Isomerization of Styrene Oxide to Phenyl Acetaldehyde over Different Modified Beta Zeolites

Article abstract of DOI:10.1007/s10562-017-2004-x

Abstract: The catalytic performance of the isomerization of styrene oxide to phenyl acetaldehyde was investigated over different modified beta catalysts under gas-phase atmosphere free of solvents. The physicochemical properties were characterized by a combination of X-ray diffraction (XRD), N₂ adsorption, Fourier transform infrared spectroscopy (FT-IR), Magic Angle Spinning Nuclear Magnetic Resonance (MAS-NMR), NH₃-Temperature programmed desorption (NH₃-TPD) and thermo gravimetric analysis (TGA), etc. Acid sites with stronger strength are responsible for the formation of condensation products and fast deactivation of catalysts. The obtained results also demonstrate that the introduction of boron atoms can create suitable acid strength and acid distribution without significantly destroying its microporous structures. Therefore, BAl-BEA sample is highly active and selective for the rearrangement of styrene oxide as well as very stable to deactivation. Graphical Abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s10562-017-2004-x

Catal Lett (2017) 147:1523–1532
DOI 10.1007/s10562-017-2004-x
Isomerization of Styrene Oxide to Phenyl Acetaldehyde
over Different Modified Beta Zeolites
Xiong‑Fei Zhang1 · Jianfeng Yao1 · Xiaoxia Yang2
Received: 20 December 2016 / Accepted: 17 February 2017 / Published online: 6 April 2017
© Springer Science+Business Media New York 2017
Abstract The catalytic performance of the isomeriza-
tion of styrene oxide to phenyl acetaldehyde was inves-
tigated over different modified beta catalysts under gas-
phase atmosphere free of solvents. The physicochemical
properties were characterized by a combination of X-ray
diffraction (XRD), N2 adsorption, Fourier transform infra-
red spectroscopy (FT-IR), Magic Angle Spinning Nuclear
Magnetic Resonance (MAS-NMR), NH3-Temperature pro-
grammed desorption (NH3-TPD) and thermo gravimetric
analysis (TGA), etc. Acid sites with stronger strength are
responsible for the formation of condensation products
and fast deactivation of catalysts. The obtained results
also demonstrate that the introduction of boron atoms can
create suitable acid strength and acid distribution without
significantly destroying its microporous structures. There-
fore, BAl-BEA sample is highly active and selective for
the rearrangement of styrene oxide as well as very stable to
deactivation.
Graphical Abstract
Keywords Styrene oxide · Beta zeolites · Acidity · Boron
modification
* Jianfeng Yao
jfyao@njfu.edu.cn
1
College of Chemical Engineering, Nanjing Forestry
University, Nanjing 210037, People’s Republic of China
2
School of Chemical Engineering and Technology, Tianjin
University, Tianjin 300072, People’s Republic of China
Vol.:(0123456789)
1 3

1524
X.-F. Zhang et al.
1 Introduction
2 Experimental
Phenyl acetaldehyde (PhAA), a versatile chemical inter-
mediate for the production of fragrances, drugs and
herbicides, is of considerable interest in the fine chem-
istry [1–3]. On an industrial scale, the main process to
produce PhAA is via the isomerization of styrene oxide
(SO) [4–7]. However, such reaction is always accom-
panied by other undesired reactions such as aldol con-
densation or aldehydes polymerization; as a result,
byproducts such as 2,4-diphenyl-2-butenal (dimer) and
2,4,6-tribenzyl-s-trioxane (trimer) are produced, leading
to low yield of PhAA. To prevent undesired by-products,
one commonly used method is to add a large amount of
solvents into feedstock to dilute reactants. Though such
method can improve the yield, new problems such as
environmental pollution, high energy consumption and
subsequent products separation occurred. Thus, there is
a need to develop environmental-friendly and energy-
saving methods to isomerizes SO to PhAA and relevant
high-performance catalysts that can work in solvent-free
system.
