The effects of pre-sintering temperature and La2O3 addition on sinterability and corrosion resistance of stoichiometric magnesium aluminate spinel

Abstract

In this study, effects of pre-sintering temperature and subsequent La2O3 addition on the densification of magnesium aluminate spinel (MgAl2O4) were investigated. Further, the role of La2O3 was evaluated in terms of sintering/mechanical properties and corrosion resistance. Results reveal that two-step sintering achieves higher relative density than one-step pressureless sintering (from 70.21% to 83.86%) because it avoids volume expansion effect of MgAl2O4, but high pre-sintering temperature leads to a decrease in the activity of the raw material (for post-sintering). Subsequent experimental data show that the introduction of La2O3 endows the sample with greater relative density (from 91.24% to 95.65%), which is attributed to the formation of LaAlO3 and defect activation effect caused by stoichiometric mismatch. Besides, corrosion resistance of MgAl2O4 gets improved due to the formation of 2CaO·4La2O3·6SiO2, CaO·2Al2O3, and CaO·6Al2O3.

Full text

Ceramics International 48 (2022) 32470–32478
Available online 21 July 2022
0272-8842/© 2022 Elsevier Ltd and Techna Group S.r.l. All rights reserved.
The effects of pre-sintering temperature and La
2
O
3
addition on sinterability
and corrosion resistance of stoichiometric magnesium aluminate spinel
Xinming Ren
a
,
b
, Beiyue Ma
a
,
b
,
*
, Jialong Tian
b
, Zhouhua Jiang
b
a
Key Laboratory for Ecological Metallurgy of Multimetallic Mineral (Ministry of Education), Shenyang, 110819, Liaoning, PR China
b
School of Metallurgy, Northeastern University, Shenyang, 110819, Liaoning, PR China
ARTICLE INFO
Keywords:
Spinel
La
2
O
3
Corrosion resistance
Cement clinker
Sintering temperature
ABSTRACT
In this study, effects of pre-sintering temperature and subsequent La
2
O
3
addition on the densication of mag-
nesium aluminate spinel (MgAl
2
O
4
) were investigated. Further, the role of La
2
O
3
was evaluated in terms of
sintering/mechanical properties and corrosion resistance. Results reveal that two-step sintering achieves higher
relative density than one-step pressureless sintering (from 70.21% to 83.86%) because it avoids volume
expansion effect of MgAl
2
O
4
, but high pre-sintering temperature leads to a decrease in the activity of the raw
material (for post-sintering). Subsequent experimental data show that the introduction of La
2
O
3
endows the
sample with greater relative density (from 91.24% to 95.65%), which is attributed to the formation of LaAlO
3
and defect activation effect caused by stoichiometric mismatch. Besides, corrosion resistance of MgAl
2
O
4
gets
improved due to the formation of 2CaO⋅4La
2
O
3
⋅6SiO
2
, CaO⋅2Al
2
O
3
, and CaO⋅6Al
2
O
3
.
1. Introduction
The spinels belong to a class of minerals with general formulation of
AB
2
O
4
, where A belongs to divalent cations such as Mg
2+
, Fe
2+
, Zn
2+
,
and Mn
2+
, and B indicates trivalent cations such as Al
3+
, Fe
3+
, Cr
3+
, and
V
3+
[1]. Among the various congurations, magnesium aluminate
spinel (MgAl
2
O
4
) has been widely studied and applied in many heavy
industries due to its high refractory degree, stable physicochemical
property, and excellent high-temperature mechanical properties [2–5].
In general, the development of an innitely near-dense microstructure is
the primary goal in MgAl
2
O
4
manufacturing as it leads to better overall
performance, such as high slag resistance (for lining applications) [6] or
lower refractive index (for optical applications) [7]. Till date, although
some new sintering technologies, such as hot (isostatic) pressing sin-
tering [7,8], ash sintering [9], spark plasma sintering [10], and electric
eld-assisted sintering [11] have been proposed for the fabrication of
MgAl
2
O
4
; based on the consideration of the total cost and process dif-
culty factors, the traditional pressureless sintering method remains an
attractive alternative. Moreover, to deal with the detrimental effect of
the volume expansion accompanying the in situ formation of MgAl
2
O
4
on
densication, the pressureless sintering was further optimized as a
two-step sintering method [12–15]. In fact, the densication and grain
growth of ceramics always accompany and inuence each other, in
particular, in the nal stage of sintering. Thus, a suitable pre-sintering
temperature is always pursued.
Furthermore, the utilization of various additive sources, including
halides (KF [16], NaF [16], AlF
3
[15], MnF
2
[17], CoF
2
[17], LiCl [16],
NaCl [16], MgCl
2
[18,19], AlCl
3
[15,18], and KCl [16]) represented by
LiF [16–18] and oxides (B
2
O
3
[20,21], SiO
2
[22], CaO [12,23], TiO
2
[21,
22,24], V
2
O
5
[21], MnO
2
[18], Cr
2
O
3
[21,25], Y
2
O
3
[26,27], ZrO
2
[28],
La
2
O
3
[26,27,29], CeO
2
[26,29,30], Sm
2
O
3
[5], Eu
2
O
3
[26], Gd
2
O
3
[29], Dy
2
O
3
[31], Yb
2
O
3
[24,29], Sm
2
O
3
–Y
2
O
3
[32], and Sm
2
O
3
–La
2
O
3
[32]) represented by rare earth oxides, has been considered as another
effective research attempt for obtaining dense MgAl
2
O
4
ceramics/re-
fractories in recent years. In most cases, additives achieve better
densication by triggering the liquid phase sintering, including particle
rearrangement (the initial stage of sintering) and plastic ow (all ha-
lides, B
2
O
3
, CaO, TiO
2
, and CeO
2
) [12,16,20,22,26]. However, owing to
the limitation of the high-temperature performance, these additives are
only suitable for the processing of MgAl
2
O
4
-based ceramics, and not for
the manufacture of MgAl
2
O
4
-based refractories. Based on the mecha-
nism of mass transfer, solid-state diffusion has been considered the main
process for densication at the nal stage of MgAl
2
O
4
sintering. This
process is governed by the rate of O
2−
, which is at least four orders of
magnitude lower than that of Mg
2+
and Al
3+
[33]. Interestingly, ac-
cording to the literature, the benecial effects of additives are attributed
* Corresponding author. No. 3–11, Wenhua Road, Heping District, Shenyang City, Liaoning Province, 110819, PR China.
