Conversion of 2,3-butanediol to butenes over bifunctional catalysts in a single reactor

Article abstract of DOI:10.1016/j.jcat.2015.07.004

Abstract In this study, Cu/ZSM-5 catalysts were used to catalyze the hydrodeoxygenation of 2,3-butanediol to butenes in a single reactor in the presence of hydrogen. The carbon selectivity of butenes increased with increasing SiO₂/Al₂O₃ ratio (lowering acidity of zeolite) and H₂/2,3-butanediol ratio. Cu/ZSM-5 with a SiO₂/Al₂O₃ ratio of 280 showed the best activity toward the production of butenes; however, Cu/ZSM-5 with SiO₂/Al₂O₃ of 23 was favorable for the dehydrogenation reaction of 2,3-butanediol even in the excess of hydrogen. On zeolite ZSM-5(280), the carbon selectivity of butenes increased with increasing copper loading and 19.2 wt% of CuO showed the highest selectivity of butenes (maximum 71%). The optimal reaction temperature is around 250 °C. Experiments demonstrated that methyl ethyl ketone (MEK) and 2-methylpropanal are the intermediates in the conversion of 2,3-butanediol to butenes. The optimal performance toward the production of butene is the result of a balance between copper and acid catalytic functions.

Full text of DOI:10.1016/j.jcat.2015.07.004

Journal of Catalysis 330 (2015) 222–237
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Conversion of 2,3-butanediol to butenes over bifunctional catalysts
in a single reactor
a
a
a
a
b
Quanxing Zheng , Michael D. Wales , Michael G. Heidlage , Mary Rezac , Hongwang Wang ,
b
a,
⇑
Stefan H. Bossmann , Keith L. Hohn
a
Department of Chemical Engineering, Kansas State University, 1005 Durland Hall, Manhattan, KS 66506, United States
Department of Chemistry, Kansas State University, 213 CBC Building, Manhattan, KS 66506, United States
b
a r t i c l e i n f o
a b s t r a c t
Article history:
In this study, Cu/ZSM-5 catalysts were used to catalyze the hydrodeoxygenation of 2,3-butanediol to
butenes in a single reactor in the presence of hydrogen. The carbon selectivity of butenes increased with
Received 11 May 2015
Revised 30 June 2015
Accepted 2 July 2015
increasing SiO
SiO /Al ratio of 280 showed the best activity toward the production of butenes; however,
Cu/ZSM-5 with SiO /Al of 23 was favorable for the dehydrogenation reaction of 2,3-butanediol even
2 2 3 2
/Al O ratio (lowering acidity of zeolite) and H /2,3-butanediol ratio. Cu/ZSM-5 with a
2
2 3
O
2
2 3
O
in the excess of hydrogen. On zeolite ZSM-5(280), the carbon selectivity of butenes increased with
increasing copper loading and 19.2 wt% of CuO showed the highest selectivity of butenes (maximum
Keywords:
2
,3-Butanediol
71%). The optimal reaction temperature is around 250 °C. Experiments demonstrated that methyl ethyl
Cu/ZSM-5
Butene
Hydrogenation
Hydrodeoxygenation
Dehydration
ketone (MEK) and 2-methylpropanal are the intermediates in the conversion of 2,3-butanediol to
butenes. The optimal performance toward the production of butene is the result of a balance between
copper and acid catalytic functions.
Ó 2015 Elsevier Inc. All rights reserved.
Dehydrogenation
1
. Introduction
Because petroleum is a finite resource, there is growing interest
2,3-BDO concentration of 95.5 g/L with productivity of 1.71 g/L/h
by fermentation of media containing 200 g/L of glucose in a 3 L
batch fermentor using Klebsiella oxytoca. Jansen et al. [14]
obtained a final concentration of 2,3-BDO of 12.63 g/L by fermen-
tation of media containing 50 g/L of xylose in a 7-L batch fermentor
using Klebsiella oxytoca ATCC 8724. Saha and Bothast [13]
obtained a yield of 2,3-BDO of 0.4 g/g arabinose with a correspond-
ing productivity of 0.63 g/L/h by fermentation of media with an ini-
tial arabinose concentration of 50 g/L.
in processes that produce hydrocarbons from renewable resources
for use as fuels [1–4]. Such processes could produce essentially the
same product currently made in petroleum refineries, eliminating
the need for modifications to vehicles or to the hydrocarbon
distribution infrastructure. A number of routes for producing
hydrocarbons from sustainable resources have been proposed
[
1,5,6]. These routes convert biomass-derived sugars to oxygenated
Once 2,3-butanediol (2,3-BDO) is produced via fermentation,
routes to convert it to hydrocarbons would be needed. However,
there is little research in this area. Dehydration of 2,3-BDO to
methyl ethyl ketone (MEK) has been well studied, and occurs read-
ily over a number of catalysts [10,15]. Zhang et al. [16] investigated
the dehydration of 2,3-BDO over zeolite HZSM-5 and HZSM-5
modified with boric acid, and studied the effect of framework
Si/Al ratio and addition of boric acid on 2,3-BDO dehydration.
They reported that high Si/Al ratio was beneficial to low-
temperature activation of 2,3-BDO and the methyl migration to
2-methylpropanal, and the addition of boric acid enhanced the
catalytic stability. Lee et al. [17] utilized in situ DRIFTS to investi-
gate the dehydration of 2,3-BDO over a series of zeolites ZSM-5,
mordenite, b- and Y-type zeolites, and found that dehydration of
2,3-BDO to MEK was favored on ZSM-5. Duan et al. [18] investi-
intermediates, which are upgraded to fuel-range hydrocarbons. The
key to developing a successful process of this nature is to select
intermediate compounds that can be selectively produced from
sugars, and can easily be converted to fuel-range hydrocarbons.
A potential intermediate compound that has not previously
been considered for the production of hydrocarbons from
biomass-derived sugars is 2,3-butanediol (2,3-BDO). 2,3-BDO is
an intriguing intermediate because it can be produced via fermen-
tation of sugars with a high productivity at high concentration by
using a variety of microorganisms, such as Klebsiella oxytoca
[
7–10], Enterobacter aerogenes [11], Bacillus licheniformis [12]
and Enterobacter cloacae [13]. Ji et al. [7] obtained a maximum
⇑
Corresponding author. Fax: +1 785 532 7372.
E-mail address: hohn@ksu.edu (K.L. Hohn).
2
gated the dehydration of 2,3-BDO over monoclinic ZrO and the
http://dx.doi.org/10.1016/j.jcat.2015.07.004
021-9517/Ó 2015 Elsevier Inc. All rights reserved.
0

Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
223
result showed that 3-buten-2-ol was produced with a maximum
selectivity of 59.0% along with major byproducts such as MEK
and 3-hydroxy-2-butanone. In addition, further dehydration of
zeolites with high copper dispersion [33,34]. The catalysts used in
this work were synthesized by the ion exchange method as fol-
lows, which is similar to the deposition precipitation (DP) method
[35,36] or ammonia evaporation (AE) methods [37,38]. First, the
ammonium-type ZSM-5 was calcined at 550 °C for 4 h to convert
2
,3-BDO yields 1,3-butadiene [19], which can be dimerized to pro-
duce the aromatic intermediate styrene (Diels–Alder reaction) [20]
and hydrogenated to butene. Sato and coworkers investigated
it to HZSM-5. Then the desired amount of Cu(NO
solved in 100 mL of deionized water at room temperature.
Ammonia was added to the solution until the pH was about 9.1
3
)
2
ꢀ3H
2
O was dis-
dehydration of 2,3-BDO to 1,3-butadiene over Sc
dehydration of other diols, such as 1,3-butanediol and
,4-butanediol over rare earth oxides [22,23], ZrO [24]
and Cu-based catalysts [25].
2
O
3
[21], and
2+
1
2
3 4 2 2
to form a dark blue cupric ammine complex [Cu(NH ) (H O) ] ,
and then water was added to make 250 mL of a copper-ammonia
complex solution. 20 g of HZSM-5 zeolite was added to the solu-
tion and then the container was capped to avoid the evaporation
of ammonia and stirred for 4 h at room temperature. After that,
the container was transferred to an oil bath and heated to 60 °C
for 2 h. Then the solution was filtered and the precipitate was
washed at least five times by water and dried at 110 °C overnight
followed by calcination at 550 °C for 4 h. Finally, the calcined cat-
alyst was pelletized, crushed and sieved to obtain a particle size
distribution in the range 40–60 mesh. To make 10%CuO/ZSM-5,
The approach reported here is to convert 2,3-BDO to butene,
which is a basic building block of fuels as well as many chemicals.
As a precursor, butene can be converted to a variety of oligomers
(dimer, trimer, tetramer, etc.) [26,27], which can be further converted
to saturated hydrocarbons through hydrogenation reaction. In this
way, butene can serve as an intermediate to produce high-grade liq-
uid fuel with specific type of saturated hydrocarbons [6,27].
The major challenge is to remove the two hydroxyl groups of
2
,3-butanediol (2,3-BDO) in a single step to produce butene. This
process involves a bifunctional pathway, in which 2,3-butanediol
is dehydrated on an acid site to methyl ethyl ketone (MEK),
the amount of Cu(NO
The content of CuO was determined to be 9.5 wt%, 9.7 wt% and
9.2 wt% on ZSM-5 with SiO /Al ratios of 23, 50 and 280, respec-
3
)
2
ꢀ3H
2
O added to the solution was 24.16 g.
2
-methylpropanal, and butadiene, which can be further hydro-
2
2 3
O
genated to butene on the metal sites. Copper is interesting for
use as the hydrogenation catalyst. Cu-containing catalysts show
high activity for vapor-phase hydrogenation reaction particularly
the selective hydrogenation of carbon–oxygen bonds; however,
copper catalysts are relatively inactive for hydrogenolysis of car-
bon–carbon bonds [28]. Guo et al. [29] investigated hydrogenolysis
of glycerol to propanediols over Cu catalysts, and found that
tively, by the inductively coupled plasma (ICP) method. Two
Cu/ZSM-5(280) catalysts with high loading of CuO were prepared
by increasing the amount of Cu(NO
)
3 2
2
ꢀ3H O to 36.24 g, and extend-
ing the time of ion exchange in the oil bath to 12 h and 24 h. For
these two catalysts, the content of CuO was determined by ICP to
be 19.2 wt% and 29.1 wt%, respectively. For the catalyst
Cu/ZSM-5(280) with low loading of CuO, the amount of
c
-Al
2
O
3
supported Cu catalysts showed excellent performance
Cu(NO
3
)
2
ꢀ3H
2
O was decreased to 12.0 g and the time of ion
(
selectivity to propanediol, 96.8%) and successfully suppressed
exchange was shortened to about 1 h in the oil bath, and the
content of CuO was determined by ICP to be 6.0 wt%.
the scission of CAC bonds. Sitthisa and Resasco [30] investigated
the hydrodeoxygenation of furfural over Cu, Pd and Ni supported
on SiO
2
, and found that the Cu catalyst mainly produced furfuryl
2.3. Catalytic reactions
alcohol via hydrogenation of the carbonyl group due to the weak
interaction of Cu with C@C. Vasiliadou et al. [31] investigated the
hydrogenolysis of glycerol to propylene glycol over highly
The catalytic reactions were performed in a conventional con-
tinuous flow fixed-bed reactor made of stainless steel (id = 8 mm)
under atmospheric pressure. Prior to reaction, the catalyst sample
2
dispersed Cu/SiO catalyst. The result showed that Cu selectively
converted glycerol to propylene glycol with selectivity of 92–97%
via consecutive dehydration–hydrogenation reactions. Sato et al.
(weight = 1.0 g) was reduced in the reactor in the H
rate of H /N = 1/5) at 300 °C for 2 h. The H flow of 24 cm /min
(standard ambient temperature and pressure, SATP) and the N
2 2
/N flow (flow
3
2
2
2
[
32] reported that reduced Cu catalyst could effectively catalyze
the dehydration of glycerol to hydroxyacetone in N , and the
hydrogenation of hydroxyacetone followed by hydrogenolysis in
2
3
2
flow of 120 cm /min (SATP) were controlled with mass-flow con-
trollers (Brooks). 2,3-BDO was fed via a micropump (Eldex 1SMP)
3
H
2
to form ethylene glycol, acetaldehyde and ethanol.
at 3 mL/h together with a H
2
flow of 67.2 cm /min (SATP) and N
2
3
Based on the excellent hydrogenation performance of copper,
flow of 15.4 cm /min (SATP). Reactor temperature was set between
200 and 300 °C. Product compositions were analyzed by an on-line
gas chromatograph (SRI 8610C) equipped with an MXT-1 column
we have studied a high copper loading catalyst supported on
ZSM-5 to convert 2,3-butanediol to butenes in a single reactor.
The impact of reaction conditions (temperature and hydrogen to
(nonpolar phase, 60 m, ID 0.25 mm, film thickness 0.25 lm), TCD
2
,3-butanediol ratio) and the Si/Al ratio of the ZSM-5 catalyst are
and FID detectors for the analysis of hydrocarbons and oxygenated
chemicals, and quantified by injecting calibration standards to the
GC system. The temperature of the tubing from the bottom of
the reactor to the inlet of GC was maintained at 230 °C to avoid
the condensation of liquid products. The products were injected
through the sample loop (0.2 mL), which was controlled by a high
temperature ten-port valve. The oven was kept at 40 °C for 5 min,
and then raised to 120 °C at a ramp rate of 40 °C/min, finally raised
to 250 °C at a rate of 20 °C/min, and held at this temperature for
10 min. As MXT-1 column is not capable of separating some
hydrocarbons, such as 1-butene and isobutene, to determine the
distribution of butenes (1-butene, isobutene, trans-2-butene and
reported and it is demonstrated for the first time that 2,3-BDO
can be converted to butenes in a single reactor at a high yield.
2
. Experimental
2.1. Materials
2 2 3
Ammonium-type ZSM-5 with SiO /Al O ratios of 23, 50 and
2
80 was obtained from Zeolyst International. ZSM-5 is referred
/Al ratio. 2,3-butanediol
>97%) was purchased from TCI America. Cu(NO O (99%)
ꢀ3H
was purchased from Fisher scientific.
to as ZSM-5(n), where n is the SiO
(
2
2 3
O
3
)
2
2
2 2 3
cis-2-butene) over catalysts with different SiO /Al O ratios, addi-
tional experiments were performed where the MXT-1 column was
replaced with an MXT-Alumina BOND/MAPD column (30 m, ID
2.2. Catalyst preparation
0
.53 mm, film thickness 10 lm), which is capable of separating
As previously reported, the ion exchange of zeolite ZSM-5 with
Cu(II) in ammonia could result in excessively exchanged copper on
the four isomers of butenes. To ensure the identification of prod-
ucts, GC–MS analyses were also carried out by using an Agilent