2.1 Catalyst Preparation
The chemicals used in this work are of analytical grade and
used without further purification. Al incorporated sam-
ple was synthesized hydrothermally from a gel mixture
of 1 SiO2:0.067 Al:0.5 TEAOH:45 H2O. Powder Al was
added to the mixture of tetraethyl ammonium hydroxide
(TEAOH) and de-ionized water under violent stirring to
obtain a solution (A) Then, fumed silica was used as sili-
con sources and slowly bubbled into remaining TEAOH to
obtain solution (B) Following that, solution B was mixed
with solution A under continuous stirring. The acquired gel
was transferred into a 100 ml stainless-steel autoclave and
crystallized at 140°C for 72 h. The solid was filtered and
washed several times with de-ionized water until neutral.
The obtained white powder was dried at 110°C for 12 h
and denoted as Al-BEA. Boron incorporated sample was
synthesized with the similar procedures by using H3BO3
as boron sources, TEAOH as structure directing agent and
with a gel mixture of 1 SiO2:0.025 H3BO3:0.5 TEAOH:30
H2O. The obtained sample was denoted as B-BEA.
As previously reported, various solid acid materials
have been investigated for the isomerization of SO, such
as natural silicates, Nafion-H, and zeolites, etc [6, 8–14].
Among them, zeolites have attracted much interest due
to its excellent hydrothermal stability, adjustable acidity
and shape selectivity [15]. Recently, we have proved that
zeolites of MFI-type are one of the potential catalysts for
the selective transformation of SO [16–18]. However,
there are two reasons responsible for the rapid deacti-
vation of the catalysts: (1) the ten-ring channel systems
(5.3 × 5.6 and 5.1 × 5.5 Å) of MFI framework are smaller
than that of SO (5.9 Å); thus most of active sites inside
MFI might not function due to the intra-crystalline diffu-
sion limitation and (2) the existence of too much strong
acid sites on MFI facilitates the occurrence of side reac-
tions and accelerates the deactivation process [19–21].
To overcome the above-mentioned obstacles, BEA-
type (β) zeolite is chosen based on the following consid-
erations: (1) the crystallographic pore dimensions of β
zeolite are about 7.3 × 6.4 Å for the straight pores along
the a and b axis, is thought to be superior to other zeo-
lites on account of the promising shape selectivity [22]
and (2) another interesting property of β zeolite is that
its acidity can be controlled by a series of synthesis and
post-synthesis methods [23–25]. Therefore, it is worth
to study the catalytic performance of beta zeolites to
isomerizes SO to PhAA. To the best of our knowledge,
this is the first time to report the use of beta zeolites in
the rearrangement of SO.
Post-treatment was performed on as-prepared Al-BEA
sample. Phosphorus modified Al-BEA sample was pre-
pared by impregnation method. The typical synthesis pro-
cedures were described as follows, at first, certain amount
of as-prepared Al-BEA sample was dissolved in H3PO4
solution under stirring at room temperature. The slurry was
mixed thoroughly, dried overnight at 110°C and calcined at
550°C for 6 h. Boron treated Al-BEA sample was prepared
with the same procedures by substituting H3PO4 solution
as H3BO3 solution. The obtained samples were named as
PAl-BEA and BAl-BEA, corresponding to phosphorus and
boron content of 3.4 and 4.6 wt%, respectively.
2.2 Catalyst Characterization
X-ray diffraction (XRD) patterns of the samples were
recorded on a Bruker D8-Advanced X-ray diffract meter
with Cu-Ka radiation operating at 40 KV, 30 mA. Chemi-
cal composition of the samples was measured by Induc-
tively Coupled Plasma-Optical Emission Spectrometry
(ICP-OES, Vista MPX, Varian Company). N2 adsorption
was performed on a Quantachrome Autosorb-1 instrument
at −196°C. NH3-Temperature programmed desorption
(NH3-TPD) experiment was carried out on a home-made
apparatus equipped with a programmable temperature
furnace and TCD detector. Physically adsorbed NH3 was
extracted by flushing with N2 at 80°C for 2 h and desorp-
tion was performed by heating the sample from 100 to
550°C at a rate of 15°C/min. FT-IR spectra of adsorbed
pyridine was investigated on a Nicolet 380 spectrometer.
1 3

Isomerization of Styrene Oxide to Phenyl Acetaldehyde over Different Modified Beta Zeolites
1525
Pyridine molecules were adsorbed in situ at 25°C and then
evacuated at 200°C for 2 h. MAS-NMR spectrometry of
the fresh and spent catalysts was carried out to character-
ize the B and Al environment in the samples. 11B spec-
tra were recorded on a Bruker Avance III 400 working at
128.07 MHz with a spinning rate of 12 kHz, a pulse length
of 2.0 s and a pulse delay of 1.0 s. 27Al MAS NMR spectra
of the samples were obtained on a Varian Infinity Plus 300
NMR spectrometer with a magic angle of spinning probe.
