Simple preparation method for Mg–Al hydrotalcites as base catalysts

Article abstract of DOI:10.1016/j.molcata.2016.07.018

In this paper, we report a very simple preparation method for Mg–Al hydrotalcites (HTs) as base catalysts via the hydration of their corresponding metal oxides (MgO and Al₂O₃) without any particular controlled variables. We also propose a reasonable explanation for a series of reaction pathways in which Mg–Al HTs are formed from their metal oxides. The Mg–Al HTs were formed by the reaction of Mg²⁺ and Al(OH)₄^?, which resulted from the hydration of both metal oxides. The prepared Mg–Al HTs retained their unique properties, including their layered hydroxide structure and memory effect and showed considerable catalytic activity in the isomerization of glucose to fructose. Consequently, this simple hydration of both metal oxides successfully synthesized Mg–Al HTs, which can be used as base catalysts or as catalyst precursors for calcined HTs in various base-catalyzed reactions.

Full text of DOI:10.1016/j.molcata.2016.07.018

Journal of Molecular Catalysis A: Chemical 423 (2016) 347–355
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Simple preparation method for Mg–Al hydrotalcites as base catalysts
Gihoon Leea, Ji Yeon Kanga, Ning Yanb, Young-Woong Suhc,∗, Ji Chul Junga,∗
a Department of Chemical Engineering, Myongji University, Yongin 17058, Republic of Korea
b Department of Chemical and Biomolecular Engineering, National University of Singapore, 4 Engineering Drive 4, 117585, Singapore
c Department of Chemical Engineering, Hanyang University, Seoul 04763, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history:
In this paper, we report a very simple preparation method for Mg–Al hydrotalcites (HTs) as base catalysts
via the hydration of their corresponding metal oxides (MgO and Al2O3) without any particular controlled
variables. We also propose a reasonable explanation for a series of reaction pathways in which Mg–Al HTs
Received 16 May 2016
Received in revised form 7 July 2016
Accepted 7 July 2016
are formed from their metal oxides. The Mg–Al HTs were formed by the reaction of Mg2+ and Al(OH)4
,
−
Available online 9 July 2016
which resulted from the hydration of both metal oxides. The prepared Mg–Al HTs retained their unique
properties, including their layered hydroxide structure and memory effect and showed considerable
catalytic activity in the isomerization of glucose to fructose. Consequently, this simple hydration of both
metal oxides successfully synthesized Mg–Al HTs, which can be used as base catalysts or as catalyst
precursors for calcined HTs in various base-catalyzed reactions.
Keywords:
Hydrotalcite
Preparation method
Metal oxide
Hydration
© 2016 Elsevier B.V. All rights reserved.
Base catalyst
1. Introduction
polymers, catalysts, electroactive materials, and pharmaceuticals
[11–13]. As such, many attempts have been made to synthesize HTs
Hydrotalcites
(HTs)
are
well-known
layered
dou-
with unique properties to meet the specific requirements of vari-
ous application areas. Accordingly, various preparation techniques
have been extensively investigated, including the co-precipitation
method, sol-gel method, hydrothermal synthesis, microwave irra-
diation, and electrochemical methods. The preparation techniques
of HTs were well reviewed by M.R. Othman et al. [14].
ble hydroxides (LDH) that have the general formula
3+
[M2+
M
x(OH)2][An− ·mH2O], where M2+ and M3+ repre-
1-x
x/n
sents metal components with divalent and trivalent cations,
respectively. For the divalent metal component, magnesium is the
most common, although many other metal species with a cation
size similar to that of Mg are also available, such as Co, Cu, Ni, Ca,
Zn, and Mn. Al generally occupies the positions of the trivalent
metal component, although Fe and Cr can be also employed
[1–5]. The isomorphous substitution of M2+ by M3+ leads to a
brucite-like (Mg(OH)2) layered structure and excessive positive
Although the co-precipitation method has been widely
employed in previous studies, in the preparation of HTs it has a
severe drawback in the excessive number of parameters that must
be considered during the co-precipitation step, which affects the
successful formation of HTs. These parameters include pH value,
temperature, metal precursor, and aging time. Moreover, keeping
the pH value of solution constant throughout the entire precip-
itation procedure is very difficult even when high-speed stirring
is employed, which leads to the low crystallinity of HT. More-
over, the urea method, which uses urea as a precipitating and
pH-control agent, is very sensitive to temperature during prepara-
tion. As such, temperature control is a key factor in obtaining HT by
the urea method. In addition to the above two typical preparation
methods, sol-gel, hydrothermal, combustion, and steam activation
methods can be applied in the preparation of HTs. However, most
preparation methods have rather complex preparation processes
with inherent problems, such as too many synthetic parameters to
achieve precise control, low reproducibility, high preparation cost,
and difficulty in mass production [14].
charges. Accordingly, some available anions, such as CO32−, NO3
,
−
Cl−, SO42−, and OH− are located in the interlayer along with water
molecules to compensate for the excessive positive charges. As a
result, HTs have a unique layered structure with positively charged
brucite-like layers and compensating anions [6–10].
Due to their unique structures, HTs have two representative and
prominent properties: the ability to capture and exchange organic
and inorganic anions and memory effect. These unique proper-
ties have resulted in these compounds having a wide range of
applications as anion exchangers, absorbing agents, stabilizers for
∗
Corresponding authors.
E-mail addresses: ywsuh@hanyang.ac.kr (Y.-W. Suh), jcjung@mju.ac.kr,
jcjung98@gmail.com (J.C. Jung).
http://dx.doi.org/10.1016/j.molcata.2016.07.018
1381-1169/© 2016 Elsevier B.V. All rights reserved.

348
G. Lee et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 347–355
To overcome these problems, here we report a simple method
2.2. Reconstruction process
for the preparation of Mg–Al HTs, which are the most representa-
tive HTs, using their corresponding metal oxides. Of course, there
have been a number of previous attempts to prepare Mg–Al HTs via
the hydration of metal oxides [15–18]. G.Q. Lu et al. demonstrated
that Mg–Al HTs could be prepared in three steps: dissociation,
deposition, and diffusion. Although interesting results regarding
the preparation of Mg–Al HTs from their corresponding metal
oxides were already reported, there has been a lack of clear discus-
sion about the reaction pathway, or, in particular, its uncommon
deposition and diffusion [15]. Furthermore, a slight change in the
initial pH value of the precursor solution greatly affects the final
product, leading to difficulty in obtaining single-phase Mg–Al HTs.