2
24
Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
7
890A GC system equipped with an Agilent 5975C MS detector and
2.4.5. Scanning electron microscopy (SEM)
HP-1 capillary column. The carbon selectivity [39,40] and conver-
sion of 2,3-BDO were calculated in the following methods.
Scanning electron microscopy (SEM) experiments were per-
formed on a FEI Nova NanoSEM 430 to identify the morphology
of Cu/ZSM-5 catalysts.
Moles of carbon in specific product
Carbon selectivity ¼
Total carbon atoms in identified products
2
.4.6. X-ray photoelectron spectroscopy (XPS)
XPS data were obtained with a PerkinElmer PHI 5400 using
ꢁ
100%
achromatic Al Ka radiation (1486.60 eV). The pressure in the anal-
ðmoles of 2; 3-BDOÞ ꢂ ðmoles of 2; 3-BDOÞ
in
out
Conversion ¼
ꢂ8
ysis chamber was typically 8.0 ꢁ 10 Torr. Cu 2p3/2 (932.7 eV) and
Au 4f7/2 (84.0 eV) were used as standards to calibrate the binding
energy (BE) range, and the binding energy of carbon 284.6 eV
was used as the BE standard to correct for charging on the sub-
strate [42]. XPS spectra were curve fitted by the software
CasaXPS based on the centered position, full width at half maxi-
mum (FWHM) and peak intensity.
ðmoles of 2; 3-butanediolÞ
in
ꢁ
100%
Two repeat runs were performed at each reaction condition and the
two trials were generally within 5% of each other. The relative dif-
ference between the sum of butene selectivity from MXT-1 column
and MXT-Alumina BOND/MAPD column is about 5–6%. The carbon
balances closed with above 90% for all runs in this paper.
2
.4.7. N
Copper surface area (SACu) and dispersion (DCu) were deter-
mined by dissociative N O adsorption method at 90 °C [43,44]
using the same system as H -TPR and NH -TPD. Prior to N
2
O adsorption
2
2
.4. Catalyst characterization
2
.4.1. NH
The surface acidity of catalysts was investigated by temperature
programmed desorption of ammonia (NH -TPD). NH -TPD was
carried out in an Altamira AMI-200 system. Prior to adsorption,
.2 g of copper catalyst was loaded in a quartz U-tube reactor
and pre-treated at 550 °C in helium for 1 h followed by cooling
to 100 °C. Then the catalyst was reduced in a flow of H /Ar
10 v/v%) at a constant rate of 10 °C/min to 300 °C and then main-
3 3
temperature-programed desorption (NH -TPD)
2
3
2
O
adsorption, the catalysts (0.1 g) were first treated at 550 °C in Ar
at a flow of 40 mL/min for an hour. After cooling, the temperature
was ramped from 50 °C to 400 °C at a ramp rate of 5 °C/min in
3
3
0
H
2
/Ar flow (10 v/v%, 40 mL/min) by the
described above and decreased to 90 °C in Ar (99.999%,
0 mL/min). In this step, the amount of hydrogen consumption
was denoted as X. Then the pre-reduced catalysts were exposed
to N O/He (5 v/v%, 40 mL/min) isothermally at 90 °C for 1 h to oxi-
dize surface copper atoms to Cu O followed by cooling to 50 °C in
Ar (30 mL/min). After this process, the second H -TPR was carried
out on the freshly oxidized catalysts from 50 °C to 400 °C at a ramp
rate of 5 °C/min in H /Ar flow (10 v/v%, 40 mL/min) in order to
reduce Cu O back to metallic Cu. The hydrogen consumption in
2
H -TPR procedure
2
3
(
tained for 2 h followed by cooling to 100 °C. 10 mL/min of ammo-
nia (anhydrous, 99.99%) was then introduced at 100 °C for 30 min.
2
Physisorbed NH
at 100 °C for 2 h. Finally, the temperature was raised to 700 °C at
0 °C/min. Desorbed ammonia was detected with a thermal con-
ductivity detector (TCD).
3
molecules were removed by flowing pure helium
2
2
1
2
2
2
.4.2. H
2
temperature-programed reduction (H
2
-TPR)
this step was denoted as Y. Copper dispersion (DCu) is calculated
as DCu = 2Y/X, which is defined as the ratio of the surface copper
atoms to the total copper atoms present in the catalyst. Copper sur-
H
2
-TPR was measured in the same system as NH -TPD. 0.1 g of
3
sample was loaded in a quartz U-tube reactor and treated at 550 °C
in Ar (99.999%) at a flow of 40 mL/min for an hour. After cooling,
the temperature was ramped from 50 °C to 900 °C at a ramp rate
2
face area SACu (m /gCu) is calculated as described in literatures
[
43,44] by the following equation:
2 2
of 5 °C/min in H /Ar flow (10 v/v%, 40 mL/min). H consumption
was recorded by a thermal conductivity detector (TCD). H
sumption was calibrated by reducing 0.03 g of pure CuO.
2
con-
2 ꢀ Y ꢀ Nav
2
SACu
¼
ꢃ 1353 Y=X ðm =gCuÞ;
X ꢀ M ꢀ SDCu
Cu
2
3
2
.4.3. Brunauer–Emmett–Teller (BET)
Surface area measurements of catalysts were conducted accord-
where Nav = Avogadro’s constant = 6.02 ꢁ 10 atoms/mol, MCu = atomic
weight of Cu = 63.546 g/mol, SDCu = copper surface density = 1.47 ꢁ
19
2
ing to the Brunauer–Emmett–Teller (BET) gas (nitrogen) adsorp-
tion method. About 0.06 g of catalyst powder was poured into
the sample cell and degassed at 350 °C for 4 h before determining
the exact mass of the sample, which was then confirmed after
degassing. The adsorption/desorption isotherms were measured
using a Quantachrome Autosorb-1 instrument at ꢂ196 °C and
analyzed with Autosorb-1 software. The total surface area was
10 atoms/m (the average value for Cu(111), Cu(110), and Cu(100)
crystal surfaces).
2.4.8. Thermogravimetric analysis (TGA)
Thermogravimetric analysis (TGA) was performed with a ther-
mogravimetric analyzer (Shimadzu TGA-50). The used catalyst
samples were collected after a specific reaction time (40 min or
determined from N
2
adsorption branch in the linear range of
2
80 min). After the feed of 2,3-BDO was stopped, the flow of
relative pressure from 0.007 to 0.03. The micropore surface area
and micropore volume were evaluated by the t-plot method [41].
The total pore volume was evaluated by single point pore volume
at a relative pressure of 0.95.
hydrogen and nitrogen (the same flow rate as reaction) was main-
tained for about 30 min to remove residual 2,3-BDO in the reactor
and the products adsorbed on the catalysts as well. Catalyst was
recovered after cooling to room temperature. Prior to TGA analy-
sis, the used catalysts were kept in the oven at 100 °C for 3 days
so that the copper in catalysts could be oxidized completely.
Typically, about 20 mg of the used catalyst sample was heated
in air (air flow: 10 mL/min) from room temperature to 600 °C at
a ramp rate of 10 °C/min. The coke content for each sample was
then determined from the weight loss between 300 °C and
600 °C [45].
2.4.4. X-ray diffraction (XRD)
XRD analysis was conducted using a Rigaku Miniflex II desktop
X-ray diffractometer. Scans of two theta angles were from 5° to 90°
for all catalysts with a step size of 0.02° and scan speed of
0
to provide a significant number of oriented particles to fulfill the
Bragg condition of reflection.
.75°/min. All samples were well ground before analysis in order

Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
225
3
. Results and discussion
2
08.7
3.1. Characterization of catalysts
3.1.1. X-ray diffraction
Fig. 1 shows the XRD patterns of the Cu/ZSM-5 catalysts with
different SiO /Al ratios and those of the parent HZSM-5 calcined
2
2 3
O
at 550 °C, and Figs. S1 and S2 display the XRD patterns and relative
crystallinity, respectively, of the calcined catalysts with different
CuO loadings on zeolite HZSM-5(280). As seen in these figures,
although a slight decrease in the intensity of the main peaks was
noticed after the introduction of copper, all the characteristic peaks
of the parent HZSM-5 were observed in Cu/ZSM-5 catalysts, which
indicated that the introduction of copper did not destroy the struc-
ture of the parent HZSM-5. In addition, two characteristic peaks
related to CuO (35.7° and 38.55°) were only observed on
2
11.4
(c)
b)
a)
(
2
73.2
(
100
200
300
400
500
600
o
Temperature ( C)
9
.5%CuO/ZSM-5(23), which indicated that Cu species were well
dispersed on HZSM-5(50) and HZSM-5(280). This is in accordance
with the SEM result (see Fig. S3), copper clusters were observed
Fig. 2. H
2
-TPR profiles of calcined Cu/ZSM-5 with different SiO
2
/Al
2
O
3
ratios. (a)
9.5%CuO/ZSM-5(23), (b) 9.7%CuO/ZSM-5(50), and (c) 9.2%CuO/ZSM-5(280).
only on 9.5%CuO/ZSM-5(23) with the size of 0.5–1.0 lm and the
weight percent of CuO in these clusters was estimated at ꢄ43%
obtained by EDS detector.
218.1
2
10.9
3
.1.2. H
In order to investigate the reducibility of Cu on zeolite ZSM-5,
-TPR measurements were performed. Fig. 2 shows the H -TPR
/Al ratios of 23, 50 and 280. It
2
-TPR
H
2
2
profiles of Cu/ZSM-5 with SiO
2
2 3
O
2
08.7
is reported that Cu/ZSM-5 catalysts undergo a stepwise reduction
process [46–49]. Generally, the peak at the low temperature
(
170–210 °C) may be attributed to the reduction of CuO in one step
(
(
d)
c)
(b)
a)
282.0
272.4
0
2+
+
to metallic Cu , and the reduction of isolated Cu ions to Cu as
well; and the high temperature peak (normally > 350 °C) may be
ascribed to the reduction of Cu to metallic Cu . However, in
Figs. 2 and 3, the high-temperature peak was not observed. As seen
in Fig. 2, catalyst 9.5%CuO/ZSM-5(23) was seen to exhibit two sep-
arate reduction peaks. The main peak at low temperature peak
208.1
+
0
(
1
00
200
300
400
500
600
o
Temperature ( C)
(
211.4 °C) is assigned to the well dispersed CuO on zeolite ZSM-5
and the small peak at higher temperature (273.2 °C) is assigned
to the reduction of bulk CuO. The assignment of the reduction
Fig. 3. H -TPR profiles of calcined Cu/ZSM-5(280) with various CuO loadings. (a)
6.0%, (b) 9.2%, (c) 19.2%, and (d) 29.1%.
2
peaks is similar to the catalyst Cu/SiO
oration (AE) method [35,38,50]. The main reduction peaks of
CuO/ZSM-5 catalysts with SiO /Al ratios of 50 and 280 (see
2
prepared by ammonia evap-
Fig. 2b and c) became sharper than on Cu/ZSM-5(23) and shift to
a lower temperature (208.7 °C). The shoulder peak at high temper-
ature for 9.7%CuO/ZSM-5(50) was smaller than that of
CuO/ZSM-5(23). For the catalyst 9.2%Cu/ZSM-5(280), only one
low-temperature reduction peak was identified. Fig. 3 exhibits
2
2 3
O
2
the H -TPR profiles of catalysts CuO/ZSM-5(280) with various
CuO loadings. With increasing CuO loadings from 6.0% to 29.1%,
the intensity of the main reduction peak was observed to increase
and the temperature was shifted from 208.1 to 218.1 °C. For all
(
(
a)
b)
trials, the H
2
consumption of catalysts is proportional to the
2
CuO loading and the ratio of H /CuO is almost close to 1 (see
(
c)
d)
supplementary information, Table S1), which indicates that Cu is
divalent. On the catalysts with CuO loadings of 19.2% and 29.1%
(
(Fig. 3c and d), the shoulder peak at high temperature related to
(
(
(
(
e)
f)
g)
h)
bulk CuO was observed. From the result of TPR measurement, it
can be concluded that most Cu species on zeolite HZSM-5 are well
dispersed Cu based on the large reduction peak at low
temperature.
1
0
20
30
40
50
60
70
80
90
3
.1.3. N
The structural properties of the zeolites and copper catalysts
can be derived from the results of N adsorption–desorption
2
adsorption
2θ (degree)
2
Fig. 1. XRD patterns of the calcined catalysts with different SiO
HZSM-5(280), (b) 9.2%CuO/ZSM-5(280), (c) HZSM-5(50), (d) 9.7%CuO/ZSM-5(50),
e) HZSM-5(23), (f) 9.5%CuO/ZSM-5(23), (g) CuO, and (h) Cu O.
2
/Al
2
O
3
ratios. (a)
measurements at ꢂ196 °C. The surface area and pore volume are
(
2
summarized in Table 1. As is shown in Table 1, both mesopores

