Applied Energy 209 (2018) 1–7

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Direct syngas conversion to liquefied petroleum gas: Importance of a MARK

multifunctional metal-zeolite interface
⁎
Peng Lua, Jian Sunb, , Dongming Shena, Ruiqin Yanga, Chuang Xinga, Chengxue Lua,
⁎
Noritatsu Tsubakia,c, Shengdao Shana,
a
School of Biological & Chemical Engineering/School of Light Industry, Zhejiang University of Science and Technology, Hangzhou 310023, PR China
b
Dalian National Laboratory for Clean Energy (DNL), Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, PR China
c
Department of Applied Chemistry, Graduate School of Engineering, University of Toyama, Toyama 930-8555, Japan

H I G H L I G H T S G RA P H I C A L AB S T R A C T

• Awell-defined
CuZnAl@H-Beta catalyst with
metal-zeolite interface
a

was fabricated.
• We realized an efficient and tandem
conversion of syngas to LPG.
• The LPG selectivity of in hydrocarbons
reaches as high as 77%.
• Alectivity
record low methane and C2 se-
is achieved (< 2.0%).

A R T I C L E I N F O A B S T R A C T

Keywords: It is challenging to fabricate a multifunctional catalyst for consecutively catalyzing multiple reactions. Herein,
Multifunctional catalyst we report a well-defined metal-zeolite interface derived from a CuZnAl@H-Beta core@shell catalyst to realize
Syngas one-step syngas conversion to liquefied petroleum gas (LPG). The multifunctional interface between CuZnAl core
LPG and H-Beta zeolite shell is composed of Cu and acid zeolite with a content gradient, through which synthesized
Interface
methanol via syngas on the core will pass. The interface is able to catalyze the tandem dehydration of methanol
Zeolite
to olefins on acid sites and olefin hydrogenation to C3–4 saturated hydrocarbons (LPG fraction) over exposed Cu
sites on the interface instead of noble metals in conventional catalysts. The selectivity of LPG in hydrocarbons
over the prepared capsule catalyst reaches as high as 77% accompanied by a record low methane and C2 se-
lectivity (< 2.0%).

1. Introduction alternative energy sources, especially clean liquid fuels. Conversion of

syngas (CO and H2) derived from coal, natural gas, waste and biomass
The increasing energy demands worldwide expedite the quest for to hydrocarbons is a well candidate for production of valuable

⁎
Corresponding authors.
E-mail addresses: sunj@dicp.ac.cn (J. Sun), 100099@zust.edu.cn (S. Shan).

