The effect of K2HPO4 and Al2(SO4)3 modified MCM-41 on the dehydration of methyl lactate to acrylic acid

Article abstract of DOI:10.1039/c4ra08738k

In this paper, Al₂(SO₄)₃ and K₂HPO₄ modified MCM-41 were developed as catalysts for efficient conversion of methyl lactate (ML) to acrylic acid (AA) and methyl acrylate (MA). By changing the loading amount of the two salts and their ratio, the catalytic performance was optimized and the highest AA + MA yield of 73.6% was achieved over the 45 wt% K₂HPO₄ and Al₂(SO₄)₃ (6:4)/MCM-41 catalyst at 410 °C. The physicochemical properties of the catalysts were investigated by various techniques including XRD, BET, NH₃-TPD, CO₂-TPD, pyridine adsorption-FTIR and CO₂ adsorption-FTIR. The 45% loading sample showed the smallest acidic amounts and a moderate basicity amount, and the acidic-basicity ratio was thought to be crucial for the AA selectivity. Further decrease or increase in the ratio would cause a drop in AA selectivity. Carbon deposition was the main reason for catalyst deactivation. The catalyst could be fully regenerated by air treatment and behaved with good activity after 10 regenerations. This journal is

Full text of DOI:10.1039/c4ra08738k

RSC Advances
View Article Online
View Journal | View Issue
PAPER
The effect of K HPO and Al (SO ) modified
2
4
2
4 3
MCM-41 on the dehydration of methyl lactate to
acrylic acid
Cite this: RSC Adv., 2014, 4, 45679
Bin Wang,* Chao li, Qiangqiang Zhu and Tianwei Tan
2 4 3 2 4
In this paper, Al (SO ) and K HPO modified MCM-41 were developed as catalysts for efficient conversion
of methyl lactate (ML) to acrylic acid (AA) and methyl acrylate (MA). By changing the loading amount of the
two salts and their ratio, the catalytic performance was optimized and the highest AA + MA yield of 73.6%
ꢀ
was achieved over the 45 wt% K
2
HPO
4
and Al
2
(SO
4
)
3
(6 : 4)/MCM-41 catalyst at 410 C. The
physicochemical properties of the catalysts were investigated by various techniques including XRD, BET,
NH -TPD, CO -TPD, pyridine adsorption-FTIR and CO adsorption-FTIR. The 45% loading sample
3
2
2
Received 16th August 2014
Accepted 8th September 2014
showed the smallest acidic amounts and a moderate basicity amount, and the acidic–basicity ratio was
thought to be crucial for the AA selectivity. Further decrease or increase in the ratio would cause a drop
in AA selectivity. Carbon deposition was the main reason for catalyst deactivation. The catalyst could be
fully regenerated by air treatment and behaved with good activity after 10 regenerations.
DOI: 10.1039/c4ra08738k
www.rsc.org/advances
In the literature, several catalysts consisting chiey of
sulphates and phosphates of the group I and II metals, sup-
1
. Introduction
Acrylic acid is an extremely valuable chemical intermediate ported phosphates salts on NaY and calcium hydroxyapatites
1
7–24
used to produce paint additives, adhesives, textile and leather have been reported.
The catalytic dehydration rate of
1–4
treating agents and other products. Oxidation of propene to alcohol over catalyst is thought to be dependent on the catalyst
acrylic acid is a commercial process which is the major surface properties. The dehydration of lactic acid over various
5
contributor for AA production. In recent years, the partial metal sulphates was investigated by Tang. BaSO
4
was found to
oxidation of propane to AA through a one-step process has show an efficient activity with 99.8% lactic acid conversion and
5–8
25
drawn much attention. However, all of the above routes are 74.0% acrylic acid selectivity. The high activity was ascribed to
based on fossil resources. In recent years, the reduction of CO2 the moderate acidity on its surface. This view was also sup-
26
emission has become a crucial problem in the chemical ported by Huang. The effect of rare earth metals (lanthanum,
industry. Given the current robust forces driving sustainable cerium, samarium, and europium) supported on NaY catalyst
chemical production and available biomass conversion tech- was investigated and decreased surface acidity density was
nologies, biomass-based routes are expected to make a signi- considered to cause the activity improvement. As the impor-
cant impact on the chemical industry for producing bulk tance of basicity property was noticed, the relationship between
9
–11
chemicals in the future.
commodity chemicals produced from renewable biomass prepared different alkaline earth phosphates by co-precipita-
27
Methyl lactate and lactic acid are activity and acidic–basicity was established. S. Loridant
12–15
resources.