2
26
Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
Table 1
Surface area, pore volume, copper dispersion and ammonia uptake of catalysts and supports.
Sample
Surface area
Pore volume
NH
uptake
(mmol/g)
3
SACu
(m /gCu
D
Cu
2
)
S
BET
2
S
micro
2
S
external
V
total
3
V
micro
3
V
meso
3
a
b
2
c
d
b
e
(
m /g)
(m /g)
(m /g)
(cm /g)
(cm /g)
(cm /g)
HZSM-5(23)
HZSM-5(50)
HZSM-5(280)
431
447
437
414
423
425
445
437
429
283
333
332
284
203
258
220
215
198
148
115
105
130
220
167
226
222
231
0.411
0.319
0.301
0.299
0.426
0.358
0.426
0.455
0.482
0.212
0.169
0.180
0.134
0.107
0.144
0.146
0.173
0.184
0.199
0.150
0.121
0.165
0.319
0.214
0.280
0.282
0.298
1.167
0.746
0.145
1.462
0.922
0.314
0.356
0.487
0.668
–
–
–
–
–
–
0.03
0.23
0.31
0.24
0.22
0.10
9
9
6
9
1
2
.5%CuO/ZSM-5(23)
.7%CuO/ZSM-5(50)
.0%CuO/ZSM-5(280)
.2%CuO/ZSM-5(280)
9.2%CuO/ZSM-5(280)
9.1%CuO/ZSM-5(280)
20.3
155.6
209.7
162.4
148.8
67.7
a
b
c
The Brunauer–Emmett–Teller (BET) surface area (SBET) was calculated from the linear part of BET plot from 0.007 to 0.03.
The micropore area (Smicro) and volume (Vmicro) were obtained by the t-plot method.
The external surface area Sexternal = SBET ꢂ Smicro
.
d
e
The total pore volume (Vtotal) was evaluated by single point total pore volume at a relative pressure of 0.95.
The mesopore volume Vmeso = Vtotal ꢂ Vmicro
.
and micropores exist for the zeolites. The external surface area and
mesopore volume of HZSM-5(23) (148 m /g, 0.199 cm /g) are
CuO loadings. The amount of ammonia uptake is given in Table 1.
NH -TPD curves with regard to the temperature can provide infor-
mation of the strength of acid sites of the zeolites. As seen in
Fig. 4a, the NH -TPD profiles of parent zeolite HZSM-5 with differ-
ent SiO /Al O ratios exhibited two distinct desorption peaks cen-
tering at around 250 and 450 °C, which are the characteristic peaks
of zeolite with MFI structure [51]. However, in Nanba’s work, the
corresponding two desorption temperature of NH3 from HZSM-5
was 200 and about 400 °C [52]. The temperature difference
(50 °C) was probably due to the different flow rate of carrier gas
2
3
3
2
3
higher than those of HZSM-5(50) (115 m /g, 0.150 cm /g) and
2
3
HZSM-5(280) (105 m /g, 0.121 cm /g). It is seen that the introduc-
tion of copper into zeolite HZSM-5(23) leads to nearly unchanged
micropore area, but lowers both external area and mesopore vol-
ume. The external surface area dropped from 148 m /g to
30 m /g and mesopore volume decreased from 0.199 cm /g to
3
2
2 3
2
2
3
1
0
3
.165 cm /g, which indicated that most of the copper species were
deposited in the mesopores, reducing the contributions of these
pores to the total surface area and total pore volume as a result.
However, introduction of copper on zeolite HZSM-5(50) and
and different ramp rate of heating when NH -TPD was performed
3
(in this work, the flow rate of He is 25 mL/min, ramp rate is
2
HZSM-5(280) lowers the micropore surface area (203 m /g and
10 °C/min). The peak at low temperature is assigned to ammonia
weakly held or physically adsorbed on the Lewis acid sites of zeo-
lite, whereas the peak at high temperature is ascribable to the des-
orption of ammonia strongly adsorbed on and/or interacting with
2
2
20 m /g for 9.7%CuO/ZSM-5(50) and 9.2%CuO/ZSM-5(280),
3
respectively), and micropore volume (0.107 cm /g and
.146 cm /g for 9.7%CuO/ZSM-5(50) and 9.2%CuO/ZSM-5(280),
3
0
+
4
respectively) (see Table 1). Moreover, as shown in Table 1, when
the addition of CuO was increased from 6.0% to 29.1% on zeolite
HZSM-5(280), the BET surface area remained almost unchanged;
the dislodged Al, and decomposition of NH on the Brönsted acid
sites [53–56]. As is shown, with increasing SiO /Al O ratios of
2
2 3
the zeolite from 23 to 280, the peak intensity of ammonia
desorption, especially the peak at low temperature decreased
dramatically and the total acid concentration dropped from
1.167 mmol/gcat to 0.145 mmol/gcat (see Table 1), which is
consistent with the idea that the acidity of a zeolite is inversely
proportional to the SiO /Al O ratio.
2
however, the micropore area dropped from 258 to 198 m /g, which
is assumed to be caused by copper deposition in the micropores of
the zeolite. The external surface area increased from 167 to
2
2
0
31 m /g and the mesopore volume increased from 0.214 to
3
.298 cm /g.
2
2 3
As is reported, copper catalyst exhibits a strong capability of
oxidizing NH to NO or N [57,58]. However, Nanba et al. demon-
strated that only N was formed over Cu species, which was
accompanied by the reduction of Cu to Cu [52]. Therefore, to
3
.1.4. NH
The acidity of the zeolites and reduced copper catalysts was
determined by NH -TPD, as shown in Fig. 4. Fig. S4 displays the
NH -TPD profiles of reduced Cu/ZSM-5(280) catalysts with various
3
-TPD
3
2
2
+
2
2
+
+
3
reduce the possibility of oxidization of NH
3
3
prior to an NH -TPD
3
(
a)
(b)
9
.5%Cu/ZSM-5(23)
9
.7%Cu/ZSM-5(50)
HZSM-5(23)
HZSM-5(50)
HZSM-5(280)
9.2%Cu/ZSM-5(280)
1
00
200
300
400
500
600
700
100
200
300
400
500
600
700
o
o
Temperature ( C)
Temperature ( C)
Fig. 4. NH
3
-TPD profiles of (a) HZSM-5 with different SiO
2
/Al
2
O
3
ratios and (b) reduced Cu/ZSM-5 with different SiO
2
2
/Al O
3
ratios.

Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
227
experiment, all copper catalysts in this work were reduced by H
temperature of 300 °C for 2 h. As seen in Fig. 4b, NH -TPD profiles
of reduced Cu/ZSM-5 with various SiO /Al ratios also exhibited
two distinct peaks, both of which shifted to higher temperatures
about 300–350 °C and 600–650 °C) compared to characteristic
2
at
(see Fig. 5a) and two peaks (932 and 935.3 eV) were displayed
3
on Cu/SiO , which is analogous to our Cu/ZSM-5 catalysts with
2
2
2
O
3
SiO /Al O ratios of 50 and 280 (Fig. 5b and c). Contarini and
2
2 3
Kevan [60] reported that the lower binding energy peak (933–
934 eV) and the higher binding energy peak (935–936 eV) were
(
assigned to the tetrahedrally and octahedrally coordinated Cu2+
,
peaks of the parent zeolites; meanwhile, the peaks of low temper-
ature became larger and broader, which can be ascribable to the
respectively, and the shake-up satellites have a stronger correla-
tion with the octahedrally species when they studied the valence
state of Cu on dehydrated and hydrated copper-exchanged X-
and Y-zeolite. In addition, they pointed out that different Si/Al
ratios in zeolites might affect the symmetry and water coordina-
tion around the exchanged Cu ion. Espinós et al. [61] pointed out
that, for the same oxidation state of Cu, the binding energy can
change depending on the dispersion degree. They suggested that
the higher binding energy (935.4 eV) and lower binding energy
combination of NH
lites and copper sites on the surface. The higher temperature peak
above 600 °C is not shown in the NH -TPD profiles of parent zeo-
lites, which indicates that some copper species strongly adsorb
NH , and is probably due to NH adsorbed on Cu that only binds
3
desorption from both Lewis acid sites of zeo-
3
3
3
to one Al [58]. The high temperature peak exhibits a slight shift
to lower temperature from 650 to 600 °C with increasing
SiO
tration of Cu/ZSM-5 catalyst is higher than that of the correspond-
ing parent zeolite and decreases with increasing SiO /Al ratio
from 1.462 mmol/gcat to 0.356 mmol/gcat, see Table 1).
2 2 3
/Al O ratios from 23 to 280. In addition, the total acid concen-
(
933.6 eV) were related to the dispersed and bulk Cu species,
2
2 3
O
respectively. Böske et al. [62] investigated the binding energy of
various cuprate crystals by high-resolution XPS and found out that
the Cu 2p3/2 peak varied due to the changes with the linking
arrangement of CuAO networks within the lattice, such as linear
chain, zigzag chain and CuO plane, which resulted in the shifts
2
in the position of binding energy and the relative intensity
between the satellite and the main peak.
(
3
.1.5. X-ray photoelectron spectroscopy
XPS analysis was employed to elucidate the chemical states of
copper on the Cu/ZSM-5 catalysts. The Cu 2p photoelectron spectra
of the calcined Cu/ZSM-5 with various SiO /Al ratios are shown
2
2 3
O
In this work, if the peak at 933.2 eV is ascribable to Cu(I), the H
2
in Fig. 5. The asymmetric peaks of Cu 2p3/2 were deconvoluted into
two peaks centering about 935.2 and 933.2 eV. It is seen from Fig. 5
that the Cu 2p3/2 peak of CuO/ZSM-5(23) is centered around
consumption of catalyst 9.5%CuO/ZSM-5(23) should be much
smaller than the other two catalysts (9.7%CuO/ZSM-5(50) and
9
.2%CuO/ZSM-5(280)). However, from the TPR results, it is clearly
seen that the H consumption area of catalyst 9.5%CuO/ZSM-
(23) is very close to the other two catalysts and the molar ratio
9
33.2 eV; however, the peaks of the other two catalysts
2
CuO/ZSM-5(50) and CuO/ZSM-(280) are at 935.2 eV. Typically,
peaks observed at 935.2 eV are assigned to well dispersed Cu(II)
species [59]. The binding energy of the bulk CuO species is
5
2
of H /Cu is close to 1 (see Table S1). Based on the results of XPS,
TPR and XRD, we can safely conclude that the difference between
the binding energy of Cu on HZSM-5 with various SiO /Al ratios
9
33.6 eV (Cu 2p3/2), and the shift to higher binding energy of well
2
2 3
O
dispersed Cu(II) species is indicative of a charge transfer from the
metal ion to the support oxide [59]. The presence of the Cu 2p
is due to the dispersion of Cu or the structural environment where
Cu is located. The binding energy of Cu on ZSM-5(23) tends to shift
to the lower energy level (see Fig. S6a for 18.6%CuO/ZSM-5(23));
2
+
shake-up satellite peak (942–944 eV) is characteristic of Cu with
9
electron configuration of (d ) [38]. However, the Cu 2p3/2 at about
however, the binding energy of Cu on ZSM-5 with SiO
of 50 and 280 shifts to the higher energy level.
2 2 3
/Al O ratios
9
33.2 eV is difficult to discriminate between Cu(I) and Cu(II). It was
reported that X-ray irradiation from XPS could cause the reduction
of the CuO particles [42] and X-ray sensitivity to metal ion reduc-
tion depends strongly on the chemical environment of the metal
ion [60]. Gervasini et al. [59] suggested that the peak of lower bind-
ing energy (Cu 2p3/2 933.15 eV) could be attributed to Cu(I) when
XPS results for Cu/ZSM-5(280) with higher CuO loadings (see
Fig. S5) show no significant change in the Cu binding energy.
Fig.
6
displays the XPS spectra of reduced and used
1
9.2%CuO/ZSM-5(280) compared to the fresh one. The satellite
peaks (943.2 eV) disappeared for the reduced and used catalyst.
For the used catalyst, the peak of Cu 2p3/2 was symmetric and
they compared the valence state of Cu on the catalysts Cu/Al
and Cu/SiO –Al . Interestingly, one peak at 933.15 eV was
observed on Cu/SiO
2 3
O
2
2 3
O
0
found at 932.4 eV, which is assigned to Cu species [38]. This
2
2
–Al O
3
, which is similar to our Cu/ZSM-5(23)
Cu 2p3/2
Cu 2p
Cu 2p
9
35.2
932.4
satellite
933.2
(c)
(b)
(a)
9
43.2
935.2
(c)
(b)
(a)
933.0
930
9
70
960
950
940
930
920
970
960
950
940
920
Binding Energy (eV)
Binding Energy (eV)
Fig. 5. XPS spectra of the calcined CuO/ZSM-5 with different SiO
2
/Al
2
O
3
ratios. (a)
Fig. 6. XPS spectra of 19.2%CuO/ZSM-5(280) catalyst. (a) Without reduction, (b)
after reduction, and (c) after reaction. Spectra were curve fitted by the software
CasaXPS.
9
.5%CuO/ZSM-5(23), (b) 9.7%CuO/ZSM-5(50), (c) 9.2%CuO/ZSM-5(280). Spectra
were curve fitted by the software CasaXPS.