http://dx.doi.org/10.1016/j.apenergy.2017.10.068
Received 12 July 2017; Received in revised form 5 October 2017; Accepted 22 October 2017
0306-2619/ © 2017 Elsevier Ltd. All rights reserved.
P. Lu et al.
chemicals and fuels [1–4]. This process is well-known as Fischer–-
Tropsch (F-T) synthesis, which is successfully commercialized for dec-
ades. The high temperature F–T technology applied by Sasol process in
South Africa is the largest commercial scale application with the mature
Sasol Advanced Synthol (SAS) technology [5,6]. However, the con-
ventional F-T synthesis, generally follows the Anderson–Schulz–Flory
distribution law, producing a very wide range of olefins, paraffins and
oxygenated compounds [7–9]. It is challenging to control hydrocarbon
selectivity with certain narrow carbon number range.
Liquefied petroleum gas (LPG), a mixture of propane and butanes
isomers, which is generally derived from refineries, is regarded as an
environmentally benign and high-combustion value fuel due to its low
carbon/hydrogen ratio, no toxicity, no corrosive activity and no aro-
matic hydrocarbons compared to traditional fuels [10–12]. It can be
widely employed as a clean fuel, chemical feed and a propellant for
aerosols. The primary driving force to the LPG application remains its
low price for the end user, low pollutant emissions (especially carbon
dioxide), will probably increase the interests in LPG as an internal
combustion engine fuel. Nowadays, there are continuously increasing
stock production of dual-fuel (gasoline–LPG) passenger car models. To
date, LPG engines has been commercially available in the EU and Asian
countries for many years [13]. The LPG available for the automotive
market comply with a standard that does not define compositions, but
limits fuel properties only [10].
There are following pathways for LPG synthesis: (1) direct synthesis
from syngas; (2) indirect or semi-indirect synthesis from syngas via
multiple stage reactors, including synthesis of methanol or DME from
syngas and conversion of methanol or DME into hydrocarbons of LPG
fraction [14,15]. From the viewpoint of energy production, the direct
route of LPG synthesis from syngas is more potential and advantageous
to large-scale application if compared to the indirect or semi-indirect
synthesis, owing to its low energy cost, low investment scale and facile
heating recycle of tail gas.
However, it is very challenging for direct synthesis of LPG from
syngas over one multiple functional catalyst comprising various cata-
lytically active sites, such as methanol synthesis, methanol dehydration
to DME, DME dehydration to olefins, and olefin hydrogenation to sa-
turated hydrocarbons. For constructing such multiple functional cata-
lysts, a general strategy is physical mixing methanol synthesis catalyst
(such as Cu/ZnO/Al2O3) and a noble metal-modified zeolite for a hy-
brid catalyst [16–19]. Both of the Cu/ZnO/Al2O3 and zeolite catalysts
have been commercially applied. The former is the most commonly
employed catalysts for methanol synthesis in industry [20]. The latter is
successfully served in the field of large-scale separation and purification
and catalytic dehydration process [21]. Unfortunately, it is difficult to
control distribution of active sites over these hybrid catalysts. The
catalytic performance is easily influenced by the contact state between
various hybrid components.
In recent years, the spherical zeolite capsule structure with metal-
based catalyst as core and zeolite layer as shell is been widely reported
as efficient catalysts in tandem catalytic reactions, such as syngas to
isoparaffins [22], syngas to DME [23,24], glycerol conversion to 1,2-
PDO and 1,3-PDO [25]. However, in above cases, only typical tandem
reactions including two steps are realized. There are no reports con-
cerning the possibility of three or more catalytic tandem reactions over
zeolite capsule catalysts.
Herein, we report one-step controlled tandem catalytic production
of LPG from syngas achieved by a core–shell capsule catalyst CZA@β
comprising Cu/ZnO/Al2O3 as core and H-Beta zeolite as shell. The in-
terface with controlled gradient of Cu and zeolite between two com-
ponents is key to the tandem process of three or more catalytic reac-
tions. The advantages of core–shell catalyst prepared with the
hydrothermal method is demonstrated by comparing with physical