E-mail address: maceramic@126.com (B. Ma).
Contents lists available at ScienceDirect
Ceramics International
journal homepage: www.elsevier.com/locate/ceramint
https://doi.org/10.1016/j.ceramint.2022.07.193
Received 16 June 2022; Received in revised form 10 July 2022; Accepted 17 July 2022

Ceramics International 48 (2022) 32470–32478
32471
to the formation of cationic vacancies, even though this assumption is
contradictory to the dominance of oxygen vacancies [24,25,29,30]. As
far as rare earth elements are concerned, they often cannot form solid
solutions with MgAl
2
O
4
because of their excessively large ionic radii;
nonetheless, they can react with the host matrix (mainly Al
2
O
3
) to form
a second phase. Therefore, their actual impact depends largely on the
stoichiometric ratio (MgO:Al
2
O
3
) of the MgAl
2
O
4
structure, which has
often been overlooked. Among the various rare earth oxide additives,
La
2
O
3
has arisen as quite a promising candidate due to its low cost (the
cheapest rare earth oxide) and excellent efciency (one of the few rare
earths that can react with both MgO and Al
2
O
3
). However, the control
mechanism of La
2
O
3
on densication has not yet been fully understood,
and it has even been reported to be detrimental [29]. Therefore, eluci-
dation of this question is expected to provide new insights into the
densication mechanism of MgAl
2
O
4
, which is of great practical sig-
nicance for the preparation of dense MgAl
2
O
4
-based refractories.
Furthermore, our recent studies have shown that La
2
O
3
possesses the
effect of purifying grain boundaries, which is likely to have a positive
impact on the corrosion resistance of MgAl
2
O
4
refractories for cement
rotary kilns applications [34].
Based on these considerations, the impact of the pre-sintering tem-
perature (1100–1500 ◦C) and La
2
O
3
addition (after determining the pre-
sintering temperature, 1–8 wt.%) on the MgAl
2
O
4
densication process
was systematically investigated. Furthermore, the phase composition,
microstructural evolution, mechanical properties, and cement clinker
corrosion resistance were also evaluated. Finally, all the recorded
benecial effects were analyzed in detail and reasonable explanations
were given.
2. Experimental
2.1. Raw materials
Commercially available calcined magnesia (MgO ≥98.50 wt%;
Sinopharm Chemical Reagent Co., Ltd., China) and sintered alumina
(
α
-Al
2
O
3
≥98.50 wt%, particle size ≤45
μ
m; Yingkou, China) were
selected as the starting materials. Lanthanum oxide (La
2
O
3
≥99.90 wt%;
Shanghai Macklin Biochemical Co., Ltd., China) and polyvinyl alcohol
(PVA, type-1750; Sinopharm Chemical Reagent Co., Ltd., China) were
selected as the additive and the binder, respectively.
2.2. Preparation processes
As a distinction for research purposes, the preparation of the samples
was divided into two parts based on two parameters, namely the pre-
sintering temperature and the added amount of La
2
O
3
. When the sam-
ples used to investigate the pre-sintering temperature were studied, rst,
the raw materials were weighed and mixed one by one according to the
formulas presented in Table 1. Second, the uniformly mixed powders
were pressed into cylindrical (Ф20 ×20 mm) green bodies by enforcing
a uniaxial press at 20 MPa for 1 min. Third, the green bodies (samples
MA–MA15) were pre-sintered at 1100–1500 ◦C for 2 h within a high-
temperature chamber furnace (KJ-M1700, China), at the heating rate
of 10 ◦C min
−1
in the temperature range of 25–1000 ◦C and 5 ◦C min
−1
in the range of 1000–1500 ◦C. Finally, the pre-sintered samples were
crushed, sieved (200 mesh), formed, and post-sintered at 1600 ◦C for 3
h. For samples that were used to investigate the La
2
O
3
addition, the
sample with optimum pre-sintering temperature was crushed, sieved
(200 mesh), weighed, and ball-milled (QM-3SP4, China; at 200 r min
−1
for 5 h) according to the formulas presented in Table 1. The uniformly
mixed powders were pressed into cylindrical (Ф20 ×20 mm) and
tabular (50 ×8 ×8 mm
3
) green bodies using a uniaxial press at 150 MPa
for 3 min. Finally, after drying at 120 ◦C for 24 h, the green bodies
(samples MA14–MA14-8L) were sintered at 1600 ◦C for 3 h in a high-
temperature chamber furnace (KJ-M1700, China), at a heating rate of
10 ◦C min
−1
from 25 to 1000 ◦C and 5 ◦C min
−1
from 1000 to 1600 ◦C.
2.3. Characterization and testing
The local phase composition, micromorphology, cold modulus of
rupture, and sintering properties (apparent porosity and bulk density) of
the sintered samples were tested by X-ray diffraction (XRD, Bruker D8,
Germany; 2◦/min, 15◦–85◦), scanning electron microscopy (SEM, Hita-
chi S4800, Japan), and the vacuum method (distilled water as the me-
dium; GB/T 25995-2010), and by using a universal testing machine (0.5
mm min
−1
, WDW-100, China; GB/T 6569-2006). The linear shrinkage
ratio (L
r
), closed porosity (P
c
), and relative density (D
r
) of the sintered
samples were calculated by using the following Eqs. (1)–(3).
Lr=La−Lb
Lb
×100% (1)
Dr=∑Dti×Ri
Db
(2)
Pc=Pt−Pa= (1−Dr) × 100% −Pa(3)
where L
a
and L
b
are the lengths of the samples after and before sintering,
P
a
and P
t
represent the apparent and true porosities of the sintered
samples (%), D
b
and D
r
indicate the bulk density (g cm
−3
) and relative
density (%) of the sintered samples, D
ti
and R
i
denote the theoretical
density (MgAl
2
O
4
, 3.58 g cm
−3
; LaAlO
3
, 6.52 g cm
−3
) of phase i and its
proportion in the test samples, respectively [35].
