Vapor phase aldol condensation of methyl acetate with formaldehyde over a Ba-La/Al2O3 catalyst: The stabilizing role of la and effect of acid-base properties

Article abstract of DOI:10.1039/c7ra10008f

Vapor phase aldol condensation of methyl acetate with formaldehyde was studied over Ba-La/Al₂O₃ with different amounts of lanthanum catalysts. The catalysts were characterized by X-ray diffraction (XRD), N₂ adsorption-deso

Full text of DOI:10.1039/c7ra10008f

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Vapor phase aldol condensation of methyl acetate
with formaldehyde over a Ba–La/Al2O3 catalyst: the
stabilizing role of La and effect of acid–base
properties†
Cite this: RSC Adv., 2017, 7, 52304
Qiang Bao,a Wanchun Zhu,a Jianbiao Yan,ab Chunlei Zhang,b Chunli Ning,b Yi Zhang,b
a
Mengmeng Haoa and Zhenlu Wang
*
Vapor phase aldol condensation of methyl acetate with formaldehyde was studied over Ba–La/Al2O3 with
different amounts of lanthanum catalysts. The catalysts were characterized by X-ray diffraction (XRD), N2
adsorption–desorption, pyridine absorption performed via Fourier transform infrared spectroscope
(Py-IR), NH3 and CO2 temperature-programmed desorption (NH3 and CO2-TPD), thermal analysis
(TG-DTA) and scanning electron microscopy (SEM). The catalytic performance was evaluated using
a fixed-bed microreactor. The results showed that bare Al2O3 was intrinsically active but poorly selective
to methyl acrylate. The addition of barium species significantly improved the catalytic activity and
selectivity. However, the Ba/Al2O3 catalyst was not stable in the continuous reaction due to a large
amount of carbon deposition on the catalyst surface. Compared with adding individual components
(BaO), the combination of the two promoters (BaO and La2O3) showed higher catalytic stability.
Although the activity of the Ba–La/Al2O3 catalyst was not obviously increased compared with the
Ba/Al2O3 catalyst, the carbon deposition was obviously suppressed in the target reaction due to the
alkaline function of La2O3. Combined with the characterization results, we found that the addition of
lanthanum species could significantly reduce the number of strong acid sites on the catalyst surface,
inhibit the generation of carbon species in the reaction process, and stabilize the catalytic activity of the
catalyst. In addition, the lifetime of the optimum 5Ba–0.5La/Al2O3 catalyst was evaluated over
a continuous period of 300 h, and the initial catalytical activity did not exhibit an obvious decrease.
Received 8th September 2017
Accepted 28th October 2017
DOI: 10.1039/c7ra10008f
rsc.li/rsc-advances
ever increasing demand for petroleum, it is becoming increas-
ingly to use propylene as a feedstock. Therefore, development of
1. Introduction
alternative routes for MA production is necessary. In recent
decades, a one-step synthesis of AA and ester via a vapor-phase
aldol condensation of acetic acid (Aa) or methyl acetate (Ma)
with formaldehyde (FA) has been studied and developed.
According to previous work, there are two types of catalysts
tested for the aldol condensation reaction: acidic catalysts and
basic catalysts. Acidic catalysts consist mainly of oxides of
metals such as vanadium(V) oxide (V2O5), phosphorus(V) oxide
(P2O5), and niobium(V) oxide (Nb2O5). Early on, Ai12–14 reported
that both the vanadium–titanium binary phosphates and V2O5–
P2O5 binary oxide can effectively catalyze the aldol condensation
of HCHO with acetic acid/methyl acetate to acrylic acid and its
derivative. Based on their work, Feng et al.15 successfully
prepared an improved VPO catalyst using benzyl alcohol and
PEG additive, activated in a butane-containing atmosphere, and
the optimal conversion of methyl acetate reached 84.2%
(including 30–35% acetic acid and 3–12% COx). More attention
has been paid to basic catalysts, which oen include the oxides
or hydroxides of alkali or alkaline earth metals supported on
As a fundamental industrial monomer, methyl acrylate (MA) is
widely used in the manufacture of paint additives, adhesives,
textiles, and leather treating agents.1–4 The traditional way of
producing MA is the acetone cyanohydrin (ACH) process.5,6 This
route requires large quantities of toxic hydrogen cyanide and
produces copious amounts of ammonium sulfate wastes, which
is unsustainable and environmentally unfriendly. Commer-
cially, MA is produced by esterication of acrylic acid (AA) and
methanol, prior to the synthesis of AA by the two-stage oxida-
tions of propylene with air.7–11 However, propylene comes from
non-sustainable sources in the petrochemical industry, and the
price of it is greatly inuenced by the crude oil price. With the
aKey Laboratory of Surface and Interface Chemistry of Jilin Province, College of
Chemistry, Jilin University, Qianjin Road 2699, Changchun, 130012, PR China.
