Cyanoethylation of alcohols by activated Mg-Al layered double hydroxides: Influence of rehydration conditions and Mg/Al molar ratio on Broensted basicity

Article abstract of DOI:10.1016/j.jcat.2011.01.018

Activated Mg-Al layered double hydroxides (LDHs) with varying Mg/Al molar ratios are tested as heterogeneous catalysts for the cyanoethylation of methanol and 2-propanol. The activation procedure is performed by calcination of the LDH precursors, followed by gas-phase rehydration on an H₂O/N ₂ flow. In this case, rehydration is carried out at 80 °C, instead of room temperature as usual. Increased temperature has a very significant effect on the reconstruction speed, shortening the process from ～12 to 0.5 h. Good conversions and yields to alkoxypropionitriles are obtained with thus activated LDHs. Correlations are established between basic strength, as determined by CDCl₃ adsorption followed by FTIR, and catalytic activity. Depending on alcohol acidity, greatest conversions are obtained over catalysts with Mg/Al molar ratios of 2 or 3. Solids are further characterized by X-ray diffraction, thermal gravimetric analyses, and ²⁷Al MAS NMR.

Full text of DOI:10.1016/j.jcat.2011.01.018

Journal of Catalysis 279 (2011) 196–204
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Cyanoethylation of alcohols by activated Mg–Al layered double hydroxides:
Influence of rehydration conditions and Mg/Al molar ratio on Brönsted basicity
a,
⇑
b
b
c
c
Jaime S. Valente , Heriberto Pfeiffer , Enrique Lima , Julia Prince , Jorge Flores
a
Instituto Mexicano del Petroleo, Eje Central 152, Mexico, D.F. 07730, Mexico
Instituto de Investigaciones en Materiales, Universidad Nacional Autonoma de Mexico, Circuito Exterior s/n, Cd. Universitaria, Mexico, D.F. 04510, Mexico
Universidad Autonoma Metropolitana-A, Av. San Pablo 180, Mexico, D.F. 02200, Mexico
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Activated Mg–Al layered double hydroxides (LDHs) with varying Mg/Al molar ratios are tested as heter-
ogeneous catalysts for the cyanoethylation of methanol and 2-propanol. The activation procedure is per-
formed by calcination of the LDH precursors, followed by gas-phase rehydration on an H O/N flow. In
2 2
this case, rehydration is carried out at 80 °C, instead of room temperature as usual. Increased temperature
has a very significant effect on the reconstruction speed, shortening the process from ꢀ12 to 0.5 h. Good
conversions and yields to alkoxypropionitriles are obtained with thus activated LDHs. Correlations are
Received 20 December 2010
Revised 16 January 2011
Accepted 19 January 2011
Available online 22 February 2011
Keywords:
3
established between basic strength, as determined by CDCl adsorption followed by FTIR, and catalytic
Cyanoethylation of alcohols
Layered double hydroxides
Meixnerite-like compounds
Basic catalysts
activity. Depending on alcohol acidity, greatest conversions are obtained over catalysts with Mg/Al molar
ratios of 2 or 3. Solids are further characterized by X-ray diffraction, thermal gravimetric analyses, and
2
7
Al MAS NMR.
Brönsted basic sites
Ó 2011 Elsevier Inc. All rights reserved.
1
. Introduction
Cyanoethylation reactions consist of the addition of compounds
molecules. LDHs are represented by the general formula
2þ
3þ
nꢁ
2+
3+
½
M
M
ðOHÞ ꢂA ꢃ mH
2
O, where M and M are metal cations
ð1ꢁxÞ
x
2
x=n
capable of occupying the octahedral interstices of a brucite-like
containing active hydrogen to the double bond of acrylonitrile.
Alcohols react with acrylonitrile to give alkoxypropionitriles,
which can be further converted to carboxylic acids or amines by
hydrolysis or reduction, respectively. Industrially, important or-
ganic compounds and drug intermediates are obtained in this
manner [1–3]. The reaction is base catalyzed, typically using
homogeneous base catalysts such as alkali hydroxides [1] or an-
ion-exchange resins [1]. Given that heterogeneous catalysts are
environment friendly, easily handled and reusable, some efforts
have been dedicated to studying basic solids suitable for this reac-
tion [2–5]. Various solids have been tested, including alkaline earth
oxides and hydroxides [2], as well as activated [3] and/or rare-
earth-modified [4,5] layered double hydroxides (LDHs).
sheet, i.e. mainly those from the third and fourth periods, while
nꢁ
A
may be virtually any organic or inorganic anion [6–8].
LDHs are also known as hydrotalcite-like compounds by refer-
ence to the naturally occurring Mg–Al–CO hydrotalcite mineral.
3
Upon calcination, it progressively suffers the loss of physisorbed
and interlayer water, decomposition of interlayer anions and dehy-
droxylation of brucite-like sheets. Above ꢀ300 °C, the layered
structure collapses, and at approximately 400 °C, the solid crystal-
3
+
lizes in MgO rock-salt like structure, with Al cations evenly dis-
tributed, i.e. Mg(Al)O. It presents relatively large specific surface
2
ꢁ1
2ꢁ
areas (>200 m g ) and strong Lewis basic sites, O
species.
Therefore, it has found application as basic catalyst in numerous
chemical reactions [9,10].
LDHs are a family of anionic clays that have received much
attention in the past decades, due to their vast applicability in di-
verse fields [6–8]. The LDH structure resembles that of brucite,
Most of the calcination products from the LDH family, including
the Mg(Al)O mixed oxide, have the remarkable capacity of recover-
ing their original layered structure when contacted with anions
and/or water at low temperatures. This property is known as mem-
ory effect [6], and it has been exploited for incorporating interlayer
anions different from the original ones, such as anionic contami-
nants [11]. When the calcined MgAl LDH is exposed to water or
2
+
6 2
where M (OH) octahedra share edges to build infinite M(OH)
sheets. An LDH is created by partial isomorphic substitution of
divalent cations for trivalent ones, such that the layer acquires a
positive charge. This charge is electrically balanced by anionic spe-
cies located in the interlayer region, along with hydration water
ꢁ
water vapor, the LDH structure is recovered, incorporating OH
as charge-balancing anions, provided that care is taken to avoid
CO
2
contamination [12]. The Mg–Al–OH LDH is commonly known
⇑
Corresponding author.
E-mail addresses: jsanchez@imp.mx, sanchezvalente@yahoo.com (J.S. Valente).
ꢁ
as meixnerite, meixnerite-like, or activated LDH. The OH groups
0
021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.01.018

J.S. Valente et al. / Journal of Catalysis 279 (2011) 196–204
197
are capable of acting as Brönsted basic sites in catalytic reactions
12,13]. Activated LDHs have shown high catalytic activity in sev-
eral reactions of industrial interest, such as aldol condensation
kept under vigorous stirring at 80 °C for 18 h. Then, it was washed
repeatedly with hot deionized water and dried at 100 °C for 24 h.