Moreover, the sessile drop method was used to study the corrosion/
adhesion behavior between the cement clinker (CaO 66.23%, SiO
2
22.21%, Al
2
O
3
4.62%, Fe
2
O
3
3.55%, MgO 2.31%, K
2
O 0.68%, TiO
2
0.31%, K
2
O 0.09%) and the prepared samples. Fig. 1 shows the sche-
matic illustration of the cement clinker test instrument. The specic test
steps are described as follows: (1) The sintered samples and cement
clinker were cut into discs (Ф15 ×3 mm) and cylinders (Ф3 ×3 mm), (2)
when the furnace temperature reached 600 ◦C, the test sample was sent
into the furnace through the automatic moving arm, (3) the sample was
heated at 1550 ◦C and soaked for 10 min, with the heating rate of 10 ◦C
min
−1
from 25 to 1000 ◦C and at 5 ◦C min
−1
from 1000 to 1550 ◦C, and
(4) after cooling down to room temperature, the test sample was xed
with resin and sliced radially for subsequent studies [36].
3. Results and discussion
3.1. Study on pre-sintering temperature
3.1.1. Microstructure
The evolution of the local microstructure can intuitively reect the
sinterability of the material. Therefore, the microstructures of different
Table 1
Formulas of different batches of samples (wt.%).
No. MgO Al
2
O
3
La
2
O
3
PVA Pre-sintering
temperature
(◦C)
Post-sintering
temperature
(◦C)
MA 28.34 71.66 – +5 – 1600
MA11 28.34 71.66 – +5 1100 1600
MA12 28.34 71.66 – +5 1200 1600
MA13 28.34 71.66 – +5 1300 1600
MA14 28.34 71.66 – +5 1400 1600
MA15 28.34 71.66 – +5 1500 1600
MA14-
1L
28.06 70.94 1 +5 1400 1600
MA14-
2L
27.77 70.23 2 +5 1400 1600
MA14-
4L
27.21 68.79 4 +5 1400 1600
MA14-
8L
26.07 65.93 8 +5 1400 1600
X. Ren et al.

Ceramics International 48 (2022) 32470–32478
32472
samples sintered at the temperature of 1600 ◦C for 3 h were recorded
and presented in Fig. 2. Comparative analysis indicates that the surface
of sample MA (Fig. 2a) that was subjected to the pre-sintering process
was covered with large pores and cracks. At the same time, for the pre-
sintered samples (Fig. 2a–f), microstructures became denser post sin-
tering with the increase of the pre-sintering temperature. In fact, as
mentioned before, during the sintering process of spinel ceramics, in
addition to eliminating the pores formed by the green-body compaction
mechanism, it is also necessary to remove the pores generated by the
volume expansion during the in situ formation of spinel. Therefore, the
microstructure of the pre-sintered samples is more ideal, as shown in
samples MA14 (Fig. 2e) and MA15 (Fig. 2f).
3.1.2. Densication and sintering properties
In order to further elucidate the impact of the pre-sintering tem-
perature on the sinterability of the samples and determine the optimal
pre-sintering temperature, the sintering properties of different samples
were tested and the corresponding results are shown in Fig. 3. More
specically, Fig. 3a illustrates the linear shrinkage ratio of the samples,
and as expected, except for sample MA, the values of the remaining
samples are concentrated in the range of 11.16%–12.51% and increase
in order. The linear shrinkage ratio of the sample MA is −1.76%, which
indicates that its volume shrinkage caused by the sintering procedure
does not even offset the volume expansion caused by the in situ
formation of spinel, and its apparent porosity is as high as 27.27%
(Fig. 3b). Owing to this improvement, the apparent porosity of the pre-
sintered samples reduced to the range of 11.79%–18.46%. Fig. 3c
demonstrates that the closed porosity of the samples rst decreases from
2.52% (MA) to 1.86% (MA11) and then increases to 4.35% (MA15). In
general, the formation of closed-pore structures and the increase of the
closed porosity are due to the growth of grains, which represent the nal
stage of the sintering process of ceramics [35]. Moreover, the increased
bulk density (from 2.51 to 3.01 g cm
−3
, Fig. 3d) and the corresponding
decreased relative density (from 70.21% to 83.66%, Fig. 3e) also
conrm that the applied pre-sintering treatment is benecial to the
subsequent densication of the samples. Apparently, the improved ef-
ciency is different for the samples that were pre-sintered at different
temperatures. For instance, compared with sample MA, the relative
density of sample MA11 increases by 9.47%, while that of sample MA15
is only 0.5% higher than that of the sample MA14. Undoubtedly, this
effect is closely related to the amount of spinel generated during the
pre-sintering stage. Fig. 3f and g exhibit XRD patterns of samples MA11
and MA12, which exhibit diffraction peaks of
α
-Al
2
O
3
(crystal face
(113), PDF# 00-010-0173) and MgO (crystal face (200), PDF#
00-004-0829), indicating that the raw materials did not react
completely. However, when the pre-sintering temperature was above
1300 ◦C, only the diffraction peaks of MgAl
2
O
4
(PDF# 00-021-1152)
existed in the recorded XRD patterns of the samples. Moreover, the
Fig. 1. Schematic showing the cement clinker corrosion test instrument.
Fig. 2. SEM images of surfaces of sample MA after sintering at 1600
◦C for 3 h (a) and the samples with different pre-sintering temperatures after post-sintering at
1600 ◦C for 3 h (b–f: 1100–1500 ◦C).
X. Ren et al.

Ceramics International 48 (2022) 32470–32478
32473
crystallite size (calculated by using Scherrer’s formula) of the samples
increased with the increase of the pre-sintering temperature (Fig. 3h),
which stems from the decrease in the surface energy and the increase in
the grain boundary energy [37]. More specically, the crystallite size of
the sample MA15 pre-sintered at 1500 ◦C increased signicantly to
3508 Å, which possibly led to an insufcient driving force for their
post-sintering process. In the light of the sinterability and energy con-
sumption, 1400 ◦C (sample MA14) is considered the best pre-sintering
temperature.