Therefore, the main aim of this paper is to present a very simple
preparation method for obtaining single-phase Mg–Al HTs using
their metal oxides without any particular controlled variables, and
to provide a reasonable explanation for a series of reaction path-
ways in which Mg–Al HTs are formed from their metal oxides. We
hope this paper will contribute to making the preparation process
of HTs easy and simple to achieve using commercially available
materials.
Researchers in the field of heterogeneous catalysts have exhib-
ited great interest in the utilization of the memory effect of HTs for
various base-catalyzed reactions, which is a property unique to HTs.
In the memory effect of HTs, as-synthesized HTs can be transformed
into their corresponding mixed metal oxides by thermal treatment
in air, and in turn, the original structure of HTs with layered double
hydroxides can be successfully recovered by immersing the mixed
metal oxides in water, which is known as the reconstruction pro-
cess. Our group has consistently studied the tuning of the basic
catalytic properties of Mg–Al HTs using this memory effect [19,20].
Accordingly, in addition to preparing Mg–Al HTs using a simple
method, in this study we also focused on their memory effect and
base catalytic properties. We employed glucose isomerization as a
model reaction to explore the catalytic base properties of Mg–Al
HTs.
To confirm the memory effect of the prepared sample, we
also conducted a reconstruction process according to the method
reported in our previous work [19]. The prepared sample was cal-
cined at 450 ◦C for 10 h in air to obtain the calcined Mg–Al HT. Next,
we added the calcined Mg–Al HT (approximately 2 g) to 100 ml of
distilled water, and then stirred the resulting solution vigorously in
an inert atmosphere at 60 ◦C for 24 h. We then filtered the solution
to obtain a solid product, and lastly, dried it overnight at 80 ◦C to
yield the reconstructed Mg–Al HT.
2.3. Characterization
We characterized the crystalline structure of the prepared sam-
ples by powder X-ray diffraction (XRD) measurement, and recorded
the XRD spectra on an X’pert-Pro PANanalytical diffractometer
using Cu-K␣ radiation (␭ = 1.54056 Å). We recorded diffraction pat-
terns within a 2␪ range of 5–80◦. We analyzed the morphological
features of the sample by scanning electron microscopy (SEM),
and obtained SEM images on a JSM-6700F field emission (FE)
microscope operating at 15 kV. We immobilized the samples on
a copper holder using conducting resin and coated them with plat-
inum prior to characterization. We also conducted ICP-AES analysis
(PerkinElmer, OPTIMA 4300DV) in order to confirm the metal con-
tents of the samples. The specific surface areas of the samples
were determined by the Brunauer-Emmett-Teller (BET) method
using N2 adsorption and desorption measurements conducted at
−196 ◦C with a constant-volume adsorption apparatus (BEL Japan,
BELSORP-mini II). In order to confirm the chemical bonds in the
samples, infrared spectra were also recorded on a Jasco FT-IR-
460 spectrometer in range of 4500–900 cm−1 using the KBr pellet
technique. Using a mass spectrometer (MS), we carried out ther-
mogravimetric (TG) analysis (NETZSCH, TG209F1) to confirm the
thermal behavior of the samples. The TG analyses were carried out
in a temperature range of 25–900 ◦C at a heating rate of 10 ◦C/min
in an air stream, and recorded MS traces of m/z = 18 and 44 for the
evolved water and CO2, respectively.
2. Experimental
2.4. Procedure for glucose isomerization and product analysis
2.1. Material preparation
We conducted the isomerization of glucose to fructose in a
batch-type glass reactor. In a typical reaction, we used 0.3 g of
glucose, 0.1 g of catalyst, and 10 ml of dimethylformamide. The
reaction was performed at 100 ◦C for 5 h. After the reaction was
complete, we cooled the reactor in an ice bath. We then filtered
off the catalysts using a syringe filter and analyzed the filtrate by
high performance liquid chromatography (Young Lin instrument,
YL 9100) using an ultraviolet-visible (UV–vis) detector and a refrac-
tive index (RI) detector. The mobile phase was 0.005 M H2SO4 at
a flow rate of 0.5 mL/min and we separated the reaction products
using a Biorad Aminex HPX87 H column at 50 ◦C. We plotted the cal-
ibration curves using a series of standard solutions to quantify the
glucose and fructose. We determined the glucose conversion and
fructose selectivity using the following equations, and calculated
the fructose yield by multiplying the conversion and selectivity.
All chemicals were used without further purification. We used
two different types of metal oxides as sources of magnesium and
aluminum. We purchased commercial metal oxides (MC: MgO
commercially provided, AC: Al2O3 commercially provided) from
Sigma-Aldrich, and we prepared as-synthesized metal oxides (MS:
MgO synthesized in the laboratory, AS: Al2O3 synthesized in the
laboratory) by calcining corresponding metal nitrates at 450 ◦C for
10 h in air. In a typical procedure, we simultaneously added known
amounts of magnesium oxide (1.6 g) and aluminum oxide (0.68 g)
to 100 ml of distilled water in a wide-mouth bottle. Then, we sealed
the bottle with its cap, and aged the resulting solution containing
the magnesium and aluminum sources while vigorously stirring
at different temperatures (60 and 80 ◦C) and for different periods
of times (6 h, 12 h, 1 d, 3 d, 5 d, and 7 d). Thereafter, we collected
the white precipitate through filtration, and dried it overnight at
80 ◦C in an oven. We denoted the prepared samples as XY Z, where
XY and Z represent the sources of the metal oxides and the aging
time, respectively. For example, MSAC 5 d indicates a sample pre-
pared via hydration for 5 d using MgO that was synthesized in
the laboratory (MS) and commercially provided Al2O3 (AC). For
comparison, we also prepared a Mg–Al hydrotalcite (HT C) with a
ratio of Mg/Al = 3 by a co-precipitation method using the procedure
reported in our previous work [19].
moles of glucose reacted
Conversion of glucose(%) =
Selectivity for fructose(%) =
× 100
× 100
moles of glucose supplied
moles of fructose formed
moles of glucose reacted

G. Lee et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 347–355
349
(b)
AS (Synthesized Al2O3)
MS (Synthesized MgO)
(a)
AS_1 d
MS_1 d
Mg(OH)2
AlO(OH)
β-Al(OH)3
10
20
30
40
50
60
70
10
20
30
40
50
60
70
2 theta (degree)
2 theta (degree)
Fig. 1. XRD patterns of samples prepared via the hydration of a single metal oxide: (a) the MS 1 d and (b) the AS 1 d.