mixed catalyst Mix-CZA-β and physically prepared core–shell catalyst
CZA@β-P in detail. In addition, obvious economy and social value will
be produced in the real application of direct conversion of syngas to
2
Applied Energy 209 (2018) 1–7
LPG from the estimation of production in a pilot-scale plant.
2. Experimental section
2.1. Catalyst preparation
The Cu/ZnO/Al2O3 (CZA) catalysts were prepared by coprecipita-
tion method. The Al(NO3)4·9H2O (2.49 g), Cu(NO3)2·3H2O (7.139 g)
and Zn(NO3)2·6H2O (8.790 g) was added into 300 mL deionized water,
then the aqueous solution and Na2(CO3) solution (0.5 M) were dropwise
(3 mL min−1) into another 300 mL deionized water with pH and tem-
perature of 8.0 and 60 °C, respectively. After aged for 12 h, the obtained
slurry was washed with deionized water for several times to remove the
excessive sodium ions, and subsequently, dried for 12 h, calcined at
350 °C for 2 h and pelletized to 20–40 mesh, this CZA as the core cat-
alyst.
The template agent TEAOH (Tetraethyl ammonium hydroxide, 25%
in water), fumed silica (99.5%), aluminium isopropoxide (99.0%), po-
tassium nitrate (99.0%) and dehydrated ethanol (99.5%) were pur-
chased from Aladdin Co. The zeolite synthesis recipe with molar ratio of
96.53 SiO2:34.55 TEAOH:1.0 Al2O3:1130 H2O:0.00148 KNO3 was used
to hydrothermal synthesis of β zeolite. This slurry stirred for 6 h at room
temperature in a Teflon container, which is defined as the mother liquid
of H-β zeolite. Then it is moved the Teflon container into homogeneous
reactor for hydrothermal synthesis with the rotation rate of 2 rpm at
155 °C for 72 h. The final samples were separated from the slurry and
ion-exchanged with 100 mL ammonium nitrate (0.5 M) for 10 h at
50 °C, dried at 120 °C for 12 h and calcined at 550 °C for 5 h to remove
the organic template, H-β zeolite was obtained.
The mother liquid of H-β zeolite was used to prepare H-β zeolite
shell on CZA core. After this slurry stirred for 6 h at room temperature
in a Teflon container, CZA core was added into the slurry and repeated
the zeolite hydrothermal synthesis, drying and calcination steps, the
catalyst obtained by this process was named “CZA@β”.
Another CZA core catalyst was impregnated in silica sol (Ludox),
then the wetted CZA catalyst was moved into a crucible with appro-
priate amount of H-β zeolite powder. The catalyst obtained after strong
swaying was named “CZA@β-P”.
The physical mixed catalyst Mix-CZA-β is prepared by mixing CZA
and H-β zeolite powders with a mortar.
All the catalysts are pelletized to 20–40 mesh. All the weight ratio of
CZA to H-β zeolite for CZA@β, CZA@β-P, Mix-CZA-β is controlled as
the same value of 10:1.
2.2. Catalyst characterization
X-ray diffraction (XRD) patterns were recorded using a D8
ADVANCE diffractometer with Cu-Kα radiation (40 kV, 40 mA) as the
X-ray source.
Scanning electron microscopy (SEM) and energy dispersive spec-
troscopy (EDS) analysis were carried out using JSM-6360LV scanning
electron microscope with 15 kV accelerating voltage.
NH3 temperature-programmed desorption (NH3-TPD) measure-
ments were performed on BELCAT-B3 instrument. Typically, 30 mg
sample was pretreated in a quartz reactor with He flow at 300 °C for 1 h,
then sample was cooled down to 373 K and switched the gas to NH3-He
(10 vol% NH3) for 1 h. Sample was flushed using He of 30 mL min−1 to
remove the gas phase NH3. The NH3-TPD experiment was performed in
He flow of 30 mL min−1 by increasing the temperature from 100 °C to
650 °C at a hating rate of 10 K min−1, the desorbed NH3 was detected
by a thermal conductivity detector (TCD).
The specific surface areas of catalyst were measured by N2 physical
adsorption experiments by Quantachrome Autosorb-IQ-C at −196 °C.
The sample was outgassed at 200 °C for 1 h before N2 physisorption.
Surface area and pore size were calculated by BET and BJH method,
respectively.
P. Lu et al.
2.3. Catalytic performance tests
LPG direct synthesis reactions were performed on a high-pressure
fixed-bed flow reactor (i.d. 9 mm). The catalyst (0.5 g) with grain sizes
of 20–40 mesh loaded into the reactor was pretreated in H2 of
30 mL min−1 at 220 °C for 2 h before reaction. After the reactor was
heating to 350 °C, the syngas of CO/H2/Ar = 1/2/1 and pressure
3.0 MPa was introduced into the reactor. All the products out from
reactor in a gaseous state were analyzed by two online gas chromato-
graphs. Ar, CO, CO2 and CH4 were analyzed by a thermal conductivity