3.2. Study on La
2
O
3
addition
3.2.1. Phase composition and microstructural evolution
The extracted XRD patterns of the samples with different amounts of
La
2
O
3
sintered at 1600 ◦C for 3 h are shown in Fig. 4a. Compared with
sample MA14, in the XRD patterns of the samples with La
2
O
3
, besides
the spinel phase, the diffraction peaks of LaAlO
3
(PDF# 00-004-0829)
are also observed. Moreover, after normalizing the XRD results, it was
found that the intensity of the main diffraction peak (crystal face (110))
of LaAlO
3
gradually increases (Fig. 4b), which indicates the formation of
more LaAlO
3
. Although the presence of MgO is also detected in sample
MA14-8L (crystal face (200)), it is negligible due to the extremely small
amount. The phase contents of different samples calculated by the
Rietveld renement method are shown in Fig. 4c, where it can be
ascertained that the proportion of LaAlO
3
increases from 1.34% for
sample MA14-1L to 9.59% for sample MA14-8L (for the convenience of
statistics, these trace impurity phases are intentionally ignored). Inter-
estingly, the widely reported phenomenon regarding the shift of the
diffraction peaks of spinel due to the introduction of additives was not
observed in this study [24,26,30]. The underlying origins of this effect
are associated with the radius difference between La
3+
and Mg
2+
(Al
3+
is smaller), which is too large (over 30%) to satisfy the conditions for
forming a substitutional solid solution. However, this effect does not
indicate that the presence of La
2
O
3
has no impact on the sintering of the
spinel samples, which is explained in detail in the next section.
Fig. 5 presents the acquired SEM images and energy dispersive
spectrometry (EDS) results of different samples sintered at 1600 ◦C for 3
h. Noteworthy, the sample MA14 discussed in this section is slightly
Fig. 3. (a) Linear shrinkage ratio, (b) apparent porosity, (c) closed porosity, (d) bulk density, and (e) relative density of the samples with different pre-sintering
temperatures and post-sintered at 1600
◦C for 3 h; (f, g) XRD patterns and (h) crystallite size of the samples pre-sintered at 1100–1500 ◦C for 2 h.
Fig. 4. (a, b) XRD patterns and (c) phase content of the samples with different amounts of La
2
O
3
after sintering at 1600 ◦C for 3 h.
X. Ren et al.

Ceramics International 48 (2022) 32470–32478
32474
different from the sample MA14 presented in section 3.1. To ensure the
consistency of the preparation process, the sample MA14 was ball milled
similar to the samples with La
2
O
3
additive. From the densication
perspective, the sample MA14 exhibited the worst results, with many
pores and cracks distributed on its surface (marked with white arrows,
Fig. 5a). Moreover, with the introduction of La
2
O
3
, the pore defects on
the surface of the samples decreased, in particular, for samples MA14-4L
and MA14-8L (Fig. 5d and e). As far as the grain growth behavior is
concerned, the addition of La
2
O
3
improves the crystallinity of the
MgAl
2
O
4
matrix, which leads to the formation of more complete grains
and clearer grain boundaries. Besides, the growth of matrix grains is
promoted to a certain extent, with a well-bonded microstructure. Fig. 5e
exhibits that the microstructure of sample MA14-8L consists of a large
number of euhedral-granular grains and small anhedral-granular grains.
On top of that, the smaller grains are uniformly lled between the larger
grains. Moreover, Fig. 5f illustrates that the small grains in the samples
with La
2
O
3
lead to the in situ formation of LaAlO
3
(point B); however, the
large grains create MgAl
2
O
4
(point A). Obviously, the above-mentioned
results indicate that La
2
O
3
has a positive effect on the sintering process
of the samples. This outcome is inconsistent with the literature results
[29]. In general, it is considered that La
2
O
3
is detrimental to the
densication process of spinel ceramics, and the difference in the crystal
structure of the secondary phase (La
0.9
Al
11.95
O
18.9
) and the spinel ma-
trix is attributed to its formation. In fact, however, the pinning effect of
the second phase has been widely veried to be benecial for the
densication of materials since it avoids abnormal grain growth [27,38,
39]. Fig. 5d and e exhibit that sample MA14-4L (~6
μ
m) with more
La
2
O
3
exhibits a smaller average grain size than the sample MA14-8L
(~3
μ
m) due to the presence of LaAlO
3
. The grain growth behavior of
different samples can be reasonably explained based on this theory;
however, the enhancement of the sintering process of the samples with
La
2
O
3
cannot still be interpreted. As mentioned earlier, due to the dif-
ference in ionic radius, La
2
O
3
and MgAl
2
O
4
cannot form a solid solution,
but MgO or Al
2
O
3
can form a substitutional solid solution with MgAl
2
O
4
.
Thus, as La
2
O
3
reacts with Al
2
O
3
to form LaAlO
3
, the MgAl
2
O
4
matrix
becomes a MgO-rich state, and by considering the charge balance, the
defect reaction can be given by using the following Eqs. (4)–(6) [40,41].
4MgO → 2Mg′
Al +Mg×
Mg +4O×
O+Mg…
i(4)
4MgO +Al×
Al→3Mg′
Al +Mg×
Mg +4O×
O+Al…
i(5)
3MgO → 2Mg′
Al +Mg×
Mg +3O×
O+V‥
i(6)
The massive formation of these defects induces an activated sintering
effect, which ultimately promotes the sintering process of the samples
with La
2
O
3
. Consequently, the optimization of the microstructure en-
dows the samples with better macro-properties.
3.2.2. Sintering and mechanical properties
Fig. 6a and b shows the linear shrinkage ratio and apparent porosity
of different samples sintered at 1600 ◦C for 3 h, respectively. After the
ball-milling procedure, the linear shrinkage of sample MA14 increased
from 12.46% to 13.01%, and the apparent porosity decreased from
12.83% to 4.29%. This effect can be attributed to the mechanical acti-
vation process, which mainly provides ner starting materials, thereby
providing more driving force for subsequent sintering [42]. Fig. 6a and b
reveal that with the addition of La
2
O
3
, the linear shrinkage of the
samples increases, while the apparent porosity decreases, which are
both attributed to the improvement of the sintering procedure.
