Applied Catalysis A: General 417–418 (2012) 236–242

Contents lists available at SciVerse ScienceDirect

Applied Catalysis A: General

journal homepage: www.elsevier.com/locate/apcata

Cu-doped zeolites for catalytic oxidative carbonylation: The role of Brønsted acids
Shouying Huang, Yue Wang, Zengzhu Wang, Bing Yan, Shengping Wang, Jinlong Gong, Xinbin Ma ∗
Key Laboratory for Green Chemical Technology of Ministry of Education, School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, China

a r t i c l e i n f o a b s t r a c t

Article history: The aim of this work was to establish the role of Brønsted acidic sites in Cu-doped zeolites for the oxidative
Received 7 November 2011 carbonylation of ethanol to diethyl carbonate (DEC). In order to eliminate the influences of other factors
Received in revised form such as channel structure, faujasite (FAU) with different SiO2 /Al2 O3 ratios and Beta before and after passi-
20 December 2011
vation were investigated. Fourier Transform infrared spectroscopy (FTIR) of pyridine adsorption indicated
Accepted 28 December 2011
that all of Brønsted acidic sites were exchanged with CuCl to form effective active sites for oxidative car-
Available online 8 January 2012
bonylation. Characterization data showed that passivation on Beta zeolites has little effect on surface
area and pore volume distribution except Brønsted acidic sites. The relationship between Brønsted acidic
Keywords:
Oxidative carbonylation
sites and catalyst activity was built on different types of zeolite supports. Diffuse reflectance infrared
Cu-doped zeolite Fourier transform spectroscopy (DIRFTS) of CO adsorption revealed that the location of Cu active sites
Brønsted acids was related to the distribution of Brønsted acidic sites, which also influenced catalytic performance of
Dialkyl carbonate oxidative carbonylation.
© 2012 Elsevier B.V. All rights reserved.

1. Introduction an important role in catalysis, ion exchange, adsorption, and other

The rapid evolution of green chemistry has spurred tremendous
interests on clean and safe synthetic routes and thus development
of new catalysts. Production and applications of organic carbon-
ates (e.g., dialkyl carbonates) are directly concerned within the
frame of this principle. They are environmentally benign com-
pounds with versatile applications in chemical synthesis [1]. Due
to their low toxicity and bio-accumulation, they have been widely
used as solvents [2] and oxygen-containing fuel additives. Oxida-
tive carbonylation has been considered as one of most promising
approaches to replace the conventional phosgene method to pro-
duce dialkyl carbonates (e.g., dimethyl carbonate (DMC) [3–7]
and diethyl carbonate (DEC) [8–10]). Recently, DEC has been pro-
posed as an attractive oxygen-containing fuel additive to replace
methyl tert-butyl ether (MTBE) because of its high oxygen content
(40.6 wt%) and favorable gasoline/water distribution coefficients,
which are superior to alternatives such as DMC and ethanol. Thus,
it has attracted increased attention to develop a gas-phase process
for producing DEC (Eq. (1)).
2CH3 CH2 OH + CO + 1/2O2 = (CH3 CH2 O)2 CO + H2 O
practical processes [12–14]. Recently, Cu-doped zeolites prepared
by solid-state ion exchange (SSIE) with CuCl have received consid-
erable interests in catalysis, particularly for oxidative carbonylation
[15]. Since Cu-doped zeolites via a conventional ion exchange with
aqueous solutions of cupric ions usually contain coexistence of cop-
per ions in different aggregation and oxidation states, it is difficult
to elucidate the structural properties of Cu active sites. By con-
trast, catalysts prepared via the SSIE have been considered as model
solids, containing only isolated copper species in a single oxidation
state [16]. Additionally, high exchange degree can be achieved in a
single step via SSIE, and Cu ions can enter narrow pores and anchor
at cavities more easily relative to exchange in aqueous solution.
It has been generally recognized that the Brønsted acidic sites on
zeolite can be exchanged by Cu+ ions during the SSIE process (Eqs.
(2) and (3)) [17–19]. The formed surface-bound Cu+ was proposed
to be active species for the carbonylation reaction.
H Cu
O O
Al Si + CuCl Al + HCl
Si
(1) (2)

Zeolites have been extensively employed as catalytic materials OH OCu

for synthesis of a variety of chemicals due to their unique structural + CuCl Al + HCl
Al
properties and tunable acidity [11]. Bridging hydroxyl groups (e.g.,
(3)
Al OH Si) on zeolites that function as Brønsted acidic sites play
Therefore, the nature of Brønsted acidic sites in zeolites, which
depends on the bond angle of Si O Al [13], composition of the
∗ Corresponding author. framework (e.g., SiO2 /Al2 O3 ratio), and degree of dispersion of
E-mail addresses: xbma@tju.edu.cn, xbma@tju.edu.cn (X. Ma). the Al ions [20], could influence the exchange degree and local