detector (TCD) with TDX-01 column. The organic compounds of hy-
drocarbons and oxygenates were analyzed by a flame ionization de-
tector (FID) with PLOT/Q column. The evaluation of catalytic activity
was carried out under the conditions of 350 °C, 3.0 MPa, and 2400 h−1.
All of the selectivity values of products in this work were calculated in
carbon molar base.
In detail, CO conversion and products selectivity in catalytic reac-
tions were calculated as the followed equations:
CO conv. %. =(MCO, in − MCO, out )/(MCO, in ) × 100%
MCO, in: Mole fraction of CO in original syngas;
MCO, out: Mole fraction of CO in post-reaction gas;
Xi, n sel. = nXi/(∑nXi)
Xi, n sel.: Carbon molar selectivity values of products;
n: number of carbons in single molecule product;
Xi: products including “CO2, CH4, C2, MeOH, DME, C3, C4 and C5+”,
all the carbon number of C5+ products are regarded as 5 in this
paper.
3. Results and discussion
3.1. Physical properties
The surface morphology of prepared catalysts is shown by SEM
images in Fig. 1a–d. The CZA catalyst prepared by the coprecipitation
method is composed of irregular particles with the Cu:Zn:Al ratio of
46.2:43.3:10.6 detected by EDS attachment. The self-prepared H-Beta
zeolite comprises well dispersed nanoparticles of ca. 200 nm with a Si/
Al ratio of 95.1:4.9. Besides, we can also control the ratio of Cu:Zn:Al in
the CZA core by employing different amount in raw materials. For in-
stance, a series of CuxZnyAlz catalysts with different molar ratios can be
controllably prepared and noted as Cu1Zn1Al0.09, Cu1Zn1Al0.13,
3
Applied Energy 209 (2018) 1–7
Cu1Zn1Al0.29, and Cu1Zn1Al0.33 by a conventional co-precipitation
method. The surface morphology and element composition (not shown
in the manuscript) are similar to the Cu1Zn1Al0.23 catalyst with an
optimized Cu:Zn:Al ratio as shown in Fig. 1.
For the physical mixed Mix-CZA-β catalyst, particles of CZA and H-
Beta components are randomly dispersed and ranged in very wide
distribution. However, the surface morphology and Si/Al ratio of hy-
drothermal CZA@β catalyst are similar to those of H-Beta zeolite,
suggesting well coverage of zeolite layer on the CZA core. It is note-
worthy that few copper element of 4.8% is detected on CZA@β surface.
Since Cu and Zn content in the core CZA catalyst is close, the possibility
of incomplete zeolite coverage on the CZA catalyst can be excluded
resulting from the absence of Zn in Fig. 1d. Accordingly, it is speculated
that part of CuO in the CZA core are dissolved in alkaline hydrothermal
raw solution for H-Beta zeolite synthesis containing organic ammonia
[22,26].
The physical adsorption by N2 at −196 °C (Fig. 1e) is employed for
comparing the BET surface area and pore struture for various catalysts.
Both of the Mix-CZA-β and CZA@β-P catalyst with the same ratio of
CZA:H-Beta of 10:1 show similar adsorption–desorption isotherm
curves to that of CZA. The hysteresis loop at a high relative pressure (P/
P0) show characteristic of mesopores, resulting from accumulation
among CuO, ZnO, Al2O3 and zeolite [27]. The CZA@β catalyst de-
monstrates an additional precipitous adsorption in a low P/P0 range
except mesoporous property, suggesting the existence of abundant
micropores. The superiority of CZA@β catalyst is apparent in terms of
possessing micropores from H-Beta shell and mesopores from CZA core.
On the other hand, the BET surface area of CZA@β reaches as high as
437 m2 g−1, close to that of H-Beta zeolite, and much higher than that
Mix-CZA-β (125 m2 g−1) and CZA@β-P (146 m2 g−1) catalyst. The high
surface area is beneficial to exposing more active sites and mass transfer
in multiple catalytic reactions, obviously promoting the catalytic per-
formance [28].
3.2. Chemical state of the synthesized core–shell catalyst
The crystalline phases of various catalysts are demonstrated by XRD
analysis (Fig. 2). Compared with the raw H-Beta and CZA catalyst, all of
the prepared three composite catalysts (CZA@β-P, Mix-CZA-β, and
CZA@β show typical H-Beta zeolite (2 theta of 8° and 23°), CuO and
ZnO phase with different intensity. For the case of CZA@β prepared by
a hydrothermal method, a well crystalline of H-Beta zeolite with strong
Fig. 1. Physical properties of various catalysts:
SEM images and EDS analysis of CZA (a), H-Beta
(b), Mix-CZA-β (c), CZA@β (d), N2 adsorption
isotherm curves (e).
P. Lu et al. Applied Energy 209 (2018) 1–7