Furthermore, the closed porosity of the samples shows a decreasing
trend, with only a slight increase for sample MA14-8L (Fig. 6c). In
general, the closed porosity of materials is not only related to the degree
of sintering, but also to the grain size of the matrix because the small
grain size is conducive to the migration and exclusion of closed pores
[27]. Therefore, for the samples containing La
2
O
3
, the closed porosity is
a combination of their sintering degree and grain growth behavior. With
the introduction of La
2
O
3
, LaAlO
3
with a higher theoretical density
(6.52 g cm
−3
) was in situ formed. Consequently, the theoretical density
of the samples increased from 3.58 (sample MA14) to 3.86 g cm
−3
(sample MA14-8L). Moreover, the bulk density of the samples increased
from 3.27 to 3.69 g cm
−3
due to the accelerated sintering process, as
shown in Fig. 6d. Finally, Fig. 6e exhibits that the relative density of the
samples increases from 91.24% to 95.65%; however interestingly, the
relative density of sample MA14-8L is only 0.19% higher than that of
sample MA14-4L. However, the cold modulus of rupture of sample
MA14-8L is 9.13% higher than that of sample MA14-4L (from 126.68 to
138.25 MPa). Fig. 6f shows that all samples with La
2
O
3
exhibit a greater
cold modulus of rupture than sample MA14 without the additive, and
the cold modulus of rupture is positively correlated to the content of
La
2
O
3
. The increase in density is undoubtedly the main reason for the
improved strength of the samples, while the grain size of the matrix
should also be considered. Specically, the changes in the mechanical
Fig. 5. (a–e) SEM images of surfaces and (f) EDS results of the samples with different amounts of La
2
O
3
after sintering at 1600 ◦C for 3 h (M: MgAl
2
O
4
; L: LaAlO
3
).
X. Ren et al.

Ceramics International 48 (2022) 32470–32478
32475
properties of different samples could be evaluated based on Ryshkewitch
and Hall–Petch experimental formulas [43,44]. Therefore, the samples
with La
2
O
3
exhibited a better cold modulus of rupture due to their low
porosity, while sample MA14-8L, with a smaller grain size, performed
particularly well.
3.2.3. Corrosion resistance
The cement clinker corrosion resistance of samples MA14, MA14-2L,
and MA14-8L (as representatives) was also investigated by the sessile
drop method, and the extracted test results are shown in Fig. 7. Analysis
of the SEM images of the radial sections of the samples indicates that the
incorporation of La
2
O
3
led to signicant improvement in the cement
clinker corrosion resistance of samples. More specically, the corrosion
depth of the sample MA14 (759
μ
m, Fig. 7a) is much larger than those of
the samples MA14-2L (607
μ
m, Fig. 7c) and MA14-8L (551
μ
m, Fig. 7c).
Higher magnication images of the reaction interface were also recor-
ded to investigate the reason for this benecial change. Fig. 7d exhibits
that the reaction interface of sample MA14 can be further divided into
the corroded and penetrated/original layers. MgO particles, which are
decomposed from the spinel matrix (Eqs. (7)–(12)), are observed in the
corrosion layer, while they remain in crystallized state because periclase
does not react with the main minerals (alite, belite, tricalcium alumi-
nate, and brownmillerite) of the cement clinker [45]. On the other hand,
due to the potential instability of the penetrated region, the penetrated
and the original layers of sample MA14 could not be accurately distin-
guished. In fact, this effect is detrimental to both the cement clinker
resistance and coating adhesion property of the samples.
MgO ⋅ Al2O3(s) + 3(3CaO ⋅ SiO2)(s)→3(2CaO ⋅ SiO2)(s)
+3CaO⋅Al2O3(s,l) + MgO(s)(7)
7(MgO ⋅ Al2O3)(s) + 12(3CaO ⋅ SiO2)(s)→12(2CaO ⋅ SiO2)(s)
+12CaO⋅7Al2O3(s,l) + 7MgO(s)(8)
MgO ⋅ Al2O3(s) + 3CaO⋅SiO2(s)→2CaO⋅SiO2(s) + CaO⋅Al2O3(s) + MgO(s)
(9)
MgO ⋅ Al2O3(s) + 3CaO(s)→3CaO⋅Al2O3(s,l) + MgO(s)(10)
7MgO ⋅ Al2O3(s) + 12CaO(s)→12CaO⋅7Al2O3(s,l) + 7MgO(s)(11)
MgO ⋅ Al2O3(s) + CaO(s)→CaO⋅Al2O3(s) + MgO(s)(12)
Fig. 7e illustrates that conversely, the reaction interface of sample
MA14-8L can be clearly divided into the corroded, penetrated, and
original layers. In the corroded layer, in addition to MgO particles, a
large number of columnar crystals can also be detected. Combined with
the EDS results (point C, in Fig. 7f), it can be argued that they are
composed of 2CaO⋅4La
2
O
3
⋅6SiO
2
, CaO⋅2Al
2
O
3
, and CaO⋅6Al
2
O
3
. Among
them, 2CaO⋅4La
2
O
3
⋅6SiO
2
is formed by the reaction between LaAlO
3
and cement clinker, and further, with the decrease of SiO
2
in the cement
clinker, CaO⋅2Al
2
O
3
, and CaO⋅6Al
2
O
3
become saturated and get
precipitated. The formation of these phases is believed to be associated
with the improved cement clinker resistance of samples with La
2
O
3
because their presence increases the viscosity of cement clinker [46]. In
the penetrated layer, the presence of these benecial phases was also
found (point D, in Fig. 7f). By lling the grain boundaries and pores of
the spinel matrix, further penetration of the cement clinker is effectively
prevented. Therefore, this region may also be referred to as the isolated
layer. In addition, the sintering properties (especially the apparent
porosity, as shown in Fig. 6b) of the samples were substantially
improved by the introduction of lanthanum oxide, which also has a
non-negligible contribution to the improvement of their corrosion
resistance [47]. Based on the above-mentioned results and their
Fig. 6. (a) Linear shrinkage ratio, (b) apparent porosity, (c) closed porosity, (d) bulk density and theoretical density, (e) relative density and cold modulus of rupture
of the samples with different amounts of La
2
O
3
after sintering at 1600 ◦C for 3 h.
X. Ren et al.

Ceramics International 48 (2022) 32470–32478
32476
analysis, the mechanism of incorporation of La
2
O
3
additive to improve
the cement clinker resistance of the samples is presented herein, as
shown in Fig. 8 and described as follows: (1) the in situ formation of
various ceramic phases (2CaO⋅4La
2
O
3
⋅6SiO
2
, CaO⋅2Al
2
O
3
, and
CaO⋅6Al
2
O
3
) leads to the increase in the viscosity of the cement clinker;
(2) the in situ ceramic phases induce a good bridging effect at the re-
action interface, which enhances the coating adhesion property of the
samples; and (3) the in situ ceramic phases increase the difculty of
cement clinker penetration by lling pores/cracks.