The spectra were recorded at a frequency of 78.13 MHz by
spinning at 8.0 kHz, a pulse width of 0.5 s and a recycle
delay of 3.0 s.
analyzed by GC (Agilent 6820, FID) and GC-MS (Agilent
6890/5973, MSD). After reaction, the reactor was purged
with N2 to remove any possible residual substances. Weight
loss of the spent catalysts was determined by TGA (Perki-
nElmer, Pyris 6) at a heating rate of 10°C/min from 30 to
800°C.
3 Results and Discussion
3.1 Physicochemical Properties
Figure 1 shows the XRD patterns of all catalysts. From
Fig. 1, it can be seen that all samples matched well with the
typical BEA-type crystalline structure and no other phases
were found, which indicated high purity of as-synthesized
samples. Relative crystallinity (RC) was obtained from the
2θ=7°–9° and 21°–25° using the parent Al-BEA as a refer-
ence (100%). As listed in Table 1, the RC of BAl-BEA and
PAl-BEA decreased to 95 and 89% respectively compared
to Al-BEA. The decreased crystallinity can be attributed to
the interaction of phosphate and borate with Al atoms and
subsequent removal of framework Al atoms during calci-
nations steps [11, 26]. In addition, no clear peaks associ-
ated with B2O3 or P2O5 were detected, representing a good
dispersion of boron and phosphorus species in the zeolite
surfaces.
2.3 Catalytic Performance Evaluation
The isomerization of SO was carried out in a continuous
fixed-bed reactor (i.d. = 9 mm). The catalyst (0.5 g; 20–30
mesh) located in the constant-temperature zone was pre-
treated at 500°C for 2 h and then cooled to reaction tem-
perature. N2 was used as carrier gas with a flow rate of
120 ml/min. The products were collected in an ice trap and
The physical properties and chemical composition of
the samples were summarized in Table 1. The Si/Al ratio
in Al-BEA (15.9) determined by ICP-OES shows a 6%
increase relative to nominal value (15), while the Si/B ratio
in B-BEA presents a 20% increase. It is more difficult for
boron atoms to incorporate into the BEA structure, imply-
ing that B-BEA may exhibit a poor framework stability.
After modification with phosphorus and boron, the Si/Al
values also display an ascending tendency, which caused by
de-alumination during the acid treatment [27]. This is con-
sistent with the observation from XRD.
Fig. 1 XRD patterns of different modified samples
Table 1 Chemical composition and physical properties of the catalysts
Sample
Nominal
Measureda
RCb SBET (m2/g) Pore volume
Pore size (Å) Acidity (mmol/g or peak area/cm)
Weak Strong Brønsted Lewis Externalc
(cm3/g)
Si/Al Si/B Si/Al Si/B
Total Micro
B-BEA
Al-BEA
BAl-BEA 15.9
PAl-BEA 15.9
–
15
40
–
–
–
48.8
98 548
100 564
95 560
89 398
0.37 0.29
0.34 0.27
0.33 0.28
0.21 0.14
5.7
5.8
5.8
5.4
0.096 trace
0.127 0.278
0.193 0.244
0.136 0.157
0.68
1.22
3.91
1.35
2.79
3.65
6.19
3.92
0.22
2.55
Trace
Trace
15.9
16.4
17.2
–
–
–
–
aDetermined by ICP-OES
bRelative crystallinity
cDetermined by FT-IR of 2,4,6-trimethylpyridine adsorption
1 3

1526
X.-F. Zhang et al.