E-mail: wzl@jlu.edu.cn; Fax: +86 431 88499140; Tel: +86 431 88499140
bShanghai Huayi (Group) Company Technology Research Institute, Longwu Road 4600,
Shanghai 200241, PR China
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c7ra10008f
52304 | RSC Adv., 2017, 7, 52304–52311
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
porous materials, for vapor phase aldol condensation because
of the easier preparation and higher catalytic activity than acidic
catalysts. Yan et al.16 used an SBA-15 mesoporous molecular
support impregnated with cesium oxide (Cs2O) to catalyze the
aldol condensation of methyl acetate with formaldehyde, and
the best catalyst demonstrated the highest (48.4%) conversion
of methyl acetate with 95.0% selectivity for methyl acrylate. Zhu
et al.17 reported a Cs–La–Sb/SiO2 catalyst and showed the
conversion of methyl acetate (20%) as well as a yield of methyl
acrylate (8%). Furthermore, they found that the activity of the
best catalyst sharply decreased as the reaction progressed due to
the loss of the Cs species. Recently, an Al2O3-supported cesium
catalyst was reported to be a good base catalyst for the vapor
phase aldol condensation of methyl acetate with formaldehyde.
Zhang et al.18 prepared a series of Cs–P/g-Al2O3 catalysts, and
the best catalyst demonstrated a 47.1% yield of methyl acrylate.
However, the activity decreased quickly because of the carbon
deposition on the surface of the catalyst. Then, they used the
same catalyst system in a uidized bed reactor, and they found
it had high catalytic stability, with no signicant decrease in
catalytic activity aer 1000 h of lifetime evaluation.19 However, it
is well known that the catalytic reaction with uidized bed has
a series of problems such as severe catalyst loss, wear of reaction
vessels and higher gas–solid separation requirements.
Previous studies have shown that the performance of
alumina, although very good, is limited by the carbon deposi-
tion attached to the catalyst surface. The activity of the catalyst
decreased quickly with the increase in carbon deposition. This
phenomenon might be due, at least in part, to the strong
surface acidity of alumina. As reported, the acidity of g-Al2O3
was related to the degree of coordination of the Al3+ species, of
which the strongest Lewis acid sites were associated with very
lowest coordination Al cations, such as tri- and tetra coordi-
nated Al3+.20–25 In addition, both strong solid acid and base sites
2. Experiment
2.1 Catalyst preparation
Methyl acetate ($99.0%), methanol (CH3OH; $99.0%), trioxy-
methylene ($98.0%), barium acetate (Ba(CH3COO)2; $99.0%),
lanthanum nitrate hexahydrate (corresponding La2O3 content
$44.0%) and alumina nitrate nonahydrate (Al(NO3)3$9H2O $
99.0%) of analytical grade were purchased from the Sinopharm
Chemical Reagent Company. The Al2O3 were prepared by
precipitation of Al(NO3)3$9H2O from an aqueous solution with
NH3. The precipitates were thoroughly washed with deionized
water at least three times, dried in an oven at 100 ꢀC overnight
and calcinꢁe1d in a muffle furnace at 500 ꢀC for 5 h with
a 2 ꢀC min heating ramp rate.
The Ba/Al2O3 catalysts were prepared with the wetness
impregnation method. Typically, Al2O3 was impregnated with
an aqueous solution with a desired amount of barium acetate,
and the mixture was later stirred for approximately 2 h at room
temperature. Aerwards, water in the mixture was evaporated at
ꢀ
a temperature of 80 C under atmospheric pressure. The dried
materials were nexꢁt 1calcined in a muffle furnace at 550 ꢀC for 5 h
with a 10 ꢀC min heating ramp rate. Ba–La/Al2O3 catalysts
were prepared in the same way except that two aqueous solu-
tions of Ba(CH3COO)2 and La(NO3)3$6H2O were mixed before
impregnation. The catalysts were denoted as xBa–yLa/Al2O3,
where x represents the content of Ba (wt%) and y represents the
content of La (wt%). The 5Ba–0.5La/Al2O3 catalyst was prepared
as follows: the Al2O3 supports (5 g) were impregnated with 4 ml
of an aqueous solution containing the desired amount of
Ba(CH3COO)2 (0.4407 g) and La(NO3)3$6H2O (0.0703 g). The
other catalysts were prepared in the same way.