Dehydration of LA or ML to AA is an environ- tion and gas phase dehydration of lactic acid was evaluated.
mental protection rout, but there still remain some difficulties: Correlation between acrylic selectivity of alkaline earth phos-
besides the dehydration of LA or ML to achieve AA, some side phates and the acid–base balance was clearly established. The
reactions such as decarbonylation/decarboxylation, condensa- AA selectivity was the highest for acid–base balance close to 1
tion, hydrogenation can also occur, which lead to low AA and decreased signicantly when this parameter increased.
16
selectivity. Moreover, compared with the dehydration reac- However, it was reported by Xu that for hydroxyapatites catalyst,
tion, decarbonylation/decarboxylation is a thermodynamic the AA formation rate showed a volcano type dependence on
priority reaction. Therefore, the development of an efficient this acidity–basicity ratio and the best ratio was 4 : 1, which was
28
catalyst for dehydration is necessary and crucial for an eco- different from Loridant. Quite recently, Jean-François Paul
friendly and cost-effective process.
reported that lattice oxygen atoms could easily abstract a
hydrogen atom from the methyl group of the lactic acid and this
reaction led to carbanion formation with a lower activation
energy than carbocation formation by direct C–OH breaking.
The simulation results indicated that the basicity sites may
Beijing Key Lab of Bioprocess, College of Life Science and Technology, Beijing
University of Chemical Technology, Beijing 100029, China. E-mail: wangbin@mail.
buct.edu.cn
29
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 45679–45686 | 45679

View Article Online
RSC Advances
Paper
facilitate a higher AA selectivity. In general, there is still some 2.4. Catalysts characterization
contradiction among all the conclusions and some new catalyst
needs to be developed.
Powder X-ray diffraction measurement was conducted on a
Japan Science D/max-RB diffractometer employing Cu Ka radi-
ation (l ¼ 0.15418 nm). The X-ray tube was operated at 45 kV
and 150 mA. The specic surface areas and pore volumes of the
catalysts were measured through nitrogen adsorption at 77 K
using a Micrometrics ASAP 2020 instrument. Prior to adsorp-
tion, the samples were treated at 200 C under vacuum for 3 h.
The specic surface area was calculated according to the Bru-
nauer–Emmett–Teller (BET) method.
In this paper, K HPO and Al (SO ) supported MCM-41
2
4
2
4 3
showed good AA selectivity and the reason was deduced from
the acidic properties. In order to further investigate the rela-
tionship between structure and activity and to provide some
guidance for catalyst design, some more catalysts with different
ꢀ
loading amount and K
2 4 2 4 3
HPO –Al (SO ) ratio were developed.
The catalysts were further examined with XRD, FTIR, CO /NH -
2
3
TPD and the relationship between acid–basic property and
selectivity was discussed.
The basic–acid properties of the samples were evaluated by
the temperature-programmed desorption (TPD) of CO
2
and
NH . The TPD proles were normalized by sample weight. The
3
sample (0.1 g) in a glass U-tube was preheated at 773 K for 1 h in
He ow (20 mL min ), and then the catalyst was cooled from
2. Experimental
ꢁ1
2.1. Materials
773 K to room temperature. CO or NH (20 kPa) was introduced
2 3
into the glass tube at room temperature for 30 min, and CO or
2
ML (analytic grade) was obtained from Sinopharm Chemical
Reagent Co. and deionized water was used for its dilution. The
alkali phosphates and sulphates were obtained from Sigma-
Aldrich. MCM-41 was purchased from the Nankai University
NH was evacuated for 30 min at 373 K. The TPD measurements
3
were conducted from 373 to 773 K at a heating rate of
ꢁ
1
10 K min
.
The FTIR spectra of pyridine (CO
2
) adsorption were recorded
(China).
ꢁ
1
in the range of 1000–4000 cm on a Nicolet 6700 spectrometer.
The samples were pressed into self-supporting wafers contain-
ing 20 mg of material, and then the wafers were mounted inside
2.2. Catalyst preparation
ꢀ
a Pyrex vacuum cell and degassed at 400 C for 1 h. Samples
There were two steps for the two salts supported catalyst prep-
aration. The rst one was the impregnation with phosphate. 4 g
of MCM-41 was added into 100 mL aqueous solution with a
desired phosphate amount (1.2 g, 30 wt% of the support). The
were allowed to cool down to room temperature (RT), and
pyridine vapor was admitted into the cell and adsorption lasted
for 1 h.