N2 adsorption and desorption isotherms (not shown
for Al-BEA is associated with its higher Al content in
framework, which can be further proved by the following
FT-IR and 27Al NMR results [31, 32]. With the introduc-
tion of phosphorus, the low temperature peak assigned to
weak acid sites gradually shifted to a lower temperature
and peak intensity increases, suggesting that its strength
of decreased as well as its concentration increased. Moreo-
ver, PAl-BEA partially lost the intensity of its strong acid
sites, which can be attributed to the removal of framework
Al species. The phosphorus species may interact with the
framework Al atoms to form polymeric phosphate, thus
neutralize the acidity of strong acid sites. After modifica-
tion with boron, the weak acid concentration increases
remarkably from 0.127 to 0.193 mmol/g. Meanwhile, the
strong acid concentrations slightly decreased from 0.278
to 0.244 mmol/g. The incorporation of boron in the BEA
framework has been verified to be a good approach to gen-
erate new weaker acid sites without significant reduction on
the concentration and strength of strong acid sites.
here) coincides with type I isotherms, corresponding to uni-
form microporous structure according to IUPAC classifica-
tion. The presence of a small quantity of mesoporous pores
represents structure defects or inter-crystalline volumes.
The BET areas and pore volumes of B-BEA, Al-BEA and
BAl-BEA are similar. This suggests that the incorporation
of boron in the BEA framework did not change the pore
structure much. However, the introduction of phospho-
rus leads to an obvious decline in BET surface area and
pore volume. It might be explained that the collapse of
BEA framework resulting from de-alumination generates
pyrophosphates and polyphosphates outside or near zeolite
channels. The micropore structures in PAl-BEA may be
irreversibly destroyed accompanied by the amorphization
of the zeolite crystals.
3.2 Acidity
NH3-TPD was utilized to investigate the concentration and
strength of acid sites by the amount of desorbed NH3 and
the location of desorption peak. To quantify the acidity of
the catalyst, H3BO3 solution was used to collect the des-
orbed NH3 followed by titration of H2SO4 solution [28–30].
Figure 2 illustrates the NH3-TPD profiles of the samples
and Table 1 gives the acid concentrations with differing
strengths. The B-BEA sample exhibits only one desorp-
tion peak centered at 170°C that can be ascribed to weakly
adsorbed NH3. Other samples produces two desorption
peaks around 180 and 310°C, corresponding to NH3 des-
orption from weak and strong acid sites.
In order to further study the acidity of different sam-
ples, FT-IR spectra of samples after adsorbing pyridine
were performed and the results were summarized in Fig. 3.
Generally, bridging hydroxyl groups (–Al–OH–Si–) in Al-
incorporated zeolites are known as Brønsted acid sites, and
silanol sgroup (–B–OH–Si–) in boron containing zeolites
are identified as Lewis acid sites. The amount of acid sites
was estimated by the integrated area (IA) of each character-
istic band. The peak at 1545, 1445 and 1490 cm−1 are des-
ignated to pyridine chemisorbed on Brønsted, Lewis acid
sites and synergistic effects of Brønsted and Lewis acid
sites [19]. Parent Al-BEA exhibits signals of both Brønsted
and Lewis acid sites, and the amount of Lewis acid sites
is higher than Brønsted acid sites (Table 1). By contrast,
boron-substituted B-BEA shows weaker peaks of Lewis
acid sites and trace amount of Brønsted acid sites [27].
Upon boric acid treatment, the amount of Lewis acid sites
and Brønsted acid sites increased by 50 and 146% com-
pared to Al-BEA. However, for PAl-BEA, the Lewis and
Brønsted acid sites just changed marginally, revealing that
the formation of aluminum phosphate creates some new
acid sites [33].
Generally, the acidity of zeolite highly depends on the
coordination state and distribution of the Al species in
framework. The higher concentration of strong acid sites
To determine the locaion of Al atoms in the samples,
27Al NMR was carried out and the spectra were shown in
Fig. 4. The chemical shifts at 53 and 0 ppm were refered
to the framework tetrahedral Al species (AlFW) and extra-
framework Al species (AlEF) in octahedral coordination.
Al-BEA sample presents a strong sharp peak at 53 ppm
and a relatively smaller peak at 0 ppm, indicating that
the majority of Al species were tetrahedrally coordinated
AlFW. After the introduction of boron, most of the AlFW
species remains intact and a small part of tetrahedral Al
atoms were extracted from the framework, forming ALEF
species, which can be verified by the variation of the area
Fig. 2 NH3-TPD profiles of the samples
1 3

Isomerization of Styrene Oxide to Phenyl Acetaldehyde over Different Modified Beta Zeolites
1527
ratio (Fig. 4b). Upon phosphorus modification, the signal
at 53 ppm decreased dramatically, confirming the occur-
rence of de-alumination in zeolite framework. However, the
peak area at 0 ppm shows a decreasing tendency and a new
peak at −14 ppm appeared that can be assigned to poly-
meric phosphates species. It can be deduced that AlEF spe-
cies turned into aluminum phosphates via interacting with
phosphorus and the framework of PAl-BEA was partially
destroyed, which is in agreement with the above XRD and
N2 adsorption [34, 35].