Table 1
Catalytic performance of syngas conversion to LPG.

Catalyst Conv./% Product Sel./% LPG

(C3–4) in
CO2 CH4 C2 MeOH DME C3–4 C5+ HCsa/%

Mix-CZA-β 34.0 55.9 10.8 4.1 0 7.0 19.2 3.0 51.7

CZA@β-P 33.7 52.4 11.7 7.1 0 0 25.1 3.7 52.7
CZA@β 50.2 48.4 6.6 3.9 0 0 37.3 3.7 72.4

a
LPG in Hydrocarbons (HCs) is calculated by the ratio of C3–4/
(CH4 + C2 + C3–4 + C5+); Reaction condition: 350 °C, 3.0 Mpa, Wcatalyst/
Fsyngas = 10 g h mol−1;

Fig. 2. XRD patterns of various catalysts.

intensity is observed, illustrating that H-Beta zeolite is well grown on

CZA particles. Furthermore, both of the intensity of CuO and ZnO are
obviously decreased for CZA@β, especially for weak ZnO phase. It is
suggested that the thickness of zeolite coverage is probably ranged in
the level of dozens of microns as limited by bulk detection depth of XRD
with Cu target [29]. This finding is well consistent with the element
analysis result from SEM.
To disclose the chemical structure of CZA@β in detail, we selected a
red line for EDS scan from left to right in the cross-section SEM image
(Fig. 3a), and the corresponding element intensity curves are shown in Fig. 4. The variation of LPG yield during a time-on-stream of 360 min.
Fig. 3b. The CZA core ranged from 0 to 20 m, comprising Cu, Zn and Al
with a similar composition to CZA catalyst. The H-Beta shell with a
selectivity in hydrocarbons (ca. 52%). For the case of CZA@β catalyst
thickness of 13 m is covered on the CAZ core ranging from ca. 27 m to
prepared by a hydrothermal method, the catalytic performance is ob-
40 m with a close Si/Al ratio to H-Beta zeolite. This result suggests the
viously increased compared to two previous catalysts. The LPG se-
formation of well-defined core@shell structure on CZA@β catalyst.
lectivity in hydrocarbons climbs to as high as 72.4% with a CO con-
Between the two components, a clear transition layer is identified with
version of 50.2%. The catalytic reactions on various catalysts were
a thickness of ca. 7 m, in which the content of Cu, Zn, Al is gradually
performed under a time-on-stream of 360 min (Fig. 4). The LPG yield on
decreased accompanied by distinct promotion of Si content along the
the CZA@β catalyst shows much higher and more stable than other
direction from core to shell. In this transition layer, the increasing in-
two.
tensity of Si content is still lower than the stage of Si content in the
The reason lies on the fact that the well-defined core–shell zeolite
shell. The interface between core and shell allows a gradient of Cu
encapsulated structure provides a confinement effect during the syngas
content, which can catalyze olefin hydrogenation to saturated hydro-
to LPG reaction as shown in Scheme 1. Generally, direct synthesis of
carbons of LPG fraction without noble metal modified zeolite.
LPG from syngas contains four major reactions: methanol synthesis on
CuZnAl core catalyst (1), methanol dehydration to dimethyl ether on
3.3. Catalytic performance of syngas conversion to LPG zeolite shell (2), further dehydration of dimethyl ether to generate
hydrocarbons (3) and the water–gas-shift (WGS) reaction (4)
The catalytic performances of syngas conversion to LPG on Mix-
CZA-β, CZA@β-P and CZA@β catalyst are compared under the same CO + 2H2 → CH3 OH; (1)
reaction conditions (350 °C, 3.0 Mpa, W/F = 10 g mol/h) as shown in
Table 1. The physical mixed catalyst and physically prepared core–shell
2CH3 OH → CH3 OCH3 + H2 O; (2)
catalyst CZA@β-P exhibit similar CO conversion (ca. 34%) and LPG

Fig. 3. EDS line scan of selected cross-section core–shell

area.

4
P. Lu et al. Applied Energy 209 (2018) 1–7

Scheme 1. Schematic of direct syngas conversion to LPG on the prepared CZA@β catalyst
with a confined metal-zeolite interface.
Fig. 5. NH3-TPD curves for various catalysts.

