MgO refractories with added
WO3 . Ceram Int 2007;33:1563–7.
3. Bartha P, Klischat HJ. Classification of magnesia bricks in rotary
cement kilns according to specification and service-ability. ZKG Int
1994;47:277–80.
4. Komatsu H, Arai M, Ukawa S. Current and future status of chrome-free
bricks for rotary cement kilns. Taikabutsu Overseas 1999;19:3–9.
5. Bartha P, Klischat HJ. Present state of the refractory lining for cement kilns.
CN Refract 1999;6:31–8.
6. Andrzej A, Wajdowicz. Correlation of basic refractory brick development
with the evolution of cement industrial kilns. In: UNITECR’93 congress.
1993. p. 173–97.
7. Mukerji J. Refractory for cement rotary kiln. Jadarpur: Central Glass and
Ceramic Research Institute; 2013. p. 125–32.
8. Griffin DJ. The application of magnesia–zirconia brick in the burning zones
of rotary cement kilns. In: UNITECR’89 proceedings. 1989. p. 350–67.
9. Partridge TJ. Cement kiln refractories the chrome-free solution. Export
Regional Sales Manager, Baker Refractories; 1996. p. 2–6.
10. Itoh K, Takemoto M, Ogata M. Recent trends in cement kiln refractories.
Shinagawa Tech Rep 2000;43:45–58.
11. Kozuka H, Kaita Y, Tuchiya Y, Honda T, Ohta S. New kind of chrome-
free (MgO–CaO–ZrO2 ) bricks for burning zone of rotary cement kiln. In:
UNITECR’93. 1993. p. 1027–37.
12. Kozuka H, Kajita Y, Tuchiya Y, Honda T, Ohta S. Further improvements of
MgO–CaO–ZrO2 refractory bricks. In: UNITECR. 1995.
13. Szczerba J. Calcium zirconate as the secondary phase of magnesia refrac-
tories for cement rotary kiln. Adv Sci Technol 2010;70:15–20.
14. Serena S, Sainz MA, De Aza S, Caballero A. Experimental determination
and thermodynamic calculation of the ZrO2 –CaO–MgO phase equilibria
diagram. Isothermal sections at 1600 ◦ C, 1700 ◦ C and 1750 ◦ C. J Am Ceram
Soc 2004;87:2268–74.
15. Serena S, Sainz MA, De Aza S, Caballero A. Thermodynamic assessment
of the system ZrO2 CaO–MgO using new experimental results. Calcula-
tion of the isopletal section dolomia–zircona. J Eur Ceram Soc 2005;25:
681–93.
16. Serena S, Sainz MA, Caballero A. The system clinker–MgO–CaZrO3 and
its application to the corrosion behavior of CaZrO3 /MgO refractory matrix
by clinker. J Eur Ceram Soc 2009;29:2199–209.
17. Álvaro Obregón A, Rodríguez-Galicia JL, López-Cuevas J, Pena P, Baudín
C. MgO–CaZrO3 -based refractories for cement kilns. J Eur Ceram Soc
2011;31:61–74.
18. Chen M, Lu C, Yu J. Improvement in performance of MgO–CaO refractories
by addition of nano-sized ZrO2 . J Eur Ceram Soc 2007;27:4633–8.
19. Rodríguez JL, Baudín C, Pena P. Relationships between phase constitution
and mechanical behaviour in MgO–CaZrO3 –calcium silicate materials. J
Eur Ceram Soc 2004;24:669–79.
20. Rodríguez-Galicia JL, Fernandez-Arguijo B, Rendón-Ángeles JC, Valle FJ,
Pena P. Reaction sintering of mexican dolomite–zircon mixtures. Boletin
Soc Esp Ceram Vídrio 2005;44:185–91.
21. Guo Z, Palco S, Rigaud M. Bonding of cement clinker onto doloma-based
refractories. J Am Ceram Soc 2005;88:1481–7.
22. Rodríguez JL, Pena P. Obtención de materiales de magnesia-circonato dicál-
cico por sinterización reactiva de mezclas dolomita – circón, Estudio del
procesamiento. Boletin Soc Esp Ceram Vídrio 2001;40:463–71.
23. Rodríguez JL, Rodríguez MA, De Aza S, Pena P. Reaction sintering of
zircon–dolomite mixtures. J Eur Ceram Soc 2001;21:343–54.
24. Rodríguez JL, De Aza S, Pena P. Effect of agglomerate and grain size
on the reaction sintering of zircon–dolomite mixtures. Brit Ceram Trans
2001;100:181–91.
25. Dávila JC. Refractarios magnesita-espinel para horno de cemento: experi-
encias, limitaciones y alternativas. In: UNITECR89 proceedings. 1989. p.
306–23.
26. Rodríguez EA, Alan-Castillo G, Guzmán A-M. Effect of spinel “MgAl2 O4 ”
on microstructural and mechano-chemical properties evolution in a sintered
MgO–CaZrO3 refractory matrix. Ind Ceram 2011;31:1–5.
27. Rodríguez EA, Alan-Castillo G, Jose Contreras E, Das TK. Microstructure
and properties of spinel–magnesia–zirconia–calcium zirconate refractory
mixtures. In: UNITECR’07 congress. 2007. p. 612.
28. Rodríguez-Galicia JL, de Aza AH, Rendón-Angeles JC, Pena P. The mech-
anism of corrosion of MgO/CaZrO3 –calcium silicate materials by cement
clinker. J Eur Ceram Soc 2007;27:79–89.
29. Serena S, Sainz MA, Caballero A. Corrosion behavior of MgO/CaZrO3
refractory matrix by clinker. J Eur Ceram Soc 2004;24:2399–406.
30. Rodríguez EA, Castillo GA, Guzmán AM. Development of a chrome-free
refractory based on electrofused magnesia–calcium zirconate technology
using spinel addition for use in cement rotary kilns. In: UNITECR’09
congress. 2009. p. 072.
31. Rodríguez E, Alan Castillo G, Contreras J, Aguilar Martínez JA, González Y,
Guzmán A-M, Laura I, Ortiz G. Desarrollo de un refractario MgO–CaZrO3
dopado con MgAl2 O4 para la industria cementera. Ciencias UANL
2011;15:23–30.
32. Contreras JE, Castillo GA, Rodríguez EA. About the influence of electro-
fused FeAl2 O4 in a refractory base on MgO–CaZrO3 by sintering on this
properties. In: UNITECR’03. 2003.
33. Contreras JE, Castillo GA, Rodríguez EA, Das TK, Guzmán AM.
Microstructure and properties of hercynite–magnesia–calcium zirconate
refractory mixtures. Mater Charact 2005;54:354–9.
34. Edén Rodrígueza E, Alan Castillo A, Contreras J, Puente-Ornelas R,
Aguilar-Martínez JA, García L, Gómez C. Hercynite and magnesium alumi-
nate spinels acting as a ceramic bonding in an electrofused MgO–CaZrO3
refractory brick for the cement industry. Ceram Int 2012;38(8):6769–75.
35. Liddle J, Brett NH. Phase equilibria in the system CaO–MgO–Al2 O3 –ZrO2 .
Trans J Br Ceram Soc 1984;84:128–34.
36. Atlas S. Verein Deutscher Eisenhuttenleute (VDEh). 2nd ed. D-Dusseldorf,
Germany: Verlag Stahleisen GmbH; 1995. ISBN 3-514-00457-9.