As-synthesized samples were labeled XMgAl, where X stands for
the Mg/Al molar ratio.
[
[
[
12,14], Claisen–Schmidt condensation [15], Knoevenagel reaction
16], and cyanoethylation of alcohols [3,4,17].
The catalyst activation was performed in a two-step process.
Typically, 0.2 g of the as-synthesized MgAl LDH was heat-treated
Given its importance, many efforts have been dedicated to
ꢁ1
ꢁ1
understanding the calcination–reconstruction process of LDHs
14,18–30]. It is generally accepted that microstructural changes
2
in flowing N (100 mL min ), with a heating rate of 10° min ,
[
up to 500 °C, where it was maintained for 5 h. Afterward, the sam-
ple was cooled down to 80 °C, and the rehydration process was car-
take place, i.e. the process is not entirely reversible. However, the
precise nature and extent of these changes, the influence of the
reconstruction procedure and conditions on the structure, and
their combined effect on catalytic activity are not fully understood
yet. Neverthless, it is accepted that the properties and catalytic
activity of meixnerite depend to a large extent on the experimental
procedure and conditions [14,24,25]. Usually, rehydration is car-
ꢁ1
2
ried out using an N flow (60 mL min ) saturated with water, for
0.5, 1, or 3 h. The deionized, decarbonated water contained in the
saturator was heated at 80 °C in order to maintain water partial
pressure constant at 47.4 kPa in the N /water vapor mixture. All
2
in and out stainless steel lines were heated in order to avoid water
condensation. Once the established rehydration time concluded,
ried out using a water-saturated N
2
flow, at room temperature
2
the reconstructed LDH was flushed with a pure N stream
ꢁ1
and for varying time periods. Alternatively, mexinerite-like com-
pounds may be prepared by liquid-phase reconstruction of the cal-
cined LDH or by direct anion-exchange of an as-synthesized LDH.
In this sense, it was recently reported that varying the relative
(100 mL min ) for 0.5 h in order to eliminate excess water. Finally,
the solid was allowed to cool down to room temperature. Activated
catalysts were labeled XMgAl-RY, where X stands for the Mg/Al
molar ratio and Y the rehydration time, in hours (0.5, 1 or 3 h).
2 2
humidity and especially the temperature of the H O/N flow has
a significant effect on reconstruction speed and on the activation
energy of the process [31]. Reconstruction was observed to proceed
faster to completion at higher temperatures. It is to be expected,
then, that shorter rehydration times will be required to reach the
greatest catalytic activity when reconstruction is carried out at
higher temperatures. Furthermore, reconstruction speed was
found to be highly dependent on the Mg/Al molar ratio [32].
Moreover, even though high catalytic activities are obtained
over activated LDHs in many industrially relevant reactions, and
despite the clear advantages of a heterogeneous catalytic system,
the industrial application of activated LDHs is not yet as wide-
spread as one may expect. This is due to several obstacles, such
2.2. Characterization techniques
The chemical composition of solids was determined in a Perkin-
Elmer model Optima 3200 Dual Vision by inductively coupled plas-
ma atomic emission spectrometry (ICP-AES).
a1
X-ray diffraction patterns were recorded using CuK radiation
(k = 1.54 Å) on a Philips X’Pert instrument operating at 45 kV and
40 mA in the 2h range of 4–80°, with a step size of 0.02° and step
scan of 0.4 s.
Thermogravimetric (TG) analyses were carried out on a Netzch
STA-409EP equipment which was operated under a nitrogen flow,
ꢁ1
at a heating rate of 10 °C min from 25 to 1000 °C. In all determi-
as the rapid deactivation upon ambient CO
2
exposure, the difficulty
nations, 100 mg of finely powdered dried LDH sample was used.
The single pulse solid-state Al MAS NMR spectra were ac-
quired under MAS conditions on a Bruker Avance 300 spectrometer,
2
7
in synthesizing LDHs on a large scale, and the lengthy catalyst reac-
tivation procedure [33]. Efforts must be made to surmount these
obstacles. Accordingly, in cyanoethylation reactions, activated
LDHs have been observed to remain active after prolonged contact
with air [3], which would greatly facilitate handling during indus-
trial operation. Furthermore, an environment friendly, easily scal-
able method for LDH synthesis has been developed and was
presented recently [34–36]. Additionally, in this work, it is demon-
strated that the catalyst (re)activation time may be considerably
shortened by performing rehydration at 80 °C instead of room tem-
perature. Hence, this may bring us one step closer to the industrial
application of activated LDHs.
Thus, this paper focuses on studying the cyanoethylation of two
alcohols, methanol and 2-propanol, using meixnerite-like catalysts
with different Mg/Al molar ratios. Rehydration of calcined LDHs
was performed at 80 °C, for varying time periods (0.5–3 h), to
investigate the effect of rehydration conditions on catalytic activ-
ity. A correlation is established between catalytic activities and
the strength of basic sites.
at 7.05 Tesla. A short pulse (
p/12) width of 1 ls was employed to
ensure quantitative reliability of the intensities observed for the
2
7
Al central transition for sites experiencing different quadrupole
couplings. Spinning speed of samples was 10 kHz, and at least
5000 scans were performed. Chemical shifts are expressed as
3
+
2 6
ppm of aqueous complex [Al(H O) ] .
Deuterated chloroform was used as a basicity probe molecule,
followed by diffuse reflectance infrared Fourier-transform spec-
troscopy (DRIFTS), which was carried out in a Bruker Equinox 55
ꢁ1
spectrometer, between 4000 and 400 cm with a resolution of
ꢁ
1
4 cm after 600 scans per spectrum. The equipment was furnished
with a Harrick Praying Mantis chamber allowing in situ treatments
up to 450 °C. Activated catalysts were rapidly transferred into the
chamber, which was previously purged with a He flow to ensure
an inert atmosphere. The samples were allowed 5 min in a He flow
ꢁ1
(5 mL min ), and then the chamber was sealed. Spectra were re-
corded at room temperature and static helium atmosphere, prior
3
to and immediately following injection of 2 lL of CDCl . Then, He
flow was restarted, and the sample was heated up to 450 °C,
recording spectra at 50° intervals, allowing 1 h stabilization at each
temperature.
2
. Experimental
2.1. Catalyst synthesis and activation
2.3. Cyanoethylation catalytic tests
MgAl LDHs with varying Mg/Al molar ratios were prepared by
the coprecipitation at low supersaturation method. An aqueous
solution (1 M) was prepared by dissolving the appropriate
The cyanoethylation reaction procedure was similar to that pre-
viously reported [3]. Typically, 80 mmol of acrylonitrile and 20 mL
of alcohol (methanol or 2-propanol) were added to a three-necked
50-mL round-bottom flask equipped with a reflux condenser and a
thermometer. The solution was magnetically stirred at 50 °C, and
the freshly activated catalyst was rapidly added to the reactor in
amounts of Mg(NO
Separately, a 2 M alkaline solution containing K
3
)
2
ꢃ6H
2
O and Al(NO
3
)
3
ꢃ9H
2
O in distilled water.