2.2 Catalyst characterization
were also ineffective in catalyzing the aldol condensation reac- X-ray diffraction patterns were measured with an Empyrean X-
tion, and the moderate acidity corresponded to higher selec- ray diffractometer using a nickel ltered Cu Ka source at
tivity.26 Thus, it is necessary to reduce the surface acidity of a wavelength of 0.154 nm. An accelerating voltage of 40 kV and
g-Al2O3 not only in inhibiting the formation of coke but also in a current of 40 mA were used. A slit width of 0.25ꢀ was used as
improving the catalytic activity. Doping of alumina with basic the source. Scans were collected using a PIXcel3D detector. N2
components could be an efficient way to improve the catalytic adsorption/desorption isotherms were measured using
stability and resistance to coking.27–29
a Micromeritics ASAP 2010N analyzer. Specic surface areas
In this paper, vapor-phase aldol condensation of methyl were calculated using the BET model. Pore size distributions
acetate with formaldehyde was studied over a series of Ba–La/ were evaluated from desorption branches of nitrogen isotherms
Al2O3 catalysts. Metal-modied catalysts were prepared for using the BJH model. The carbon deposition on spent catalysts
inhibiting carbon accumulation and stabilizing the catalytic was examined by thermal analysis, using a Germany NETZSCH
activity. The physical–chemical properties of the catalysts were thermoanalyzer (model STA 499F3). The spent samples (10 mg)
studied by N2-adsorption and X-ray diffraction (XRD). The were heated at a rate of 10 ꢀC minꢁ1 from room temperature to
surface acidity and basicity of the catalysts were studied by CO2- 900 ꢀC in an air ow of 10 cm3 minꢁ1. Scanning electron
TPD, NH3-TPD and pyridine adsorption IR. Catalytic tests were microscopy observations of the fresh and the spent catalyst
carried out in a xed-bed microreactor. Additionally, the effect samples were performed using a Hitachi SU8020. Temperature
of the lanthanum loading amounts of the aldol condensation of programmed desorption of NH3 and CO2 was used to charac-
Ma and FA were investigated under the same conditions. The terize the acidic and basic sites over the studied catalysts,
carbon deposition of the used catalysts with or without respectively. The TPD experiments were conducted using
lanthanum promotion was studied by thermal analysis (TG-
a ChemBET Pulsar TPR/TPD instrument (Quantachrome
DTA) and scanning electron microscopy (SEM). The stability Instruments) with a built-in TCD detector. Typically, a 50 mg
and regeneration of the optimum catalyst were also catalyst was used in each measurement. The catalyst was rst
investigated.
purged with He (UHP grade, Airgas) at 550 ꢀC for 0.5 h with
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 52304–52311 | 52305

View Article Online
RSC Advances
Paper
a 10 ꢀC minꢁ1 heating ramp rate, then cooled down to 50 ꢀC. A
ow of CO2 (research grade, Airgas) or NH3 (electronic grade,
Airgas) was introduced into the tubular catalyst bed for 30 min
at 50 ꢀC for CO2 adsorption and 50 ꢀC for ammonia adsorption.
Aer purging the catalyst bed for approximately 30 min with He
to evacuate the physisorbed NH3 or CO2, the catalyst was heated
to 550 ꢀC with a 10 ꢀC minꢁ1 heating ramp rate. The change in
thermal conductivity due to the concentration change of NH3 or
CO2 in the effluent was recorded on the TCD detector. The TPD
prole was deconvoluted using a Gaussian multipeak tting
function built-in OriginPro7.5 soware. IR spectra were recor-
ded on a Nicolet Impact spectrometer at 4 cmꢁ1 optical reso-
lution. Prior to the measurements, 30 mg of the catalyst was
pressed in self-supporting discs and activated in the IR cell
attached to a vacuum line at 250 ꢀC for 0.5 h. The adsorption of
Fig. 1 N2 adsorption–desorption isotherms and pore-size distribution
of studied catalysts.
ꢀ
pyridine (Py) was performed at 50 C for 30 min. The excess of
shows that both Al2O3 and metal-modied Al2O3 were type IV
with an H1-type hysteresis loop at high relative pressure. All of
the textural properties of Ba–La/Al2O3 samples are depicted in
Table 1. From Table 1, we can see the specic surface area of
Al2O3 decreased obviously with the addition of 5 wt% BaO.
Although the pore volume changed minimally, the pore diam-
eter increased signicantly, indicating that Ba species produced
changes in the textural properties of the alumina support. For
Ba–La-modied catalysts, the specic surface area of Ba–La/
Al2O3 progressively decreased with increasing lanthanum
content. The pore volume of the catalysts did not change
signicantly, whereas the pore diameter increased gradually.
This nding suggests that during the impregnation procedure,
there was some dissolution and precipitation of oxides with
blockage of smaller pores. However, it should be noted that the
specic surface area of Ba–La-modied Al2O3 catalysts was
obviously higher than that of the 5Ba/Al2O3 catalyst, indicating
that the addition of small quantities of lanthanum species may
help to limit the decrease in the SBET of Ba/Al2O3.
The XRD patterns of the samples are presented in Fig. 2.
From Fig. 2, we can see that the diffraction peaks of the Al2O3
support were broad and diffuse, which means the amorphous
alumina is the main existing form. For the 5Ba/Al2O3 sample, no
diffraction peaks due to Ba species are detected, which is
consistent with prior studies.30 As the content of lanthanum
increased from 0 to 3.0 wt%, the XRD patterns of Ba–La/Al2O3
showed only the reections of the Al2O3 support material. This
nding means that lanthanum species are highly dispersed on
the surface of Al2O3.
ꢀ
pyridine was further evacuated at 50 C for 0.5 h. Aer elimi-
nation of the physical absorption the sample was cooled to
room temperature for registration of the IR spectra, and the
sample was returned to reaction conditions and heated to
a further temperature.