CH3 OCH3 → Hydrocarbons + H2 O; (3)

CO + H2 O→ CO2 + H2;
Total equation: CO + H2 → CO2 + Hydrocarbons
CO comes from raw material gas, and H2O comes from Eqs. (2) and
(3). CO2 is generated by the WGS reaction as a by-product catalyzed by
the Cu-based core catalyst. The methanol synthesis on the CZA core (1)
is the key reaction in the reaction system constrained by thermo-
dynamic equilibrium. The generated methanol is dehydrated into di-
methyl ether (2) and even hydrocarbons (3) on acidic sites of zeolite
layer. This process will break the thermodynamic equilibrium of me-
thanol direct synthesis, promoting the positive shift of the reaction
balance with water generated at the same time. Enough CO and water
promotes the generation of CO2 over Eq. (4).
It is important to point out that the CO2 decreases from 55.9% on
the Mix-CZA-β catalyst to 48.4% on the core–shell catalyst. The lowest
CO2 selectivity over CZA@β among three catalysts demonstrates that
compared to the random dispersed Cu and acidic sites on physical
mixed catalysts, the well-defined core–shell structure can efficiently
and quickly separate water produced from zeolite dehydration (2) and
(3), producing a low H2O/CO ratio in the metal-zeolite interface. This
will prevent its excessive contact to CO and suppress the WGS reaction
(4). Besides, CO2 selectivity will enhance with the increasing reaction
temperature below 350 °C since the WGS reaction rate is promoted,
while the exothermic WGS is obviously suppressed if the temperature is
continuously increased to 400 °C.
For the core–shell catalyst, the Cu-Zn-Al core can firstly catalyze
methanol synthesis from CO and H2. Secondly, methanol produced in
the CZA core must pass through the interface with a Cu gradient and
acidic zeolite layer, in which methanol dehydration to DME and DME
dehydration to olefins occurs in sequence on acidic centers. Finally,
olefin hydrogenation to saturated C3–4 hydrocarbons on exposed Cu
sites of interface layer. Totally, two different kinds of copper phases
exist in the capsule catalyst. One of those is present in the CZA core,
catalyzing methanol synthesis assisted with Zn promoter and Al2O3
support, while another is exposed on the interface layer instead of noble
metal in conventional composite catalysts, promoting olefin hydro-
genation to final LPG product.
In addition, it is noteworthy that a suitable acidity is also very im-
portant for the control of product distribution in syngas conversion
probably attributed to strong acidic sites or ammonia adsorption on
exposed and unencapsuled Cu-Zn sites on surface. These strong ad-
(4)
sorbed sites can accelerate the water gas shift (WGS) reaction to CO2,
which is also one of the reason why the CO2 selectivity on CZA@β is
less than that on those two physical prepared catalysts.
A series of catalytic reactions with different reaction temperature
(260–400 °C) on the CZA@β catalyst for comparing the effect of reac-
tion temperature in detail. From Table 2, CO conversion is lower than
30% and DME is the primary product with the temperature below
300 °C. When the temperature climbs to a higher temperature
(320–400 °C), CO2, methane and C2 product is clearly enhanced owing
to the excessive hydrocracking despite an increase of CO conversion.
However, a clear decrease of CO conversion at a temperature higher
than 350 °C is observed owing to the thermodynamic restraint of exo-
thermic reaction of methanol synthesis and methanol dehydration. The
hydrocracking of heavy hydrocarbons on zeolite is also enhanced at
higher temperature, producing more low carbon product, such as me-
thane and C2. On the other hand, the dehydration ability of H-Beta
zeolite is slightly weakened at a low temperature of 300 °C, bringing
about few methanol (0.7%) and DME (5.0%) intermediates in the
product. However, the selectivity towards methane or C2 is also clearly
decreased, the sum of which is only 2.0%, reaching the lowest value
among reported literatures. In parallel, more LPG with a selectivity in
hydrocarbons of as high as 77% is produced, suggesting the hydro-
cracking reaction of hydrocarbons with a carbon number higher than 3
is severely suppressed. Therefore, the optimized reaction temperature is
confirmed as 300 °C with the highest LPG selectivity of 77.0% in hy-
drocarbons in this work. It is notable that the optimized reaction con-
dition, especially the temperature, is different in the LPG production
work of related literatures as the chemical structure and physical
composition vary in multiple catalysts.
Table 2
Effect of temperature on catalytic performance of syngas conversion to LPG.
Catalyst Conv./% Product Sel./% LPG
(C3–4)
CO2 CH4 C2 MeOH DME C3–4 C5+ in
HCs/%