ꢀ
mixture was stirred at 80 C for 3 h, and the resulting slurry was
ꢀ
subjected to oven with air ow at 110 C for 10 h, and the dried
3
. Results
ꢀ
samples were then calcined at 550 C for 3 h with a heating rate
ꢀ
ꢁ1
3.1. ML dehydration activity
of 10 C min . In the second step, the obtained catalyst from
step one was added into 100 mL aqueous solution with a
desired sulphate amount (0.8 g 20 wt% of the support). The
The catalytic activities of the phosphate and sulphate modied
MCM-41 catalysts in ML conversions were listed in Table 1. The
conversions of ML by these catalysts over the controlled reac-
tion conditions gave mainly AA, MA, acetaldehyde and other
small amounts of gaseous by-products such as carbon
monoxide and hydrogen. It could be seen that the modication
ꢀ
mixture was stirred at 80 C for 3 h, and the resulting slurry was
ꢀ
subjected to oven with air ow at 110 C for 10 h. The dried
ꢀ
ꢀ
samples were calcined at 550 C with a heating rate of 10 C
ꢁ
1
min and then pressed at 20 MPa for 10 min and crushed into
particles of 20–40 mesh.
with both K
compared with the single salt supported. Based on this, the
effect of different K HPO to Al (SO ratio on catalyst activity
was to be investigated. The K HPO to Al (SO ratio was
The evaluation of catalyst activity was carried out in an upright chosen from 10 to 0 with the whole loading amount of 45 wt%
xed-bed quartz reactor which was 8 mm in inner diameter and and the results were shown in Fig. 1. Only the AA production
00 mm in length. The reaction was operated at atmospheric was depicted since the amount of MA was too small compared
2 4 2 4 3
HPO and Al (SO ) showed better AA selectivity
2
4
2
4 3
)
2.3. Catalyst evaluation
2
4
2
4 3
)

3
pressure. A 2 g portion of catalyst was charged into the reactor, with AA for each sample which were around 1–2%. For each
and the space above the catalyst bed was lled with quartz chips ratio, the ML conversion was above 96% and there was a
to ensure preheating of the in-coming ML-containing liquid. signicant improvement in AA selectivity from the ratio 10 : 0 to
Before introduction of the feedstock, the sample was heated up 6 : 4. Further decrease of the K
2
HPO
4 2 4 3
–Al (SO ) ratio would
ꢁ
1
in a ow of pure N
2
ꢁ1
(30 mL min ) to a desired temperature at a result in a drop of AA selectivity. The selectivity to MA did not
ꢀ
rate of 10 C min and kept at this temperature for 0.5 h. Then change remarkably within the whole range of the ratios. The
a ow of LA liquid (1.2 mL h ) was introduced, and the prod- highest AA selectivity could be obtained at the 6 : 4 ratio
ucts were collected at a cold trap. The analysis of the collected (K HPO –Al (SO ) ).
ꢁ
1
2
4
2
4 3
species was conducted using a gas chromatograph equipped
with FID and HP-FFAP capillary column (0.32 mm 25 m), and was then investigated with the K
The effect of the two salts loading amount on catalyst activity
2
HPO to Al (SO ratio of 6 : 4
4
2
4 3
)
lactic acid butyl ester was adopted as internal standard.
and the results were shown in Fig. 2. The ML conversion
45680 | RSC Adv., 2014, 4, 45679–45686
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
a
Table 1 Catalytic activity of modified MCM-41 for dehydration of ML to AA and MA
Product selectivity
AA (%)
Catalyst
MCM-41
ML conversion (%)
MA (%)
Acetaldehyde (%)
99
93
90
95
1
43
70
0.8
0.2
2
2
70
45
22
72
K
K
2
HPO
HPO
4
/MCM-41
+ Al (SO
/MCM-41
2
4
2
4
)
3
/MCM-41
Al
2
(SO
4
)
3
0.1
a
ꢀ
ꢀ
ꢁ1
Conditions: calcination temperature 550 C, reaction temperature 400 C, catalyst: 2 mL, particle size: 20–40 meshes, carrier gas N
2
: 30 mL min
,
ꢁ1
feed ow rate: 1.2 mL h ML feedstock: 60 wt% in water.
dependence appeared between the AA selectivity and the
loading amount, and the highest AA selectivity was obtained at
45 wt% loading sample with the AA selectivity 72%. Further
increase of the loading amount would result in a drop of AA
selectivity.
3
.2. Catalyst characterization
.2.1. XRD. The X-ray diffraction (XRD) patterns of K
and Al (SO modied MCM-41 catalysts were shown in Fig. 3
and 4 with different K HPO and Al (SO ratios and different
3
2 4
HPO
2
4 3
)
2
4
2
4 3
)
loading amounts respectively. MCM-41 only showed a wide
ꢀ
diffraction peak around 22 ascribed to SiO as shown in Fig. 3.