3. Results and discussion
MgO
Mg–Al hydrotalcite
Mg(OH)2
3.1. Formation of hydrotalcites via hydration of metal oxides
First, we attempted to synthesize Mg–Al hydrotalcites (HTs)
using both metal oxides such as MS (MgO synthesized in the labora-
tory) and AS (Al2O3 synthesized in the laboratory), which could be
easily obtained by the calcination of corresponding metal nitrate
precursors in air. Prior to the synthesis of Mg–Al HTs using both
metal oxides, we examined the samples obtained via the hydration
of a single metal oxide (MS or AS). As shown in Fig. 1(a), all the
MgO diffraction peaks (JCPDS #87-0653) completely disappeared
after the hydration process, wherein MS was immersed in water
and was then vigorously stirred at 60 ◦C for 1 d. However, in the
XRD patterns of the hydrated sample (MS 1 d), there were evi-
dent diffraction peaks for Mg(OH)2 (JCPDS #86-0441), which is a
well-known mineral as brucite. Considering that positively charged
brucite-like layers of metal hydroxides form the basis of the Mg–Al
HT structure, these XRD results suggest the high feasibility of a
simple preparation method for Mg–Al HTs from metal oxides. In
the case of AS, we found two crystalline phases such as bayerite
(␤-Al(OH)3, JCPDS #77-0250) and boehmite (AlO(OH), JCPDS #76-
1871) in the XRD pattern of the AS 1 d sample (Fig. 1(b)). This
indicates that amorphous alumina was converted into boehmite
and bayerite during the hydration process in water being vigor-
ously stirred at 60 ◦C for 1 d. These results are in good agreement
with previous reports that alumina could be transformed into alu-
minum hydroxide species via the hydration process and different
aluminum hydroxides such as bayerite (␤-Al(OH)3), gibbsite (␣-
Al(OH)3), and boehmite (AlO(OH)) could be formed depending on
the hydration conditions [21–24].
Fig. 2 shows XRD patterns of the samples prepared by the hydra-
tion of both metal oxides (MS and AS) in water for different times.
For the purposes of comparison, the XRD pattern of the HT C, the
Mg–Al HT prepared by a co-precipitation method, is also shown.
The XRD peaks for MgO (2␪ = 42.8 and 62.2◦) disappeared with an
increase in the hydration reaction time. Interestingly, the char-
acteristic peaks of Mg–Al HT at 2␪ = 11.4 (003), 23.0 (006), and
34.5◦ (009) (JCPDS #41-1428) gradually developed with increas-
ing hydration time. In single metal oxide (MS or AS) hydrations
(Fig. 1), we obtained the corresponding metal hydroxides after the
hydration for 1 d, while we found no noticeable peaks for the metal
hydroxides (Mg(OH)2, ␤-Al(OH)3, and AlO(OH)) in the samples pre-
pared via the hydration of both metal oxides. Instead, it appeared as
if the two metal oxides (MgO and Al2O3) directly transformed into
Mg–Al HT via the hydration reaction (Fig. 2). In the meantime, the
MSAS_1 d
MSAS_3 d
MSAS_5 d
MSAS_7 d
HT_C
70
10
20
30
40
50
60
2 theta (degree)
Fig. 2. XRD patterns of the MSAS X samples with different hydration times (X).
XRD peaks for MgO gradually decreased with increasing hydration
time.
Upon closer examination of the XRD pattern of the MSAS 1
d obtained via the 1 d-hydration reaction of both metal oxides,
we found that the XRD peaks for brucite (Mg(OH)2), bayerite (␤-
Al(OH)3), and Mg–Al HT occurred at the same time (Fig. 3). This
suggests that Mg–Al HT could be prepared from both metal oxides
passing through corresponding hydroxide forms. We can also infer
that aluminum oxide could be incorporated into the hydrotalcite
structure by way of bayerite (␤-Al(OH)3) rather than boehmite
(AlO(OH)). The preference for bayerite can be explained by the pH
value of the solution during the hydration process. As mentioned
earlier, three types of aluminum hydroxides (boehmite, bayerite,
and gibbsite) can be obtained from the hydration of aluminum
oxide. Also, the hydration reaction conditions strongly affect the
formation of aluminum hydroxide, and the pH value of the solution

350
G. Lee et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 347–355
−
hydration of both metal oxides. Finally, the Al(OH)4 might react
with Mg2+ by the following equation to form Mg–Al HT under basic
conditions:
Mg–Al hydrotalcite
Mg(OH)2
MSAS_1d
β-Al(OH)3
+ 8OH− + 2A + xH2O → Mg Al2 OH 162A− · xH2O.
−
−
6Mg2+ + 2Al OH
(
)
(
)
6
4
−
The formation of Al(OH)4 during aluminum oxide hydration
is also supported by ICP-AES analysis, which we conducted using
samples with different hydration reaction times (Table S1). The
Mg/Al ratios in the samples decreased with increasing hydration
times, and finally reached 3.4 after the 1-week hydration reaction.
From the ICP-AES analyses and the XRD and pH value variation
results, we can easily infer that the aluminum component existed in
the form of Al(OH)4− from the initial stage of the hydration of both
metal oxides, leading to high Mg/Al ratios compared to the amount
of metal oxide (Mg/Al = 3) employed in the hydration process. With
10
15
20
25
30
an increasing hydration time, Al(OH)4 and Mg2+ were incorpo-
−
2 theta (degree)
rated into a hydrotalcite structure, and accordingly, the Mg/Al ratio
gradually decreased. However, the final Mg/Al ratio after the 7 d-
hydration reaction did not reach the designed value of 3, which we
attributed to the formation of Mg(OH)2 as a minor phase. Therefore,
both the hydration reaction time and the molar ratio of the metal
oxides used in the hydration reaction must be precisely controlled
when using this method in order to synthesize single-phase Mg–Al
HT.