0926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2011.12.043
 S. Huang et al. / Applied Catalysis A: General 417–418 (2012) 236–242
environment of Cu species in the catalysts. We note that the local
environment and concentration of Cu [21], residual Brønsted acid-
ity [22,23], and channel structure [24] of the zeolite supports are
all crucial factors to determine catalytic performance. Therefore,
architecture of channel and acidity are always considered in par-
allel. In order to solely determine the role of Brønsted acidity, it
is necessary to design experiments for eliminating the influences
of other factors such as channel structure. This paper, there-
fore, describes an investigation exploring the relationship between
Brønsted acidity of zeolite supports and catalytic performance for
oxidative carbonylation.
2. Experimental
2.1. Chemical reagents
Na-FAU zeolites with different SiO2 /Al2 O3 molar ratios were
purchased from XinNian Petrochemical Additives Company
(Shanghai, China) and Na-␤ was provided by NanKai Catalyst
(Tianjin, China). Reagents include cuprous chloride (AR, GuangFu
Chemical Research Institute), NH4 NO3 (≥99%, Standard Technology
Company, Tianjin, China), ethanol, tetraethyl orthosilicate, pyridine
and n-hexane (AR, Tianjin Kermel Chemical Co., Ltd.), potassium
bromide (Spectrosol, VWR International, Inc.), and nitrogen, oxy-
gen, carbon monoxide, helium (high purity, Tianjin Sixon Gas Co.,
Ltd.).
3. Catalyst preparation
The ammonium form of zeolites were obtained via exchanging
Na-zeolites (SiO2 /Al2 O3 = 2.5, 7.1, 10.1, 19) with a 0.5 M NH4 NO3
solution twice at 333 K and dried at 393 K for 4 h under vacuum.
The obtained NH4 -zeolites were calcined at 773 K for 3 h in air at
2 K/min. The mixture of protonated zeolites and fresh CuCl with a
mass ratio of 2 was exposed to N2 (99.999%) flowing at 60 ml/min,
and the temperature was ramped at 2 K/min to 773 K and kept for
6 h. The sample was then cooled down to room temperature slowly.
All samples were stored in a dry-box and then pressed and sieved
to 20–40 mesh before catalytic tests.
The Na␤ (protonated and passivated) samples were exchanged
with CuCl under the same conditions mentioned above and then
used for catalytic measurement.
3.1. Passivation with TEOS
Na␤ was fist protonated under the progress mentioned above.
Subsequently, TEOS (0.2 g/g of H␤ zeolite) was added to the sus-
pension of ␤ in n-hexane and the mixture was stirred vigorously
for 1.5 h and washed by deionized water and filtered repeatedly.
The sample was dried at 373 K for 5 h, and then calcinated in air at
823 K, followed by exchange with CuCl at the same procedure as
mentioned above.
3.2. Catalytic performance test
Measurements of catalytic activity and selectivity were car-
ried out in a computer-controlled continuous microreactor system
(WFS-3015) with a quartz tubular reactor of 4 mm inner diameter. A
back pressure regulator (26-1700 series, Tescom) was used to main-
tain the reaction pressure at 0.7 MPa. In a typical run, 1 ml (∼0.6 g)
of catalyst was loaded into the reactor. Before being exposed to
reactants, the catalyst was heated to 413 K at a ratio of 3 K/min in
a stream of N2 . Ethanol was fed into the main stream by a syringe
pump (Elite, P230) via an evaporator and permanent gases were
controlled using mass flow controllers. The feed was a gaseous
237
mixture of ethanol/CO/N2 /O2 with a molar ratio of 4.8/10/6.5/1
and a total flow rate was 89.2 ml/min. The reaction products were
directly analyzed by an on-line coupled gas chromatography (Agi-
lent 7890A GC), which was equipped with DB-624 capillary column
connected to a FID and two Porapack Q and one Molecular sieve
13X columns connected to a TCD detectors at 30 min intervals. The
byproducts observed including acetal and acetaldehyde (Eqs. (4)
and (5), respectively), and the selectivity of DEC was determined
by Eq. (6):
3CH3 CH2 OH + 1/2O2 = (CH3 CH2 O)2 CH2 CH3 + 2H2 O (4)
CH3 CH2 OH + 1/2O2 = CH3 CHO + H2 O (5)