CO and KOH
2
3
was prepared. Both solutions were added simultaneously to a glass
reactor, maintaining the pH constant at 9.0. The precipitate was

1
98
J.S. Valente et al. / Journal of Catalysis 279 (2011) 196–204
2
order to minimize exposure to atmospheric CO . Once the reaction
started, aliquots were periodically taken from the reaction mixture,
filtered and analyzed on an HP-6890 GC equipped with a 60-m
HP-5 (5%-phenyl-methylpolysiloxane) capillary column. Conver-
sion was calculated following the decrease in acrylonitrile
concentration.
3
. Results and discussion
3
3
.1. Catalyst characterization
.1.1. As-synthesized LDHs
The chemical composition of the solids and their Mg/Al molar
ratios, displayed in Table 1, are close to the nominal values, within
experimental error.
X-ray diffraction patterns of as-synthesized LDHs, presented in
Fig. 1, reveal a pure LDH phase. XRD patterns are characterized by
strong, sharp peaks at low 2h values, belonging to basal spacing
Fig. 1. X-ray diffraction patterns of as-synthesized LDHs.
0
0 l reflections, and smaller in-plane diffraction peaks at high an-
gles. Cell parameters c and a of the triple hexagonal unit cell, cal-
culated assuming a 3R polytype, are presented in Table 1.The
c = 3d003 parameter is indicative of interlayer distance, and the
a = 2d110 parameter is related to the average cation–cation distance
in the brucite-like sheets.
Weight percentage losses, as measured by TGA, correspond to
approximately 19 and 22 wt.% for the first and second stages,
respectively.
Only small deviations in weight percentage losses with varying
Mg/Al molar ratios were detected on the TGA profiles (data not
shown). The samples with a higher Mg/Al molar ratio, 3.5MgAl
and 4MgAl, contain larger amounts of hydration water molecules,
evidenced by a greater weight loss in the first thermal event (25–
1
Both unit cell parameters clearly increase along with the Mg/Al
molar ratio. The increase in interlayer distance is due to the decre-
ment in layer charge density. With lower Mg/Al molar ratio, i.e. lar-
3
+
ger Al substitution degree x, there is a stronger electrostatic
attraction between positively charged sheets and interlayer anions.
Therefore, the layers are packed closer together in LDHs with lower
Mg/Al ratio. Furthermore, interlayer distances are well in agree-
2
50 °C). This phenomenon could be attributed to the lower amount
of interlayer anions present in samples with greater Mg/Al molar
ratio, which leaves more room for the intercalation of water mole-
cules. Contrarily, the amount of weight lost during the second ther-
mal event increases with decreasing Mg/Al ratio, due to the larger
amount of interlayer anions.
2
3
ꢁ
ment with CO intercalated LDHs in all cases.
Correspondingly, the a cell parameter is known to obey
Vegard’s law for solid solutions. The average cation–cation dis-
Furthermore, as may be observed in DTG profiles in Fig. 2, the
temperature at which the first thermal event occurs is dependent
on the Mg/Al ratio. This peak becomes broader and shifts to lower
temperatures with increasing molar ratio, due to the larger amount
of water, in agreement with previously discussed TGA profiles.
Moreover, it is worth noting that the interlayer region of LDHs is
made up of an intricate system of hydrogen bonds, between layer
OH groups, interlayer anions, and water molecules [8]. Then, the
shift to lower temperatures could be due to the lower layer charge
density and fewer interlayer anions, as this would reduce the
strength of the hydrogen bonding network and facilitate water
departure. Also, the larger interlayer distance in samples with
higher Mg/Al molar ratio (Table 1) implies less diffusional barriers
for departing molecules. On the other hand, the temperature of the
second thermal event, at which interlayer anions are lost, shows no
clear dependence on the Mg/Al molar ratio.
2
+
tance decreases with the isomorphous substitution of Mg by
3
+
Al (lower Mg/Al molar ratio), given that their Shannon octahedral
coordination ionic radii are 0.72 Å and 0.535 Å, respectively [37].
Therefore, the observed increase in a values corroborates the in-
tended variation of Mg/Al molar ratios. Additionally, a values agree
well with those reported for MgAl samples with similar chemical
compositions [38], confirming the complete incorporation of the
metal cations to the brucite-like sheets.
3.1.2. Thermal decomposition of as-synthesized LDHs
3
The thermal decomposition process of MgAl–CO LDHs is well
documented [6–8,39]. Initially, between room temperature and
50 °C, physisorbed water molecules are lost, followed by inter-
layer water molecules, possibly along with small amounts of
weakly bound OH groups and physisorbed CO . From 250 °C and
up to 450 °C, CO is evolved from interlayer carbonate, and water
2
2
2
3
3
.1.3. Activated LDHs
.1.3.1. Structural analyses. It is well known that upon calcination,
and OHs from the dehydroxylation of the brucite-like sheets.
the layered structure collapses and the solid then crystallizes in
MgO rock-salt like structure [6–8,30]. Accordingly, XRD patterns
of calcined samples revealed pure MgO phase (data not shown).
When the calcined samples are placed in a decarbonated water-
Table 1
Unit cell parameters and chemical compositions of as-synthesized MgAl LDHs with
varying molar ratios.
saturated N
2
flow, they recover their original layered structure,
Sample
Unit cell
parameters (Å)
General formula
Mg/
Al
molar
ratio
ꢁ
incorporating OH groups as charge-balancing anions. The ob-
tained Mg–Al–OH LDH is commonly known as activated LDH, mei-
xnerite, or meixnerite-like. The reconstruction process is
traditionally carried out at room temperature and lasts an average
of 12 h. However, it was recently reported that process conditions,
such as relative humidity and temperature, have a major influence
in the process’ activation energies and the overall reconstruction
speed [31,32].
c
a
2
2
3
3
4
MgAl
22.804 3.046 ½Mg0:675Al0:324ðOHÞ ꢂðCO
3
3
3
3
3
Þ
Þ
Þ
Þ
Þ
ꢃ 0:77H
ꢃ 0:80H
ꢃ 0:80H
ꢃ 0:98H
ꢃ 1:05H
2
2
2
2
2
O
O
O
O
O
2.08
2.75
2.93
3.40
3.78
2
0:162
0:133
0:127
0:114
0:105
.5MgAl 23.173 3.054 ½Mg :733Al0:267ðOHÞ ꢂðCO
0
2
MgAl
23.367 3.059 ½Mg :745Al0:255ðOHÞ ꢂðCO
0
2
.5MgAl 23.707 3.070 ½Mg0:773Al0:227ðOHÞ ꢂðCO
2
MgAl
23.802 3.072 ½Mg0:790Al0:210ðOHÞ ꢂðCO
2

J.S. Valente et al. / Journal of Catalysis 279 (2011) 196–204
199
Fig. 3. X-ray diffraction patterns of activated LDH catalysts: (a) 2MgAl-R0.5, (b)
2
4
MgAl-R1, (c) 2.5MgAl-R0.5, (d) 3MgAl-R0.5, (e) 3MgAl-R1, (f) 3.5MgAl-R0.5, (g)
MgAl-R0.5, (h) 4MgAl-R1.