Fig. 3. XRD pattern of the MASA 1d.
is a particularly important factor in the aluminum hydroxide for-
mation process [21,23,24]. For such a reasons, we examined the pH
values of the solution during the hydration of both metal oxides (MS
and AS) to prepare the MSAS 7 d, and the obtained results were pre-
sented in Fig. S1. From the variation of pH values, it could be inferred
that aluminum oxide first transformed into bayerite, which is easily
formed in the high pH value solutions (pH > 9) of our experimental
conditions, and is subsequently incorporated into the layered struc-
tures that form Mg–Al HT. Consequently, the XRD analysis results
clearly indicated that Mg–Al HT was successfully prepared via the
hydration of both metal oxides (MS and AS) through the formation
of their metal hydroxides (brucite and bayerite, respectively).
The process for magnesium oxide hydration can be explained
by its general aspect and the pH value variation of the solution dur-
ing the hydration process (Fig. S1). During the hydration process,
MgO first reacts with water molecules to form MgOH+ and OH−,
and then, these ions transform into Mg(OH)2 on the surface of the
MgO [25,26]. The Mg(OH)2 then detaches from the MgO surface and
simultaneously dissociates into Mg2+ and OH− in an aqueous solu-
tion. With respect to the pH value variation of the solution during
the hydration process, the initial pH value of the solution was mild
basic (ca. 8.3), and increased with the hydration time. Interestingly,
the pH value increased for 5 d, and thereafter maintained a constant
value (ca. 10.5). We infer from the XRD results (Fig. 2) and the pH
value variation (Fig. S1) that magnesium oxide first changed into
Mg(OH)2, and subsequently, the Mg(OH)2 dissociated into Mg2+
and OH−, which led to the increase in the solution’s pH value during
the hydration process. This process is also in good agreement with
the reported mechanism for magnesium oxide hydration [27–30].
The dissociated Mg2+ was finally incorporated into the layered
structure with the aluminum species to form Mg–Al HT. However,
the dissociation of Mg(OH)2 seems to be thermodynamically lim-
ited at too high a pH value (around 10.5), resulting in the formation
of Mg(OH)2 as a minor phase in the samples obtained after 5 days
of the hydration process. This result also implies that the amount
of metal oxides and the time employed in the hydration process
might be important factors in obtaining single-phase Mg–Al HTs.
Moreover, the increase in the pH value might clearly explain the
reaction pathway for the formation of Mg–Al HT via the hydration
of both metal oxides. As confirmed by XRD analysis, the aluminum
component was incorporated into the Mg–Al HT structure by way
of bayerite (␤-Al(OH)3) under basic conditions [22]. Interestingly,
bayerite retains its amphoteric property, and accordingly, is ionized
into Al(OH)4− in basic solutions [31]. Therefore, we assume that the
bayerite obtained by the hydration of aluminum oxide was directly
ionized with OH−, which is provided sufficiently from−the magne-
sium oxide hydration process, to finally form Al(OH)4 during the
We also noted that a new XRD peak for (006) of Mg–Al HT
appeared after the 5 d-hydration reaction (MSAS 5 d), which might
be explained by the presence of interlayer anions. It is well known
2−
that CO3 anions are thermodynamically most favorable as com-
pensating interlayer anions in the Mg–Al HT structure [32,33]. In
our experiments, however, OH− could be consistently supplied by
the dissociation of Mg(OH)2 at the initial stage of hydration. Accord-
ingly, once positively charged hydroxide layers form, it is assumed
that OH− ions in the solution could be quickly adsorbed on the
surface of these layers to form an anion layer. These results are
in good agreement with previous reports [15]. On the other hand,
the increase in the solution’s pH value during the hydration pro-
cess could result in the enhanced solubility of carbon dioxide in
the atmosphere, leading to abundant CO32− species in the solution
[34,35]. We also expect this environment to facilitate the formation
of an anion layer of CO32−. This means that the distance between
the positively charged hydroxide layers could increase compared
to that of Mg–Al HT with an anion layer of OH−. Evidence for this
situation was provided by the appearance of a shoulder peak for
(003) and a lower angle peak for (006) in the samples obtained after
5 days of the hydration process (MSAS 5 d and MSAS 7 d). Addition-
2−
ally, the interlayer anions could be completely replaced with CO3
by the reconstruction process (Fig. 5).
Based on the above experimental findings, here, we propose a
reaction pathway for the simple preparation method for Mg–Al
HTs via the hydration of both metal oxides (Scheme 1). In sum-
mary, Mg2+ was formed from the dissociation of Mg(OH)2, which
was obtained on the surface of the magnesium oxide via the hydra-
tion of the magnesium oxide. Through this process, the magnesium
oxide consistently served as a magnesium source, supplying Mg2+
for the Mg–Al HT until the pH value of the solution became high
enough (ca. 10.5) to initiate the formation of Mg(OH)2 as a minor
phase. In the meantime, ␤-Al(OH)3, which was formed via alu-
minum oxide hydration in the basic solution, was changed into
Al(OH)4 by abundant the OH− coming−from the dissociation of
−
Mg(OH)2. Finally, the Mg2+ and Al(OH)4 were incorporated into
the Mg–Al HT structure.
3.2. Thermal properties of the prepared hydrotalcite
Next, to confirm the successful formation of Mg–Al HT, we
examined the thermal properties of the MSAS 7 d prepared via

G. Lee et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 347–355
351
MgO
Al2O3
Mg–Al hydrotalcite
MgO
+ H2O
+ H2O
Mg(OH)+ + OH−
Hydration @ basic condition
Deposition
MSAS_7 d
Mg(OH)2@MgO
Dissociation
β-Al(OH)3
Ionization @ basic condition
MSAS_7 d_C
MSAS_7 d_R
Mg2+ + 2OH−
Al(OH)4
−
10
20
30
40
50
60
70
80
90
2 theta (degree)
Fig. 5. Memory effect of the MSAS 7 d: XRD patterns of the MSAS 7 d, the MSAS 7
d C, and the MSAS 7 d R.
Mg-Al hydrotalcite
Scheme 1. Proposed reaction pathway for the simple preparation method for Mg–Al
HTs via the hydration of both metal oxides.
second temperature region, significant dehydroxylation and decar-
bonization occur. In this step, the OH groups that are bonded to
metal species are detached, and evolve into the form of water. The
removal of OH groups in the brucite-like layers results in a loss
of cation charges, which allows carbonate and hydroxide ions to
evolve as CO2 and water molecules, respectively, at the same time.