2.3 Catalyst evaluation
The vapor phase aldol condensation reaction of methyl acetate
with formaldehyde was conducted in a xed-bed reactor under
atmospheric pressure. The reason why we choose a xed bed
reactor is that it has a small amount of catalyst and a small
reaction volume to achieve greater production capacity. And it
can strictly control the time of stay, reaction temperature and
other parameters, which are more conducive to the study of the
performance of the catalyst. Typically, 500 mg of catalyst was
loaded at the center of a stainless steel tube reactor that was
30 cm long with a 0.8 cm inside diameter and supported by
a quartz frit. The mixed solution of methyl acetate and form-
aldehyde was fed into the reactor using an advection pump with
a typical ow rate of 0.06 mL minꢁ1. The products were iden-
tied and analyzed by gas chromatography (GC-8A, FID)
equipped with a 60 m capillary column DB-WAX. Some GC data
for the catalytic reaction have been added to ESI (in Fig. S1†).
The molar ratio of Ma and FA was xed at 1/2 unless other-
wise indicated. The catalysts were evaluated based on their
catalytic performance. The denition of methyl acetate
conversion and product selectivity was as follows:
Ma conversion (%) ¼ (Main ꢁ Maout)/Main ꢂ 100
MA selectivity (%) ¼ MA/(Main ꢁ Maout) ꢂ 100
MA yield (%) ¼ Ma conversion (%) ꢂ MA selectivity (%)
Table 1 Textural properties of the Al2O3 and metal-modified Al2O3
catalysts
SBET
Pore volume
Mean pore diameter
(nm)
Catalysts
(m2 gꢁ1
)
(cm3 gꢁ1
)
Al2O3
5Ba/Al2O3
5Ba–0.1La/Al2O3
5Ba–0.5La/Al2O3
5Ba–1.0La/Al2O3
5Ba–3.0La/Al2O3
307.7
272.3
299.4
292.6
290.6
286.4
0.42
0.41
0.44
0.44
0.43
0.45
5.4
6.0
5.8
6.0
5.9
6.2
3. Results and discussion
3.1 Physical properties
The surface area and pore size distribution of Ba–La/Al2O3
catalysts were analyzed using N2 adsorption–desorption. Fig. 1
52306 | RSC Adv., 2017, 7, 52304–52311
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
(103.6 mmol gꢁ1 cat.) was observed over Al2O3, probably due to the
rich Lewis acid sites from coordinative unsaturated aluminum
ions on the catalyst surface. The alumina supported BaO catalyst
had a dramatically decreased NH3 uptake at a temperature of
100–330 ꢀC (90.3 mmol gꢁ1 cat. for bare Al2O3 and 48.4 mmol gꢁ1
cat. for 5Ba/Al2O3), indicating that a large number of weak acid
sites were reduced when BaO was supported on Al2O3. According
to previous reports, for g-Al2O3, bulk Al atoms display either
tetrahedral or octahedral coordination. Exposed surface Al sites
can display three-, four-, or 5-fold coordination, and exhibit Lewis
acidity, of which the weakest Lewis acid sites were associated
with very high coordination Al cations, such as 5-coordinated
Al3+.25,29 Thus, the reduction of weak acid sites may be due to the
interaction between BaO and 5-coordinated Al3+. J. Kwak and co-
workers found the BaO could selectively anchor the penta-
coordinated Al3+ sites on the (100) facets of g-Al2O3.31 This
nding was consistent with our experimental results. For Ba–La/
Al2O3 catalysts, the total NH3 uptake over 5Ba–0.1La/Al2O3 was
approximately 75.4 mmol gꢁ1 cat., slightly more than the NH3
uptakes for 5Ba/Al2O3 (61.4 mmol gꢁ1 cat.), indicating that adding
a small amount of lanthanum could increase total acid content.
More interestingly, the NH3 uptake over 5Ba–0.5La/Al2O3 at
a temperature of 215–250 ꢀC increased about 7.6 mmol gꢁ1 catalyst
compared to 5Ba/Al2O3. However, the NH3 uptake in other
Fig. 2 The XRD patterns of Al2O3 and metal-modified Al2O3 catalysts.
3.2 Acid–base properties
The acid–base properties of the Al2O3 and metal-modied Al2O3
were characterized by the TPD of NH3–CO2 in Fig. 3 and Fig. 4.
Table 2 lists the total CO2 and NH3 uptakes on the basis of
catalyst weight and the corresponding concentrations for the
different types of acid/base sites. In Fig. 3, all NH3-TPD proles
were integrated (deconvoluted) in six xed temperature ranges
(100–215 ꢀC, 215–250 ꢀC, 250–290 ꢀC, 290–330 ꢀC, 330–365 ꢀC
and 365 ꢀC ꢃ) in order to compare acidic sites of the same
strength. For support, a very large amount of NH3 uptake
ꢀ
temperature ranges (especially at a temperature of 330–365 C)
decreased dramatically. It might be the result of an interaction
between lanthanum species and acid sites on the catalyst surface,
leading to an increase in weak acid sites and a decrease in strong
acid sites. With a further increase of lanthanum species, the NH3
uptake at all the temperature ranges did not changed signi-
cantly. Fig. 4 shows the desorption proles of CO2 from Ba–La/
Al2O3 catalysts with different amounts of lanthanum. From Fig. 4,
we can see that all catalysts exhibited a prominent CO2 desorption
peak in the range of 100–350 ꢀC and the basic sites on these
catalysts were arbitrarily considered weak. Adding barium species
slightly increased the total CO2 uptake (10.1 mmol gꢁ1 cat.)