Fig. 2. Differential thermal gravimetric profiles of indicated samples.
IV
of Al species is expected to migrate to tetrahedral sites (Al ). Then,
during reconstruction, if sufficient time is given, all Al atoms re-
cover their initial octahedral coordination [14]. Therefore, the
In this case, samples were activated at 100% relative humidity
and 80 °C, for 0.5, 1, or 3 h. In a very short time, only 0.5 h, the lay-
ered structure is almost entirely recovered, as may be appreciated
in the XRD patterns of Fig. 3. The reconstruction rate is dependent
on the Mg/Al molar ratio, in agreement with a previous report [32].
Sample 4MgAl-R0.5 shows the slowest reconstruction rate. Its
most prominent diffraction peak, located at ꢀ43° 2h, belongs to
MgO. Also, it is interesting to notice that for a given chemical com-
position, there are no major changes in XRD patterns when the
rehydration time is increased from 0.5 to 1 h (Fig. 3).
27
reconstruction degree may also be estimated by means of Al
MAS NMR spectroscopy. The spectra presented in Fig. 4 show a
very intense NMR signal close to 0 ppm, indicating Al species in
octahedral coordination (Al ) on brucite-like layers. A second,
smaller resonance signal close to 70 ppm is also appreciated,
denoting a small fraction of Al in tetrahedral coordination (Al ).
All MgAl LDHs present Al , after 0.5 and 1 h reconstruction, indi-
cating that the LDH structure has not been fully recovered. The rel-
ative percentages of Al and Al are presented in Table 2.
The amount of tetrahedral aluminum decreases when the rehy-
dration time is increased from 0.5 to 1 h, for a given chemical com-
position. In addition, if samples with the same rehydration time are
compared (0.5 h), a clear dependency on Mg/Al molar ratio is
observed. The rapidity with which aluminum regains octahe-
dral coordination is 3MgAl-R0.5 > 3.5MgAl-R0.5 > 2MgAl-R0.5 ꢄ
VI
IV
IV
IV
VI
Cell parameters of reconstructed samples are presented in Table
. The c cell parameter is slightly decreased, compared to the as-
2
synthesized samples. Interlayer distances coincide with those re-
ꢁ
ported for OH intercalated LDHs [6]. The low c value is related
ꢁ
to the similarity of OH and H
2
O ionic diameters, and to the strong
ꢁ
hydrogen bridges among water and OH , leading to the best close-
packed arrangement [6].
2
.5MgAl-R0.5 ꢅ 4MgAl-R0.5.
Furthermore, comparing the a cell parameter of as-synthesized
and reconstructed LDHs with a given molar ratio, one notices a
small decrease in all cases. As mentioned previously, the a param-
eter is related to the Mg/Al molar ratio in the brucite-like sheets.
Then, this decrease is associated with a higher relative amount of
aluminum in the layers. It may be interpreted as an incomplete
Total reconstruction of the layered structure may not be re-
quired for catalytic purposes. Indeed, the most favorable recon-
struction time varies from one catalytic reaction to the other.
This may be related to the strength of the basic sites created at dif-
ferent rehydration stages.
2
+
incorporation of Mg to the LDH phase. Accordingly, an MgO
phase is observed in most of these samples’ XRD patterns and is
probably still present even when MgO is no longer detected by
3
.1.3.3. Basicity measurements. The quantification of the basic
strength of Brönsted sites in reconstructed LDHs may be performed
by adsorption of suitable probe molecules followed by FTIR spec-
troscopy [40,41]. Deuterated chloroform may be used as IR probe
of basic surface sites, since it is a weak acid that should not suffer
important chemical transformations at the tested conditions, other
than the acid–base associative interactions [42,43]. It is a relatively
small molecule capable of accessing the basic sites that may partic-
3
+
XRD. However, it is interesting to remark that apparently, Al re-
turns to the LDH phase faster than Mg2 . Moreover, comparing sets
of samples with the same chemical composition and different
rehydration times (e.g. 2MgAl-R0.5 and 2MgAl-R1), it is clear that
+
2
+
Mg continues to incorporate itself to the brucite-like layers, given
the increase in a parameter values (Table 2).
ipate in the cyanoethylation reaction. The main
mCD stretching
ꢁ1
vibration band, which appears at 2264 cm on the gas phase, does
not overlap the bands of surface OH groups, enabling straightfor-
3
has the ability of establishing hydrogen
bond interactions with basic sites, which are manifested by a
27
3
.1.3.2. Al MAS NMR. Typically, an as-synthesized MgAl LDH pre-
VI
sents only aluminum species in octahedral coordination (Al ). In
the course of thermal treatment, a fraction (approximately 20%)
ward identification. CDCl

2
00
J.S. Valente et al. / Journal of Catalysis 279 (2011) 196–204
Table 2
Unit cell parameters, relative populations of tetrahedral (AlIV) and octahedral (AlVI
aluminum species from deconvoluted Al MAS NMR spectra, and estimated pKa
)
2
7
values of indicated catalysts.
Sample
Unit cell
parameters (Å) population
%)
Relative
Estimated pKa of basic sitesa
(
c
a
AlIV
AlVI
2
2
2
3
3
3
4
4
MgAl-R0.5
MgAl-R1
.5MgAl-R0.5 22.751 3.049
MgAl-R0.5
MgAl-R1
.5MgAl-R0.5 22.534 3.052
22.278 3.035
22.528 3.042
9
4
9
5
4
6
91
96
91
95
96
94
83
87
–
+14.6
+15.3, +19.2
+15.5, +19.5
+10.6, +15.5, +19.5
+11.6
–
–
22.703 3.047
22.788 3.052
MgAl-R0.5
MgAl-R1
22.703 3.051 17
22.541 3.053 13
a
Estimated by the
3
Dm of the C–D vibration band of adsorbed CDCl [42,43].
red-shifting of the
the basic strength; a correlation between
of the basic centers was established by Paukshtis et al. [42,43].