Until the second step, the structure of the Mg–Al HTs preserves
its layered character while transforming the Mg–Al mixed oxides.
The mixed oxides can be reconstructed into their original structure
by immersing them in water, which is well-known as the mem-
ory effect of hydrotalcite. During heating in the final temperature
region above 750 ◦C, Mg–Al HTs finally change into mixed phases
of MgO and MgAl2O4, and lose their unique memory effect.
the hydration of both metal oxides for 7 d. For comparison, we
also examined the Mg–Al HT sample (HT C) prepared by the
co-precipitation method. Wide catalytic applications of calcined
Mg–Al HTs have led to many investigations of the thermal decom-
position of Mg–Al HTs [2,3]. According to the results reported in
literatures [36,37], thermal decomposition processes, such as dehy-
dration, dehydroxylation, and decarbonization, mainly proceed
during the thermal treatment of HTs. Many researchers also agree
that there are three prominent temperature intervals: room tem-
perature −250 ◦C, 250–750 ◦C, and above 750 ◦C, although opinions
differ regarding the exact temperature and mechanism of thermal
decomposition. Water molecules mainly evolve in the first temper-
ature region from room temperature to 250 ◦C, due to a reversible
loss of the interlayer and surface-adsorbed water molecules. In the
We observed the general aspect of Mg–Al HTs during thermal
treatment in our TG-analysis (Fig. 4). We found a typical weight loss
of Mg–Al HTs with temperature elevation. Also, the H2O and CO2
(a)
100
90
80
70
60
50
40
30
20
(b)
MSAS_7 d
HT_C
MSAS_7 d
HT_C
m/z = 18
Loss of interlayer and
surface-adsorbed water molecules
Dehydroxylation and decarbonization
m/z = 44
Formation of mixed phase of
MgO and MgAl2O4
100
200
300
400
500
600
700
800
900
100
200
300
400
500
600
700
800
900
Temperature (oC)
Temperature (oC)
Fig. 4. Thermogravimetric analysis of the MSAS 7 d and the HT C: (a) TG curves of the MSAS 7 d and the HT C, (b) selective MS profiles (m/z = 18 and 44).

352
G. Lee et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 347–355
molecules evolved within the appropriate temperature regions.
With respect to the water molecules (m/z = 18), we attribute the first
peak to the surface-adsorbed and interlayer water molecules, while
the water molecules formed during the recombination of hydroxyl
groups from the brucite-like layer contributed to the second peak.
Of course, we partially attribute the second peak to the evolution
of the OH− interlayer anions. As expected, we also detected inter-
layer carbonate ions in the form of CO2 within the temperature
region of 300–500 ◦C. Along with the XRD results, therefore, the
TG results strongly support the proposal that Mg–Al HT can be
successfully prepared via the hydration of both metal oxides. As
shown in Fig. 4(a), both the MSAS 7 d and HT C samples showed
similar trends in weight loss with increasing temperature, although
there were differences in the amount of weight loss. This might be
due to the different crystallinity of the two samples (Fig. 2), result-
ing in different amounts of interlayer and surface-adsorbed water
molecules. This result accords with the fact that the first water
molecules evolved a little earlier in the HT C sample due to its rel-
atively less-developed layered structure (Fig. 4(b)). Moreover, we
assume that the difference in the amount of weight loss was par-
tially due to the different interlayer anions of the MSAS 7 d (OH−
and CO32−) and HT C (CO32−).
the Mg–Al HT were slightly shifted to the left, and we found only
a single peak for (006) after the reconstruction process. This sug-
gests that the interlayer anions were mainly composed of CO32− in
the MSAS 7 d R and more interlayer water molecular were inserted
between the layers during the reconstruction process. Ultimately,
this led to an increase in the distance between the positive hydrox-
ide layers in the MSAS 7 d R.
We also note that the characteristic XRD peak for Mg(OH)2 in
the MSAS 7 d completely disappeared during the reconstruction
process. As a result, we could obtain single-phase Mg–Al HT in the
MSAS 7 d R. As mentioned above, both the hydration reaction time
and the molar ratio of the metal oxides employed in the hydration
reaction should be precisely controlled to avoid the formation of
Mg(OH)2 as a minor phase. Alternatively, we also confirmed that
single-phase Mg–Al HT can be easily obtained by the reconstruction
of the sample prepared via the hydration of both metal oxides.
A variation of specific surface areas of the Mg–Al hydrotalcite
samples during the calcination and the reconstruction processes
was also examined. The BET surface areas of MSAS 7 d, MSAS 7
d C, and MSAS 7 d R were found to be 24.6, 65.4, and 37.2 m2/g,
respectively. This trend is in good agreement with the previous
reports [19,20]. Especially, we assume that an increase in the BET
surface area between MSAS 7 d and MSAS 7 d R was attributed
to the facile vertical breaking and exfoliation of hydrotalcite layers
during the reconstruction process, as reported in our previous work
[20].
3.3. Memory effect of the prepared hydrotalcite
As mentioned earlier, memory effect is the most prominent
characteristic of HTs, so it is very important to check whether the
sample prepared via the hydration of both metal oxides retains
its memory effect. To confirm the memory effect of the MSAS 7
d, we conducted a general reconstruction process using a method
reported in our previous work [19]. First, we transformed the
MSAS 7 d into a calcined sample (MSAS 7 d C) by thermal treat-
ment in air at 450 ◦C for 10 h. Next, we obtained the reconstructed
sample (MSAS 7 d R) by immersing the calcined sample in water
and vigorously stirring it under nitrogen atmosphere at 60 ◦C for
24 h. As shown in the XRD patterns of these three samples (Fig. 5),
the MSAS 7 d prepared via the hydration of both metal oxides
retained its unique memory effect, and thus retained the capability
of reverting into its original layered structure through successive
calcination and reconstruction processes. This provides further key
evidence that we successfully synthesized Mg–Al HT in this study,
using this simple method involving the hydration of both metal
oxides.
We also confirmed the layered structures of the samples pre-
pared via the hydration of both metal oxides by FE-SEM analysis
(Fig. 6). For comparison, we present them with an SEM image of
HT C, which is well-known for its layered structure. These SEM
images clearly demonstrate that both the MSAS 7 d and the MSAS 7
d R retained their unique morphologies with characteristically lay-
ered structures. Consequently, all the characterizations performed
in this study lead us to conclude that Mg–Al HT can be successfully
prepared using this simple method of the hydration of both metal
oxides.