compared to the Al2O3 support (8.3 mmol gꢁ1 cat.). It should be
noted that adding lanthanum could also increase the total CO2
uptake from 13.3 mmol gꢁ1 cat. for 5Ba–0.1La/Al2O3 to
18.4 mmol gꢁ1 cat. for 5Ba–3.0La/Al2O3. Furthermore, the CO2
adsorption capacity of Ba–La/Al2O3 catalysts gradually increased
with a further increase in lanthanum species, indicating some new
stronger base sites formed on the catalyst surface.
Fig. 3 NH3-TPD patterns of the studied catalysts.
To determine the nature of the acidic sites, the acid prop-
erties of the Al2O3 and metal-modied Al2O3 were further
characterized by FTIR spectra of pyridine adsorption. In Fig. 5,
several feature bands were observed at 1446, 1490 and
1614 cmꢁ1 on these samples, and their intensities became
weaker when adding metal to the Al2O3 supports. The bands at
1446 and 1614 cmꢁ1 could be attributed to pyridine adsorbed
on Lewis acid sites. The band at 1490 cmꢁ1 was also attributed
to Lewis acid sites because there was no evidence for the
formation of pyridinium ions at 1540 cmꢁ1. The IR results
indicated that the surface of the Al2O3 and metal-modied
Al2O3 catalysts mainly consist of Lewis acid sites. Further-
more, the relative amount of these sites declined when adding
metal to the Al2O3 supports.
Fig. 4 CO2-TPD patterns of the studied catalysts.
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 52304–52311 | 52307

View Article Online
RSC Advances
Paper
Table 2 List of the total CO2 and NH3 uptakes on the catalysts surface and the corresponding concentrations for the different types of acid/base
sites
Catalyst
Al2O3
5Ba/Al2O3
5Ba–0.1La/Al2O3
5Ba–0.5La/Al2O3
5Ba–1.0La/Al2O3
5Ba–3.0La/Al2O3
Total (mmol NH3 per g cat.)
T1 (100–215 ꢀC)
103.6
8.1
25.5
36.9
19.8
8.7
61.4
4.6
8.1
24.1
11.6
8.8
75.4
6.5
38.5
3.8
15.7
6.6
4.2
5.9
39.9
3.8
16.0
6.8
4.7
6.0
40.2
3.8
15.7
6.9
4.8
5.9
T2 (215–250 ꢀC)
11.2
30.0
13.7
10.1
3.9
T3 (250–290 ꢀC)
T4 (290–330 ꢀC)
T5 (330–365 ꢀC)
T6 (>365 ꢀC)
4.6
4.2
2.3
2.6
3.1
Total (mmol CO2 per g cat.)
8.3
8.3
10.1
10.1
13.3
13.3
14.4
14.4
15.0
15.0
18.4
18.4
T1 (100–350 ꢀC)
peaks at 400 and 480 ꢀC for Al2O3 samples, two exothermic
peaks at 395 and 460 ꢀC for 5Ba/Al2O3 samples, and one broader
exothermic peak at 385–450 ꢀC for 5Ba–0.5La/Al2O3 samples.
These ndings indicate that two different carbon deposition
species on these catalysts were oxidized and that the carbon
deposition species on the Al2O3 with BaO and La2O3 promoters
were more easily oxidized than that on the Al2O3 catalyst. Thus,
the temperature of the exothermic peaks for carbon deposition
oxidation on the Al2O3 with BaO and La2O3 promoters was lower
than that on the Al2O3 catalyst.
Coke formation was also conrmed by SEM images as shown
in Fig. 7. Fig. 7 shows SEM images of the Al2O3 catalyst (Fig. 7a
and b), the 5Ba/Al2O3 catalyst (Fig. 7c and d), and the 5Ba–0.5La/
Al2O3 catalyst (Fig. 7e and f) before and aer the condensation
reaction for 10 h at 390 ꢀC and atmospheric pressure. From
Fig. 5 FTIR spectra of pyridine adsorbed on Al2O3, 5Ba/Al2O3 and
5Ba–0.5La/Al2O3.