4. Conclusions
The MgAl
2
O
4
spinel with improved sintering performance and
corrosion resistance was successfully prepared by an optimized two-step
pressureless sintering method (pre-sintering at 1400 ◦C +post-sintering
at 1600 ◦C). Based on the results of this study, the following conclusions
can be drawn:
(1) Study on the optimization of pre-sintering temperature indicates
that with the increase in the temperature from 1100 to 1500 ◦C,
the apparent porosity of the samples decreases from 18.46% to
11.79% (the value for the blank sample is 27.27%), and the
corresponding bulk density increases from 2.58 to 3.01 g cm
−3
(density of the blank sample is 2.51 g cm
−3
). This effect can not
only be attributed to the elimination of the volume expansion
during the pre-sintering stage, but also leads to a certain reduc-
tion in the activity of the raw material for the post-sintering stage.
The latter is manifested by a gradually increasing crystallite size
(from 449 to 3508 Å).
(2) Furthermore, when the added amount of La
2
O
3
increases from 1
to 8 wt%, the apparent porosity of the samples decreases from
2.07% to 0.94% (the blank sample is 4.29%), the bulk density
increases from 3.38 to 3.69 g cm
−3
(bulk density of the blank
sample is 3.27 g cm
−3
), and the cold modulus of rupture increases
from 114.33 to 138.25 MPa (the value for the blank sample is
93.84 MPa). All these effects could be attributed to the activated
sintering triggered by the stoichiometric mismatch caused by the
reaction between La
2
O
3
and the matrix. Moreover, the corrosion
depth of the samples decreases from 759 (blank sample) to 607
μ
m (2 wt% of La
2
O
3
) to 551
μ
m (8 wt% of La
2
O
3
), which is
Fig. 7. (a–c) SEM images of the reaction interface between cement clinker and samples; EDS results of samples (d) MA14 and (e, f) MA14-8L after cement clinker
corrosion test.
X. Ren et al.

Ceramics International 48 (2022) 32470–32478
32477
benecial for the in situ formation of 2CaO⋅4La
2
O
3
⋅6SiO
2
,
CaO⋅2Al
2
O
3
, and CaO⋅6Al
2
O
3
.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Acknowledgements
The authors greatly acknowledge the nancial support from the
National Natural Science Foundation of China (Grant nos. U1908223,
U21A2057, and U20A20239) and the Fundamental Research Funds for
the Central Universities (Grant no. N2125002).
References
[1] J.B. Ma, B. Zhao, H.M. Xiang, F.Z. Dai, Y. Liu, R. Zhang, Y.C. Zhou, High-entropy
spinel ferrites MFe
2
O
4
(M =Mg, Mn, Fe, Co, Ni, Cu, Zn) with tunable
electromagnetic properties and strong microwave absorption, J. Adv. Ceram. 11
(2022) 754–768, https://doi.org/10.1007/s40145-022-0569-3.
[2] W. Yan, G.Y. Wu, S.B. Ma, S. Schaff¨
oner, Y.J. Dai, Z. Chen, J.T. Qi, N. Li, Energy
efcient lightweight periclase-magnesium alumina spinel castables containing
porous aggregates for the working lining of steel ladles, J. Eur. Ceram. Soc. 38
(2018) 4276–4282, https://doi.org/10.1016/j.jeurceramsoc.2018.05.002.
[3] L.H. Lv, G.Q. Xiao, D.H. Ding, Y. Ren, S.L. Yang, P. Yang, X. Hou, Combustion
synthesis of C/MgAl
2
O
4
composite powders using magnesium oxalate as carbon
source, Int. J. Appl. Ceram. Technol. 16 (2019) 1253–1263, https://doi.org/
10.1111/ijac.13134.
[4] Y.B. Wen, X.H. Liu, X.Y. Chen, Q.L. Jia, R.H. Yu, T. Ma, Effect of heat treatment
conditions on the growth of MgAl
2
O
4
nanoparticles obtained by sol-gel method,
Ceram. Int. 43 (2017) 15246–15253, https://doi.org/10.1016/j.
ceramint.2017.08.061.
[5] Z.H. Quan, Z.F. Wang, X.T. Wang, H. Liu, Y. Ma, Effects of Sm
2
O
3
addition on
sintering behavior of pre-synthesized magnesia-rich magnesium aluminate spinel,
J. Rare Earths 39 (2021) 1450–1454, https://doi.org/10.1016/j.jre.2021.01.020.
[6] X.M. Ren, B.Y. Ma, S.M. Li, H.X. Li, G.Q. Liu, W.G. Yang, F. Qian, S.X. Zhao, J.
K. Yu, Comparison study of slag corrosion resistance of MgO–MgAl
2
O
4
, MgO–CaO
and MgO–C refractories under electromagnetic eld, J. Iron Steel Res. Int. 28
(2021) 38–45, https://doi.org/10.1007/s42243-020-00421-0.
[7] C. Gajdowski, J. B¨
ohmler, Y. Lorgouilloux, S. Lemonnier, S. d’Astorg, E. Barraud,
A. Leriche, Inuence of post-HIP temperature on microstructural and optical
properties of pure MgAl
2
O
4
spinel: from opaque to transparent ceramics, J. Eur.
Ceram. Soc. 37 (2017) 5347–5351, https://doi.org/10.1016/j.
jeurceramsoc.2017.07.031.
[8] L.L. Zhu, Y.J. Park, L. Gan, S.I. Go, H.N. Kim, J.M. Kim, J.W. Ko, Fabrication of
transparent MgAl
2
O
4
from commercial nanopowders by hot-pressing without
sintering additive, Mater. Lett. 219 (2018) 8–11, https://doi.org/10.1016/j.
matlet.2018.02.010.
[9] H. Yoshida, P. Biswas, R. Johnson, M.K. Mohan, Flash-sintering of magnesium
aluminate spinel (MgAl
2
O
4
) ceramics, J. Am. Ceram. Soc. 100 (2017) 554–562,
https://doi.org/10.1111/jace.14616.