Representative spectra for sample 2.5MgAl-R0.5 in the CD re-
gion are shown in Fig. 5. A reference spectrum was first recorded
and then the spectrum after adsorption of 2 L CDCl . As may be
observed, upon CDCl adsorption, there are two appreciable differ-
ences in the spectra. On one hand, there appears a small, broad,
m
CD band. The magnitude of the shift is related to
Fig. 5. FTIR spectra of catalyst 2.5MgAl-R0.5, taken prior (A) and subsequent (—) to
CDCl adsorption, and then at indicated temperatures during thermal decomposi-
tion of CDCl -adsorbed 2.5MgAl-R0.5.
D
m
CD and the pKa values
3
3
m
ꢁ1
l
3
vibrational frequency is 2416 cm [44]. In this case, because of in-
creased relaxation of the molecule, the band is blue-shifted,
3
ꢁ1
approximately 100 cm . The appearance of an O–D vibration im-
ꢁ1
ꢁ
two-component band with maxima at 2173 and 2119 cm , which
is ascribed to the CD stretching band. As expected, the band has
plies that a proton-exchange reaction, between an OH basic site
m
and deuterium, has taken place. This provides further evidence of
the strong interaction that occurs between chloroform and the
Brönsted sites. A similar phenomenon was observed previously
with LDH-derived MgGa mixed oxides with similar basic strengths
[45].
shifted to lower frequencies due to hydrogen bond interactions
with basic sites. The two maxima must then correspond to basic
sites of different strength. According to the correlations established
ꢁ
1
by Paukshtis et al., DmCD of 91 and 145 cm correspond to proton
ꢁ1
affinities of ca. 965 and 1000 kJ mol , and the estimated pKa val-
Once the adsorption spectrum was recorded, the sample was set
ꢁ1
ues are +15 and +19, respectively [42,43].
in a He flow (5 mL min ) and heat-treated in situ, to investigate
The other spectral modification is seen in the OH region. Specif-
thermal behavior. As soon as the He flow was started, the mCD
ꢁ1
ꢁ1
ically, a shoulder appears between ꢀ2600 and 2400 cm . This
band is ascribed to an O–D stretching vibration. In solid HDO,
where all molecules are tetrahedrally hydrogen-bonded, the O–D
vibration bands close to 2200 cm
disappeared, in agreement
with previous observations [24,45].
ꢁ
1
In LDHs, the broad OH vibration band in the 3800–2400 cm
region is in fact the result of two or even three overlapping OH
stretching vibrations, plus a vibration due to interlayer water [7].
Upon heating, physisorbed water molecules, interlayer water,
hydroxyls from brucite-like layers, and charge-compensating
ꢁ
OH groups are progressively lost. Remarkably, the O–D vibration,
ꢁ1
centered at ꢀ2550 cm , persists at temperatures as high as
2
50 °C. Therefore, it is reasonable to assume that the proton-ex-
ꢁ
change reaction took place with interlayer OH anions, which are
lost approximately at this temperature. These anions are likely
the strongest Brönsted basic sites of the catalyst.
The interactions between the catalysts and CDCl
straightforwardly observed by subtracting the spectra acquired
prior and subsequent to CDCl adsorption. Fig. 6 presents differen-
tial spectra of CDCl adsorbed on several catalysts. Also, the pKa
values of the Brönsted basic sites, estimated by the magnitude of
the red shift of the CD stretching band [42,43], are summarized
3
may be
3
3
m
in Table 2. Spectra of all analyzed catalysts, in the
mCD region,
may be found in the Supplementary data.
The previously discussed observations regarding CDCl
3
adsorp-
tion on sample 2.5MgAl-R0.5 are unmistakably observed in the dif-
ꢁ
1
ferential spectrum, i.e. the
mOD stretching vibration at 2551 cm
ꢁ1
and the shifted CD bands at 2173 and 2119 cm (Fig. 6). The other
m
catalysts present somewhat similar behaviors. In 2MgAl-R0.5,
there is a single, strong, narrow band in the differential spectra,
ꢁ1
centered at 2530 cm , assigned to the stretching O–D vibration.
No other bands that could be attributed to a shifted CD band are
detected. This may be understood as a very strong interaction of
CDCl with the basic sites of this sample, so that they completely
m
Fig. 4. 27Al MAS NMR spectra of indicated catalysts.
3

J.S. Valente et al. / Journal of Catalysis 279 (2011) 196–204
201
ꢁ
1
undergo proton-exchange reactions. Thus, it is inferred that
MgAl-R0.5 has the strongest Brönsted basic sites of the analyzed
sity (Supplementary data). The 2213 cm
band is ascribed to
2
newly created, weaker Brönsted basic sites, with estimated pKa va-
lue of +10.6. Hence, it may be concluded that prolonging rehydra-
tion time could engender new, weaker basic sites. The sites that are
first hydrated are the most reactive and thus the ones with stron-
gest basicity. Increasing water amounts will progressively generate
more basic sites, until all available sites are hydrated. From this
point on, further rehydration will very likely poison basic sites.
catalysts. It is worth mentioning that the thermal behavior of this
O–D band is very similar to that previously described for 2.5MgAl-
R0.5 (see Supplementary data).
In sample 3MgAl-R0.5, two peaks at 2171 and 2114 cmꢁ1, as-
signed to the shifted
mCD band, are appreciated. These bands also
correspond to basic sites with approximate pKa values of +15
ꢁ1
and +19 (Table 2). There is also a weak band around 2534 cm
,
which is ascribed to a proton-exchange reaction that occurs only
to a very limited extent. This may indicate that sample 3MgAl-
R0.5 has fewer strong basic sites than the catalysts with lower
3.2. Cyanoethylation of methanol and 2-propanol
3.2.1. Reaction mechanism
Mg/Al molar ratio or that these sites are less accessible to CDCl
For sample 3.5MgAl-R0.5, the interaction with CDCl was very
weak. A very small band, attributed to the CD vibration, appeared
, the estimated pKa of
these basic sites is +11.6. Samples 4MgAl-R0.5 and 4MgAl-R1,
apparently have no interaction with CDCl (see Supplementary
data), as no spectral differences were detected before and after
CDCl adsorption. This may be attributed to a lower basicity of
3
.
The cyanoethylation of alcohols reaction is thought to proceed
in a 3-step process. The mechanism presented in Scheme 1 was de-
rived for a homogeneous system and adapted by analogy to heter-
ogeneous catalysts [2]. In step 1, the Brönsted basic site abstracts a
proton from the hydroxyl group of the alcohol, producing an ad-
sorbed alkoxide anion. This anion is stabilized on a surface acid
site. Given that both Brönsted and Lewis acid sites are present in
these catalysts, the acidic site is depicted by A. Afterward, in step
2, the adsorbed alkoxide anion reacts with the double bond of acry-
lonitrile to form an adsorbed 3-alkoxypropanenitrile anion. This
3
m
ꢁ1
at 2205 cm . Based on the magnitude of Dm
3
3
these samples. Accordingly, it has been reported that higher Mg/
Al molar ratios induce higher acidity and lower basicity in both cal-
cined [46] and activated [13] MgAl LDHs. In activated LDHs, sam-
ples with larger Mg/Al molar ratio (lower layer charge density)
+
anion takes an H from the basic site in step 3, to yield 3-alkoxy-
propanenitrile and conclude the catalytic cycle.