2
28
Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
indicates that the valence state of Cu on the catalyst did not change
during reaction.
Next, reactions were performed over reduced catalysts with
ꢄ10 wt% of CuO loaded on ZSM-5 zeolites with silica to alumina
(
2 2 3
SiO /Al O ) ratios of 23, 50 and 280 at the same reaction condi-
3
.1.6. N
Copper surface area (SACu) and dispersion (DCu) were calculated
by the N O decomposition method [43,44]. As is shown in Table 1,
the catalyst 9.5%CuO/ZSM-5(23) has the worst copper dispersion
2
O adsorption
tions (feed rate of 2,3-butanediol of 3.0 mL/h, hydrogen to
2,3-butanediol molar ratio of 5, and a reaction temperature
250 °C). The selectivities of the major reaction products as a func-
2
2 2 3
tion of time on stream for the catalysts with different SiO /Al O
2
(
0.03) and lowest copper surface area (20.3 m /gCu), which is in
agreement with the results of H -TPR, SEM (Fig. S3) and XPS. On
ZSM-5(280), the dispersion of copper decreases slightly from
.31 to 0.22 with increasing CuO loadings from 6.0% to 19.2%,
and the copper surface area decreases from 209.7 to
ratios are shown in Fig. 7, and the conversion of 2,3-butanediol
and selectivities to the products taken at 40 min and 280 min are
shown in Table 3.
2
0
It can be seen in Table 3 that the conversion of 2,3-butanediol
on all three catalysts was extremely high, especially in the begin-
ning of the reaction, because the dehydration reaction of
2,3-butanediol occurs readily on acid catalysts [15–18]. It has been
reported that dehydration of diols will occur on silica-supported
copper catalysts since Cu is a Lewis acid [63]. Hence, Cu is active
for both hydrogenation [28–32] and dehydration reactions.
Interestingly, it is observed that the conversion of 2,3-butanediol
on 9.5%CuO/ZSM-5(23) is 98.95% at 40 min, which is lower than
that of the other two catalysts (almost 100% at 40 min) and the
control data of HZSM-5(23) as well (see Table 2). This is probably
due to the deactivation of the Cu/ZSM-5 catalyst.
2
1
2
48.8 m /gCu. However, when the CuO loading was increased to
9.1%, the copper dispersion and the copper surface area drasti-
2
cally decreased to 0.10 and 67.7 m /gCu, respectively. This is in
accordance with the result of H
per is not favorable for the copper dispersion [63].
2
-TPR (Fig. 3). High loading of cop-
3
.2. Reaction of 2,3-butanediol over catalysts with different SiO
2 2 3
/Al O
ratios
Experiments were performed on the parent zeolites HZSM-5
with different SiO /Al ratios (23, 50 and 280) in the absence
of hydrogen (N flow: 82.6 cm /min, SATP) and the presence of
hydrogen (H : 67.2 cm /min, SATP; N
respectively, at temperature 250 °C. The conversion of 2,3-butanediol
and selectivities of the main products taken at 40 min and
O
3
It can be observed that the selectivities of MEK (Fig. 7b) and
2-methylpropanal (Table 3), both of which are the main products
from dehydration reaction of 2,3-butanediol (see Table 2) by a
pinacol rearrangement [15–18,65], decreased with increasing
2
2
3
2
3
3
2
2
: 15.4 cm /min, SATP),
SiO
inconsistent with the report that high SiO
for the high yield of MEK and high selectivity of 2-methylpropanal
[16]. As we can see, on the catalyst with a SiO /Al ratio of 23,
2
/Al
2
O
3
ratio in the beginning of reaction (40 min), which is
1
00 min are shown in Table 2. As is reported [15–18], the dehydra-
2
/Al ratio is favorable
2 3
O
tion reaction of 2,3-butanediol occurs readily on acid catalysts. As
seen in Table 2, the conversion of 2,3-butanediol over HZSM-5
under all conditions is very high (>99.0%), and the main products
are MEK, 2-methylpropanal and 1,3-butadiene, which is in
accordance with the result reported in the literature [16]. Minor
products including 2-methyl-1-propanol, 3-hydroxy-2-butanone
and 2-ethyl-2,4,5-trimethyl-1,3-dioxolane were detected, but the
selectivities were much lower than the main products mentioned
above (see Table 2). 3-hydroxy-2-butanone was produced via
dehydrogenation of 2,3-butanediol, while 2-methyl-1-propanol
was from hydrogenation of 2-methylpropanal. The formation
of the cyclic ketal 2-ethyl-2,4,5-trimethyl-1,3-dioxolane was
reported from the intermolecular condensation of 2,3-butanediol
and MEK [64]. However, the selectivities of butenes, which are of
interest, are negligible for conversion of 2,3-butanediol over
HZSM-5. Moreover, it can be seen that running the reactions of
2
2 3
O
the MEK (Fig. 7b) selectivity decreased slightly from 35% to 27%
over time and the selectivity of 2-methylpropanal (Table 3) was
extremely high (10.65% at 40 min, 9.03% at 280 min); however,
on catalyst with SiO /Al O of 280, the selectivity of MEK was
2 2 3
found to be about 20% and the selectivity of 2-methylpropanal
was negligible even at 280 min of time on stream. In addition, as
2 2 3
seen in Fig. 7a, the catalyst with SiO /Al O of 280 was found to
have significantly higher butene selectivity, which is the sum of
the selectivities of 1-butene, cis-2-butene, trans-2-butene and iso-
butene, than the other two catalysts; butene selectivity increased
from 48% during the initial 10 min to 65% at 100 min and then
tended to be relatively stable. However, the highest butene selec-
2 2 3
tivity over catalysts with SiO /Al O ratios of 50 and 23 was
approximately 50% and 40%, which dropped slightly to 45% and
dramatically to 10%, respectively, in 280 min of time on stream.
Table 3 also shows the distribution of the different butene iso-
mers made over all catalysts. All four isomers of butene (1-butene,
trans-2-butene, cis-2-butene, and isobutene) were detected. The
2
,3-butanediol over each HZSM-5 in the absence or presence of
hydrogen led to almost the same result, which indicates that
hydrogen is not involved in the conversion of 2,3-butanediol over
HZSM-5.
Table 2
Conversion of 2,3-butanediol (%) and carbon selectivity of main products (%) on the parent HZSM-5 with different SiO
parentheses).
2
/Al
2
O
3
ratios in 40 min and 100 min (shown in
)b
Zeolite (without H
HZSM-5(23)
2
)a
Zeolite (with H
2
HZSM-5(50)
HZSM-5(280)
HZSM-5(23)
HZSM-5(50)
HZSM-5(280)
1
,3-Butadiene (C
Butenes (C
MEK (C
4
H
6
)
10.05 (10.42)
0.85 (0.65)
55.30 (54.69)
27.95 (27.30)
0.53 (1.23)
0.71 (1.68)
0.36 (0.62)
4.25 (3.41)
100 (99.65)
11.11 (11.32)
0.64 (0.51)
56.21 (54.36)
29.25 (28.21)
0.68 (0.98)
0.60 (1.04)
0.24 (1.71)
1.27 (1.87)
100 (99.70)
13.87 (15.36)
0.80 (0.50)
52.77 (49.87)
29.38 (26.00)
0.68 (1.02)
0.76 (1.30)
0.38 (3.39)
1.36 (2.56)
100 (99.50)
10.61 (11.06)
0.85 (0.67)
56.32 (55.1)
28.79 (28.62)
0.77 (1.24)
0.64 (1.37)
0.10 (0.53)
1.92 (1.41)
100 (99.71)
10.42 (12.08)
0.67 (0.48)
56.13 (53.65)
29.80 (28.44)
0.71 (1.02)
0.62 (1.01)
0.29 (1.36)
1.36 (1.96)
100 (99.81)
14.00 (14.48)
0.81 (0.48)
53.69 (51.51)
28.41 (25.59)
0.67 (0.95)
0.81 (1.27)
0.36 (3.24)
1.25 (2.48)
100 (99.00)
4
8
H )
4
H
8
O)
-Methylpropanal (C
-Methyl-1-propanol (C
-Hydroxy-2-butanone (C
-Ethyl-2,4,5-trimethyl-1,3-dioxolane (C
2
2
3
2
4
H
8
O)
10O)
4
H
4 8 2
H O )
8 16 2
H O )
c
Others
Conversion
a
b
c
3
Reaction condition: feed rate of 2,3-butanediol, 3.0 mL/h; catalyst weight, 1.0 g; temperature, 250 °C; N
Reaction condition: feed rate of 2,3-butanediol, 3.0 mL/h; catalyst weight, 1.0 g; temperature, 250 °C; H
Other products: acetone, tetramethylfuran, 3,4,5-trimethyl-2-cyclopentenone and aromatics.
2
flow, 82.6 cm /min.
3
3
2
flow, 67.2 cm /min; N
2
flow, 15.4 cm /min.

Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
229
7
6
5
4
3
2
1
0
0
0
0
0
0
0
50
(
a)
(b)
4
3
2
1
0
0
0
0
0
0
0
50
100
150
200
250
300
0
50
100
150
200
250
300
Time (min)
Time (min)
3
2
2
1
1
0
5
0
5
0
16
(
c)
(d)
1
1
1
4
2
0
8
6
4
2
0
5
0
0
50
100
150
200
250
300
0
50
100
150
200
250
300
Time (min)
Time (min)
Fig. 7. Catalytic results as a function of time on stream for the conversion of 2,3-butanediol over reduced copper supported on ZSM-5 with different SiO
2
/Al
2
O
3
: (j) 9.5%CuO/
ZSM-5(23), (s) 9.7%CuO/ZSM-5(50), (N) 9.2%CuO/ZSM-5(280). Carbon selectivity to main products (a) butene, (b) MEK, (c) 3-hydroxy-2-butanone, and (d) 2-methyl-1-
propanol. Reaction conditions: feed rate of 2,3-butanediol, 3.0 mL/h; catalyst weight, 1.0 g; H /2,3-butanediol (molar ratio), 5:1; temperature: 250 °C.
2
Table 3
2 2 3
/Al O
ratios in 40 min and 280 min (shown in parentheses).a
Conversion of 2,3-butanediol (%) and carbon selectivity of the products (%) on reduced Cu/ZSM-5 with different SiO
Catalysts
9
.5%CuO/ZSM-5(23)
9.7%CuO/ZSM-5(50)
9.2%CuO/ZSM-5(280)
Ethylene (C
Propylene (C
Isobutane and butane (C
2
H
4
)
0.04 (0.01)
0.38 (0.13)
0.04 (0.03)
24.30 (7.23)
0.16 (0.02)
0.05 (0.02)
0.10 (0.02)
0.62 (0.27)
33.07 (26.31)
10.65 (9.03)
10.23 (11.40)
0.21 (1.41)
5.14 (24.53)
0.87 (1.64)
1.84 (3.52)
0.45 (0.79)
0 (0)
0.36 (0.13)
2.80 (0.79)
0.32 (0.09)
44.13 (42.41)
6.02 (0.66)
0.16 (0.34)
0.52 (0.19)
3.37 (1.61)
25.62 (32.52)
0.03 (4.21)
0.80 (3.46)
0.35 (0.02)
1.24 (0.20)
1.00 (0.41)
0.05 (0.03)
1.07 (0.71)
1.12 (0.58)
2.75 (7.25)
0 (0.10)
0.21 (0.15)
1.90 (0.75)
0.18 (0.11)
58.65 (62.84)
3.74 (0.79)
2.10 (0.12)
0.38 (0.19)
1.24 (0.72)
19.06 (24.30)
0 (0.17)
3
6
H )
4
10
H )
Butenes (C
4
H
8
)
C
C
C
C
5
6
7
8
olefins (C
5 10
H )
6 12
H )
7 14
H )
8 16
H )
olefins (C
olefins (C
olefins (C
4
MEK (C H
8
O)
-Methylpropanal (C
-Methyl-1-propanol (C
,3-Butanedione (C
-Hydroxy-2-butanone (C
-Butanol (C 10O)
-Ethyl-2,4,5-trimethyl-1,3-dioxolane (C
2
2
2
3
2
2
4
H
8
O)
10O)
4
H
)
0 (0.46)
0.14 (0.01)
0.22 (0)
4
H
6
O
2
4 8 2
H O )
4
H
0.57 (0.22)
0.15 (0.01)
0.55 (0.57)
0.72 (0.43)
0.98 (3.43)
0.19 (0.08)
0.19 (0.09)
100 (99.67)
8 16 2
H O )
Ethylbenzene (C
p-Xylene (C
Tetramethylfuran (C
1
1
8 10
H )
H )
10
8
8
H
12O)
-Ethyl-3-methyl-benzene (C
,3,5-Trimethyl-benzene (C
9.36 (8.37)
0.03 (0.88)
0.12 (0.07)
98.95 (91.69)
9 12
H )
9
H
12
)
0.42 (0.15)
100 (99.80)
Conversion of 2,3-butanediol
Distribution of butenes
1
-Butene (C
Isobutene (C
Trans-2-butene (C
Cis-2-butene (C
4
H
8
)
3.85 (2.40)
6.82 (0.43)
7.58 (2.27)
5.97 (1.80)
4.61 (3.67)
7.62 (7.97)
4
H
8
)
11.03 (10.90)
14.45 (13.17)
12.20 (12.95)
11.54 (12.80)
20.22 (21.43)
16.91 (18.12)
4
H
)
8
)
4
H
8
a
2
Reaction conditions: feed rate of 2,3-butanediol, 3.0 mL/h; catalyst weight, 1.0 g; H /2,3-butanediol (mol ratio), 5:1; temperature, 250 °C.
selectivity of 1-butene is much smaller than the other three iso-
mers (isobutene, trans-2-butene and cis-2-butene) in the initial
dehydration of 2,3-butanediol (see Table 2), was not detected, sug-
gesting that either it is not formed over the Cu/ZSM-5 catalysts, or
it was immediately hydrogenated to butenes after forming. In
4
0 min. In this work, 1,3-butadiene, which is a major product from