11B MAS NMR was performed to explore the boron
environment of B-BEA and BAl-BEA. As shown in Fig. 5,
one notable resonance at −3 ppm was observed on B-BEA,
indicating the existence of the well ordered tetrahedrally
framework coordinated boron (BO4) whereas the character-
istic bands centered at 7.1 and 16.6 ppm for extra frame-
work trigonal boron (BO3) were not found. These results
confirm that the majority of the boron species in the coex-
istence state of the BO4. The spectrum of BAl-BEA shows
Fig. 3 Pyridine IR spectra of the parent and modified catalysts
Fig. 4 27Al NMR spectra (a), deconvolution of Al-BEA (b), BAl-BEA (c), and PAl-BEA (d)
1 3

1528
X.-F. Zhang et al.
Fig. 5 11B NMR spectra of B-BEA and BAl-BEA catalysts
Fig. 6 Conversion of styrene oxide with TOS of the catalysts (reac-
tion conditions: T=200°C, P=1 atm, catalyst loading=0.5 g, flow-
rate of N2 =120 ml/min, WHSV=3h−1
)
noticeable differences with B-BEA, dominated by a broad
line which exhibits a lineshape which typical for second
order quadrupolar interaction of BO3. The chemical shift
around 10 ppm is assigned to BO3 sites in the framework
with three –B–O–Si– connectivity [23]. The broad peak
around 2.6 ppm was designated to BO4 in a distorted envi-
ronment or BO3 in separate amorphous phases. Therefore,
post-treatment of BEA boric acid can successfully incor-
porated boron into zeolite framework, resulting in the
variation of acid properties with the structural properties
untouched.
can be concluded that the weak acid sites exists in isolation
is not capable of achieving high activity. The deactivation
(denoted by the dashed line in Fig. 6) of Al-BEA sample
was observed after 20 h TOS, higher than 8 h of B-BEA,
which verifies the positive effects of strong acid sites on the
conversion of SO.
Upon boron modification, the initial conversion of BAl-
BEA remained unchanged relative to Al-BEA, but the cata-
lyst lifetime remarkably prolonged to 38 h.
Previous studies reported that condensation products
and coke were prioritized to form on strong acid sites.
NH3-TPD tests showed that Al-BEA exhibited too much
strong acid sites and the introduction of boron exerted
pronounced impacts on both the acid strength and con-
centration. The better stability performance of BAl-BEA
over PAl-BEA can be ascribed to the weak Brönsted acid
sites derived from –B–OH–Si– bands and Lewis acid sites
derived from BO3 species. Moreover, as measured by N2
adsorption, the surface area and volume of PAl-BEA pre-
sents an obvious decrease, which may narrow or block the
zeolite pores.
3.3 Isomerization of SO
According to our previous studies, the proper evaluation
condition was estimated at temperature of 200°C, pressure
of 1 atm, catalyst loading of 0.5 g, N2 flow rate of 120 ml/
min and WHSV of SO of 3 h−1 [18–20]. Under above con-
ditions, the reaction can reach a steady stage at 1 h of time
on stream (TOS). No remarkable SO uptake was observed
in the blank test without catalyst. And the initial conver-
sions (TOS=1 h) of SO over all samples were all higher
than 97.0%, indicating that this reaction was a catalysis
process. To obtain a quantitative evaluation of the catalytic
stability, the catalyst lifetime is defined as the time at which
the conversion reduces to 95.0%.
As depicted in Scheme 1, the acid-catalyzed reac-
tion is possibly to proceed via a carbenium-ion mecha-
nism. The isomerization is mainly catalyzed by the strong
Brønsted acid sites (i.e., Si–OH–Al), so we think that the
Lewis acid sites have little role in this catalytic system.