ꢁ
have fewer charge-balancing OH anions and therefore less Brön-
Both step 1 and step 2 may be the rate-limiting reaction steps. If
the catalysts have relatively weak basic sites, the slowest step will
be the proton abstraction in step 1. In this case, reaction rate is
determined by the acidity of the alcohol, which determines the
sted basic sites. Furthermore, bonding between octahedral layers
and interlayers involves a combination of electrostatic effects and
hydrogen bonding. Hydroxyl groups bonded to trivalent cations
are strongly polarized and interact with interlayer anions [8]. Thus,
layer OH groups in solids with lower layer charge density are less
polarized, and the negative charge density carried by interlayer an-
+
ease of abstraction of H . On the other hand, on catalysts with
strong basic sites, proton abstraction should not represent an
obstacle. Reaction rate would then be determined by the stability
of the alkoxide anion in step 2.
ꢁ
ions is also lowered. Then, the basic strength of the interlayer OH
Brönsted sites is decreased when the Mg/Al ratio rises.
The effect of increasing rehydration time on basic strength is
observed when the differential spectra of samples 2MgAl-R0.5
3.2.2. Influence of Mg/Al molar ratio
Cyanoethylation of methanol and 2-propanol was carried out to
determine the catalytic activity of the activated LDHs with varying
Mg/Al molar ratios. Selectivities of 100% to the corresponding
alkoxypropionitrile were obtained with both alcohols in all cases.
Initial reaction rates, as well as yields to the corresponding alkoxy-
propionitrile, are presented in Table 3.
Fig. 7 presents results from the cyanoethylation of methanol by
several activated LDHs, all with 0.5 h reconstruction time. As may
be observed, the highest activity was displayed by catalyst 3MgAl-
R0.5, closely followed by 2.5MgAl-R0.5, then 2MgAl-R0.5. Catalyst
and 2MgAl-R1 are compared (Fig. 6). In the latter, the mOD band is
ꢁ
1
broader, and a small peak is detected at 2178 cm , which corre-
sponds to basic sites with pKa of approximately +15. For catalyst
3
MgAl, when rehydration time was increased to 1 h, a small band
ꢁ1
ꢁ1
appeared at 2213 cm . The bands at 2171 and 2114 cm ob-
served on 3MgAl-R0.5 are still present, albeit with reduced inten-
3
.5MgAl-R0.5 presents considerably less activity, while 4MgAl-
R0.5 was inactive (not shown). According to the reaction mecha-
nism, for methanol, being the most acidic aliphatic alcohol, proton
abstraction is most likely not an obstacle. Step 2, related to the
stability of the methoxide anion (Scheme 1), should be the
rate-limiting step. Catalytic activity decreases as 3MgAl-
R0.5 > 2.5MgAl-R0.5 > 2MgAl-R0.5. Thus, among the strongest ba-
sic catalysts, methoxide reactivity is inversely proportional to basic
strength, as determined by CDCl
tial reaction rates are 26, 23, 13, and 2 ꢆ 10 mol g min for
MgAl-R0.5, 2.5MgAl-R0.5, 2MgAl-R0.5, and 3.5MgAl-R0.5, respec-
3
adsorption (Section 3.1.3.3). Ini-
ꢁ3
ꢁ1
ꢁ1
3
tively (Table 3). It is further interesting to notice that conversions
continue to increase with reaction time, particularly on the least
active catalysts. At 60 min reaction, yields are above 90% for the
three most active catalysts (Table 3); meanwhile, 3.5MgAl-R0.5 at-
tains 92% conversion at 85 min (Fig. 7).
On the other hand, when these catalysts were tested for the
cyanoethylation of 2-propanol, a nearly inverse activity trend
was observed (Fig. 8). The most active catalyst was 2MgAl-R0.5,
followed by 2.5MgAl-R0.5, then 3MgAl-R0.5, 3.5MgAl-R0.5, and
4
MgAl-R0.5. The observed trend is well in agreement with the ba-
Fig. 6. Differential FTIR spectra of CDCl
3
adsorbed on indicated catalysts.
sic strength determined by CDCl adsorption (Section 3.1.3.3) and
3

2
02
J.S. Valente et al. / Journal of Catalysis 279 (2011) 196–204
Fig. 7. Cyanoethylation of methanol over catalysts with varying Mg/Al molar ratios
and 0.5 h rehydration time.
3
MgAl, rehydrated for different times: 0.5, 1, and 3 h. It was ob-
served that increasing rehydration time from 0.5 to 1 h had a min-
or beneficial effect on catalytic activity. Meanwhile, if the
rehydration step is prolonged for 3 h, there is a somewhat detri-
mental effect on the initial reaction rate. In all cases, conversions
above 95% are reached in less than 1 h (Fig. 9); nonetheless, initial
reaction rates are 26, 33, and 24 ꢆ 10 mol g min (Table 3).
Since no significant differences in yields are apparent, rehydration
of 0.5 h is considered the optimum.
In the case of 2-propanol cyanoethylation over 3MgAl, an in-
creased rehydration time had a more significant effect. As may
be observed in Fig. 9, acrylonitrile conversions at 120 min reaction
time are 74% and 58% over catalysts rehydrated for 1 and 0.5 h,
respectively. Corresponding yields at 60 min are 63.4% and 48.2%
Scheme 1. Proposed reaction mechanism for the cyanoethylation of alcohols over
rehydrated MgAl LDHs.
with the reaction mechanism (Section 3.2.1) discussed previously.
In the case of 2-propanol, the rate-limiting step likely is proton
abstraction in Step 1 (Scheme 1), since it is a weaker acid than
methanol. Then, stronger basic sites are required, and the reaction
proceeds faster on 2MgAl-R0.5 and decreases with decreasing basic
strength and decreasing Mg/Al molar ratio.
It is worth noting that 3.5MgAl-R0.5 and 4MgAl-R0.5 are the
least active catalysts in the cyanoethylation of both alcohols. This
3
is in agreement with observations from CDCl adsorption followed
ꢁ3
ꢁ1
ꢁ1
by FTIR, which indicate that samples with highest Mg/Al molar ra-
tio have the weakest basic sites. Furthermore, these solids require
less charge-balancing anions and therefore have fewer Brönsted
basic sites. In all cases, greater conversions were obtained with
methanol than with 2-propanol, with the exception of 4MgAl-
R0.5, which is almost not reactive in either case.