2
30
Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
addition, the selectivities of butane and isobutane, which could be
produced by further hydrogenation of the butenes, were very small
due to the formation of coke [66,67], which is a non-desorbed pro-
duct that prevents access to the acid sites of catalysts. To deter-
mine the coke content of the used catalysts, a quantitative
analysis of coke formation over the used catalysts after the reaction
was investigated by thermogravimetric analysis (TGA). Fig. 8
shows the TGA results of the catalysts after 280 min of reaction.
As is shown, the coke content on 9.5%CuO/ZSM-5(23) was 5.81%,
which was higher than that on 9.7%CuO/ZSM-5(50) (4.97%) and
9.2%CuO/ZSM-5(280) (2.57%). This indicates that the acidity of cat-
alyst is an important cause in coke formation. Indeed, with more
acid sites, the coke formation will occur faster. The coke content
on 9.5%CuO/ZSM-5(23) was 4.29% after 40 min of reaction, which
was much higher than the other two catalysts (see Fig. S7).
Hence, the catalyst will deactivate sooner as a result. For this rea-
(
see Table 3), which indicates that copper catalysts are not favor-
able for the hydrogenation of butenes to C alkanes in this work.
As discussed above, there is a trend that increasing the
SiO /Al ratio (lowering acidity of HZSM-5) increases butene
selectivity. However, it was also observed that during the first
0 min of reaction, the selectivity of butene over catalyst with
SiO /Al of 23 (40%) is higher than with a ratio of 50 (25%).
4
2
2 3
O
1
2
2 3
O
This is likely due to the influence of dimerization and cracking
reactions that occurred during this period of time. As we can see
in Table 3, the catalyst with a SiO
highest selectivity of C olefins (3.37%), which is from dimerization
of butenes [26], and highest selectivities of propylene (2.80%) and
2 2 3
/Al O ratio of 50 showed the
8
C
3
5
olefins (6.02%, the sum of the selectivities of 2-methyl-1-butene,
-methyl-1-butene and 2-methyl-2-butene), both of which are
2 2 3
son, the catalyst with a SiO /Al O ratio of 23 deactivated faster
than the other two catalysts. As is shown in Table 3, the conversion
of 2,3-butanediol over 9.5%CuO/ZSM-5(23) decreased from 98.95%
at 40 min to 91.69% at 280 min, while conversion over the other
two catalysts remained above 99.0% even after 280 min. In addi-
tion, deactivation of catalysts results in the loss of acid sites,
decreasing the possibility of dehydration reaction over acid sites;
from cracking reactions [26]. It seems that catalyst with modest
acidity (SiO /Al = 50, acidity 0.746 mmol/g) is beneficial for
the oligomerization of butenes to form dimers, with subsequent
cracking reaction to form ethylene, propylene, C , C and C olefins,
which reduced the production of butenes especially at the begin-
ning of reaction. It has to be mentioned that the selectivities of
products from dimerization and cracking reactions decreased over
time due to the deactivation of catalysts.
2
2 3
O
5
6
7
therefore, the selectivity of MEK over the catalyst with
a
SiO /Al ratio of 23 decreased over time. Meanwhile, the selec-
2
2 3
O
tivity of 2-methyl-1-propanol increased dramatically from 3.5%
to 12.0% with time on stream (Fig. 7d), which was accompanied
by the decrease of isobutene selectivity from 6.82% at 40 min to
0.43% at 280 min (see Table 3). Consequently, the selectivity of
butenes on 9.5%CuO/ZSM-5(23) showed a decreasing trend over
time. It has been reported that copper catalysts are also active in
dehydrogenation [25]. Hence, as is seen in Fig. 7c and Table 3,
when deactivation occurred, Cu on 9.5%CuO/ZSM-5(23) turned into
the active sites in conversion of 2,3-butanediol to 3-hydroxy-2-
butanone and 2,3-butanedione, both of which showed an increas-
ing trend in selectivity with time on stream.
As mentioned above, a small amount of 3-hydroxy-2-butanone
can be seen in the control experiment of conversion of
2
,3-butanediol over HZSM-5 (see Table 2), which is a product from
the dehydrogenation reaction of 2,3-butanediol [18]. Over
Cu/ZSM-5 catalyst, 3-hydroxy-2-butanone was also observed (see
Fig. 7c) even in the excess of hydrogen present in the reactor
(
2 2 2 3
molar ratio of H /2,3-BDO = 5). Over the catalyst with SiO /Al O
of 23, the selectivity of 3-hydroxy-2-butanone increased dramati-
cally with increasing time on stream (increased from almost 0%
to about 25%), which is probably due to deactivation of catalyst.
However, the catalyst with SiO
ity for dehydrogenation reaction only in the first 70 min on stream,
while the selectivity of 3-hydroxy-2-butanone over the catalyst
with SiO /Al O ratio of 280 was negligible. The reaction mecha-
2 2 3
nism will be discussed later.
Over the catalyst with a SiO
2
/Al
2
O
3
ratio of 50 presented activ-
As is mentioned above, large copper clusters were observed on
the surface of 9.5%CuO/ZSM-5(23) (see SEM image, Fig. S3), which
is in accordance with the copper dispersion result (0.03) shown in
Table 1. It is possible that poor copper dispersion and large Cu par-
ticle sizes could affect the catalytic reaction, as shown in the liter-
ature [31,68]. However, in this work, we believe that the
2
2 3
/Al O ratio of 23, the increase of
2
-methyl-1-propanol selectivity is a result of the deactivation of
2 2 3
differences noted for the different SiO /Al O ratios are due to dif-
catalyst (Fig. 7d). Deactivation of zeolite-based catalysts is mainly
ferences in catalyst acidity rather than Cu dispersion or Cu size,
since the activity of catalyst 9.7%CuO/ZSM-5(50) was not as good
as catalyst 9.2%CuO/ZSM-5(280) even though it exhibited a good
dispersion of Cu on the surface (0.23, see Table 1). Over a catalyst
2 2 3
with SiO /Al O of 280, only a trace amount of 2-methyl-1-
1
00
propanol and 3-hydroxy-2-butanone was found in 280 min of time
on stream. Also, it was seen to exhibit the highest selectivity of
butenes and lowest selectivities to other byproducts, such as aro-
matics and tetramethylfuran (Table 3). Based on the discussion
9
9
9
9
8
6
4
2
2 2 3
above, it can be concluded that zeolite ZSM-5 with SiO /Al O ratio
of 280 can be chosen as the best support for catalyzing the
hydrodeoxygenation of 2,3-butanediol to butene in a single reac-
tor. Next, we will focus on this zeolite to examine other reaction
parameters.
-
2.57%
9
9
.5% CuO/ZSM-5(23)
.7% CuO/ZSM-5(50)
9
.2% CuO/ZSM-5(280)
-4.97%
3.3. Effect of copper content
-
5.81%
Fig. 9 shows the impact of copper loading on catalytic conver-
sion of 2,3-butanediol to the main products as a function of time
on stream under the same reaction conditions as used in Fig. 7
and Table 3 for ZSM-5(280) with four different CuO loadings:
1
00
200
300
400
o
500
600
Temperature( C)
6
.0%, 9.2%, 19.2%, and 29.1%. The conversion of 2,3-butanediol
and selectivities to the products are shown in Table S2. As is seen,
the conversion of 2,3-butanediol was high (>99%) over all four
catalysts.
Fig. 8. Thermogravimetric profile of catalysts after 280 min of reaction. Reaction
conditions: feed rate of 2,3-butanediol, 3.0 mL/h; catalyst weight, 1.0 g; H /2,3-
2
butanediol (molar ratio), 5:1; temperature: 250 °C.

Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
231
80
70
60
50
40
30
20
10
0
(a)
(b)
3
2
1
0
0
0
0
0
50
100
150
200
250
300
0
50
100
150
200
250
300
Time (min)
Time (min)
2.5
2.0
1.5
1.0
0.5
0.0
10
8
(c)
(d)
6
4
2
0
0
50
100
150
200
250
300
0
50
100
150
200
250
300
Time (min)
Time (min)
Fig. 9. Catalytic results for the conversion of 2,3-butanediol to butene over different copper loadings on ZSM-5 (SiO
ZSM-5(280), (j) 19.2%CuO/ZSM-5(280), (N) 29.1%CuO/ZSM-5(280). Carbon selectivity to main products (a) butene, (b) MEK, (c) 3-hydroxy-2-butanone, and (d) 2-methyl-1-
propanol. Reaction conditions: feed rate of 2,3-butanediol, 3.0 mL/h; catalyst weight, 1.0 g; H /2,3-butanediol (molar ratio), 5:1; temperature: 250 °C.
2
/Al
2
O
3
= 280). (4) 6.0%CuO/ZSM-5(280), (h) 9.2%CuO/
2
As is shown in Fig. 9a, the selectivity of butene increased with
increasing copper loading, and the catalyst with 19.2 wt% of CuO
showed the highest catalytic activity toward the production of
butene; the selectivity gradually increased from 60% in the initial
2-methylpropanal, leading to high selectivities of 2-methyl-1-
propanol and 2-methylpropanal.
The selectivity of 3-hydroxy-2-butanone with time on stream is
depicted in Fig. 9c. As seen in Fig. 9c, in the initial 10 min, the cat-
alyst with the lowest CuO loading (6 wt%) exhibited the highest
selectivity of 3-hydroxy-2-butanone (1.8%), however, catalyst with
19.2 wt% of copper loading presented the lowest, 0.2%. The selec-
tivities of 3-hydroxy-2-butanone tended to decrease with time
on stream, reaching 0 after 100 min, with the exception of the cat-
alyst with 29.1 wt% CuO, on which selectivity was observed to
decrease from 0.5% to 0.1% at 100 min, and then increased steadily
to 1.2% at 310 min, which probably can be attributed to low copper
dispersion. Torresi et al. investigated the conversion of
1,3-butanediol by dehydrogenation and dehydration reactions on
CuO/SiO , and found that dehydrogenation predominated over cat-
2
alysts with high copper loading [63], which is similar to our finding
in this work.
In addition, over catalysts with different copper loadings, it
can be seen that the selectivities of C , C , C and C olefins
8 7 5 3
1
0 min to reach a maximum of approximately 71% at 70 min and
then dropped slightly to 65% after 310 min on stream. However,
it does appear that the catalyst with highest weight loading
(
29.1 wt%) of CuO deactivated faster than the catalysts with lower
loadings of CuO (9.2 wt% and 19.2 wt%). This suggests that it is not
necessary to have excessively high loadings of CuO to get high
selectivities of butenes.
As seen in Fig. 9b, the selectivity of MEK showed a decreasing
trend with increasing weight loading of copper. As copper is the
active site for hydrogenation, higher copper loading is expected
to favor the hydrogenation of MEK to butenes, resulting in a lower
selectivity of MEK. It can also be seen that the selectivity of MEK
over catalysts showed a general tendency to increase gradually
with increasing time on stream, which can be ascribed to the deac-
tivation of catalysts.
Fig. 9d shows the selectivity of 2-methyl-1-propanol over cata-
lysts with time on stream. With the exception of the catalyst with
and 2-butanol (Table S2) were higher on the catalyst with lower
CuO loadings (6.0% and 9.2%), which indicates that lower copper
loadings favored the dimerization of butenes, and subsequent
cracking reaction and the formation of 2-butanol (the intermedi-
ate to form butenes), especially in the first 40 min of stream.
The optimal amount of copper is not yet clear. As discussed in
2
9.1 wt% CuO, all catalysts showed similar behaviors toward the
selectivity of 2-methyl-1-propanol, which increased slightly from
% to about 0.6% with time on stream. However, over the catalyst
0
with highest loading of CuO (29.1 wt%) in this work, the selectivity
of 2-methyl-1-propanol increased dramatically from 0% to 8.5%. In
addition, the selectivity of 2-methylpropanal (Table S2) was negli-
gible except on the catalyst with highest CuO loading (29.1 wt%).
These trends may be due to the low copper dispersion (0.10, see
Table 1) on the catalyst with 29.1% CuO, which is not favorable
for dehydration of alcohols to butenes and hydrogenation of
2
the literature [63], on SiO , the copper monolayer surface
coverage is about 13.5 wt% of Cu (e.g. 16.9 wt% of CuO), which
is close to the CuO loading (19.2 wt%) on the catalyst that gave
the highest butene selectivity. The optimal performance is
the result of
functions.
a balance between copper and acid catalytic