2[CH3 CH2 OCOOCH2 CH3 ]
SDEC = (6)
[CH3 CHO] + 3[CH3 CH(OCH2 CH3 )2 ] + 2[CH3 CH2 OCOOCH2 CH3 ]
3.3. Catalytic characterization
FTIR spectra of H␤ and H␤-passivated using KBr dilution were
recorded on a Thermo Scientific Nicolet 6700 in the range of
4000–450 cm−1 (32 scans, 4 cm−1 ). For acidity determination, sam-
ples were pressed into 13 mm diameter self-supporting wafers and
placed into an in situ cell equipped with CaF2 windows. Before
absorbate exposure, all samples were evacuated at approximately
10−4 Pa and heated at 623 K for 30 min. After saturated by pyri-
dine at 423 K for 30 min, samples were outgassed to eliminate the
physically adsorbed pyridine. All scans were acquired at 423 K and
concentration of Brønsted and Lewis acid were calculated from the
integrated area of the ∼1540 and ∼1450 cm−1 , respectively. The
molar absorption coefficients were used as reported earlier [25].
The in situ DIRFTS of CO adsorption was carried out on the
same spectrometer using a high temperature/vacuum chamber
with ZnSe windows connected to a gas-dosing and evacuation sys-
tem. Before adsorbed CO, all samples were placed in the holder
and pretreated at 723 K for 30 min under a helium gas stream, and
then cooled down to 413 K (the optimized reaction condition). After
obtained the background spectrum, CO was adsorbed until satura-
tion followed by purging He gas for 30 min at the same temperature.
Four scans were averaged using a resolution of 4 cm−1 .
For acidity evaluation, NH3 -TPD measurements were run on a
Micromeritics Autochem II 2910 instrument equipped with a TCD.
The sample (∼50 mg) was pretreated at 473 K for 1 h in a flow of Ar
followed by NH3 adsorption at room temperature. Desorbed NH3
was monitored at a heating rate of 10 K/min from room tempera-
ture to 823 K.
Surface morphology and composition of catalysts was examined
by a Philips XL-30 scanning electron microscopy (SEM) equipped
with an energy dispersive X-ray spectrometer (EDX). The acceler-
ating voltage of the microscope was 20.0 kV.
X-ray diffraction patterns were recorded using Cu K␣ radiation
( = 1.54056 Å) on a Rigaku C/max-2500 diffractometer at 40 kV and
200 mA in the 2 range from 3◦ to 50◦ with a scanning rate of 8◦ /min.
Pore distribution and specific surface areas of the samples were
determined by nitrogen adsorption at 77 K on a Micromeritics
ASAP-2020 analyzer. Prior to the measurement, the samples were
degassed at 573 K for 16 h.
4. Results and discussion
FAU zeolites offer good activity for oxidative carbonylation as
supports among microporous materials, which possess character-
istics such as a wide range ratio of SiO2 /Al2 O3 , relative large pore
diameter, and unique 3-dimensional channel structure. Therefore,
to exclude the influence of crystal structure and geometry, we
first prepared Cu-doped FAU zeolites with varied SiO2 /Al2 O3 molar
ratios (i.e., tuning surface acidity) employed the SSIE method, on
238 S. Huang et al. / Applied Catalysis A: General 417–418 (2012) 236–242
0.1
SiO2/Al2O3=2.5
SiO2/Al2O3=7.1
Absorbance
SiO2/Al2O3=10.1
SiO2/Al2O3=19
1700 1600 1500
Wavenumber (cm )
Fig. 1. FTIR spectra of pyridine adsorbed on H-FAU (a) and Cu-FAU (b) with different SiO2 /Al2 O3 ratios.
which catalytic behaviors were examined for oxidative carbonyla-
tion of ethanol. Generally, zeolites with SiO2 /Al2 O3 ratios greater
than 20 possess the similar acidity [26], thus we chose SiO2 /Al2 O3
ratios of 2.5, 7.1, 10.1, and 19 for detailed examination. The effects
of pore distribution and specific surface area have been taken into
account. The results of N2 adsorption at low temperature (not pre-
sented) showed that pore width and pore volume calculated using
the Horvath–Kawazoe method varied in a small range and the trend
was not associated with the catalytic performance.
To investigate the nature of acid sites in zeolites, several basic
probe molecules with different sizes were employed such as
ammonia, d3 -acetonitrile, trimethylamine, and pyridine. Pyridine