2
Fig. 1 ML dehydration activity of two salts modified MCM-41 with Aer the impregnation of Al (SO ) or K HPO , some new peaks
2
4 3
2
4
ꢀ
different ratio. Conditions: reaction temperature 410 C, catalyst: 2
ascribed to Al (SO ) or K P O appeared, which indicated that
2
4 3
4
2
7
ꢁ1
mL, particle size: 20–40 meshes, carrier gas N
2
: 30 mL min
ML feedstock: 60 wt% in water.
4 3
) /MCM-41 (45 wt%).
,
K
2
HPO
change of Al
AlPO and K
reaction between phosphate and sulphate at high temperature.
It could be seen that the peak intensity of K SO varied as the
K HPO –Al (SO ) ratio changed and the maximum appeared
4
condensed to K
(SO
SO
4
P
2
O
7
aer calcinations with no obvious
ꢁ1
feed flow rate: 1.2 mL h
HPO –Al (SO
,
2
4
)
3
. To the catalysts with two salts supported,
K
2
4
2
4
2
4
generated, which should attribute to the
2
4
2
4
2
4 3
when K HPO –Al (SO ) ratio was 6 : 4. The increase in the peak
2
4
2
4 3
intensity could be related to the amount of K SO4 and the
2
increase of crystallinity, which needed further explore.
As provided in Fig. 4, the effect of salts loading amounts on
catalysts structure was investigated. As the loading amount
Fig. 2 ML dehydration activity of two salts modified MCM-41 with
ꢀ
different loading amount. Conditions: reaction temperature 410 C,
catalyst: 2 mL, particle size: 20–40 meshes, carrier gas N
2
: 30 mL
min , feed flow rate: 1.2 mL h , ML feedstock: 60 wt% in water.
HPO –Al (SO
¼ 6 : 4.
ꢁ1
ꢁ1
K
2
4
2
4 3
)
declined slowly as the loading amount increased in the range of
5–80 wt% and the conversions were all over 96%. When the
1
loading amount was higher than 80%, there was a remarkable
decrease of the conversion and the 120% loading sample
showed the lowest conversion of 76%. A volcano type
Fig. 3 XRD patterns of MCM-41 and modified catalysts with different
K HPO –Al (SO ) ratio.
2 4 2 4 3
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 45679–45686 | 45681

View Article Online
RSC Advances
Paper
3
Fig. 5 NH -TPD profiles of MCM-41 and modified catalysts.
Fig. 4 XRD patterns of modified MCM-41 catalysts with different
loading.
lower temperature. These observations suggested that the
and introduction of the two salts eliminated the sites of medium-
SO
strong acidity and reduced the density and strength of the sites
increased from 15% to 80%, the peak intensities of K
2 4
SO
AlPO increased accordingly, indicating that more K
4
2
4
generated. However, there was no more increase when the of weak acidity. It also should be pointed out that compared
loading amount was higher than 80%. Meanwhile, some more with the other loading amounts, the supported catalyst with
ꢀ
peaks such as the peaks centered at 18, 28 ascribed to some 45% loading amount showed the lowest acidic density. A higher
complex sulphate and phosphate appeared when the loading or lower loading amount would lead to a larger acidic amounts.
amount was higher than 45%. Combining the fact that the 45%
Shown in Fig. 6 are the CO -TPD proles recorded over the
2
loading sample showed the highest AA selectivity, it may indi- two salts modied MCM-41 with different loadings. The
cate that the formation of complex sulphate and phosphate was unmodied MCM-41 sample showed three desorption peaks
ꢀ
detrimental for the selectivity. In other words, when the loading centered at 485, 643, and 676 C and the basic amount was 266
ꢁ
1
amount was 45%, the two precursors could completely react mmol g . With the two salts addition at elevated loading, the
ꢀ
with each other and generate AlPO
4
and K
2
SO
4
.
485 C desorption peak diminished and only one peak in the
ꢀ
The BET surface area and pore structure of all samples were range of 600–700 C remained for all the modied samples.
shown in Table 2. MCM-41 sample showed the largest surface Based on the CO
-TPD result, the 15% loading amount sample
2
2
ꢁ1
3
ꢁ1
ꢁ1
area of 1003 m g and the largest pore volume of 0.55 cm g
.
showed the largest basic amount of 441 mmol g . There was a
The introduction of two salts with any amount would cause a remarkable decrease of the basic density as the loading amount
decrease of surface area and pore volume. As the two salts increased, and the 120% loading amount sample showed the
loading amount increased from 15% to 120%, the BET surface smallest basic amount.
area and pore volume decreased accordingly.