[24,30]. The NH3-TPD curves as shown in Fig. 5 demonstrate that the CZA@β-260 16.1 19.6 0.7 3.8 2.1 54.4 12.8 6.6 53.5
medium strong acid at an ammonia desorption temperature of ca. CZA@β-280 21.3 28.6 1.0 2.1 0.6 32.8 26.9 8.0 70.7
350–500 °C responsible for deep dehydration of methanol to hydro- CZA@β-300 30.3 44.6 1.1 0.9 0.7 5.0 38.3 9.5 77.0
carbons is maintained, while weak acid sites below 250 °C responsible CZA@β-320 43.6 45.7 7.5 2.1 0 0 36.7 8.0 67.6
CZA@β-350 50.2 48.4 6.6 3.9 0 0 37.3 3.7 72.4
for methanol initial dehydration to DME are decreased if compared to
CZA@β-400 25.1 38.6 13.5 14.3 0 0 29.6 4.1 48.2
raw H-Beta zeolite [31]. However, the physical prepared CZA@β-P and
Mix-CZA-β reserved stronger and sharper adsorption at higher tem- Note. CZA@β-T stands for the catalyst used in the temperature of T °C. Reaction condi-
perature in comparison to the raw CZA and CZA@β catalyst, which is tion: 3.0 Mpa, Wcatalyst/Fsyngas = 10 g mol/h.