2
32
Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
3.4. Effect of hydrogen to 2,3-butanediol ratio
2
The impact of H /2,3-BDO ratio toward MEK selectivity is
shown in Fig. 10b. It is observed that the selectivity of MEK
decreases with increasing H /2,3-BDO ratios from 2 to 5, and
/2,3-BDO ratio of 2 exhibits the highest and most stable selectiv-
ity (about 30–34%) of MEK. At H /2,3-BDO ratios of 0 and 1, the
The impact of hydrogen to 2,3-butanediol ratio on the catalytic
performance of 19.2 wt%CuO/ZSM-5(280) for conversion of
,3-butanediol to main products with time on stream is depicted
2
H
2
2
2
in Fig. 10 and the selectivities to all products are shown in
Table S3. All conditions exhibit high conversion of
selectivities of MEK decreased dramatically over time due to the
deactivation of catalysts which lead to decreasing catalytic activity
for dehydration. Meanwhile, the selectivities of 3-hydroxy-2-
butanone and 2,3-butanedione increased dramatically over time
in such conditions. This indicates that the dehydrogenation reac-
tion became dominant at the conditions with low hydrogen.
The selectivity of 2-methyl-1-propanol with time on stream is
depicted in Fig. 10d. This selectivity increases with decreasing
2
,3-butanediol, though these is a general increasing trend with
increasing hydrogen to 2,3-butanediol ratio.
As expected, the dehydrogenation reaction is an important
2
reaction at low H /2,3-BDO ratio, especially when the ratio is
below 2 (see Fig. 10c and 2,3-butanedione in Table S3). The selec-
tivity of 3-hydroxy-2-butanone generally increased with increas-
ing time on stream and decreased when H
increased from 0 to 5. The selectivities of 2,3-butanedione at
0 min and 280 min also decreased with increasing H /2,3-BDO
ratios. At H /2,3-BDO of 5, it can be seen that the dehydrogenation
was suppressed, as indicated by the negligible amount of
2
/2,3-BDO ratio was
2
H /2,3-BDO ratio from 5 to 2. It is well known that hydrogen can
improve the catalytic stability of zeolite catalysts due to the inhibi-
tion effect of the hydrogen on coke formation [69–71] by reacting
with carbenium ions to limit the formation of carbonaceous com-
pounds responsible for deactivation, which is in agreement with
4
2
2
3
-hydroxy-2-butanone and 2,3-butanedione present.
2
the TGA results (see Fig. S8). With decreasing H /2,3-BDO ratio,
The main trend of interest is that butene selectivity increases as
H /2,3-BDO ratio increases (see Fig. 10a), which is attributed to the
2
more coke is formed (Fig. S8), and faster catalyst deactivation is
observed, resulting in increasing selectivity of 2-methyl-1-
fact that hydrogen has a positive impact on catalytic activity
toward the hydrogenation reaction. Also, it was observed that
the selectivity of butene decreases with increasing time on stream,
which is due to the deactivation of catalysts especially with low
2
propanol over time. At the H /2,3-BDO ratio of 5, the catalyst
exhibited extremely high catalytic activity for hydrogenation and
dehydration reactions, resulting in negligible selectivity of 2-methyl-
1-propanol, which is expected to be converted to isobutene. As is
H
2
/2,3-BDO ratios of 0 and 1. Interestingly, butenes can be formed
even in the absence of H (H /2,3-BDO = 0), which are higher than
the control data of HZSM-5(280) without H (see Table 2). As
mentioned above, the dehydrogenation reaction became the main
reaction at low H /2,3-BDO ratios. We suggest that H formed in
2
shown in Fig. 10d, however, when the H /2,3-BDO ratio is decreased
2
2
from 1 to 0, the selectivity of 2-methyl-1-propanol decreases. This is
because dehydrogenation of 2,3-butanediol to 3-hydroxy-2-
butanone and 2,3-butanedione becomes the dominant reaction
pathway under the conditions with low hydrogen. As is seen in
2
2
2
the process of dehydrogenation of 2,3-BDO is involved in the
hydrogenation reactions to produce the butenes.
Table S3, with different H
2
/2,3-BDO ratios, it can be seen that lower
H
2
partial pressures are not favorable for the cracking reaction.
80
70
60
50
40
30
20
10
0
(a)
(b)
40
30
20
10
0
0
50
100
150
200
250
300
0
50
100
150
200
250
300
Time (min)
Time (min)
5
4
3
2
1
0
12
(c)
(d)
1
0
8
6
4
2
0
0
0
0
0
0
0
50
100
150
200
250
300
0
50
100
150
200
250
300
Time (min)
Time (min)
Fig. 10. Catalytic results for the conversion of 2,3-butanediol to butene over 19.2 wt% copper supported on ZSM-5 (Si/Al
s) 1, (d) 2, (4) 3, (.) 4, (j) 5. Carbon selectivity to main products (a) butene, (b) MEK, (c) 3-hydroxy-2-butanone, and (d) 2-methyl-1-propanol. Reaction conditions: feed
rate of 2,3-butanediol, 3.0 mL/h; catalyst: 19.2 wt%CuO/ZSM-5(280); catalyst weight, 1.0 g; temperature, 250 °C.
2 3
O = 280) at a hydrogen to 2,3-BDO ratio of: (h) 0,
(

Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
233
Another product from dehydration reaction, 2-methylpropanal,
exhibited decreasing trend of selectivity with increasing
/2,3-BDO ratios from 0 to 5, decreasing from 13.15% to 0% at
from 0 during the initial 10 min to 26% at 310 min, indicating cat-
alyst deactivation. As we can see, low temperature (230 °C) favored
the production of 2-methyl-1-propanol over time (Fig. 11d) and
formation of 2-methylpropanal (Table S4), accompanied by low
selectivity of butenes. This is also due to the deactivation of cata-
lyst. From TGA results (see Fig. S9), we can see the sharp weight
loss (ꢂ0.8%) between 220 °C and 300 °C on the used catalyst after
reaction at 230 °C, which is reported due to the formation of heavy
oligomers from butenes that do not evaporate on zeolites [26].
However, at higher temperature, the heavy products can evaporate
or be cracked into smaller molecules, reducing coke deposition
a
H
2
40 min, which is the source of 2-methyl-1-propanol and isobutene.
In conclusion, hydrogenation reactions are essential toward get-
ting high butene selectivities. Higher H /2,3-BDO ratios are better
2
for catalyzing hydrogenation reactions because hydrogen can
improve catalytic stability of zeolite catalysts.
3.5. Effect of temperature
[
26]. And at extremely higher temperatures (270 °C and 300 °C),
The impact of temperature on the selectivities of the main prod-
the deactivation of catalyst is due to hard coke formation at high
temperature [66]. From the TGA (Fig. S9), we can see the weight
loss on the used catalysts after reaction at 270 °C and 300 °C occurs
mainly between 400 °C and 600 °C, which are due to the combus-
tion of ‘‘hard coke’’ [72–74].
The selectivity of 3-hydroxy-2-butanone at various tempera-
tures with time on stream is depicted in Fig. 11c. Clearly, low tem-
peratures favor the dehydrogenation reaction of 2,3-butanediol to
lose one hydrogen atom from one hydroxyl group, especially after
ucts on 19.2 wt%CuO/ZSM-5(280) as functions of time on stream is
depicted in Fig. 11. The conversion of 2,3-butanediol and selectiv-
ities of all products in 40 min and 310 min are shown in Table S4.
All temperatures except 230 °C exhibit stable and high conversion
of 2,3-butanediol, with conversions of nearly 100%. However, at
lower temperature (230 °C), the conversion is relatively lower at
3
10 min (93.11%) due to deactivation.
As seen in Fig. 11a, with the exception of 230 °C, the selectivity
of butene decreased with increasing temperature, which is mainly
due to oligomerization of butenes, and subsequent cracking reac-
tions, resulting in lower selectivity of butene and higher selectivi-
1
2
00 min of stream. However, at higher temperatures such as
70 °C and 300 °C, 2,3-buanediol is likely to lose two hydrogen
atoms from both hydroxyl groups to form 2,3-butanedione (see
Table S4).
3 5 6 7 8
ties of C , C , C , C and C olefins (see Table S4). In particular, at a
temperature of 250 °C, the catalyst exhibits the highest selectivity
of butene. A high butene selectivity (55%) is initially observed for
the lowest temperature (230 °C), but it dramatically decreased to
As discussed above, higher temperatures (270 °C and 300 °C)
are beneficial for the oligomerization and cracking reactions,
3 5
resulting in significantly higher selectivities of C olefins.
and C+
1
0% over 310 min.
Moreover, it should be noted that higher temperatures lead to
higher selectivities of heavy products, such as aromatic compounds
Fig. 11b displays the selectivity of MEK at various reaction tem-
peratures. The initial selectivity of MEK decreases with increasing
temperature. However, at 300 °C, the selectivity of MEK increased
4
(Table S4) and C alkanes (isobutene and butane, see Table S4).
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
40
35
30
25
20
15
10
5
0
(
a)
(
b)
0
0
50
100
150
200
250
300
0
50
100
150
200
250
300
Time (min)
Time (min)
1
1
1
4
2
0
8
6
4
2
0
(
c)
(d)
1
6
2
8
4
0
1
0
50
100
150
200
250
300
0
50
100
150
200
250
300
Time (min)
Time (min)
Fig. 11. Catalytic results for the conversion of 2,3-butanediol to butene over 19.2 wt% CuO supported on ZSM-5 (Si:Al
different reaction temperatures: (4) 230 °C, (j) 250 °C, (h) 270 °C, (N) 300 °C. Carbon selectivity to main products, (a) butene, (b) MEK, (c) 3-hydroxy-2-butanone, and (d) 2-
methyl-1-propanol. Reaction conditions: feed rate of 2,3-butanediol, 3.0 mL/h; catalyst: 19.2 wt%CuO/ZSM-5(280); catalyst weight, 1.0 g; H /2,3-BDO (molar ratio)=5.
2 3
O = 280) at a hydrogen to 2,3-BDO ratio of 5 at
2

2
34
Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
80
60
40
20
0
80
60
40
20
0
1
9
9
8
8
00
5
1
9
9
8
8
00
5
(
b)
(
a)
0
0
5
5
0
0
0
100
200
300
400
500
0
100
200
300
400
500
Time (min)
Time (min)
80
60
40
20
0
1
9
9
8
8
00
5
0
5
0
(
c)
0
100
200
300
400
500
Time (min)
Fig. 12. Catalytic results for the conversion of 2,3-butanediol over reduced catalysts (a) fresh catalyst 19.2%CuO/ZSM-5(280), (b) 19.2%CuO/ZSM-5(280) after first
regeneration, (c) 19.2%CuO/ZSM-5(280) after second regeneration. Selectivity to products: (j) butene, (d) MEK, (N) 3-hydroxy-2-butanone, (5) 2-methyl-1-propanol, (s) 2-
methylpropanal, (h) conversion of 2,3-butanediol. Reaction conditions: feed rate of 2,3-butanediol, 3.0 mL/h; catalyst weight, 1.0 g; H
temperature, 250 °C.
2
/2,3-BDO (molar ratio) = 5; reaction
3
.6. Regeneration of catalysts
the catalytic results of conversion of MEK over 19.2 wt%CuO/
ZSM-5(280) with time on stream. As we can see, the selectivity
of butene is high (about 50%), although it is not as high as when
2,3-butanediol was the reactant. Interestingly, the results show
Since deactivation of the catalyst was shown to be an issue, we
have investigated whether the catalyst can be regenerated by heat-
ing in air. Deactivation by coke is reversible, because generally
coke can be removed by oxidation [75,76]. This would be expected
to regenerate the catalyst if coke deposition is the main source of
deactivation [77], but not if sintering or other structural rearrange-
ment was responsible. Experiments were run where 19.2% CuO on
5 8
that the selectivities of C –C olefins are significantly higher in
comparison with the result from 2,3-butanediol. Similar results
are also observed in hydrogenation reaction of 2-methylpropanal,
in which, the oligomerization and cracking reactions became dom-
inant catalytic processes (see Fig. S10), with selectivities of C
5 8
–C
2 3
ZSM-5 (280, Si/Al O ratio) was regenerated twice under air with
olefins higher than butenes. The reason why the formation of
3
air flow 120 cm /min at 550 °C for 3 h. Fig. 12a shows the initial
catalytic performance, while Fig. 12b and c show the performance
after the first and second regenerations, respectively. As seen by
comparison of these three figures, regeneration under air is capable
to yielding a catalyst with almost identical performance as the ini-
tial fresh catalyst. It is remarkable that high conversion of
80
60
40
20
0
100
9
0
2
,3-butanediol (almost 100%) is exhibited by the catalysts, even
ones displaying signs of deactivation. However, after a second
regeneration, the catalytic activity, especially the selectivity of
butene, is shown to drop faster than the fresh or used catalyst after
the first regeneration, which is probably due to the formation of
stable or hard coke species agglomerated on the zeolite in the first
and second runs which cannot be removed by the regeneration
procedure.
80
70
6
5
0
0
3.7. Hydrogenation of MEK
40
300
0
50
100
150
200
250
As discussed above, the results suggest that MEK and
-methylpropanal are the intermediates in the conversion of
,3-butanediol to butene. To explore the roles of the intermediate
in conversion of 2,3-butanediol, hydrogenation reactions of MEK
and 2-methylpropanal were conducted under similar reaction con-
ditions to those used for 2,3-butanediol conversion. Fig. 13 shows
Time (min)
2
2
Fig. 13. Catalytic results for the conversion of MEK over reduced catalyst 19.2%CuO/
ZSM 5(280) with time on stream. Selectivity to products: (j) butene, (s) propylene,
8
, (d) 2-butanol, (h) conversion of MEK. Reaction
conditions: feed rate of MEK, 3.0 mL/h; catalyst weight, 1.0 g; H
=
=
(
4) pentene, (.)
C
6
ꢄ C
2
/MEK (molar
ratio) = 5.

Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
235
Scheme 1. Probable reaction pathways in the hydrodeoxygenation of 2,3-butanediol to products.
C
6
–C
8
olefins is favorable when MEK and 2-methylpropanal are
2-methylpropanal are the intermediates in the conversion of
2,3-butanediol to butenes. The optimal performance toward the
production of butene is the result of a balance between copper
and acid catalytic functions.
used as reactants is likely because no acid sites are required to
dehydrate 2,3-BDO to MEK and 2-methylpropanal, so more acid
sites are available for dehydration of 2-methyl-1-propanol and
2
-butanol to butenes, and the subsequent oligomerization and
cracking reactions as well.
Acknowledgments
Based on the results discussed above, we summarize the prob-
able reaction pathways in the hydrodeoxygenation of
Financial support from Navy SBIR Phase II Contract
#NG8335-13-C-0174, in collaboration with TekHolding, Inc., is
gratefully acknowledged. We would like to acknowledge Phillip
Defoe for the help in the ICP analysis of samples and Dr. Daniel
Boyle for help in SEM measurement.
2
,3-butanediol in Scheme 1. The primary pathways involve dehy-
dration and hydrogenation reactions. Over acid sites (including
Cu sites), 2,3-butanediol is dehydrated to form primarily MEK
and 2-methylpropanal [16]. MEK and 2-methylpropanal are con-
verted to 2-butanol and 2-methyl-1-propanol, respectively, over
copper sites through hydrogenation. Finally, 2-butanol is dehy-
drated to form 1-butene, trans-2-butene and cis-2-butene and
Appendix A. Supplementary material
2
-methyl-1-propanol is converted to isobutene through dehydra-
tion. 1,3-butadiene, which is a product from dehydration of
,3-BDO [16], is undetectable in this work. For this reason, a route
XRD patterns and XPS spectra of catalysts with different
CuO loadings on ZSM-5(280); SEM images; NH -TPD profiles of
3
reduced Cu/ZSM-5(280) with different CuO loadings; three TGA
figures of used catalysts; a figure presenting catalytic results
for the conversion of 2-methylpropanal over reduced catalyst
9.2%CuO/ZSM-5(280); H consumption summarized from TPR;
2
three tables showing the conversion of 2,3-butanediol and selectiv-
ities to the products as functions of CuO loading, temperature and
: 2,3-butanediol ratio. Supplementary data associated with this
article can be found, in the online version, at http://dx.doi.org/10.
016/j.jcat.2015.07.004.
2
to butenes via 1,3-butadiene as an intermediate is possible, but
likely not very important. Once butene compounds are formed,
they can be oligomerized to form dimers, trimers, etc., via a
carbenium-ion mechanism [78], which can further be cracked to
other olefins such as propylene and pentene. In addition,
1
2
2
,3-butanediol can be converted to 3-hydroxy-2-butanone and
,3-butandione via dehydrogenation reactions on copper sites.
H
2
1
4
. Conclusions
References
The conversion of 2,3-butanediol to butenes in a single reactor
[
1] G.W. Huber, J.N. Chheda, C.J. Barrett, J.A. Dumesic, Production of liquid alkanes
by aqueous-phase processing of biomass-derived carbohydrates, Science 308
has been demonstrated with a 70% yield over Cu/ZSM-5. Enhanced
butene yields were achieved with a high SiO /Al ratio, with the
best results achieved at SiO /Al = 280. This is attributed to the
(
2005) 1446–1450, http://dx.doi.org/10.1126/science.1111166.
2
2 3
O
[
2] D. Liu, E. Chen, Integrated catalytic process for biomass conversion and
upgrading to C12 furoin and alkane fuel, ACS Catal. 4 (2014) 1302–1310.
2
2 3
O
lower acidity of the high silica catalysts. Increasing copper loading
had a relatively minor impact on catalytic results, though in gen-
eral butene yield increased with copper loading. Deactivation of
the catalysts due to coke formation was noted on all catalysts as
product selectivities changed over a 5-h run time, but the original
activity could be restored by heating the catalyst in air.
The impact of reaction conditions showed that the optimal tem-
perature is around 250 °C. Lower temperature (230 °C) is beneficial
for dehydrogenation of 2,3-butanediol; higher temperatures
[3] Y.-T. Cheng, G.W. Huber, Chemistry of furan conversion into aromatics and
olefins over HZSM-5: a model biomass conversion reaction, ACS Catal. 1 (2011)
611–628, http://dx.doi.org/10.1021/cs200103j.
[
[
[
4] P. Azadi, R. Carrasquillo-Flores, Y.J. Pagán-Torres, E.I. Gürbüz, R. Farnood, J.A.
Dumesic, Catalytic conversion of biomass using solvents derived from lignin,
Green Chem. 14 (2012) 1573–1576, http://dx.doi.org/10.1039/c2gc35203f.
5] G.W. Huber, S. Iborra, A. Corma, Synthesis of transportation fuels from
biomass: chemistry, catalysts, and engineering, Chem. Rev. 106 (2006) 4044–
4098, http://dx.doi.org/10.1021/cr068360d.
6] J.Q. Bond, A.A. Upadhye, H. Olcay, G.A. Tompsett, J. Jae, R. Xing, et al.,
Production of renewable jet fuel range alkanes and commodity chemicals from
integrated catalytic processing of biomass, Energy Environ. Sci. 7 (2014) 1500–
1523, http://dx.doi.org/10.1039/c3ee43846e.
(
270 °C and 300 °C) favor the oligomerization and cracking reac-
+
tions, resulting in significantly higher selectivities of C
3
and C
5
ole-
[7] X.-J. Ji, H. Huang, J. Du, J.-G. Zhu, L.-J. Ren, N. Hu, et al., Enhanced 2,3-butanediol
production by Klebsiella oxytoca using a two-stage agitation speed control
strategy, Bioresour. Technol. 100 (2009) 3410–3414, http://dx.doi.org/
fins. The results show that hydrogenation reactions are essential
toward getting high butene selectivities, and higher hydrogen to
10.1016/j.biortech.2009.02.031.
2
,3-butanediol ratios are favorable for catalyzing hydrogenation
reactions. Dehydrogenation reactions become dominant for condi-
tions with low hydrogen, especially when H /2,3-BDO ratios are
below 2. The reaction mechanism demonstrated that MEK and
[8] X.-J. Ji, H. Huang, P.-K. Ouyang, Microbial 2,3-butanediol production: a state-
of-the-art review, Biotechnol. Adv. 29 (2011) 351–364, http://dx.doi.org/
10.1016/j.biotechadv.2011.01.007.
2
[
9] X.-J. Ji, H. Huang, J. Du, J.-G. Zhu, L.-J. Ren, S. Li, et al., Development of an
industrial medium for economical 2,3-butanediol production through co-