was most used because of its sensitivity in IR spectrum. However,
not all of acidic sites are accessible to absorb pyridine considering
the steric hindrance. The framework structure of FAU zeolite con-
sists of supercage with 12-ring window (7.4 Å), sodalite cage with
6-ring (2.2–2.6 Å), and hexagonal prism, and the kinetics diame-
ter of pyridine is about 5 Å. It has been shown that pyridine can
only probe the acidic sites in supercages of FAU zeolites, while the
acid sites located in the sodalites and prisms units do not interact
with this probe molecule. On the other hand, FAU has three gen-
eral cation-exchange sites which can be occupied by Cu. Among
them, only those copper cations in supercages are accessible to
reactants, and hence contribute to the activity of catalyst in synthe-
sis of DMC [21]. For oxidative carbonylation of ethanol, the reactant
and product molecules are both relatively larger, so that geometric
constraints could be expected. Therefore, pyridine is an appropriate
probe molecule to detect the acidic properties of protonated FAU
in this work, although the number of Brønsted acidic sites deter-
mined in this manner is fewer than that expected from the Si/Al
ratio. Considering inaccessibility of Cu located in sodalite cage and
hexagonal prism for reactant molecule, it is more reasonable to
establish relationship between actual catalytic capability and the
concentration of Brønsted acidic sites rather than the Cu content
in Cu-doped zeolite. The similar phenomenon was also reported
on the relationship between zeolite acidity and activity in aro-
matic transformations [27]. The obtained spectra in the frequency
range of 1400–1700 cm−1 were shown in Fig. 1a. All the samples
revealed the bands due to pyridine adsorbed on strong Lewis acidic
sites (∼1450 and 1617 cm−1 ), Brønsted acidic sites (∼1540 and
∼1630 cm−1 ) and a band at 1490 cm−1 which can be assigned to
the contribution of pyridine associated with both Brønsted and
a 0.5 b
SiO2/Al2O3=2.5
SiO2/Al2O3=7.1
SiO2/Al2O3=10.1
SiO2/Al2O3=19
1400 1700 1600 1500 1400
-1
Lewis sites. The bands at ∼1450 cm−1 and ∼1540 cm−1 are often
used to determine the number of acid sites quantitatively, which
attributed to the C C stretching vibration of pyridine complex on
Lewis sites and pyridinium ion on Brønsted sites. With increasing
of SiO2 /Al2 O3 ratio, intensity of the band at 1540 cm−1 increased
first followed by a decrease. The NH3 -TPD profiles of protonated
FAU with different SiO2 /Al2 O3 ratios were shown in Fig. 2 and
the total peak areas were listed in Table 2. In aluminosilicate-type
zeolites, the substitution of Si atoms for Al atoms at tetrahedral
positions creates excess negative charge and then performs the
nature of acid sites, and hence the concentration of acid sites deter-
mined by NH3 -TCD signal varies linearly with the reduction of
SiO2 /Al2 O3 ratio. The trend is somewhat different from the result
from FTIR spectroscopy of pyridine adsorption due to inaccessibil-
ity of acidic sites in sodalite cages and hexagonal prism. Due to
the steric hindrance, the amounts of Brønsted acidic sites detected
by pyridine is less than the total quantity of Brønsted acidic sites
over HY. For synthesis of organic carbonates by oxidative car-
bonylation, the accessibility to active sites on various locations in
Cu-doped zeolite catalyst is also a crucial factor of consideration.
Upon proton exchange with Cu ions via the SSIE process, the band
at 1540 cm−1 and 1630 cm−1 (Fig. 1b) characteristic of Brønsted
sites disappeared completely. Bell and co-workers [17] reported
that complete exchange of Cu+ cations with Brønsted acid pro-
tons was achieved on ZSM-5 zeolite only when SSIE temperature
SiO2/Al2O3=2.5
TCD Signal (a.u.)
SiO2/Al2O3=7.1
SiO2/Al2O3=10.1
SiO2/Al2O3=19
100 200 300 400 500 600
Temperature (ºC)
Fig. 2. NH3 -TPD profiles of protonated H-FAU with different SiO2 /Al2 O3 ratios.
 S. Huang et al. / Applied Catalysis A: General 417–418 (2012) 236–242 239