First, the oxygen atoms was protonated after the adsorp-
tion of SO on the acid sites. Then, the epoxy ring open-
ing was induced to acquire carbocation. After that, PhAA
was obtained by 1,2-shift of a hydrogen atom and pro-
ton loss. Different directions of ring cleavage leads to
different products. Theoretically, the neighboring aryl
group favors the formation of carbocation by stabilizing
The conversion of SO over each sample is presented in
Fig. 6. An initial conversion (1 h) of 97.4% was observed
on B-BEA sample, however, the conversion drops dramati-
cally below 95.0% within 8 h. All of other samples exhib-
ited initial conversions over 99.0% and lasted a longer time
along the TOS. From acidity characterization results, it is
found that B-BEA just had weak acid sites, while other
three samples possessed both weak and strong acid sites. It
1 3

Isomerization of Styrene Oxide to Phenyl Acetaldehyde over Different Modified Beta Zeolites
1529
Scheme 1 Mechanism of the
CH
CH2
CH
CH2
rearrangement of SO over beta
zeolite
CH
CH2
O
O
O
H
H
H
O
H
O
H
H
O
Si
Al
Si
Al
O
O
O
O
positive charge on α-carbon atom. The main by-products
in this reaction were dimer and trimer, and the distribu-
tion of main products were given in Fig. 7 (as a function
of TOS). The dimer was formed through aldol condensa-
tion of PhAA, while the trimer was derived from trimeri-
zation of PhAA.
As shown in Fig. 7b, the initial selectivity (1 h) to
PhAA over Al-BEA sample was just 76.4%. Interestingly,
the selectivity of PhAA increased gradually up to a maxi-
mum value of 99.5% and correspondingly the selectivity to
trimer decreased to 0. After the introduction of phospho-
rus, trimer formation was significantly suppressed. The
Fig. 7 Main products distribution with TOS on different samples: a B-BEA; b Al-BEA; c PAl-BEA; and d BAl-BEA (reaction conditions:
T=200°C, P=1 atm, catalyst loading=0.5 g, flow-rate of N2 =120 ml/min, WHSV=3 h−1
)
1 3

1530
X.-F. Zhang et al.
selectivity to PhAA over BAl-BEA sample towards drop-
ping from 99.1 to 97.2% and maintaining at this level along
with TOS. Generally, the trimer favors to form in the zeo-
lite surface, where adsorbed precursor molecules were dis-
tributed in adjacent external acid sites. External acid sites
located outside the channels were readily accessed by spe-
cies without spatial constraints, leading to undesired bulky
hydrocarbons. As the reaction proceeded, the external acid
sites were gradually deactivated and the trimer was only
detected in the early stage of the reaction. The selectivity to
dimer kept below 2.0% from beginning to end of the reac-
tion on all catalysts, suggesting that the aldol condensation
might be a spontaneous process controlled by thermody-
namics and is not sensitive to acidity variation. While some
other minor by-products, such as phenylethanol, phenyle-
thanediol, styrene and so on, were also detected by GC-MS
with total selectivity of less than 0.5%.
3.4 Deactivation Behavior
The investigation of deactivation behavior of the spent
catalysts after running for 10 h has a certain guiding sig-
nificance for the catalytic performance improvement.
There are mainly two mechanisms to explain the deactiva-
tion of Al containing zeolites: (1) the removal of Al spe-
cies from framework and (2) the active sites are poisoned
by carbonaceous deposits. The 27Al NMR spectra of the
fresh and spent catalysts are given in Fig. 9. For the sam-
ple of spent Al-BEA, the peak of tetrahedral AlFW species
became weaker and shifted to lower frequency, indicating
that the AlFW species were extracted irreversiblly from zeo-
lite framework. The spent PAl-BEA exhibits a broader peak
of AlFW species relative to fresh PAl-BEA, which could be
ascribed to coke species i nteracting with the framework
AlO4 groups. However, the peak for BAl-BEA just showed
negligible differences before and after reaction, suggesting
that its deactivation is mainly caused by carbon deposition
rather than the removal of AlFW species [18]. This implies
that BAl-BEA sample can be easily recycled and regener-
ated without a significant decrease of catalytic activity.
Considering that coke deposition is the main reason for
deactivation, weight loss of the spent catalysts was inves-
tigated by TG analysis (Fig. 10). The TG curves can be
divided into three stages according to temperature, corre-
sponding to the evaporation of adsorbed water and vola-
tile species (50–200°C), the desorption of soft coke spe-
cies (200–400°C), and the decomposition of hard coke
deposits (400–700°C), respectively. The weight loss of soft
coke was in well agreement with the number of weak acid
sites as determined by NH3-TPD. Al-BEA showed higher
amount of hard coke, implying that the strong acid sites
was a more significant contributor for catalyst deactiva-
tion [36, 37]. Upon boron modification, the weight loss in
400–700°C distinctly reduced from 7.2 to 1.3%. This may
be associated with the predominant moderate acidity of
BAl-BEA, which limit the secondary reactions.