3.3 The characterization of used catalysts
The TG-DTA analysis of the catalyst samples used for 10 h in the
aldol condensation of methyl acetate with formaldehyde at
390 ꢀC and atmospheric pressure is presented in Fig. 6. TG
analysis showed that weight losses were approximately 47%,
28% and 24% for the Al2O3, 5Ba/Al2O3 and 5Ba–0.5La/Al2O3
catalysts, respectively, indicating a signicant amount of
carbon deposition on the Al2O3. The addition of BaO to Al2O3
obviously improved the carbon formation on the Al2O3 catalyst,
which might be due to the reduction of acid sites. The smallest
amount of carbon deposition was found on the 5Ba–0.5La/Al2O3
catalyst aer the reaction for 10 h, indicating that adding
a small amount of lanthanum could further prevent carbon
deposition for aldol condensation of methyl acetate with
formaldehyde. In the DTA curves, there were two exothermic
Fig. 7 SEM micrographs of the catalysts before and after reaction for
10 h at 390 ꢀC (a and b) before and after reaction on Al2O3; (c and d)
before and after reaction on 5Ba/Al2O3; (e and f) before and after
Fig. 6 TG and DTA curves of Al2O3, 5Ba/Al2O3 and 5Ba–0.5La/Al2O3. reaction on 5Ba–0.5La/Al2O3.
52308 | RSC Adv., 2017, 7, 52304–52311
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
Fig. 7, it can be seen that the surface of the fresh catalyst is reduce the amount of strong acid, but also produce more weak
porous and rough. The hole is clearly visible. However, the acid sites corresponding to NH3 uptake at a temperature of 215–
surface of the spent catalyst is relatively smooth, and the 250 ꢀC in our study. The TG-DTA analysis showed that the 5Ba–
channel is seriously blocked. Furthermore, there are some 0.5La/Al2O3 catalyst had the least amount of carbon produced
particles of new, uniform material produced on the surface of aer 10 h compared with the Al2O3 and 5Ba/Al2O3 catalysts. It is
the spent catalyst. The reason for these results is that the well known that carbon deposition (or coking) is one of the
catalyst is covered with a large amount of carbon deposition principal factors that may deactivate catalysts.34–36 Zhu et al.17
during the reaction, thereby blocking the pore structure of the studied the compositions of soluble coke on the Cs–La–Sb/SiO2
catalyst surface.
catalyst. They found that the composition of soluble coke in
deactivated catalysts was mainly aromatic hydrocarbon.
Toluene, DMBs, TriMBs and hextraMB were all detected.
Furthermore, they also found that the addition of lanthanum
species was conducive to the inhibition of carbon formation
due to covering strong acid sites and providing weak acid sites.
This conclusion is exactly consistent with our present study.
Therefore, combined with the catalytic activity and character-
ization results, we can conclude that the addition of lanthanum
species reduced the number of strong acid sites on the catalyst
surface, inhibited the generation of carbon species in the
reaction process, and stabilized the catalytic activity of the
catalyst.
To investigate the effect of the lanthanum loading amounts
on the aldol condensation reaction of methyl acetate with
formaldehyde, the Ba–La/Al2O3 catalysts with different
lanthanum content were also prepared by the same preparation
method. The evaluation results of the studied catalysts are
shown in Fig. 9. From Fig. 9, we can see that with the increase in
lanthanum content, the activity and selectivity of the catalysts
did not change much, but the stability of the Ba–La/Al2O3
catalysts had been obviously improved. When adding 0.5 wt%
La2O3 to 5Ba/Al2O3, the stability of the obtained catalyst was the
best. The specic stability order of studied catalysts was as
follows: 5Ba–0.5La/Al2O3 > 5Ba–1.0La/Al2O3 > 5Ba–0.1La/Al2O3
z 5Ba–3.0La/Al2O3. From the CO2-TPD analysis, with the
addition of lanthanum species, the number of basic sites on the
catalyst surface gradually increased, and the center of the
desorption peak was shied toward the high-temperature zone,
which indicated that some new stronger basic sites were
generated on the catalyst surface. In our previous studies, the
proper acid–base properties, such as the number and strength
of the catalysts, were critical to promoting this aldol conden-
sation reaction.33 Therefore, too much or too little load may
cause the acid–base environment of the catalyst to not be suit-
able for the target reaction.
3.4 Activity tests
Different dopants including rare earth metals, alkali metals and
other elements were also prepared via the same impregnation
method. The evaluation result is shown in Table S1.† It can be
observed that the 5Ba–0.5La/Al2O3 catalyst showed the best
catalytic performance. Garbarino et al.32 studied the ethanol
conversion reaction on the La–alumina catalysts. They found
that the addition of lanthanum could stabilize alumina with
respect to sintering and loss of surface area, as well as against
carbon laydown. Furthermore, they also suggested that the
La3+–O2ꢁ couples interact specically with the strongest acido–
basic sites of alumina reducing the Lewis acidity but providing
moderate acido–basic couples. And in our previous studies, we
have shown that the medium acid–base sites is more suitable
for the aldol condensation reaction.33 Therefore, we only re-
ported the Ba–La-modied Al2O3 catalysts. The evaluation
results of the studied catalysts are presented in Fig. 8. From
Fig. 8, bare Al2O3 was found to be intrinsically active but poorly
selective to methyl acrylate. With the addition of 5 wt% BaO to
the Al2O3 support, the 5Ba/Al2O3 catalyst showed an obvious
increase in catalytic activity to methyl acetate. Furthermore, the
selectivity for methyl acrylate increased 13.4% compared to the
bare Al2O3. This result is agreement with our previous
research.33 When adding 0.5 wt% La2O3 to 5Ba/Al2O3, the 5Ba–
0.5La/Al2O3 catalyst did not show a marked increase in catalytic
activity and selectivity. However, it should be noted that,
compared with adding individual components, a combination
of the two promoters showed higher catalytic stability since the
decline in catalytic activity in 5Ba–0.5La/Al2O3 was not so sharp
as in 5Ba/Al2O3 and Al2O3. From the NH3-TPD analysis, the
addition of lanthanum species signicantly altered the acid
environment of the catalyst surface. An appropriate amount of
lanthanum species (5Ba–0.5La/Al2O3) can not only effectively
Fig. 8 The conversation of Ma and the selectivity to MA on the Al2O3, Fig. 9 The conversation of Ma and the selectivity to MA on the Ba–La/
5Ba/Al2O3, and 5Ba–0.5La/Al2O3 catalysts.