To double check the successful formation of the Mg–Al HTs pre-
pared in this study, infrared spectroscopy was carried out over
the MSAS 7 d, the MSAS 7 d R, and the HT C, the obtained spec-
tra were shown in Fig. S2. As expected, all tested samples showed
the characteristic IR peaks for Mg–Al HT. The broad IR peaks at
around 3500 cm−1 indicate the hydroxyl-stretching in brucite-like
layers and hydrogen bonding between water and carbonate anion
in the interlayer. Also, the infrared adsorption peaks at ca. 1635
and 1380 cm−1 correspond to interlayer water molecules (bend-
ing mode) and carbonate anions (v3 mode), respectively. These
characteristic IR peaks of the samples are consistent with previous
literatures [38–40]. Especially, all IR peaks were identically found
in the MSAS 7 d R and the HT C spectra, indicating the successful
In the XRD patterns of the calcined sample (MSAS 7 d C), we
found very broad diffraction peaks only for the MgO. This sug-
gest that both MgO and Al2O3 were completely mixed without any
phase segregation in the MSAS 7 d C, which would make possible
the recovery of the original layered structure by the hydration pro-
cess. Interestingly, the XRD peaks for the (003), (006), and (009) of
Fig. 6. SEM images of (a) the MSAS 7 d, (b) the MSAS 7 d R, and (c) the HT C.

G. Lee et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 347–355
353
[44], the fructose yields obtained by the Mg–Al HTs prepared in this
study seem to be significantly high. Therefore, we expect that the
Mg–Al HTs prepared via the hydration of both metal oxides may act
as efficient base catalysts in various base-catalyzed reactions, espe-
cially those requiring the weak base properties of solid catalysts.
Furthermore, we expect this simple method to be an efficient and
simple way to prepare precursors for calcined HTs, as they exhibit
significant potential in various base-catalyzed reactions.
100
90
80
70
60
50
40
30
20
10
0
Conversion of glucose
Selectivity for fructose
Yield for fructose
For comparison, the catalytic activities of magnesium oxide (MS)
and aluminum oxide (AS) were also examined under the same
reaction conditions for the prepared Mg–Al HTs. Both materials
exhibited very little catalytic performance in this reaction (yield for
fructose: 2.5% (MS) and 1.6% (AS)). Once again, these results sup-
port that the Mg–Al HTs prepared via the hydration of both metal
oxides were highly selective for glucose isomerization reaction.
To check the feasibility of recycling the Mg–Al HTs prepared
via the hydration of both metal oxides, recycle experiments were
conducted using the MSAS 7 d R. In a typical recycle experiment,
the reactor was cooled in an ice bath immediately after the cat-
alytic reaction. Then, the solid catalyst was collected by filtration
and washed with ethanol and water to remove adsorbed organic
impurities. The obtained catalyst was employed in the next run.
Finally, the MSAS 7 d R was reused three times in the same way,
and the obtained results were summarized in Table 1. Even after
the third run, the MSAS 7 d R exhibited a considerable catalytic
activity without significant loss in the fructose yield. This imply
that the Mg–Al HTs prepared in this study could serve as efficient
base catalysts in the isomerization of glucose to fructose.
MSAS_7 d
MSAS_7 d_R
HT_C
Fig. 7. Catalytic activities in the isomerization of glucose to fructose.
Table 1
Catalyst recycling tests with the MSAS 7d R.a
Cycle No.
Conversion of
glucose (%)
Selectivity for
fructose (%)
Yield for fructose
(%)
1
2
3
27.3
25.1
24.7
73.9
72.2
70.5
20.2
18.1
17.4
a
Glucose: 0.3 g, catalyst: 0.1 g, solvent: dimethylformamide (10 ml).
formation of single-phase Mg–Al HT in the MSAS 7 d R. On the other
hand, the MSAS 7 d exhibited an unique peak at 3698 cm−1, which
was attributed to the stretching of the hydroxyl group in brucite
layers (Mg(OH)2) [15,41]. This means that the MSAS 7 d retains a
little amount of Mg(OH)2 as minor phase, which is in good agree-
ment with the XRD analysis (Fig. 2). Once again, IR analysis support
the conclusion that Mg–Al HT can be successfully prepared using
the hydration of both metal oxides.
We also investigated the catalytic stability of MSAS 7 d R by
comparing XRD patterns before and after the catalyst recycling
test. For this purpose, the MSAS 7 d R used was recovered after
the third run in the recycle experiment and the change in structure
of the MSAS 7 d R was examined. As show in Fig. S3, no changes
were found in the XRD patterns of the MSAS 7 d R between before
and after the third run reactions. In other words, the MSAS 7 d R
maintained its initial crystalline phase even after the third run reac-
tions. From the above results, we could confirm that the Mg–Al HTs
prepared in this study were stable during the catalytic reaction.
3.4. Catalytic properties of the prepared hydrotalcites as base
catalysts
3.5. Preparation of hydrotalcites using various metal oxide
sources
We examined the catalytic activity of the samples prepared via
the hydration of both metal oxides using the catalytic isomeriza-
tion of glucose into fructose. Although Mg–Al HTs are well-known
to be efficient and base catalysts, many investigations have focused
on calcined Mg–Al HTs with respect to various base-catalyzed
reactions, such as Knoevenagel condensation and Michael conden-
sation, in consideration of their strong basic properties and high
reaction temperatures [42,43]. In this study, therefore, to investi-
gate the catalytic properties of Mg–Al HTs with layered structures,
we chose glucose isomerization as a model reaction, which can be
performed at low reaction temperatures on heterogeneous cata-
lysts with relatively weak base sites [19,20].
As shown in Fig. 7, the MSAS 7 d and MSAS 7 d R exhibited con-
siderable fructose yields, although the maximum fructose yield was
obtained by the HT C prepared using the co-precipitation method.