FT-IR with collidine adsorption was used to explore the
amount of external acid sites (figures not shown here). The
results evidenced that external acid sites were substantially
deactivated by phosphorus or boron modification. To fur-
ther study the effects of external acid sites, the external
acid sites in Al-BEA was poisoned by triphenylphosphine
(TPP) and denoted as Al-BEA-TPP. Figure 8 displays the
results of activity tests in Al-BEA-TPP, the conversion of
SO maintains above 99.1% for 7 h and then drops to 93.7%
at 25 h. As expected, Al-BEA-TPP with trace amount of
external acid sites produced no trimer, verifying that the
trimerization only occurs at the external acid sites.
4 Conclusion
In the present study, a series of samples, e.g., Al-BEA,
B-BEA (without Al), BAl-BEA (boric acid treatment) and
PAl-BEA (phosphoric acid treatment) with the same BEA
framework type but different acid properties were synthe-
sized. They were characterized for physicochemical prop-
erties and then evaluated in the fixed-bed reactor under a
gas-phase atmosphere free of solvents. Partial or complete
substitution of Al atoms by other elements is verified to be
a good method to tune the acidity. High catalytic perfor-
mance was observed on all the samples, except B-BEA,
whose acidity is too weak. The deactivation of weakly
Fig. 8 Conversion of SO and selectivity to the products on Al-
BEA-TPP (reaction conditions: T=200°C, P=1 atm, flow-rate of
N2 =120 ml/min, WHSV=3 h−1
)
1 3

Isomerization of Styrene Oxide to Phenyl Acetaldehyde over Different Modified Beta Zeolites
1531
Fig. 9 27Al MAS NMR spectra of the samples before and after reaction: a Al-BEA; b BAl-BEA; c PAl-BEA
acidic B-BEA sample was observed only after 8 h TOS.
Other samples with both weak and strong acid sites revealed
a remarkable enhancement on stability. The rearrangement
of SO to PhAA is mainly influenced by the strong acid sites
whereas the lifetime of catalyst are depended on both acid
strength and concentration. The dimer formed by aldol con-
densation is independent of acidity properties, while the
trimerization mainly occurs at external acid sites without
much steric hindrance. The presence of excessive strong
acid sites or external acid sites was detrimental to the cata-
lytic performance. Owing to the reduction of the external
acid sites density by phosphorus modification, the selectiv-
ity to PhAA was improved by suppressing the side reac-
tion of trimerization. However, phosphorus treatment led
to significant destruction of microporous structures, thus
inhibiting products diffusion and accelerating the catalyst
deactivation. Introduction of boron caused a decrease of
strong acid strength and external acid concentration, while
the microporous structures of zeolite is almost not changed.
Fig. 10 TG curves of the Al-BEA and BAl-BEA samples after run-
ning for 10 h TOS
1 3

1532
X.-F. Zhang et al.
17. Gou ML, Wang R, Qiao Q, Yang X (2014) Micropor Mesopor
Mater 206:170–176
Therefore, BAl-BEA sample with suitable acid prosperities
is more suitable for this reaction.
18. Gou ML, Wang R, Qiao Q, Yang X (2014) Catal Commun
56:143–147
Acknowledgements The authors are grateful for the financial sup-
port of Natural Science Key Project of the Jiangsu Higher Education
Institutions (15KJA220001), Jiangsu Specially-Appointed Professor
Program and Priority Academic Program Development of Jiangsu
Higher Education Institutions (PAPD).