Table 1
The chemical composition of Cu-FAU with different SiO2 /Al2 O3 ratios.

EDX results (% atomic)

SiO2 /Al2 O3 19 10.1 7.1 2.5

O 64.53 63.69 67.45 65.89 66.88 67.07 67.89 69.68

Al 1.75 1.74 1.84 1.99 5.60 5.36 6.54 6.50
Si 32.03 32.61 28.64 29.91 21.78 21.34 19.25 17.45
Cu 1.69 1.96 2.07 2.21 5.74 6.23 6.32 6.37
Cu/Al 0.97 1.13 1.13 1.11 1.03 1.16 0.97 0.98

Table 2
Amount of acid sites of FAU zeolites and apparent activation energies for oxidative carbonylation of ethanol over corresponding Cu-FAU.

Zeolite IR NH3 -TPD Peak area Apparent activation

energy (Ea ) (kJ/mol)

PyH+ (mmol/g)a PyL (mmol/g)b Total (mmol/g)

FAU(2.5) 4.37 4.34 8.71 423.5 51.6 ± 0.8

FAU(7.1) 5.26 5.16 10.42 359.0 47.6 ± 0.9
FAU(10.1) 6.75 4.28 11.03 336.9 42.3 ± 0.6
FAU(19) 3.36 5.33 8.69 130.4 61.4 ± 1.4
a
Pyridinium ion on Brønsted sites, calculated from the intensity of ∼1540 cm−1 band.
b
Pyridine complex on Lewis sites, calculated from the intensity of ∼1450 cm−1 band.

is over 1023 K and Cu/Al ratio reached approximately 1. Compara-
tively, Anderson et al. [23] considered that Cu-Y has a small amount
of residual acidity even when the Cu/Al ratio is 1. In our case,
the absence of 1540 cm−1 band indicated that complete exchange
between H proton in supercages of H-FAU and CuCl occurred at
823 K. EDX analysis was carried out to analyze the chemical com-
position of Cu-FAU catalyst. We put forward the results from two
different points on catalyst surface for reliability of data. As shown
in Table 1, Cu/Al molar ratios are between 0.97 and 1.13, which indi-
cated that bridging hydroxyl groups functioning as Brønsted acid
has been shown that the deposition of thin silica layers by CVD/CLD
from TEOS can neutralize external sites without modifying the
internal surface of zeolite [31]. For this reason, influence of Brønsted
acidity can be investigated by reducing Brønsted acidic sites. Upon
passivation, no obvious differences were observed in position and
intensity of the characteristic peaks in FTIR (Fig. 4) and XRD spectra
(Fig. 5), which reveals that zeolite framework keeps intact under the
treatment. A slight increase in the intensity of the peak at 2 = 7.6◦
Ethanol conversion,DEC yield (%)

was all occupied by Cu+ upon SSIE no matter where they located 2.5
(supercages, sodalite cage or hexagonal prism). It should be noted Ethanol conversion
that the content of Cu increased with the reduction of SiO2 /Al2 O3 , 2.0
which shows a good agreement with the acidity evaluation by NH3 -
TPD, but not FTIR spectroscopy of pyridine adsorption, since partial 1.5
Cu species enters into sodalite cage and hexagonal prism [21]. Con- DECyield
sidering accessibility and agglomeration of Cu species, it is not 1.0
reasonable to relate all exchanged-Cu sites to the formation of DEC.
Additionally, the band around 1440–1450 cm−1 became sharp and 0.5
a blue shift was observed upon the Cu exchange, indicating the
strong pyridine-Cu+ interaction, as also shown in previous inves-
0.0
tigations [28,29]. Thus, the intensities of 1450 cm−1 could reflect 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0
the degree of proton exchange for Cu+ . Concentration of Brønsted Amount of B acid sites(mmol/g)
and Lewis acidic sites obtained from the intensity of corresponding
bands is summarized in Table 2. Fig. 3. The relationship between Brønsted acidic sites on H-FAU and catalytic activ-
Based on the mechanism of CuCl exchange with H-zeolite ity of Cu-FAU catalysts.
reported earlier and the results of characterization presented
above, we presumed that it is more appropriate to establish a cor-
relation between Brønsted acidic sites detected by pyridine and
activity rather than Cu content in catalyst. Catalytic activities were 0.2
tested using Cu-FAU with various SiO2 /Al2 O3 ratios in a continu-
Absorbance