We assumed that the enhanced catalytic activity of HT C could be
attributed to its abundant surface base sites from its less-developed
layered structure, which was confirmed by XRD analysis (Fig. 2). It
is also interesting that the fructose selectivity over the MSAS 7 d R
were higher than that over the HT C, indicating that the simple
preparation method via the hydration of both metal oxides might
be better than the typical co-precipitation method in synthesizing
HTs as selective base catalysts. Consequently, the MSAS 7 d and
MSAS 7 d R served as efficient base catalysts in this reaction. In
particular, when considering that the fructose yield is limited by
the thermodynamic equilibrium in this reaction (ca. 42% at 50 ◦C)
We also used commercial metal oxides that are readily supplied
from chemical companies to attempt to synthesize Mg–Al HTs via
the hydration of both metal oxides. We assumed that the use of
commercial metal oxides would make the synthesis route for HT
even easier and simpler. Prior to this attempt, we had investigated
the effects of reaction temperature on hydration, with the goal of
reducing the hydration time for obtaining Mg–Al HTs. For this rea-
son, we conducted the hydration process at an elevated reaction
temperature (60 → 80 ◦C). As shown in Fig. S4, we obtained Mg–Al
HT only after a 3-d hydration time at 80 ◦C, while it took 7 d to
obtain Mg–Al HT via hydration at 60 ◦C. This may be due to the
enhanced reaction rates at 80 ◦C in the hydration of metal oxides
and the formation of Mg–Al HT. Accordingly, to reduce the time
needed to prepare Mg–Al HTs via the hydration of metal oxides, in
this study, we then employed a hydration temperature of 80 ◦C.
We used commercial metal oxides along with synthesized metal
oxides as our metal oxide sources in the hydration processes
for synthesizing Mg–Al HTs, specifically: commercial magnesium
oxide (MC), commercial aluminum oxide (AC), synthesized mag-
nesium oxide (MS), and synthesized aluminum oxide (AS). The
obtained results are summarized in Table 2 and Fig. 8 provides the
corresponding XRD patterns in detail. Mg–Al HTs were successfully
formed as the main phases in all attempts, regardless of the source
of the metal oxide, although the times required for obtaining Mg–Al

354
G. Lee et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 347–355
Mg–Al hydrotalcite
Mg(OH)2
(a) MCAC
(b) MCAS
MgO
MCAC_1 d
MCAC_3 d
MCAS_1 d
MCAS_3 d
MCAC_5 d
MCAC_7 d
MCAS_5 d
MCAS_9 d
MCAS_9 d_R
MCAC_7 d_R
(c) MSAC
(d) MSAS
MSAC_1 d
MSAS_1 d
MSAS_3 d
MSAC_3 d
MSAC_5 d
MSAS_5 d
MSAC_5 d_R
MSAS_5 d_R
10
20
30
40
50
60
70
80 10
20
30
40
50
60
70
80
2 theta (degree)
Fig. 8. XRD patterns of the samples prepared using various metal oxide sources.
HTs differed slightly. We can confidently infer that the differences
in the hydration time may be attributed to the different hydration
reaction rates of the metal oxides. In particular, we found the hydra-
tion reaction rate of commercial magnesium oxide to be relatively
low compared to that of synthesized magnesium oxide, which led
to an increased hydration time for the characteristic XRD for MgO
to completely disappear, thus extending the formation time of the
Mg–Al HTs (MCAC and MCAS samples).
We detected Mg(OH)2 as minor phases in most samples. As
mentioned earlier, the dissociation of Mg(OH)2 can be thermody-
namically limited under basic conditions (pH = ca. 10.5), resulting
in the formation of Mg(OH)2 as a minor phase. Once again, this
result implies that the amount of metal oxides and time employed
in the hydration process might be important factors in obtaining
single-phase Mg–Al HTs. As expected, all the samples prepared via
hydration using various metal oxide sources retained their unique
memory effects and, thus, were capable of reverting to their original
layered structures by reconstruction processes. More importantly,
we could obtain single-phase Mg–Al HTs after the reconstruction
process for all samples prepared via hydration using various metal
oxide sources. Consequently, we conclude that this simple prepa-
ration method for Mg–Al HTs can be successfully carried out using
various metal oxide sources.
4. Conclusions
In this study, we prepared Mg–Al hydrotalcites (HTs) via the
simple hydration of their corresponding metal oxides (MgO and
Al2O3). During the hydration process of magnesium oxide, Mg(OH)2
formed on the surface of the MgO, and subsequently dissociated
into Mg2+ and OH−. Meanwhile, ␤-Al(OH)3 obtained via aluminum
oxide hydration in a basic solution and OH− resulted from the dis-
sociation of Mg(OH)2 created Al(OH)4−. Finally, Mg2+ and Al(OH)4
were incorporated into the Mg–Al HT structure. The prepared
−

G. Lee et al. / Journal of Molecular Catalysis A: Chemical 423 (2016) 347–355
355
Table 2
Synthesis of Mg–Al HTs via the hydration of both metal oxides using various metal oxide sources.
Sample
Precursor
Hydration time (day)
Phasea
Major
Minor
MCAC
1
3
5
7
MgO
MgO
MgO, Mg–Al HT
Mg–Al HT
Mg–Al HT
MgO
Mg–Al HT
Mg–Al HT
Mg(OH)2
Mg(OH)2
–
Mg–Al HT
Mg–Al HT
–
Mg(OH)2
–
MgO, Mg(OH)2
Mg(OH)2
Mg(OH)2
–
Commercial MgO
+
Commercial Al2O3
Reconstruction
1
3
5
MCAS
Commercial MgO
+
Synthesized Al2O3
MgO
MgO, Mg–Al HT
Mg–Al HT
Mg–Al HT
Mg–Al HT
Mg–Al HT
Mg–Al HT
Mg–Al HT
Mg–Al HT
Mg–Al HT
Mg–Al HT
Mg–Al HT
9
Reconstruction
1
3
5
Reconstruction
1
3
5
MSAC
MSAS
Synthesized MgO
+
Commercial Al2O3
MgO, Mg(OH)2
MgO, Mg(OH)2
Mg(OH)2
–
Synthesized MgO
+
Synthesized Al2O3
Reconstruction
a
Formation of phases was confirmed by XRD analysis (Fig. 8).
Mg–Al HTs retained their unique properties of a layered structure
and memory effect, indicating the successful formation of Mg–Al
HTs. Furthermore, Mg–Al HTs prepared by this simple method
served as efficient base catalysts in the isomerization of glucose
to fructose. This method was also used to successfully synthesize
Mg–Al HTs using a variety of metal oxide sources, such as commer-
cial and synthesized oxides. Consequently, the simple hydration of
metal oxides can make the synthesis route for Mg-Al HTs easier and
simpler, and we expect the prepared Mg–Al HTs to apply to various
base-catalyzed reactions.