3.2.3. Pyridine adsorption FTIR. FTIR spectra of pyridine
3
.2.2. The acid–base properties. To estimate the surface adsorption on parent and two salts modied MCM-41 were
acidity and basicity of the prepared catalysts, the TPD of CO
2
collected for the determination of Bronsted and Lewis acid
and NH were carried out aer N pretreatments at 773 K. The sites, and the results were shown in Fig. 7. All the samples had
3
2
ꢁ1
intensity of each prole was normalized by the sample weight. bands around 1440 and 1600 cm , attributable to pyridine
Shown in Fig. 5 are the NH -TPD proles recorded over the two adsorption on Lewis acid sites. The MCM-41 samples exhibited
3
ꢁ
1
ꢁ1
salts modied MCM-41 with different loading amounts. The strong bands at 1450, 1490 and 1600, cm . The 1490 cm
unmodied MCM-41 sample showed three desorption peaks band can be ascribed to the Lewis and Bronsted acid sites.
centered at 177, 311, and 495 C. With the two salts addition at was noted that the characteristic band of pyridine adsorption
elevated loading, the 311 C desorption peak disappeared on Bronsted acid sites at 1540 cm was absent, so in the
3
0–33
It
ꢀ
ꢀ
ꢁ1
ꢀ
ꢁ1
quickly; meanwhile, the peak at 495 C also diminished. It was present case, the 1489 cm band should be ascribed solely to
noted that the position of the weak acidic peaks shied toward
Table 2 Physicochemical properties of the modified catalysts with different loading amounts
2
ꢁ1
3
ꢁ1
ꢁ1
ꢁ1
Catalyst
S
BET (m g
)
Pore volume (cm g
)
Acid amount (mmol g
)
Basic amount (mmol g
)
A/B
0
1
4
1
%
1003
287
48
0.55
0.398
0.29
720
164
86
266
441
45
2.7
5%
5%
20%
0.27
1.91
5.83
49
0.266
140
24
45682 | RSC Adv., 2014, 4, 45679–45686
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
2
Fig. 6 CO -TPD profiles of MCM-41 and modified catalysts.
Fig. 8 FTIR spectra of CO
modified MCM-41.
2
adsorption collected over unmodified and
34
largest basic sites. However, there is some contradiction
between the CO -TPD and the IR results. Take the 15% sample
2
as an example, according to the CO -TPD result in Fig. 6, it
2
showed the largest basicity amount; but based on the CO -IR
2
result it was nearly zero. One explanation of the contradiction
could be that the IR result mainly reected the surface infor-
mation and the TPD result is thought to be the whole amount.
3.3. Catalytic activity optimization
3.3.1. Effect of reaction temperature. The conversion of ML
could be critically inuenced by reaction temperature and the
effect of reaction temperature on catalyst performance was
Fig. 7 FTIR spectra of pyridine adsorption collected over unmodified
and modified MCM-41.
shown in Fig. 9. The reactions were carried out in the
ꢀ
3
20–450 C range over the phosphate and sulphate modied
catalyst with an optimized loading of 45 wt%. A notable increase
in ML conversion was observed when the reaction temperature
Lewis acid sites. This was a clear signal that the MCM-41 had
mainly Lewis acid sites.
ꢀ
raised from 320 to 380 C. Further increase of the temperature
ꢀ
to 450 C had little effect on ML conversion since the conversion
With the two salts supported together, the patterns varied
with the loading amount. It could be seen that the 15% sup-
ported sample was similar to MCM-41. At higher loadings, the
ꢀ
was higher than 99% over 400 C. The effect of reaction
temperature on the AA selectivity was quite different from the
ꢁ1
intensity of the 1440 cm peak declined, suggesting that the
Lewis acid sites associated with the framework of MCM-41 were
blocked by the impregnation of two salts. To the 120% sup-
ported sample, it showed a broad peak around 1450 and 1600
ꢁ
1
cm , indicating that some new Lewis acids associated with the
two salts appeared besides the Lewis acid related to MCM-41. It
should be pointed out that the bands intensities of 45% sup-
ported sample were really low compared with all the other
samples, indicating that this sample had the smallest surface
acid sites. This was consistent with the NH -TPD results.
3
3
.2.4. CO
spectra of CO
2
adsorption FTIR. Fig. 8 showed the infrared
adsorbed on the catalyst in the region 1800–1000
2
ꢁ
1
cm for the four representative samples. It could be seen from
the spectra that there was little surface basic sites on MCM-41
and 15% supported sample. Some basic sites (1460, 1322, 1114
ꢁ
1
cm ) which correspond to the monodentate and bulk-like Fig. 9 ML conversion as well as AA selectivity. Conditions: catalyst: 2
ꢁ
1
2
ꢁ
mL, particle size: 20–40 meshes, carrier gas N : 30 mL min ,
feed flow rate: 1.2 mL h
carbonates adsorbed on the O anions appeared when the
loading amount increased, and the 120% sample showed the
2
ꢁ1
, ML feedstock: 60 wt% in water.