ous fixed bed microreactor. Under the same reaction conditions,

both ethanol conversion and DEC yield improved with increase of
Brønsted acid sites on the surfaces of zeolites (Fig. 3).
To further verify the role of Brønsted acidic sites of zeolite
supports in this reaction, we selectively passivated the external Hβ-pass ivated
surface acidity of zeolite ␤ by deposition of tetraethylorthosiloxane
(TEOS). The zeolite contains a mutually perpendicular intersecting Hβ
three-dimensional 12-ring channel system with pore dimensions of
5.6 × 5.6 Å and 6.6 × 7.6 Å, and thus pyridine can diffuse though the 3500 3000 2500 2000 1500 1000 500
structure to react with acidic sites. Beta zeolite possesses Brønsted Wavenumber(cm-1)
acidic sites on both external and internal surfaces, while Lewis
acidic sites are predominantly present on internal surface [30]. It Fig. 4. Comparison of FTIR spectra of H␤ zeolite before and after passivation.
240
Table 3
Amount of acid sites of H␤ and TEOS-treated H␤ and catalytic activity of Cu-␤ on oxidative carbonylation of ethanol.
Zeolite IR
PyH+ (mmol/g)
␤ 0.215
␤-Passivated 0.086
is due to the formation of Si O Si bonds by silanol groups react-
ing with adjacent ones. The similarity of N2 adsorption/desorption
curves of H␤ and H␤-passivated zeolites indicates that the passiva-
tion has little effect on surface area and pore volume distribution
(Fig. 6). These results suggest that TEOS was finely dispersed on
zeolite surfaces and formed silica layer was thin and uniform. Thus,
inner pore system could be still accessible for diffusion of reactants
and products after passivation [31]. Fig. 7 presents the pyridine
adsorbed IR spectra of untreated and passivated H␤ at 423 K. A dra-
matic decrease in band intensity at 1540 cm−1 and 1490 cm−1 was
observed, which correspond to Brønsted acidic sites and overlap of
both Brønsted and Lewis acidic sites, respectively. Comparatively,
the band at around 1450 cm−1 increased to a certain extent, which
might be ascribed to the formation of tricoordinated framework
silicon [32]. According to a study by Kunkeler et al. [31], a rough
external surface makes microcrystalline ␤ less favorable for the
formation of Si layer compared to macrocrystalline ␤. However, we
show here that the TEOS-treatment successfully reduced external
Brønsted surface acidity although the sample is microcrystalline
particle. Therefore, the change of acidity is consistent with the dis-
tribution of Brønsted and Lewis acidic sites on the H␤ zeolite [30].
The acid properties of H␤ before and after TEOS deposition and the
results of catalytic performance tests are shown in Table 3. Passiva-
tion resulted in a loss of ∼50% of ethanol conversion and DEC yield
Intensity
10 20 30
2θ (º)
Fig. 5. XRD patterns of H␤ zeolite before and after passivation.
0.07
0.06
dv/dp(cm3/g·Å)
0.05
0.04
0.03
0.02
0.01
0.00
0 20 40
dp (Å)
Fig. 6. Effect of passivation on H␤ pore volume distribution.
S. Huang et al. / Applied Catalysis A: General 417–418 (2012) 236–242
Ethanol conversion (%) DEC yield (%)
PyL (mmol/g)
3.248 1.71 0.95
8.926 0.93 0.50
compared to the sample without treatment. The variation was also
in line with the reduction of Brønsted acidic sites. It indicates that
Brønsted acidic sites on both internal and external surface of H␤
could form active sites for oxidative carbonylation.
From these results, it is apparent that Brønsted acidity of zeolite
supports, which presents in large cavities and accessible to pyri-
dine molecule absorbed, dominates the catalytic performance of
Cu-exchanged zeolites on oxidative carbonylation quantitatively.
Therefore, it is rational to calculate turnover frequencies (TOFs)
of DEC formation based on the amount of Brønsted acidic sites
determined using pyridine adsorption infrared spectroscopy.
The temperature dependence on reactivity in the form of the
Arrhenius behavior is shown in Fig. 8, which is comparable to that
of Cu-Y catalysts for the synthesis of DMC as reported previously
[21]. The calculated apparent activation energies (Ea ) (Table 2) of
the DEC formation were in a range of 30–60 kJ/mol, consistent with
those reported for relevant systems [6,33,34]. Under the same reac-
tion conditions, Ea decreased in the order of Cu-FAU (10.1) < Cu-FAU
(7.1) < Cu-FAU (2.5) < Cu-FAU (19), which follows the same trend as
order of amounts of Brønsted acidic sites on the surface of zeolites.
We should also mention that distribution and location of
Brønsted acidic sites have effects on the formation of metal active
sites and consequently the catalytic activity and stability in metal-
exchanged zeolite catalyst. CO prefers to interact with Cu(I) cations
at ambient temperature, but not adsorb on Cu(II) sites and form sta-
ble Cu–CO species [35]. In our case, Cu+ cations have been proved
to exchange with Brønsted acidic sites, therefore we used DIRFTS of
CO adsorption to discriminate the location and state of active sites
in different positions inside metal-exchanged zeolites. The FTIR
spectra of CO adsorbed on FAU zeolites with varying SiO2 /Al2 O3
were presented in Fig. 9. After purging with helium for 30 min, the
bands in the range of 2140 cm−1 to 2160 cm−1 did not vanish, which
demonstrated the presence of Cu(I)–CO species. CO adsorption
Hβ-passivated spectra of all samples lead to the appearance of two overlapping
band at ∼2162 cm−1 and 2148 cm−1 , which can be assigned to Cu+
Hβ cations positioned in sites II* and II of FAU, respectively [36]. The
relative intensity at the ∼2160 cm−1 band increased first and then
decreased with an increase of Si2 O3 /Al2 O3 ratio obviously. The
40 50
trend is in good agreement with the concentration of Brønsted
acidic sites and catalytic activity. It is very likely that, upon CO
0.1
Absorbance
Hβ
Hβ-passivated
Hβ
Hβ-passivated
60 80 100 1750 1700 1650 1600 1550 1500 1450 1400 1350
-1
Wavenumber (cm )
Fig. 7. FTIR spectra of pyridine adsorbed on H␤ before and after passivation.
 S. Huang et al. / Applied Catalysis A: General 417–418 (2012) 236–242 241