Al2O3 catalysts with different La content.
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 52304–52311 | 52309

View Article Online
RSC Advances
Paper
modied Al2O3 catalysts mainly consisted of Lewis acidic
sites. Furthermore, the relative amounts of these sites declined
when adding metal to the Al2O3 supports. The results of the
NH3-TPD showed that the addition of lanthanum species can
cause an increase in weak acid sites and a decrease in strong
acid sites, making the surface acidity of the catalyst more suit-
able for the condensation reaction. The CO2-TPD analysis
showed that adding lanthanum species could also increase total
CO2 uptake, indicating that some new base sites formed on the
catalyst surface. Furthermore, the quantitative results further
demonstrated the changes in the acid–base environment on the
catalyst surface. Combining the catalytic activity and selectivity
with the characterization results, it can be concluded that the
addition of lanthanum species signicantly reduced the
number of strong acid sites on the catalyst surface, inhibited
the generation of carbon species in the reaction process, and
stabilized the catalytic activity of the catalyst.
Fig. 10 The conversation of Ma and the selectivity to MA on the 5Ba–
0.5La/Al2O3 catalyst. Note: the volume of feedstock is 0.03 ml minꢁ1
.
3.5 Catalyst stability and reusability
The stability and regeneration of the 5Ba–0.5La/Al2O3 catalyst
were also investigated at 390 ꢀC for about 300 h, and the results
are shown in Fig. 10. The catalyst was regenerated by calcining
in a stream of air at 550 ꢀC for 5 h and studied under the same
conditions. From Fig. 10, we can see that Ma conversion
declined by nearly 60% aer reaction at 60 h. In fact, the catalyst
turned brown and black aer the reaction. According to the
results of TG-DTA and SEM analysis, we can conclude that the
deactivation of catalyst can be due to the formation of carbon
deposition on the surface of the catalyst.18,34,36
Although the activity (based on the conversation of Ma)
decreased gradually with reaction time as the result of the
formation of carbonaceous deposits, the regenerated catalyst,
by removing carbonaceous deposits at high temperatures in
a stream of air, showed nearly the same initial catalytic activity
as the fresh catalyst. This result indicated that the effective
components of the regenerated catalyst were not destroyed in
the process of regeneration, and the activity of the 5Ba–0.5La/
Al2O3 catalyst was relatively stable aer a total reaction time of
300 h. Thus, the 5Ba–0.5La/Al2O3 catalyst really had good
reusability.
Conflicts of interest
There are no conicts to declare.
Acknowledgements
This work was supported by the technology institute of
Shanghai Huayi (Group) Company and Jilin Province Science
and Technology research plan (key scientic research project).
(No. 20150204020GX).
Notes and references
¨
1 H. Danner, M. Durms, M. Gartner and R. Braun, Appl.
Biochem. Biotechnol., 1998, 70, 887.
2 X. Xu, J. Lin and P. Cen, Chin. J. Chem. Eng., 2006, 14, 419.
3 K. Nagai, Appl. Catal., A, 2001, 221, 367.
4 P. Nagai and L. Lochmann, Prog. Polym. Sci., 1999, 24, 793–
873.
5 E. Hauptman, S. Sabo-Etienne, P. White, M. White, J. Gamer,
P. Fagan and J. Calabrese, J. Am. Chem. Soc., 1994, 116, 8038–
8060.
6 M. Bacaa, M. Aouine, J. Dubois and J. Millet, J. Catal., 2005,
233, 234.
7 B. Jo, S. Kum and S. Moon, Appl. Catal., A, 2010, 378, 76–82.
4. Conclusions
Vapor phase aldol condensation of methyl acetate and formal-
dehyde was investigated over Ba–La/Al2O3 catalysts in a xed-
bed reactor. The properties of the catalysts and the carbon
deposition performance on the catalyst surface were surveyed
by XRD, N2 adsorption–desorption, TG-DTA, and SEM. The
acid–base properties of the catalysts were characterized by TPD
¨
8 R. d'Alnoncourt, L. Csepei, M. Havecker, F. Girgsdies,
¨
M. Schustera, R. Schlogl and A. Trunschke, J. Catal., 2014,
311, 369–385.