K
2
HPO –Al (SO /MCM-41 (45 wt%).
4
2
4 3
)
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 45679–45686 | 45683

View Article Online
RSC Advances
Paper
ML conversion and there were three periods depending on the concentration in solution. Therefore, the effect of ML concen-
reaction temperature. Similar to the ML conversion prole, tration on reaction performance was also studied, and the
there was an obvious AA selectivity increase in the period of 320 results were depicted in Table 4. It could be seen that the ML
ꢀ
ꢀ
to 370 C. During the 380–410 C, the AA selectivity almost conversions stayed 98.5% from 15 to 75 wt%; meanwhile, AA
ꢀ
remained unchanged and it reached a maximum at 410 C with selectivity rst gradually increased from 56.1% to 71.0% and
the AA selectivity of 72%. When the temperature was higher then decreased to 68.3%. There was an optimal LA concentra-
ꢀ
than 410 C, there was a notable drop of AA selectivity. Why the tion in solution, namely, 60 wt%, at which the highest AA yield
AA selectivity behaved in this way should be explained from the of 70.8% could be obtained.
point of the competition of the dehydration reaction with other
side reactions. When the temperature was higher than 410 C, on stream was studied at 410 C over the two salts supported
3.3.4 .Catalyst stability. The stability of catalyst with time
ꢀ
ꢀ
ꢁ
1
more side reactions and further reaction of AA would be MCM-41 with 45 wt% loading amount (LHSV ¼ 0.6 h and ML
enhanced since at elevated temperature the stability of carboxyl concentration ¼ 60 wt%). In our stability test of 20 h period, the
group decreased, thus the AA selectivity decreased.
ML conversion decreased from 99% to 75%, but the AA selec-
3.3.2. Effect of liquid hourly space velocity (LHSV). The tivity remained unchanged except a drop from 71% to 63% in
inuence of LHSV on reaction performance was shown in Table the rst 5 hours. The reason for its deactivation may be attrib-
ꢀ
3
. The reactions were conducted at 410 C with ow rate of ML uted to the ML polymerization and carbon deposit on the
ꢁ
1
changed from 0.4 to 5 mL h (corresponding LHSV ¼ 0.2–2.5 catalyst since the color of the catalyst became black. Similar to
ꢁ1
h
). Other operation conditions such as the ow rate of N (30 SAPO-34 used in MTO reaction, our catalysts could also be
2
ꢁ
1
ꢀ
mL min ) and concentration of LA (60 wt%) were retained. It regenerated by a simple treatment at 410 C with air for 2 hours.
was observed from Table 2 that ML conversion decreased from The results could be seen in Fig. 10 that with a short treatment
8.0% to 59.9% as LHSV increased from 0.2 to 2.5 h . At aer 8 h test, the catalyst could be fully regenerated and the ML
ꢁ
1
9
elevated LHSV, the contact time of ML with catalyst reduced, conversion and the AA selectivity mainly remained unchanged
resulting in low ML conversion. However, AA selectivity for 10 times reuse.
increased from 59.2% to 71.0% as LHSV increased from 0.2 to
In order to investigate the reason for the catalyst deactiva-
ꢁ
1
ꢁ1
0
.6 h , and declined to 64.4% at LHSV of 2.5 h . There was an tion, the fresh and used catalyst and the regenerated sample
ꢁ
1
optimal LHSV of 0.6 h at which the highest AA yield of 72%
could be achieved. The decrease in the AA selectivity at low
LHSV may indicate the presence of diffusion limitations.
3.3.3. Effect of LA concentration. ML conversion and AA
yield were affected not only by LHSV but also by ML
a
Table 3 Reaction performance at various LHSV
LHSV
ML Conversion
(%)
AA selectivity
(%)
Acetaldehyde
(%)
AA yield
(%)
ꢁ
1
(h )
0
0
0
1
1
2
.2
.4
.6
98.1
99.8
98.5
95.2
83.5
59.9
59.2
62.3
70.8
66.6
64.8
63.3
15.9
17.5
16.0
15.9
14.8
13.4
58.1
62.2
70.0
63.4
54.1
37.9
.5
.5
Fig. 10 Catalytic activity of the modified MCM-41 catalyst with 45%
a
ꢀ
loading amount after 10 times regeneration for dehydration of ML to
Conditions: temperature: 410 C, catalyst: 2 mL, particle size: 20–40
ꢁ1
ꢀ
AA. Conditions: reaction temperature 410 C, catalyst: 2 mL, particle
2
meshes, carrier gas N : 30 mL min , ML feedstock: 60 wt% in water.