(
Table 3). It is worth noting that with methanol, the maximum
conversion is reached very fast, in ꢀ25 min. In contrast, with 2-
propanol, conversion continually increases with time (Fig. 9). This
observation further supports the hypothesis that there is a differ-
ent rate-limiting step of the reaction mechanism (Scheme 1) for
each of the studied alcohols.
Similarly, cyanoethylation of 2-propanol was carried out over
catalysts with Mg/Al ratio of 2 and varying rehydration times. As
observed in Fig. 10, conversion is slightly higher over 2MgAl-R1
than over 2MgAl-R0.5, while it is significantly lower over 2MgAl-
R3. Therefore, basic strength diminishes when the catalyst is rehy-
drated for 3 h, due to the poisoning effect of excess water, as dis-
cussed previously (Section 3.1.3.3). Yields at 60 min are 65%, 69%,
and 50% for 2MgAl-R05, 2MgAl-R1 and 2MgAl-R3. Initial reaction
3
.2.3. Influence of rehydration time
The effect of rehydration time was evaluated on several cata-
lysts. Cyanoethylation of methanol was carried out over catalyst
Table 3
Initial reaction rates and yields obtained for the cyanoethylation of methanol and 2-
propanol over indicated catalysts.
rates reveal
a
similar trend, increasing from 12 to
ꢁ3
ꢁ1
ꢁ1
Catalyst
Initial reaction rate
Yield to
alkoxypropionitrile (%)
13 ꢆ 10 mol g min for catalysts rehydrated for 0.5 and 1 h,
ꢁ
1
ꢁ1
3
a
ꢁ3
ꢁ1
ꢁ1
(mol g min ꢆ 10 )
and then decreasing to 8 ꢆ 10 mol g min for catalyst rehy-
drated for 3 h.
Methanol
2-Propanol
Methanol
2-Propanol
On the other hand, increasing rehydration time up to 3 h on the
less active catalysts, 3.5MgAl and 4MgAl, had a marked positive ef-
fect on conversion. As an example, Fig. 11 presents results of cya-
noethylation of methanol over 3.5MgAl-R0.5 and 3.5MgAl-R3. On
catalyst with 0.5 h rehydration, conversion continues to increase
with time, reaching 92% after 85 min. In a shorter time, 50 min,
3.5MgAl-R3 has reached 90%, and at 90 min, conversion is nearly
99%. It is also interesting to notice that these two catalysts show
an activation stage of about 5 min, during which time there is no
catalytic activity. Likewise, among the catalysts with Mg/Al ratio
of 4, the only one that displayed slight activity for cyanoethylation
2
2
2
2
3
3
3
3
3
4
4
MgAl-R0.5
MgAl-R1
MgAl-R3
.5MgAl-R0.5 23
MgAl-R0.5
MgAl-R1
13
n.d.
n.d.
12
13
8
6
6
7
n.d.
2
n.d.
2
n.d.
91
65
69
50
58
48
63
n.d.
41
n.d.
5
n.d.
n.d.
100
97
100
100
75
95
0
5
26
33
24
2
7
0
MgAl-R3
.5MgAl-R0.5
.5MgAl-R3
MgAl-R0.5
MgAl-R3
0
n.d.
a
Calculated at 60 min reaction time.

J.S. Valente et al. / Journal of Catalysis 279 (2011) 196–204
203
Fig. 10. Cyanoethylation of 2-propanol over catalyst 2MgAl with varying rehydra-
tion times.
Fig. 8. Cyanoethylation of 2-propanol over catalysts with varying Mg/Al molar
ratios and 0.5 h rehydration time.
of methanol was the one rehydrated for 3 h. Furthermore, the acti-
vation period for this catalyst was considerably longer, of ꢀ30 min
(Fig. 11).
Cyanoethylation activities displayed by the LDHs rehydrated
under H O/N flow at 80 °C are superior to many of the solids
2
2
studied thus far, and the catalysts present several advantages.
For instance, conversions above 80% are obtained for the cyanoeth-
ylation of 2-propanol over CaO and La O , but the catalysts are
2 3
nearly inactive for the cyanoethylation of methanol [2]. Similar
phenomena are observed with other basic solids, such as
Ba(OH)
2
ꢃ8H
2
O and Sr(OH)
2
ꢃ8H
2
O, which are highly active for cya-
noethylation of ethanol and unreactive toward 2-propanol [2].
BaO shows conversions above 60% for methanol, ethanol and 2-
propanol, but requires activation pretreatment at very high tem-
perature (1000 °C) [2]. With rare-earth modified LDHs, for the cya-
noethylation of ethanol, conversions of ꢀ40% are obtained, with
ꢀ
90% selectivity to b-ethoxypropionitrile [4]. Conversely, a MgAl
LDH activated at room temperature was found to be highly active
for the cyanoethylation of several alcohols [3]. Nonetheless, the
Fig. 11. Cyanoethylation of methanol over indicated catalysts.
activation procedure is too time-consuming for industrial applica-
tions, an obstacle that may be surmounted by following the proce-
dure here presented.
4
. Conclusions
MgAl LDHs with varying Mg/Al molar ratios were calcined and
rehydrated to obtain highly active basic catalysts. Rehydration was
carried out on the gas phase, using an H O/N flow at 80 °C. In this
2
2
way, the process is shortened from the average 12 h at room tem-
perature to only 0.5 h. Solids thus obtained presented a regener-
ꢁ
ated layered structure, with OH groups acting as interlayer
charge-balancing anions and as Brönsted basic sites. The strength
of these sites was assessed by CDCl adsorption followed by FTIR.
3
It was found that basic strength decreases with increasing Mg/Al
molar ratio. The catalysts were tested for the cyanoethylation of
alcohols (methanol and 2-propanol), an industrially important
reaction from which carboxylic acids and amines are obtained.
Excellent conversions were obtained with 100% selectivity. A
correlation between catalytic activities and basic strengths was
proposed. It appears that for cyanoethylation of 2-propanol and
Fig. 9. Cyanoethylation of methanol or 2-propanol over catalyst 3MgAl with
varying rehydration times.

2
04
J.S. Valente et al. / Journal of Catalysis 279 (2011) 196–204
[15] M.J. Climent, A. Corma, S. Iborra, A. Velty, J. Catal. 221 (2004) 474.
larger alcohols, particularly of branched alcohols, solids with very
strong basic sites should be designed. Furthermore, the faster cat-
alyst regeneration process presented here may facilitate industrial
application of activated LDHs.
[
[
[
16] K. Ebitani, K. Motokura, K. Mori, T. Mizugaki, K. Kaneda, J. Org. Chem. 71 (2006)
440.
5
17] O.D. Pavel, R. Bîrjega, M. Che, G. Costentin, E. Angelescu, S. Sß erban, Catal.
Commun. 9 (2008) 1974.
18] J.A. van Bokhoven, J.C.A.A. Roelofs, K.P. de Jong, D.C. Koningsberger, Chem. Eur.