9 Z. Zhai, A. Getsoian and A. Bell, J. Catal., 2013, 308, 25–36.
of NH3–CO2 and FTIR spectra of pyridine adsorption. The 10 W. Fang, Q. Ge, J. Yu and H. Xu, Ind. Eng. Chem. Res., 2011,
results of the catalyst evaluation showed that the 5Ba–0.5La/ 50, 1962–1967.
Al2O3 catalyst exhibited the best catalytic activity and selectivity 11 M. Havecker, S. Wrabetz, J. Krohnert, L. Csepei,
¨
¨
¨
on the aldol condensation of methyl acetate with formaldehyde,
and the conversion of Ma achieved 39.5% and the selectivity of
R. d'Alnoncourt, Y. Kolen'ko, F. Girgsdies, R. Schlogl and
A. Trunschke, J. Catal., 2012, 285, 48–60.
Ma achieved 93.9%. Although the activity of the 5Ba–0.5La/ 12 M. Ai, J. Catal., 1987, 107, 201–208.
Al2O3 catalyst was not obviously increased compared with 5Ba/ 13 M. Ai, J. Catal., 1988, 112, 194–200.
Al2O3 catalyst, the carbon deposition was greatly suppressed in 14 M. Ai, Stud. Surf. Sci. Catal., 1992, 72, 101–108.
the target reaction due to the alkaline function of La2O3. The IR 15 X. Feng, B. Sun, Y. Yao, Q. Su, W. Ji and C.-T. Au, J. Catal.,
results indicated that the surface of the Al2O3 and metal-
2014, 314, 132–141.
52310 | RSC Adv., 2017, 7, 52304–52311
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
16 J. Yan, C. Zhang, C. Ning, Y. Tang, Y. Zhang, L. Chen, S. Gao, 26 J. Tai and R. Davis, Catal. Today, 2007, 123, 42–49.
Z. Wang and W. Zhang, J. Ind. Eng. Chem., 2015, 25, 344. 27 B. Bloch, B. Ravi and R. Chaim, Mater. Lett., 2000, 42, 61–65.
17 Y. Wang, X. Lang, G. Zhao, H. Chen, Y. Fan, L. Yu, X. Ma and 28 J. Kwak, J. Hu, A. Lukaski, D. Kim, J. Szanyi and C. Peden,
Z. Zhu, RSC Adv., 2015, 5, 32826. J. Phys. Chem. C, 2008, 112, 9486–9492.
18 G. Zhang, H. Zhang, D. Yang, C. Li, Z. Peng and S. Zhang, 29 S. Bai, Q. Dai, X. Chu and X. Wang, RSC Adv., 2016, 6, 52564.
Catal. Sci. Technol., 2016, 6, 6417.
19 S. Jiang, C. Li, H. Chen, D. Yang and S. Zhang, Ind. Eng.
Chem. Res., 2017, 56, 9322–9330.
20 M. Digne, P. Sautet, P. Raybaud, P. Euzen and H. Toulhoat,
J. Catal., 2002, 211, 1–5.
21 M. Digne, P. Sautet, P. Raybaud, P. Euzen and H. Toulhoat,
J. Catal., 2004, 226, 54–68.
30 J. Kwak, D. Mei, C. Yi, D. Kim, C. Peden, L. Allard and
J. Szanyi, J. Catal., 2009, 261, 17–22.
31 J. Kwak, J. Hu, D. Kim, J. Szanyi and C. Peden, J. Catal., 2007,
251, 189–194.
32 G. Garbarino, C. Wang, I. Valsamakis, S. Chitsazan, P. Riani,
E. Finocchio, M. Flytzani-Stephanopoulos and G. Busca,
Appl. Catal., B, 2017, 200, 458–468.
22 T. Phung, A. Lagazzo, M. Crespo, V. Escribano and G. Busca, 33 Q. Bao, T. Bu, J. Yan, C. Zhang, C. Ning, Y. Zhang, M. Hao,
J. Catal., 2014, 311, 102–113. W. Zhang and Z. Wang, Catal. Lett., 2017, 147, 1540–1550.
23 G. Jenness, M. Christiansen, S. Caratzoulas, D. Vlachos and 34 J. Cueto, L. Faba, E. Dıaz and S. Ordonez, Appl. Catal., B,
R. Gorte, J. Phys. Chem. C, 2014, 118, 12899–12907. 2017, 201, 221–231.
24 J. Hu, S. Xu, J. Kwak, M. Hu, C. Wan, Z. Zhao, J. Szanyi, 35 N. Miletic, U. Izquierdo, I. Obregon, K. Bizkarra,
´
´
´
´
X. Bao, X. Han, Y. Wang and C. Peden, J. Catal., 2016, 336,
85.
I. Agirrezabal-Telleria, L. Bario and P. Arias, Catal. Sci.
Technol., 2015, 5, 1704–1715.
25 M. Christiansen, G. Mpourmpakisv and D. Vlachos, ACS 36 C. Sararuk, D. Yang, G. Zhang, C. Li and S. Zhang, J. Ind. Eng.
Catal., 2013, 3, 1965–1975.
Chem., 2017, 46, 342–349.
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 52304–52311 | 52311