[10] K. Morita, B.N. Kim, H. Yoshida, K. Hiraga, Y. Sakka, Spectroscopic study of the
discoloration of transparent MgAl
2
O
4
spinel fabricated by spark-plasma-sintering
(SPS) processing, Acta Mater. 84 (2015) 9–19, https://doi.org/10.1016/j.
actamat.2014.10.030.
[11] W. Qin, D.N.F. Muche, R.H.R. Castro, K. Van Benthem, The effect of electric elds
on grain growth in MgAl
2
O
4
spinel, J. Eur. Ceram. Soc. 38 (2018) 5512–5518,
https://doi.org/10.1016/j.jeurceramsoc.2018.08.030.
[12] D. Han, J. Zhang, P. Liu, G. Li, S.W. Wang, Effect of CaO on the optical quality and
microstructure of transparent MgO⋅1.5Al
2
O
3
spinel ceramics prepared by reactive
sintering, J. Eur. Ceram. Soc. 38 (2018) 3261–3267, https://doi.org/10.1016/j.
jeurceramsoc.2018.03.032.
[13] Y.Q. Jing, Q. Liu, X.Y. Mao, S. Su, X.Y. Li, X. Liu, T.F. Xie, J. Li, Inuence of
presintering temperature on magnesium aluminate spinel transparent ceramics
fabricated by solid-state reactive sintering, Int. J. Appl. Ceram. Technol. 19 (2021)
367–374, https://doi.org/10.1111/ijac.13888.
[14] H. Zhang, H. Wang, H.G. Gu, X. Zong, B.T. Tu, P.Y. Xu, B. Wang, W.M. Wang, S.
Y. Liu, Z.Y. Fu, Preparation of transparent MgO⋅1.8Al
2
O
3
spinel ceramics by
aqueous gelcasting, presintering and hot isostatic pressing, J. Eur. Ceram. Soc. 38
(2018) 4057–4063, https://doi.org/10.1016/j.jeurceramsoc.2018.04.057.
[15] I. Ganesh, S. Bhattacharjee, B.P. Saha, R. Johnson, Y.R. Mahajan, A new sintering
aid for magnesium aluminate spinel, Ceram. Int. 27 (2001) 773–779, https://doi.
org/10.1016/S0272-8842(01)00029-3.
[16] V. Neˇ
cina, W. Pabst, Comparison of the effect of different alkali halides on the
preparation of transparent MgAl
2
O
4
spinel ceramics via spark plasma sintering
(SPS), J. Eur. Ceram. Soc. 40 (2020) 6043–6052, https://doi.org/10.1016/j.
jeurceramsoc.2020.06.056.
[17] A. Talimian, V. Pouchly, K. Maca, D. Galusek, Densication of magnesium
aluminate spinel using manganese and cobalt uoride as sintering aids, Materials
13 (2019) 102, https://doi.org/10.3390/ma13010102.
[18] S.K. Mohan, R. Sarkar, A comparative study on the effect of different additives on
the formation and densication of magnesium aluminate spinel, Ceram. Int. 42
(2016) 13932–13943, https://doi.org/10.1016/j.ceramint.2016.05.206.
[19] F. Mohammadi, S. Otroj, M.R. Nilforushan, Effect of MgCl
2
addition on the
sintering behavior of MgAl
2
O
4
spinel and formation of nano-particles, Sci. Sinter.
46 (2014) 157–168, https://doi.org/10.2298/sos1402157m.
[20] G. Bhattacharya, S. Zhang, M.E. Smith, D.D. Jayaseelan, W.E. Lee, Mineralizing
magnesium aluminate spinel formation with B
2
O
3
, J. Am. Ceram. Soc. 89 (2006)
3034–3042, https://doi.org/10.1111/j.1551-2916.2006.01189.x.
Fig. 8. Schematic showing the permeation process of cement clinker to the samples with and without additive La
2
O
3
.
X. Ren et al.

Ceramics International 48 (2022) 32470–32478
32478
[21] R. Sarkar, S.K. Das, G. Banerjee, Effect of additives on the densication of reaction
sintered and presynthesised spinels, Ceram. Int. 29 (2003) 55–59, https://doi.org/
10.1016/S0272-8842(02)00089-5.
[22] T. Kim, D. Kim, S. Kang, Effect of additives on the sintering of MgAl
2
O
4
, J. Alloys
Compd. 587 (2014) 594–599, https://doi.org/10.1016/j.jallcom.2013.10.250.
[23] M.T.T. Camargo, Q. Jacques, L.B. Caliman, J. Miagava, D. Hotza, R.H.R. Castro,
D. Gouvˆ
ea, Synthesis of Ca-doped spinel by ultrasonic spray pyrolysis, Mater. Lett.
171 (2016) 232–235, https://doi.org/10.1016/j.matlet.2016.02.114.
[24] S. Sinhamahapatra, C. Ghosh, H.S. Tripathi, S. Mukhopadhyay, Effect of Yb
2
O
3
and
TiO
2
on reaction sintering and properties of magnesium aluminate spinel, Ceram.
Int. 47 (2021) 27372–27385, https://doi.org/10.1016/j.ceramint.2021.06.159.
[25] G.H. Sun, Q.L. Zhang, J.Q. Luo, W.P. Liu, X.F. Wang, S. Han, L.L. Zheng, W.M. Li,
D.L. Sun, Bulk crystal growth of Cr-doped MgAl
2
O
4
spinel by Czochralski method
and properties characterization, Mater. Chem. Phys. 204 (2018) 277–281, https://
doi.org/10.1016/j.matchemphys.2017.10.049.
[26] Y. Zhang, Z.H. Guo, Z.Y. Han, X.Y. Xiao, Effect of rare earth oxides doping on
MgAl
2
O
4
spinel obtained by sintering of secondary aluminium dross, J. Alloys
Compd. 735 (2018) 2597–2603, https://doi.org/10.1016/j.jallcom.2017.11.356.
[27] M.M. Hasan, P.P. Dholabhai, S. Dey, B.P. Uberuaga, R.H.R. Castro, Reduced grain
boundary energies in rare-earth doped MgAl
2
O
4
spinel and consequent grain
growth inhibition, J. Eur. Ceram. Soc. 37 (2017) 4043–4050, https://doi.org/
10.1016/j.jeurceramsoc.2017.04.073.