1.2E-3
1E-3
8E-4 Cu-Y(SiO2/Al 2O3=2.5) Cu-Y(SiO2/ Al 2O 3=7.1)
6E-4

4E-4

TOF (s-1)
2E-4
1.2E-3
1E-3 Cu-Y(SiO2/Al 2O3=10.1) Cu-Y(SiO2/ Al 2O 3=1 9)
8E-4
6E-4

4E-4

2E-4
2.35 2.40 2.45 2.50 2.55(2.35) 2.40 2.45 2.50 2.55
-1
1000/T (K )

Fig. 8. Arrhenius plot of turnover frequencies of DEC formation obtained over Cu-FAU catalysts.

active sites for the reaction of oxidative carbonylation. In addition,

2148
the distribution of Brønsted acidic sites also has an influence on
0.1 2162
the coordination and localization of Cu, and thus on catalytic per-
formance. Clarification of the effect of Brønsted acidity on solid
Absorbance

SiO /Al O =2.5

acid supports for the synthesis of Cu-based catalysts could also
help to understand mechanism of other relevant reactions, and pro-
SiO /Al O =7.1
vide important information for the design of catalytic materials in
general.
SiO /Al O =10.1

SiO /Al O =19

Acknowledgments

2250 2200 2150 2100 2050 Financial support from the National Natural Science Founda-
wavenumber (cm-1) tion of China (20876112, 20936003), Specialized Research Fund
for the Doctoral Program of Higher Education (SRFDP) (grant no.
Fig. 9. FTIR spectra of CO adsorption on Cu-FAU with varying Si2 O3 /Al2 O3 ratios. 20090032110021), the Program for New Century Excellent Tal-
ents in University (NCET-04-0242), Seed Foundation of Tianjin
University (60303002), and the Program of Introducing Talents of
adsorption, Cu+ migrates from sites II to more-exposed positions Discipline to Universities (B06006) are gratefully acknowledged.
(site II*) near the supercage [37]. Therefore, it indicated that the
more intensive distribution of Brønsted acidic sites in supercage
made Cu migration more easily, which was also relative to activ- References
ity of synthesis of diethyl carbonate. Furthermore, two shoulders
[1] A.A.G. Shaikh, S. Sivaram, Chem. Rev. 96 (1996) 951–976.
at 2111 cm−1 and 2190 cm−1 were also registered on the sample [2] I.T. Horváth, Green Chem. 10 (2008) 1024–1028.
with Si/Al ratio of 19. The band at 2111 cm−1 was also observed on [3] G. Rebmann, V. Keller, M.J. Ledoux, N. Keller, Green Chem. 10 (2008) 207–213.
[4] Y. Cao, J.C. Hu, P. Yang, W.L. Dai, K.N. Fan, Chem. Commun. (2003) 908–909.
partly reduced CuO samples [38]. Jana et al. assigned this band to
[5] S.T. King, Catal. Today 33 (1997) 173–182.
Cu+ species in oxide-like ambiance. Excess CuCl agglomerates to [6] Y.H. Zhang, I.J. Drake, D.N. Briggs, A.T. Bell, J. Catal. 244 (2006) 219–229.
form CuOx in FAU (19), due to its low concentration of Brønsted [7] J.L. Gong, X.B. Ma, S.P. Wang, Appl. Catal. A 316 (2007) 1–21.
acidic sites, which cannot offer enough exchangeable protons [36]. [8] P.B. Zhang, S.Y. Huang, Y. Yang, Q.S. Meng, S.P. Wang, X.B. Ma, Catal. Today 149
(2010) 202–206.
The results of CO adsorption indicated that the nature of Brønsted [9] D.N. Briggs, K.H. Lawrence, A.T. Bell, Appl. Catal. A 366 (2009) 71–83.
acidic sites influenced the state and location of Cu active sites and [10] D.N. Briggs, G. Bong, E. Leong, K. Oei, G. Lestari, A.T. Bell, J. Catal. 276 (2010)