J. 7 (2001) 1258.
Acknowledgments
[
[
19] T. Hibino, A. Tsunashima, Chem. Mater. 10 (1998) 4055.
20] J.C.A.A. Roelofs, J.A. van Bokhoven, A.J. van Dillen, J.W. Geus, K.P. de Jong, Chem.
Eur. J. 8 (2002) 5571.
This work was financially supported by Instituto Mexicano del
Petroleo and Universidad Autonoma Metropolitana. Authors thank
PAPIIT-UNAM, Conacyt and SEP-PROMEP for grants IN107110,
[
21] M. Bellotto, B. Rebours, O. Clause, J. Lynch, D. Bazin, E. Elkaim, J. Phys. Chem.
1
00 (1996) 8535.
[22] W. Yang, Y. Kim, P.K.T. Liu, M. Sahimi, T.T. Tsotsis, Chem. Eng. Sci. 57 (2002)
945.
2
6
0868 and UAM-PTC-104, respectively. J. P. thanks Conacyt for a
[
[
23] E. Kanezaki, Sol. State Ionics 106 (1998) 279.
24] F. Winter, X. Xia, B.P.C. Hereijgers, J.H. Bitter, A.J. van Dillen, M. Muhler, K.P. de
Jong, J. Phys. Chem. B 110 (2006) 9211.
Ph.D. scholarship. Alejandra Santana’s technical support is grate-
fully acknowledged.
[
25] R.J. Chimentão, S. Abelló, F. Medina, J. Llorca, J.E. Sueiras, Y. Cesteros, P. Salagre,
J. Catal. 252 (2007) 249.
Appendix A. Supplementary material
[
[
26] F. Millange, R.I. Walton, D. O’Hare, J. Mater. Chem. 10 (2000) 1713.
27] B. Rebours, J.B. d’Espinose de la Caillerie, O. Clause, J. Am. Chem. Soc. 116
(
1994) 1707.
3
Spectra taken prior and subsequent to CDCl adsorption, and
[
[
28] N.S. Puttaswamy, P.V. Kamath, J. Mater. Chem. 7 (1997) 1941.
29] T. Stanimirova, T. Hibino, V. Balek, J. Therm. Anal. Calorim. 84 (2006) 473.
then during thermal decomposition, of all analyzed catalysts,
may be found online as supplementary material. Supplementary
data associated with this article can be found, in the online version,
at doi:10.1016/j.jcat.2011.01.018.
[30] J.S. Valente, E. Lima, J.A. Toledo-Antonio, M.A. Cortes-Jacome, L. Lartundo-
Rojas, R. Montiel, J. Prince, J. Phys. Chem. C 114 (2010) 2089.
31] H. Pfeiffer, E. Lima, V. Lara, J.S. Valente, Langmuir 26 (2010) 4074.
32] H. Pfeiffer, L. Martínez-dlCruz, E. Lima, J. Flores, M.A. Vera, J.S. Valente, J. Phys.
Chem. C 114 (2010) 8485.
[
[
[
[
33] S. Abelló, D. Vijaya-Shankar, J. Pérez-Ramírez, Appl. Catal. A 342 (2008)
References
119.
34] J. Sanchez-Valente, E. Lopez-Salinas, M. Sanchez-Cantu, US Patent 7 807 128
B2, 2010, to Instituto Mexicano del Petróleo.
35] J.S. Valente, M.S. Cantu, F. Figueras, Chem. Mater. 20 (2008) 1230.
36] J.S. Valente, M. Sanchez-Cantu, E. Lima, F. Figueras, Chem. Mater. 21 (2009)
[
[
[
[
1] M.J. Astle, R.W. Etherington, Ind. Eng. Chem. 44 (1952) 2871.
2] H. Kabashima, H. Hattori, Catal. Today 44 (1998) 277.
3] P.S. Kumbhar, J. Sanchez-Valente, F. Figueras, Chem. Commun. (1998) 1091.
4] E. Angelescu, O.D. Pavel, M. Che, R. Birjega, G. Constentin, Catal. Commun. 5
[
[
5809.
(
2004) 647.
5] R. Birjega, O.D. Pavel, G. Costentin, M. Che, E. Angelescu, Appl. Catal. A 288
2005) 185.
[
[
37] R.D. Shannon, Acta Cryst. A32 (1976) 751.
38] I. Pausch, H.H. Lohse, K. Schürmann, R. Allmann, Clays Clay Miner. 34 (1986)
[
(
507.
[
[
6] F. Cavani, F. Trifiro, A. Vaccari, Catal. Today 11 (1991) 173.
7] V. Rives (Ed.), Layered Double Hydroxides: Present and Future, Nova Science
Publishers, Inc., New York, 2001.
[
39] J.S. Valente, J. Prince, A.M. Maubert, L. Lartundo-Rojas, P. del Angel, G. Ferrat,
J.G. Hernandez, E. Lopez-Salinas, J. Phys. Chem. C 113 (2009) 5547.
40] J.C. Lavalley, Catal. Today 27 (1996) 377.
41] H. Knözinger, S. Huber, J. Chem. Soc. Faraday Trans. 94 (1998) 2047.
42] E.A. Paukshtis, N.S. Kotsarenko, L.G. Karakchiev, React. Kinet. Catal. Lett. 12
[
[
[
[
[
8] X. Duan, D.G. Evans (Eds.), Layered Double Hydroxides, Struc. & Bonding, vol.
1
19, Springer-Verlag, Berlin Heidelberg, Germany, 2006.
9] P.S. Kumbhar, J. Sanchez-Valente, J. Lopez, F. Figueras, Chem. Commun. (1998)
35.
(
1979) 315.
5
[
[
43] E.A. Paukshtis, E.N. Yurchenko, Russ. Chem. Rev. 52 (1983) 242.
44] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination
Compounds, fifth ed., John Wiley & Sons, New York, 1997.
45] E. Lopez-Salinas, M. García-Sánchez, M.E. Llanos-Serrano, J. Navarrete-Bolaños,
J. Phys. Chem. B 101 (1997) 5112.
[
[
10] F. Figueras, Top. Catal. 29 (2004) 189.
11] J.S. Valente, F. Tzompantzi, J. Prince, J.G.H. Cortez, R. Gomez, Appl. Catal. B 90
(
2009) 330.
[
[
[
[
[
12] K.K. Rao, M. Gravelle, J.S. Valente, F. Figueras, J. Catal. 173 (1998) 115.
13] F. Prinetto, G. Ghiotti, R. Durand, D. Tichit, J. Phys. Chem. B 104 (2000) 11117.
14] J.C.A.A. Roelofs, D.J. Lensveld, A.J. van Dillen, K.P. de Jong, J. Catal. 203 (2001)
46] P. Kustrowski, D. Sułkowska, L. Chmielarz, R. Dziembaj, Appl. Catal. A 302
(
2006) 317.
1
84.