[28] S. Mandal, S. Mondal, K. Das, Microstructure and phase evolution of Indian
magnesite-derived MgAl
2
O
4
as a function of stoichiometry and ZrO
2
doping, Int. J.
Appl. Ceram. Technol. 15 (2018) 161–170, https://doi.org/10.1111/ijac.12759.
[29] S. Sinhamahapatra, K. Dana, S. Mukhopadhyay, H.S. Tripathi, Role of different rare
earth oxides on the reaction sintering of magnesium aluminate spinel, Ceram. Int.
45 (2019) 11413–11420, https://doi.org/10.1016/j.ceramint.2019.03.007.
[30] Z.H. Quan, Z.F. Wang, X.T. Wang, H. Liu, Y. Ma, Effect of CeO
2
addition on the
sintering behavior of pre-synthesized magnesium aluminate spinel ceramic
powders, Ceram. Int. 45 (2019) 488–493, https://doi.org/10.1016/j.
ceramint.2018.09.194.
[31] H.S. Tripathi, S. Singla, A. Ghosh, Synthesis and densication behaviour of
magnesium aluminate spinel: effect of Dy
2
O
3
, Ceram. Int. 35 (2009) 2541–2544,
https://doi.org/10.1016/j.ceramint.2009.02.011.
[32] X.M. Ren, B.Y. Ma, G.L. Zhang, G.F. Fu, J.K. Yu, G.Q. Liu, Preparation and
properties of MgAl
2
O
4
spinel ceramics by double-doped Sm
2
O
3
–(Y
2
O
3
, Nb
2
O
5
and
La
2
O3), Mater. Chem. Phys. 252 (2020), 123309, https://doi.org/10.1016/j.
matchemphys.2020.123309.
[33] I. Reimanis, H.J. Kleebe, A review on the sintering and microstructure
development of transparent spinel (MgAl
2
O
4
), J. Am. Ceram. Soc. 92 (2009)
1472–1480, https://doi.org/10.1111/j.1551-2916.2009.03108.x.
[34] X.M. Ren, B.Y. Ma, H. Liu, Z.F. Wang, C.J. Deng, G.Q. Liu, J.K. Yu, Designing low-
carbon MgO–Al
2
O
3
–La
2
O
3
–C refractories with balanced performance for ladle
furnaces, J. Eur. Ceram. Soc. 42 (2022) 3986–3995, https://doi.org/10.1016/j.
jeurceramsoc.2022.03.051.
[35] X.M. Ren, B.Y. Ma, J.H. Tang, Y.W. Li, J.K. Yu, Fabrication of energy-saving MgO
with large grain size and low thermal conductivity: towards a new type of
magnesia for high-temperature furnaces, Construct. Build. Mater. 342 (2022),
128097, https://doi.org/10.1016/j.conbuildmat.2022.128097.
[36] C.Y. Guo, E.H. Wang, X.M. Hou, J.Y. Kang, T. Yang, T.X. Liang, G.P. Bei,
Preparation of Zr
4+
doped calcium hexaaluminate with improved slag penetration
resistance, J. Am. Ceram. Soc. 104 (2021) 4854–4866, https://doi.org/10.1111/
jace.17859.
[37] S.H. Tamin, S.B.R.S. Adnan, M.H. Jaafar, N.S. Mohamed, Effects of sintering
temperature on the structure and electrochemical performance of Mg
2
SiO
4
cathode
materials, Ionics 24 (2017) 2665–2671, https://doi.org/10.1007/s11581-017-
2391-4.
[38] S.J. Bennison, M.P. Harmer, Effect of MgO solute on the kinetics of grain growth in
Al
2
O
3
, J. Am. Ceram. Soc. 66 (1983) 90–92, https://doi.org/10.1111/j.1151-
2916.1983.tb10065.x.
[39] M. Hillert, Inhibition of grain growth by second-phase particles, Acta Metall. 36
(1988) 3177–3181, https://doi.org/10.1016/0001-6160(88)90053-3.
[40] C.M. Ni, H.W. Fan, X.D. Wang, M. Yao, Thermal conductivity prediction of
MgAl
2
O
4
: a non-equilibrium molecular dynamics calculation, J. Iron Steel Res. Int.
27 (2020) 500–505, https://doi.org/10.1007/s42243-020-00364-6.
[41] C.J. Ting, H.Y. Lu, Defect reactions and the controlling mechanism in the sintering
of magnesium aluminate spinel, J. Am. Ceram. Soc. 82 (1999) 841–848, https://
doi.org/10.1111/j.1151-2916.1999.tb01844.x.
[42] M.K. Beyer, H. Clausen-Schaumann, Mechanochemistry: the mechanical activation
of covalent bonds, Chem. Rev. 105 (2005) 2921–2948, https://doi.org/10.1021/
cr030697h.
[43] E. Ryshkewitch, Compression strength of porous sintered alumina and zirconia,
J. Am. Ceram. Soc. 36 (1953) 65–68, https://doi.org/10.1111/j.1151-2916.1953.
tb12837.x.
[44] E.D. Jin, L. Yuan, J.K. Yu, D.H. Ding, G.Q. Xiao, Enhancement of thermal shock and
slag corrosion resistance of MgO–ZrO
2
ceramics by doping CeO
2
, Ceram. Int. 48
(2022) 13987–13995, https://doi.org/10.1016/j.ceramint.2022.01.283.
[45] Z.Q. Guo, S. Palco, M. Rigaud, Reaction characteristics of magnesia–spinel
refractories with cement clinker, Int. J. Appl. Ceram. Technol. 2 (2005) 327–335,
https://doi.org/10.1111/j.1744-7402.2005.02027.x.
[46] S. Wright, L. Zhang, S.Y. Sun, S. Jahanshahi, Viscosities of calcium ferrite slags and
calcium alumino-silicate slags containing spinel particles, J. Non-Cryst. Solids 282
(2001) 15–23, https://doi.org/10.1016/S0022-3093(01)00324-6.
[47] Y.S. Liu, E.H. Wang, L.C. Xu, T. Yang, Z.J. He, T.X. Liang, X.M. Hou, Synthesis of
CA
6
/AlON composite with enhanced slag resistance, Int. J. Miner., Metall. Mater.
(2022), https://doi.org/10.1007/s12613-022-2435-2. Online.
X. Ren et al.