[12] Z. Yong, A.E. Rodrigues, Energy Convers. Manage. 43 (2002) 1865.
[13] D. Basu, A. Das, K.W. Stöckelhuber, U. Wagenknecht, G. Heinrich, Prog. Polym.
Sci. 39 (2014) 594.
[14] M.R. Othman, Z. Helwani, Martunus, W.J.N. Fernando, Appl. Organomet.
Chem. 23 (2009) 335.
[15] Z.P. Xu, G.Q. Lu, Chem. Mater. 17 (2005) 1055.
[16] G. Ye, T. Troczynski, Ceram. Int. 32 (2006) 257.
[17] K.G. Ahari, J.H. Sharp, W.E. Lee, J. Eur. Ceram. Soc. 22 (2002) 495.
[18] E.S. Martin, US Patent 5 728 364 to Aluminum Company of America (1998).
[19] S. Yu, E. Kim, S. Park, I.K. Song, J.C. Jung, Catal. Commun. 29 (2012) 63.
[20] G. Lee, Y. Jeong, A. Takagaki, J.C. Jung, J. Mol. Catal. A Chem. 393 (2014) 289.
[21] R.A. Pethrick, D. Hayward, K. Jeffrey, S. Affrossman, J. Mater. Sci. 31 (1996)
2623.
[22] G. Lefèvre, M. Duv, P. Lepeut, R. Caplain, M. Fèdoroff, Langmuir 18 (2002) 7530.
[23] X. Carrier, E. Marceau, J.-F. Lambert, M. Che, J. Colloid Interface Sci. 308 (2007)
429.
Acknowledgements
[24] A. Violante, P.M. Huang, Clays Clay Miner. 41 (1993) 590.
[25] J.J. Thomas, S. Musso, I. Prestini, J. Am. Ceram. Soc. 97 (2014) 275.
[26] S.D. Rocha, M.B. Mansur, B.S. Ciminelli, J. Chem. Technol. Biotechnol. 79
(2004) 816.
[27] L.F. Amaral, I.R. Oliveira, R. Salomão, E. Frollini, V.C. Pandolfelli, Ceram. Int. 36
(2010) 1047.
This research was supported by the New & Renewable Energy
Core Technology Program of the Korea Institute of Energy Technol-
ogy Evaluation and Planning (KETEP), granted financial resource
from the Ministry of Trade, Industry & Energy, Republic of Korea
(No. 20143030091040).
[28] O. Fruhwirth, G.W. Herzog, I. Hollerer, A. Rachetti, Surf. Technol. 24 (1985)
301.
[29] G.L. Smithson, N.N. Bakhshi, Can. J. Chem. Eng. 47 (1969) 508.
[30] D.A. Vermilyea, J. Electronchem. Soc. 116 (1969) 1179.
[31] K.H. Gayer, L.C. Thompson, O.T. Zajicek, Can. J. Chem. 36 (1958) 1268.
[32] Y. Ono, H. Hattori, in: A.W. Castleman Jr., J.P. Toennies, K. Yaamanouchi, W.
Zinth (Eds.), Solid Base Catalysis, Tokyo Institute of Technology Press,
Springer, 2001, p. 160.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.molcata.2016.07.
018.
[33] D.G. Costa, A.B. Rocha, W.F. Souza, S.S.X. Chiaro, A.A. Leitão, Appl. Clay Sci. 56
(2012) 16.
[34] T.-J. Kim, H. Vrålstad, M. Sandru, M.-B. Hägg, Energy Procedia 37 (2013) 986.
[35] K. Al. -Anezi, N. Hilal, Desalination 204 (2007) 385.
[36] J. P-Ramíez, S. Abelló, Thermochim. Acta 444 (2006) 75.
[37] V. Vágvölgyi, S.J. Palmer, J. Kristóf, R.L. Frost, E. Horváth, J. Colloid Interface Sci.
318 (2008) 302.
[38] J.T. Kloprogge, R.L. Frost, J. Solid State Chem. 146 (1999) 506.
[39] M.J. Hernandez-Moreno, M.A. Ulibarri, J.L. Rendon, C.J. Serna, Phys. Chem.
Miner. 12 (1985) 34.
[40] F. Rey, V. Fornés, J. Chem. Soc. Faraday Trans. 88 (1992) 2233.
[41] R.L. Frost, J.T. Kloprogge, Spectrochim. Acta A 55 (1999) 2195.
[42] U. Costantino, M. Curini, F. Montanari, M. Nocchetti, O. Rosati, J. Mol. Catal. A
Chem. 195 (2003) 245.
References
[1] Q. Wang, D. O’Hare, Chem. Rev. 112 (2012) 4124.
[2] D.P. Debecker, E.M. Gaigneaux, G. Busca, Chem. Eur. J. 15 (2009) 3920.
[3] T. Stanimirova, I. Vergilov, G. Kirov, N. Petrova, J. Mater. Sci. 34 (1999) 4153.
[4] C. Zha, X. Sun, N. Li, S. Yu, Korean J. Chem. Eng. 31 (2014) 1859.
[5] M. Khitous, Z. Salem, D. Halliche, Korean J. Chem. Eng. 33 (2) (2016) 638.
[6] V. Rives, M.A. Ulibarri, Coord. Chem. Rev. 181 (1999) 61.
[7] K. Parida, J. Das, J. Mol. Catal. A Chem. 151 (2000) 185.
[8] Z.P. Xu, H.C. Zeng, J. Phys. Chem. B 105 (2001) 1743.
[9] A. Tsujimura, M. Uchida, A. Okuwaki, J. Hazard. Mater. 143 (2007) 582.
[10] D.G. Costa, A.B. Rocha, W.F. Souza, S.S.X. Chiaro, A.A. Leitão, J. Phys. Chem. B
115 (2011) 3531.
[43] M. León, E. Díaz, S. Ordón˜ez, Catal. Today 164 (2011) 436.
[44] A.M. Dehkordi, M.S. Tehrany, I. Safari, Ind. Eng. Chem. Res. 48 (2009) 3271.
[11] U. Costantino, M. Nocchetti, M. Sisani, R. Vivani, Z. Kristallogr. 224 (2009) 273.