19. Trudu F, Tabacchi G, Gamba A, Fois E (2007) J Phys Chem A
111:11626–11637
20. Kalbasi RJ, Ghiaci M, Massah AR (2009) Appl Catal A 353:1–8
21. Forni L, Fornasari G, Trifiro F, Aloise A, Katovic A, Giordano
G, Nagy J (2007) Micropor Mesopor Mater 101:161–168
22. Tong J, Li W, Bo W, Wang H, Hu Y, Zhang Z, Mahboob A
(2016) J Catal 344:474–481
23. Duarte TA, Estrada AC, Simões MM, Santos IC, Cavaleiro AM,
Neves MGP, Cavaleiro JA (2015) Catal Sci Technol 5:351–363
24. Sasu A, Dragoi B, Ungureanu A, Royer S, Dumitriu E, Hulea V
(2016) Catal Sci Technol 6:468–478
References
1. Sheldon RA, van Bekkum H (2001) Fine chemicals through het-
erogeneous catalysis. Wiley-VCH, Weinheim, pp 217–231
2. Salla I, Bergadà O, Salagre P, Cesteros Y, Medina F, Sueiras JE,
Montanari T (2005) J Catal 232:239–245
25. Starzyk FT, Stan I, Abelló S, Bonilla A, Thomas K, Fernandez
C, Gilson JP, Ramírez JP (2009) J Catal 264:11–14
26. Barbera K, Sorensen S, Bordiga S, Skibsted J, Fordsmand H,
Beato P, Janssens TVW (2012) Catal Sci Technol 2:1196–1206
27. Water LGA, Waal JC, Jansen JC, Maschmeyer T (2004) J Catal
223:170–178
3. Smuda H, Hoelderich W, Goetz N, Recker HG (1990) Patent US
4929765
4. Zaccheria F, Psaro R, Ravasio N, Sordelli L, Santoro F (2011)
Catal Lett 141:587–591
28. Reddy JK, Motokura K, Koyama TR, Miyaji A, Baba T (2012) J
Catal 289:53–61
5. Smith K, El-Hiti GA, Al-Shamali M (2006) Catal Lett 109:77–82
6. Hoelderich W, Goetz N, Hupfer L, Kropp R, Theobald H, Wolf
B (1993) Patent US 5225602
29. Han OH, Kim CS, Hong SB (2002) Angew Chem Int Ed
41:469–472
7. Costa VV, Bhatia KA, Kozhevnikov IV, Gusevskaya EV (2010)
Appl Catal A 383:217–220
30. Gu X, Jiang T, Tao H, Zhou S, Liu X, Ren J, Wang Y, Lu G,
Schmidt W (2011) J Mater Chem 21:880–886
8. Bhatia KA, Eash KJ, Leonard NM, Oswald MC, Mohan RS
(2001) Tetrahedron Lett 42:8129–8132
31. Brickson JLE, Grammer GN (1958)
80:5466–5469
J Am Chem Soc
9. Bergadà O, Boix E, Salagre P, Cesteros Y, Medina F, Sueiras JE
(2009) Appl Catal A 368:163–169
32. Zhang D, Wang R, Yang X (2008) Catal Lett 124:384–391
33. Starzyk FT, Stan I, Abelló S, Bonilla A, Thomas K, Fernandez
C, Gilson JP, Pérez-Ramírez J (2009) J Catal 264:11–14
34. Serrano DP, Griekena RV, Melero JA, Garc A, Vargasa C (2010)
J Mol Catal A Chem 318:68–74
10. Gonzàlez MD, Cesteros Y, Salagre P, Medina F, Sueiras JE
(2009) Micropor Mesopor Mater 118:341–347
11. Kochkar H, Clacens JM, Figueras F (2002) Catal Lett 78:91–94
12. Robinson MWC, Pillinger KS, Graham AE (2006) Tetrahedron
Lett 47:5919–5921
35. Grieken RV, Serrano DP, Melero JA, Garc A (2005) J Catal
236:122–128
13. Gou ML, Wang R, Qiao Q, Yang X (2014) Appl Catal A 482:1–7
14. Davies TE, Kondrat SA, Nowicka E, Kean JL, Harris CM, Socci
JM, Apperley DC (2015) Appl Catal A 493:17–24
15. Tao L, Chen L, Yin SF, Luo SL, Ren YQ, Li WS, Zhou XP, Au
CT (2009) Appl Catal A 367:99–107
36. Holm MS, Svelle S, Joensen F, Beato P, Christensen CH, Bor-
diga S, Bjørgen M (2009) Appl Catal A 356:23–30
37. Hoang PH, Dien LQ (2015) Chem Eng J 262:140–145
16. Qiao Q, Wang R, Gou ML, Yang X (2014) Micropor Mesopor
Mater 195:250–257
1 3