2
36
Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
fermentation of glucose and xylose by Klebsiella oxytoca, Bioresour. Technol.
00 (2009) 5214–5218, http://dx.doi.org/10.1016/j.biortech.2009.05.036.
10] A. Multer, N. Mcgraw, K. Hohn, P. Vadlani, Production of methyl ethyl ketone
from biomass using a hybrid biochemical/catalytic approach, Ind. Eng. Chem.
Res. 52 (2012) 56–60.
11] P. Perego, A. Converti, A. Del Borghi, P. Canepa, 2,3-Butanediol production by
Enterobacter aerogenes: selection of the optimal conditions and application to
food industry residues, Bioprocess. Eng. 23 (2000) 613–620, http://dx.doi.org/
2
[38] L. Chen, P. Guo, M. Qiao, S. Yan, H. Li, W. Shen, et al., Cu/SiO catalysts prepared
1
by the ammonia-evaporation method: texture, structure, and catalytic
performance in hydrogenation of dimethyl oxalate to ethylene glycol, J.
Catal. 257 (2008) 172–180, http://dx.doi.org/10.1016/j.jcat.2008.04.021.
[39] H. Zhang, Y.-T. Cheng, T.P. Vispute, R. Xiao, G.W. Huber, Catalytic conversion of
biomass-derived feedstocks into olefins and aromatics with ZSM-5: the
hydrogen to carbon effective ratio, Energy Environ. Sci. 4 (2011) 2297–2307,
http://dx.doi.org/10.1039/c1ee01230d.
[
[
1
0.1007/s004490000210.
[40] Z. Xiao, S. Jin, X. Wang, W. Li, J. Wang, C. Liang, Preparation, structure and
catalytic properties of magnetically separable Cu–Fe catalysts for glycerol
hydrogenolysis, J. Mater. Chem. 22 (2012) 16598–16605, http://dx.doi.org/
10.1039/c2jm32869k.
[
12] P. Perego, A. Converti, M. Del Borghi, Effects of temperature, inoculum size and
starch hydrolyzate concentration on butanediol production by Bacillus
licheniformis, Bioresour. Technol. 89 (2003) 125–131, http://dx.doi.org/
1
0.1016/S0960-8524(03)00063-4.
[41] A. Gervasini, Characterization of the textural properties of metal loaded ZSM-5
zeolites, Appl. Catal. A Gen. 180 (1999) 71–82, http://dx.doi.org/10.1016/
S0926-860X(98)00333-0.
[42] C.C. Chusuei, M.A. Brookshier, D.W. Goodman, Correlation of relative X-ray
photoelectron spectroscopy shake-up intensity with CuO particle size,
Langmuir 15 (1999) 2806–2808.
[
[
13] B.C. Saha, R.J. Bothast, Production of 2,3-butanediol by newly isolated
Enterobacter cloacae, Appl. Microbiol. Biotechnol. 52 (1999) 321–326.
14] N. Jansen, M.C. Flickinger, G.T. Tsao, Production of 2,3-butanediol from D-
xylose by Klebsiella oxytoca ATCC 8724, Biotechnol. Bioeng. (1984) 362–369.
XXVI.
[
[
15] A.V. Tran, R.P. Chambers, The dehydration of fermentative 2,3-butanediol into
methyl ethyl ketone, Biotechnol. Bioeng. 27 (1987) 343–351.
[43] C.J.G. Van Der Grift, A.F.H. Wielers, B.P.J. Joghi, J. Van Beijnum, M. De Boer, M.
Versluijs-Helder, et al., Effect of the reduction treatment on the structure and
reactivity of silica-supported copper particles, J. Catal. 131 (1991) 178–189,
http://dx.doi.org/10.1016/0021-9517(91)90334-Z.
[44] J.J. Bravo-Suárez, B. Subramaniam, R.V. Chaudhari, Ultraviolet–visible
spectroscopy and temperature-programmed techniques as tools for
16] W. Zhang, D. Yu, X. Ji, H. Huang, Efficient dehydration of bio-based 2,3-
butanediol to butanone over boric acid modified HZSM-5 zeolites, Green
Chem. 14 (2012) 3441–3450, http://dx.doi.org/10.1039/c2gc36324k.
17] J. Lee, J.B. Grutzner, W.E. Walters, W.N. Delgass, The conversion of 2,3-
butanediol to methyl ethyl ketone over zeolites, Stud. Surf. Sci. Catal. 130
[
[
x
structural characterization of Cu in CuMgAlO mixed metal oxides, J. Phys.
(
2000) 2603–2608.
18] H. Duan, D. Sun, Y. Yamada, S. Sato, Dehydration of 2,3-butanediol into 3-
buten-2-ol catalyzed by ZrO Catal. Commun. 48 (2014) 1–4, http://
Chem. C 116 (2012) 18207–18221, http://dx.doi.org/10.1021/jp303631v.
[45] S.J. Jong, A.R. Pradhan, J.F. Wu, T.C. Tsai, S.B. Liu, On the regeneration of coked
H-ZSM-5 catalysts, J. Catal. 174 (1998) 210–218.
2
,
dx.doi.org/10.1016/j.catcom.2014.01.018.
[46] E.A. Urquieta-Gonzalez, L. Martins, R.P.S. Peguin, M.S. Batista, Identification of
extra-framework species on Fe/ZSM-5 and Cu/ZSM-5 catalysts typical
microporous molecular sieves with zeolitic structure, Mater. Res. 5 (2002)
321–327.
[
[
19] M. Winfield, The catalytic dehydration of 2,3-butanediol to butadiene: II.
Adsorption equilibria, Aust. J. Sci. Res. Ser. A Phys. Sci. 3 (1949) 290–305.
20] B.O. Palsson, S. Fathi-Afshar, D.F. Rudd, E.N. Lightfoot, Biomass as a source of
chemical feedstocks: an economic evaluation, Science 213 (1981) 513–517,
http://dx.doi.org/10.1126/science.213.4507.513.
[47] A. Sultana, T. Nanba, M. Haneda, M. Sasaki, H. Hamada, Influence of co-cations
+
on the formation of Cu species in Cu/ZSM-5 and its effect on selective
[
[
[
[
[
[
[
21] H. Duan, Y. Yamada, S. Sato, Efficient production of 1,3-butadiene in the
catalytic dehydration of 2,3-butanediol, Appl. Catal. A Gen. 491 (2015) 163–
x 3
catalytic reduction of NO with NH , Appl. Catal. B Environ. 101 (2010) 61–67,
http://dx.doi.org/10.1016/j.apcatb.2010.09.007.
1
69.
22] S. Sato, R. Takahashi, T. Sodesawa, A. Igarashi, H. Inoue, Catalytic reaction of
,3-butanediol over rare earth oxides, Appl. Catal. A Gen. 328 (2007) 109–116,
[48] Y. Zhang, I.J. Drake, A.T. Bell, Characterization of Cu-ZSM-5 prepared by solid-
state ion exchange of H-ZSM-5 with CuCl, Chem. Mater. 18 (2006) 2347–2356.
[49] M.K. Neylon, C.L. Marshall, A.J. Kropf, In situ EXAFS analysis of the
temperature-programmed reduction of Cu-ZSM-5, J. Am. Chem. Soc. 124
(2002) 5457–5465.
1
http://dx.doi.org/10.1016/j.apcata.2007.05.033.
23] H. Gotoh, Y. Yamada, S. Sato, Dehydration of 1,3-butanediol over rare earth
oxides, Appl. Catal.
j.apcata.2010.01.025.
A
Gen. 377 (2010) 92–98, http://dx.doi.org/10.1016/
2
[50] Z. Huang, F. Cui, J. Xue, J. Zuo, J. Chen, C. Xia, Cu/SiO catalysts prepared by
hom- and heterogeneous deposition–precipitation methods: texture,
structure, and catalytic performance in the hydrogenolysis of glycerol to 1,2-
propanediol, Catal. Today 183 (2012) 42–51, http://dx.doi.org/10.1016/
j.cattod.2011.08.038.
24] N. Yamamoto, S. Sato, R. Takahashi, K. Inui, Synthesis of homoallyl alcohol
from 1,4-butanediol over ZrO catalyst, Catal. Commun. 6 (2005) 480–484,
2
http://dx.doi.org/10.1016/j.catcom.2005.04.011.
25] S. Sato, R. Takahashi, H. Fukuda, K. Inui, Dehydrogenation of 1,3-butanediol
over Cu-based catalyst, J. Mol. Catal. A: Chem. 272 (2007) 164–168, http://
dx.doi.org/10.1016/j.molcata.2007.03.034.
26] A. Coelho, G. Caeiro, M.a.N.D.a. Lemos, F. Lemos, F.R. Ribeiro, 1-Butene
oligomerization over ZSM-5 zeolite: Part 1 – Effect of reaction conditions,
Fuel 111 (2013) 449–460, http://dx.doi.org/10.1016/j.fuel.2013.03.066).
27] M.E. Wright, B.G. Harvey, R.L. Quintana, Highly efficient zirconium-catalyzed
batch conversion of 1-butene: a new route to jet fuels, Energy Fuels 22 (2008)
[51] Y. Li, S. Liu, Z. Zhang, S. Xie, X. Zhu, L. Xu, Aromatization and isomerization of 1-
hexene over alkali-treated HZSM-5 zeolites: improved reaction stability, Appl.
Catal.
A
Gen.
338
(2008)
100–113,
http://dx.doi.org/10.1016/
j.apcata.2007.12.026.
[52] T. Nanba, S. Masukawa, A. Ogata, J. Uchisawa, A. Obuchi, Active sites of Cu-
ZSM-5 for the decomposition of acrylonitrile, Appl. Catal. B Environ. 61 (2005)
288–296, http://dx.doi.org/10.1016/j.apcatb.2005.05.013.
[53] K. Kubo, H. Iida, S. Namba, A. Igarashi, Ultra-high steaming stability of Cu-ZSM-
5 zeolite as naphtha cracking catalyst to produce light olefin, Catal. Commun.
29 (2012) 162–165, http://dx.doi.org/10.1016/j.catcom.2012.10.004.
3
299–3302.
[
[
28] D.S. Brands, E.K. Poels, A. Bliek, Ester hydrogenolysis over promoted Cu/SiO
catalysts, Appl. Catal. A Gen. 184 (1999) 279–289.
2
0
[54] D. Jin, Z. Hou, L. Zhang, X. Zheng, Selective synthesis of para–para -
29] L. Guo, J. Zhou, J. Mao, X. Guo, S. Zhang, Supported Cu catalysts for the selective
hydrogenolysis of glycerol to propanediols, Appl. Catal. A Gen. 367 (2009) 93–
dimethyldiphenylmethane over H-beta zeolite, Catal. Today 131 (2008) 378–
384, http://dx.doi.org/10.1016/j.cattod.2007.10.049.
9
8, http://dx.doi.org/10.1016/j.apcata.2009.07.040.
3
[55] F. Lonyi, J. Valyon, On the interpretation of the NH -TPD patterns of H-ZSM-5
[
[
[
[
[
[
30] S. Sitthisa, D.E. Resasco, Hydrodeoxygenation of furfural over supported metal
catalysts: a comparative study of Cu, Pd and Ni, Catal. Lett. 141 (2011) 784–
and H-mordenite, Microporous Mesoporous Mater. 47 (2001) 293–301.
[56] M. Niwa, S. Nishikawa, N. Katada, IRMS–TPD of ammonia for characterization
of acid site in b-zeolite, Microporous Mesoporous Mater. 82 (2005) 105–112,
http://dx.doi.org/10.1016/j.micromeso.2005.03.002.
7
91, http://dx.doi.org/10.1007/s10562-011-0581-7.
31] E.S. Vasiliadou, T.M. Eggenhuisen, P. Munnik, P.E. De Jongh, K.P. De Jong, A.A.
Lemonidou, Synthesis and performance of highly dispersed Cu/SiO catalysts
2
[57] E.-Y. Choi, I.-S. Nam, Y.G. Kim, TPD study of mordenite-type zeolites for
for the hydrogenolysis of glycerol, Appl. Catal. B Environ. 145 (2014) 108–119.
32] S. Sato, M. Akiyama, R. Takahashi, T. Hara, K. Inui, M. Yokota, Vapor-phase
reaction of polyols over copper catalysts, Appl. Catal. A Gen. 347 (2008) 186–
selective catalytic reduction of NO by NH
[58] O. Mihai, C.R. Widyastuti, S. Andonova, K. Kamasamudram, J. Li, S.Y. Joshi,
et al., The effect of Cu-loading on different reactions involved in NH -SCR over
3
, J. Catal. 161 (1996) 597–604.
3
1
91, http://dx.doi.org/10.1016/j.apcata.2008.06.013.
Cu-BEA catalysts, J. Catal. 311 (2014) 170–181, http://dx.doi.org/10.1016/
j.jcat.2013.11.016.
[59] A. Gervasini, M. Manzoli, G. Martra, A. Ponti, N. Ravasio, L. Sordelli, et al.,
Dependence of copper species on the nature of the support for dispersed CuO
catalysts, J. Phys. Chem. B. 110 (2006) 7851–7861, http://dx.doi.org/10.1021/
jp056604c.
[60] S. Contarini, L. Kevan, X-ray photoelectron spectroscopic study of copper-
exchanged X- and Y-type sodium zeolites: resolution of two cupric ion
components and dependence on dehydration and x-irradiation, J. Phys. Chem.
90 (1986) 1630–1632, http://dx.doi.org/10.1021/j100399a035.
[61] J.P. Espinós, J. Morales, A. Barranco, A. Caballero, J.P. Holgado, A.R. Gonza,
33] M. Iwamoto, H. Yahiro, Y. Torikai, T. Yoshioka, N. Mizuno, Novel preparation
method of highly copper ion-exchanged ZSM-5 zeolites and their catalytic
activities for NO decomposition, Chem. Lett. (1990) 1967–1970.
34] M. Schreier, S. Teren, L. Belcher, J.R. Regalbuto, J.T. Miller, The nature of
‘
‘overexchanged’’ copper and platinum on zeolites, Nanotechnology 16 (2005)
S582–S591, http://dx.doi.org/10.1088/0957-4484/16/7/036.
35] Y.-Y. Zhu, S.-R. Wang, L.-J. Zhu, X.-L. Ge, X.-B. Li, Z.-Y. Luo, The influence of
2
copper particle dispersion in Cu/SiO catalysts on the hydrogenation synthesis
of ethylene glycol, Catal. Lett. 135 (2010) 275–281, http://dx.doi.org/10.1007/
s10562-010-0298-z.
[
[
36] J. Gong, H. Yue, Y. Zhao, S. Zhao, L. Zhao, J. Lv, et al., Synthesis of ethanol via
Interface effects for Cu, CuO, and Cu
determination of the valence state of copper in Cu/SiO
J. Phys. Chem. B. 106 (2002) 6921–6929.
2
O deposited on SiO
2
and ZrO
2
. XPS
0
+
syngas on Cu/SiO
2012) 13922–13925, http://dx.doi.org/10.1021/ja3034153.
37] S. Zhao, H. Yue, Y. Zhao, B. Wang, Y. Geng, J. Lv, et al., Chemoselective synthesis
of ethanol via hydrogenation of dimethyl oxalate on Cu/SiO enhanced
stability with boron dopant, J. Catal. 297 (2013) 142–150, http://dx.doi.org/
0.1016/j.jcat.2012.10.004.
2
catalysts with balanced Cu –Cu sites, J. Am. Chem. Soc. 134
2
and Cu/ZrO catalysts,
2
(
[62] T. Böske, K. Maiti, O. Knauff, K. Ruck, M.S. Golden, G. Krabbes, et al., Cu–O
network-dependent core-hole screening in low-dimensional cuprate
2
:
systems:
a high-resolution X-ray photoemission study, Phys. Rev. B. 57
1
(1998) 138–141.

Q. Zheng et al. / Journal of Catalysis 330 (2015) 222–237
237
[
[
63] P.A. Torresi, V.K. Díez, P.J. Luggren, J.I. Di Cosimo, Conversion of diols by
dehydrogenation and dehydration reactions on silica-supported copper
catalysts, Appl. Catal. A Gen. 458 (2013) 119–129, http://dx.doi.org/10.1016/
j.apcata.2013.03.031.
64] I. Bucsi, A. Molnar, M. Bartok, G.A. Olah, Transformation of 1,2-diols over
perfluorinated resinsulfonic acid (Nafion-H), Tetrahedron 50 (1994) 8195–
[71] M.J. Wulfers, F.C. Jentoft, Identification of carbonaceous deposits formed on H-
mordenite during alkane isomerization, J. Catal. 307 (2013) 204–213, http://
dx.doi.org/10.1016/j.jcat.2013.07.011.
[72] H. Zhang, S. Shao, R. Xiao, D. Shen, J. Zeng, Characterization of coke deposition
in the catalytic fast pyrolysis of biomass derivates, Energy Fuels 28 (2014) 52–
57, http://dx.doi.org/10.1021/ef401458y.
8
202.
[73] S.K. Sahoo, S.S. Ray, I.D. Singh, Structural characterization of coke on spent
hydroprocessing catalysts used for processing of vacuum gas oils, Appl. Catal.
A Gen. 278 (2004) 83–91, http://dx.doi.org/10.1016/j.apcata.2004.09.028.
[74] I.V. Kozhevnikov, S. Holmes, M.R.H. Siddiqui, Coking and regeneration of
[
[
65] R. Emerson, Kinetics of dehydration of aqueous 2, 3-butanediol to methyl ethyl
ketone, Ind. Eng. Chem. Prod. Res. Dev. 21 (1982) 473–477.
66] B. Wang, G. Manos, A novel thermogravimetric method for coke precursor
characterisation, J. Catal. 250 (2007) 121–127, http://dx.doi.org/10.1016/
j.jcat.2007.05.018.
H
3
2
PW12O40/SiO catalysts, Appl. Catal. A Gen. 214 (2001) 47–58, http://
dx.doi.org/10.1016/S0926-860X(01)00469-0.
[
[
67] H.S. Cerqueira, G. Caeiro, L. Costa, F. Ramôa Ribeiro, Deactivation of FCC
catalysts, J. Mol. Catal. A: Chem. 292 (2008) 1–13, http://dx.doi.org/10.1016/
j.molcata.2008.06.014.
[75] C.-J. Jia, Y. Liu, W. Schmidt, A.-H. Lu, F. Schüth, Small-sized HZSM-5 zeolite as
highly active catalyst for gas phase dehydration of glycerol to acrolein, J. Catal.
269 (2010) 71–79, http://dx.doi.org/10.1016/j.jcat.2009.10.017.
[76] M. Guisnet, P. Magnoux, Organic chemistry of coke formation, Appl. Catal. A
Gen. 212 (2001) 83–96, http://dx.doi.org/10.1016/S0926-860X(00)00845-0.
[77] D.P. Serrano, J. Aguado, J.M. Rodríguez, A. Peral, Catalytic cracking of
polyethylene over nanocrystalline HZSM-5: catalyst deactivation and
regeneration study, J. Anal. Appl. Pyrolysis. 79 (2007) 456–464, http://
dx.doi.org/10.1016/j.jaap.2006.11.013.
68] Y. Zhang, N. Zheng, K. Wang, S. Zhang, J. Wu, Effect of copper nanoparticles
2
dispersion on catalytic performance of Cu/SiO catalyst for hydrogenation of
dimethyl oxalate to ethylene glycol, J. Nanomater. 2013 (2013) 1–6.
69] C. Bearez, F. Chevalier, M. Guisnet, Mechanism of butane transformation on H
mordenite I kinetic study, React. Kinet. Catal. Lett. 22 (1983) 405–409.
70] G. Caeiro, R.H. Carvalho, X. Wang, M.A.N.D.A. Lemos, F. Lemos, M. Guisnet,
[
[
et al., Activation of C
Mol. Catal. A: Chem. 255 (2006) 131–158, http://dx.doi.org/10.1016/
j.molcata.2006.03.068).
2
–C
4
alkanes over acid and bifunctional zeolite catalysts, J.
[78] Y. Kissin, Chemical mechanisms of catalytic cracking over solid acidic
catalysts: alkanes and alkenes, Catal. Rev. 43 (2001) 85–146.