K
ꢁ
1
size: 20–40 meshes, carrier gas N
1
2
: 30 mL min , feed flow rate:
2
HPO
4
–AL
2
(SO
4
)
3
/MCM-41 (45 wt%).
ꢁ
1
.2 mL h ML feedstock: 60 wt% in water. K HPO –Al (SO
2
4
2
4 3
)
¼ 6 : 4.
a
Table 4 Reaction performance at various ML concentrations in solution
ML concentration (%)
ML conversion (%)
AA selectivity (%)
Acetaldehyde (%)
AA yield (%)
1
3
4
6
7
5
0
5
0
5
98.6
98.5
98.5
98.5
98.0
56.1
59.2
65.1
70.8
68.3
22
21
19
17
19
55.3
58.3
64.1
70.0
66.9
a
ꢀ
ꢁ1
ꢁ1
Conditions: temperature: 410 C, catalyst: 2 mL, particle size: 20–40 meshes, carrier gas N
(SO /MCM-41 (45 wt%).
2
: 30 mL min , feed ow rate: 1.2 mL h
2 4
K HPO –
AL
2
4 3
)
45684 | RSC Adv., 2014, 4, 45679–45686
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
could be adjusted and used in ML dehydration reaction. 72% AA
selectivity was achieved over 6 : 4 ratio of K HPO and Al (SO )
4 3
2
4
2
and 45 wt% supported MCM-41 catalyst. The higher selectivity
for acrylic acid should be correlated to the synergistic effect of
the weaker acid site density and moderate basic site density as
well as the acidic–basicity ratio. The catalyst showed a good
stability and could be regenerated aer air treatment.
Acknowledgements
The authors would like to acknowledge the Ministry of Science
and Technology, PR China for the nancial support of Project
2013CB733600. We would also thank China Postdoctoral
Science Foundation (2013M530515).
Fig. 11 XRD patterns of modified catalyst of fresh, reaction and recycle
sample.
References
1
L. F. Boesel and R. L. Reis, Prog. Polym. Sci., 2008, 33, 180–
90.
were examined with XRD (Fig. 11). It could be seen that the
main peaks of the three samples were quite similar except the
new peak centered at 26.6 belonging to graphite appeared for
the reaction one. Aer the regeneration, the graphite peak
disappeared and the prole became the same as the fresh one,
indicating that the catalyst could be totally regenerated.
Besides, it was proved that the dehydration reaction and the
regeneration process had little effect on the catalyst structure,
indicating that the samples could be promising industrial
catalysts with its good stability.
1
ꢀ
2 C. M. Tang, Y. Zeng, X. G. Yang, Y. C. Lei and G. Y. Wang, J.
Mol. Catal. A: Chem., 2009, 314, 15–20.
3
X. Xu, J. Lin and P. Cen, Chin. J. Chem. Eng., 2006, 14, 419–
27.
4
4
V. Grabovac, D. Guggi and A. B. Schn u¨ rch, Adv. Drug Delivery
Rev., 2005, 57, 1713–1723.
M. M. Lin, Appl. Catal., A, 2001, 207, 1–16.
5
6
X. J. Yang, R. M. Feng, W. J. Ji and C. T. Au, J. Catal., 2008,
253, 57–65.
7
M. M. Lin, Appl. Catal., A, 2003, 250, 305–318.
4
. Discussion
8 P. Botella, J. M. L. Nieto, B. Solsona, A. Mifsud and
F. M ´a rquez, J. Catal., 2002, 209, 445–455.
It has been known that the catalytic dehydration rate of alcohol
over catalyst would be dependent on the catalyst surface acid–
base property. An attempt is made to correlate the rates of ML
conversion and AA formation with the surface acidity and
basicity of the modied catalysts. The acidic amounts and
basicity amounts calculated by NH (CO )-TPD as well as the
9
X. Tong, Y. Ma and Y. Li, Appl. Catal., A, 2010, 385, 1–13.
1
1
1
1
1
1
0 P. Sun, D. H. Yu, Z. C. Tang, H. Li and H. Huang, Ind. Eng.
Chem. Res., 2010, 49, 9082–9087.
1 P. Sun, D. H. Yu, K. M. Fu, M. Y. Gu, Y. Wang, H. Huang and
H. H. Ying, Catal. Commun., 2009, 10, 1345–1349.
2 M. H. Gao, M. Hiratu, E. Toorisaka and T. Hano, Bioresour.
Technol., 2006, 97, 2414–2420.