thus catalytic activity. 215–228.
[11] S. Telalović, J.F. Ng, R. Maheswari, A. Ramanathan, G.K. Chuah, U. Hanefeld,
Chem. Commun. 47 (2008) 4631–4633.
[12] Y.J. Gong, T. Dou, S.J. Kang, Q. Li, Y.F. Hu, Fuel. Process. Technol. 90 (2009)
5. Conclusions
122–129.
[13] N. Katada, K. Suzuki, T. Noda, G. Sastre, M. Niwa, J. Phys. Chem. C 113 (2009)
In summary, we have established a quantitative relationship 19208–19217.
[14] H. Huo, L. Peng, C.P. Grey, J. Phys. Chem. C 115 (2011) 2030–2037.
between amount of Brønsted acidic sites and catalytic activities
[15] N. Keller, G. Rebmann, V. Keller, J. Mol. Catal. A: Chem. 317 (2010) 1–18.
for oxidative carbonylation. Considering aggregation and accessi- [16] C. Lamberti, A. Zecchina, E. Groppo, S. Bordiga, Chem. Soc. Rev. 39 (2010)
bility of Cu species on zeolites, it is more reasonable to relate the 4951–5001.
Brønsted acidic sites to catalytic activity rather than Cu contents. [17] I.J. Drake, Y.H. Zhang, A.T. Bell, Chem. Mater. 18 (2006) 2347–2356.
[18] Z. Li, K.C. Xie, R.C.T. Slade, Appl. Catal. A 209 (2001) 107–115.
Brønsted acidic sites on both internal surface and external surface [19] A.Z.G. Spoto, S. Bordiga, G. Ricchiardi, G. Martra, Appl. Catal. B 3 (1994) 151–172.
of zeolite, which exchange with Cu+ during the SSIE process, form [20] Y. Kuroda, T. Mori, Y. Yoshikawa, Chem. Commun. (2001) 1006–1007.
242 S. Huang et al. / Applied Catalysis A: General 417–418 (2012) 236–242
[21] Y.H. Zhang, I.J. Drake, D. Briggs, B. Lim, T. Chau, T. Alexis Bell, J. Phys. Chem. B
110 (2006) 11654–11664.
[22] T. Blasco, M. Boronat, P. Concepción, A. Corma, D. Law, J.A. Vidal-Moya, Angew.
Chem. Int. Ed. 46 (2007) 3938–3941.
[23] S. Anderson, S. Manthata, T. Root, Appl. Catal. A 280 (2005) 117–124.
[24] Y.H. Zhang, D.N. Briggs, E. Desmit, A.T. Bell, J. Catal. 251 (2007) 443–452.
[25] C.A. Emeis, J. Catal. 141 (1993) 347–354.
[26] G. Ertl, H. Knözinger, F. Schüth, J. Weitkamp, Handbook of Heterogeneous Catal-
ysis, Wiley, 2008, 1103.
[27] N. Žilková, M. Bejblová, B. Gil, S.I. Zones, A.W. Burton, C.Y. Chen, Z. Musilová-
Pavlačková, G. Košová, J. Čejka, J. Catal. 266 (2009) 79–91.
[28] M. Jiang, H.G. Karge, J. Chem. Soc. Perkin Trans. 91 (1995) 1845–1851.
[29] F. Benaliouche, Y. Boucheffa, P. Ayrault, S. Mignard, P. Magnoux, Microporous
Mesoporous Mater. 111 (2008) 80–88.
[30] J.C. Jacobus, E.J. Creyghtonb, S.L. Njoa, H.V. Koningsveld, H.V. Bekkum, Catal.
Today 38 (1997) 205–212.
[31] P.J. Kunkeler, D. Moeskops, H.V. Bekkum, Microporous Mesoporous Mater. 11
(1997) 313–323.
[32] J. Cejka, A. Corma, S. Zones, Zeolites and catalysis, Synthesis, Reactions and
Applications, Wiley, 2010, 527.
[33] I.J. Drake, K.L. Fujdala, A.T. Bell, T.D. Tilley, J. Catal. 230 (2005) 14–27.
[34] X.B. Zheng, J. Phys. Chem. C 112 (2008) 5043–5047.
[35] K. Hadjiivanov, H. Knözinger, Phys. Chem. Chem. Phys. 3 (2001) 1132–1137.
[36] J. Engeldinger, C. Domke, M. Richter, U. Bentrup, Appl. Catal. A 382 (2010)
303–311.
[37] G.T. Palomino, S. Bordiga, A. Zecchina, G.L. Marra, C. Lamberti, J. Phys. Chem. B
104 (2000) 8641–8651.
[38] B. Guido, J. Mol. Catal. 43 (1987) 225–236.