3 M. A. Abdel-Rahman, Y. Tashiro and K. Sonomoto,
Biotechnol. Adv., 2013, 31, 877–902.
4 M. A. Abdel-Rahman, Y. Tashiro and K. Sonomoto, J.
Biotechnol., 2011, 156, 286–301.
5 M. Jarvinen, L. Myllykoski, R. Keiski and J. Sohlo,
Bioseparation, 2000, 9, 163–166.
3
2
ratio between acidic and basicity were shown in Table 1. MCM-
1 showed the lowest AA selectivity due to its strong acidic sites.
4
For the other three samples, the AA selectivity was negatively
related to the acidic density and 45% sample showed the lowest
acidic density and highest AA selectivity. Basic sites should play
an important role in the dehydration reaction since it can
dissociate H–C bond and enhance the dehydration reaction. In
our study, the best AA selectivity appeared when the acid–basic
1
1
6 Y. Fan, C. Zhou and X. Zhu, Catal. Rev., 2009, 51, 293–324.
7 Y. Matsuura, A. Onda, S. Ogo and K. Yanagisawa, Catal.
Today, 2014, 226, 192–197.
¼
1.91 on the 45% sample, which was different from the result
of Xu (acidic–basicity ¼ 4) and Loridant (acidic–basicity ¼ 1).
The difference may come from different catalysts and the
precursors between LA and ML. In general, specic to this
catalyst, the acid–basic z 1.9 was best for the AA selectivity. The
best acidic–basicity ratio varied on catalysts may attribute to the
ML adsorption or consecutive reactions.
1
1
2
2
8 Y. Matsuura, A. Onda and K. Yanagisawa, Catal. Commun.,
2014, 48, 5–10.
9 C. M. Tang, J. S. Peng, G. C. Fan, X. L. Li, X. L. Pu and W. Bai,
Catal. Commun., 2014, 43, 231–234.
0 J. Yan, D. H. Yu, P. Sun and H. Huang, Chin. J. Catal., 2011,
3
2, 405–411.
1 J. Yan, D. H. Yu, H. Li, P. Sun and H. Huang, J. Rare Earths,
010, 28, 803–806.
ratio, 22 J. F. Zhang, Y. L. Zhao, M. Pan, X. Z. Feng, W. J. Ji and
5
. Conclusion
2
By modifying the loading amount and K
2
HPO
4
–Al
2
(SO
4
)
3
the surface acid–basic properties of MCM-41 based catalyst
C. T. Au, ACS Catal., 2011, 1, 32–41.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 45679–45686 | 45685

View Article Online
RSC Advances
Paper
2
2
2
2
2
2
3 V. C. Ghantani, S. T. Lomate, M. K. Dongare and 29 C. Hammaecher and J. F. Paul, J. Catal., 2013, 300, 174–182.
S. B. Umbarkar, Green Chem., 2013, 15, 1211–1217.
4 J. F. Zhang, Y. L. Zhao, X. Z. Feng, M. Pan, J. Zhao, W. J. Ji and
C. T. Au, Catal. Sci. Technol., 2014, 4, 1376–1385.
30 B. Chakraborty and B. Viswanathan, Catal. Today, 1999, 49,
253–260.
31 X. Shao, X. Zhang, W. Yu, Y. Wu, Y. Qin, Z. Sun and L. Song,
Appl. Surf. Sci., 2012, 263, 1–7.
5 J. Peng, X. Li, C. Tang and W. Bai, Green Chem., 2014, 16,
1
08–111.
32 T. D. Conesa, J. M. Hidalgo, R. Luque, J. M. Campelo and
A. A. Romero, Appl. Catal., A, 2006, 299, 224–234.
33 C. Cai, H. Wang and J. Han, Appl. Surf. Sci., 2011, 257, 9802–
9808.
34 F. Wang, N. Ta and W. J. Shen, Appl. Catal., A, 2014, 475, 76–
81.
6 H. J. Wang, D. H. Yu, P. Sun, J. Yan, Y. Wang and H. Huang,
Catal. Commun., 2008, 9, 1799–1803.
7 E. Blanco, P. Delichere, J. M. M. Millet and S. Loridant, Catal.
Today, 2014, 226, 185–191.
8 B. Yan, L. Z. Tao, Y. Liang and B. Q. Xu, ACS Catal., 2014, 4,
1931–1943.
45686 | RSC Adv., 2014, 4, 45679–45686
This journal is © The Royal Society of Chemistry 2014