5
P. Lu et al.
3.4. State of active sites after catalytic reaction
To demonstrate the state of active sites after catalytic reactions, the
XRD patterns are investigated over used Mix-CZA-β, used CZA@β-p,
and used CZA@β in Fig. 6a. All the three used catalysts are composed of
H-Beta, metallic Cu and ZnO phases with different diffraction intensity.
The used CZA@β catalyst seems to exhibit sharper ZnO diffraction
peaks compared with the other two, especially for the (1 0 0), (0 0 2),
(1 0 1) lattice planes at a 2 theta range from 30 to 50 degrees. Si-
multaneously, the primary peak of metallic Cu at ca. 43.2 degree on the
CZA@β catalyst is slightly widen in contrast to the other two. Herein,
we employ the Scherrer formula for estimating the particle size of
metallic Cu and ZnO after catalytic reactions. Based on the XRD peak at
43.2 degree for Cu and that at 56.9 degree for ZnO, the histogram of
particle sizes is compared and shown in Fig. 6b. Interestingly, the
CZA@β catalyst presents the smallest metallic Cu particle size of 9.1 nm
and the biggest ZnO particle size of 5.6 nm among three used catalysts.
This finding suggests that the embedding structure of part Cu nano-
particles (NPs) in the metal-zeolite interface, rather than that of phy-
sically mixing and physically coating, is obviously beneficial to con-
fining these Cu species against aggregation in a long time catalytic
reaction.
As for the ZnO specie as a promoter in Cu-based catalysts, the XRD
diffraction intensity is enhanced as compared to the fresh catalyst
(Fig. 2), suggesting that the ZnO NPs is grown in the catalytic reactions
without the zeolite confinement effect like Cu in the CZA@β catalyst.
One of the possible reason lies in more interspace is formed for ZnO in
the core after the migration of Cu NPs from the core into the meso-
porous H-Beta zeolite layer.
3.5. Perspective in real applications
In the future real applications, a pilot-scale LPG plant with a pro-
duction scale of ten thousand tons per year will produce more than
12,750 ton LPG over one ton catalyst in light of the optimized catalytic
reaction conditions (Wcatalyst/Fsyngas = 10 g h mol−1, one-pass conver-
sion of 30.3%, LPG selectivity in all product of 38.3%, average LPG
molecular weight of 51 g mol−1). The volume of flammable tail gas is
ca. 48,384 Nm3 over one ton catalyst per year with a composition of is V
(methane): V(C2): V(methanol): V(DME): V(C5+) = 6.4% : 5.2% : 4.1%
: 29.1% : 55.2%. According to the combustion heat value of above gas,
co-production of electric power of 2372 kWh can be available if these
combustion heat can be utilized (the electric generating efficiency is
assumed to 30%). Therefore, the obvious economy and social value will
be produced in the real application of direct conversion of syngas to
LPG.
4. Conclusions
In summary, a CuZnAl@H-Beta core@shell catalyst with a well-
6
Applied Energy 209 (2018) 1–7
Fig. 6. XRD patterns (a) and the particle size of
Cu and ZnO calculated by the Scherrer formula
(b) on different catalysts after catalytic reactions.
defined metal-zeolite interface was successfully fabricated by a hydro-
thermal method, realizing a highly-efficient and tandem conversion of
syngas to LPG. The metal-zeolite interface is a transition layer between
CuZnAl core and H-Beta zeolite shell with a thickness of 7 m, com-
prising Cu and acid zeolite with a content gradient. The Cu phases in
the hydrothermal catalyst exist in two types. One of those is in the
CuZnAl core, catalyzing CO and H2 to produce methanol, and another is
embedded in the metal-zeolite interface, hydrogenating C3–4 olefins
obtained from successive dehydration of methanol and DME on acidic
sites into saturated C3–4 LPG hydrocarbons. This concept solves the
problem of noble metal dependence in conventional ternary catalysts,
providing novel strategy for designing multifunctional catalysts.
Acknowledgements
This work was partially supported by National Natural Science
Foundation of China (21528302), Zhejiang Province Natural Science
Foundation (LQ16B060002) and Open Research Fund Program of
Zhejiang Provincial Key Lab. for Chem. & Bio. Processing Technology of
Farm Products / Zhejiang Provincial Collaborative Innovation Center of
Agricultural Biological Resources Biochemical Manufacturing
(2016KF008, 2016KF009). Dr. Jian Sun would like to thank the support
of the Hundred-Talent Program of Dalian Institute of Chemical Physics,
Chinese Academy of Sciences.
References
[1] Chen X, Jiang J, Li K, Tian S, Yan F. Energy-efficient biogas reforming process to
produce syngas: The enhanced methane conversion by O2. Appl Energ 2017;185,
Part 1:687–97.
[2] Abdul Mujeebu M. Hydrogen and syngas production by superadiabatic combustion
– A review. Appl Energ 2016;173:210–24.
[3] Jiao F, Li J, Pan X, Xiao J, Li H, Ma H, et al. Selective conversion of syngas to light
olefins. Science 2016;351:1065–8.
[4] Al-Rahbi AS, Williams PT. Hydrogen-rich syngas production and tar removal from
biomass gasification using sacrificial tyre pyrolysis char. Appl Energ
2017;190:501–9.
[5] Steynberg AP, Espinoza RL, Jager B, Vosloo AC. High temperature Fischer-Tropsch
synthesis in commercial practice. Appl Catal A-Gen 1999;186:41–54.
[6] Kim Y-D, Yang C-W, Kim B-J, Moon J-H, Jeong J-Y, Jeong S-H, et al. Fischer–tropsch
diesel production and evaluation as alternative automotive fuel in pilot-scale in-
tegrated biomass-to-liquid process. Appl Energ 2016;180:301–12.
[7] Henrici-Olivé G, Olive S. The Fischer-Tropsch synthesis: molecular weight dis-
tribution of primary products and reaction mechanism. Angew Chem Int Ed
1976;15:136–41.
[8] Torres Galvis HM, Bitter JH, Khare CB, Ruitenbeek M, Dugulan AI, de Jong KP.
Supported iron nanoparticles as catalysts for sustainable production of lower ole-
fins. Science 2012;335:835–8.
[9] Tsubaki N, Fujimoto K. Product control in Fischer-Tropsch synthesis. Fuel Process
Technol 2000;62:173–86.
[10] Masi M. Experimental analysis on a spark ignition petrol engine fuelled with LPG
(liquefied petroleum gas). Energy 2012;41:252–60.
[11] Liu Y, Sotelo-Boyás R, Murata K, Minowa T, Sakanishi K. Hydrotreatment of ve-
getable oils to produce bio-hydrogenated diesel and liquefied petroleum gas fuel
over catalysts containing sulfided Ni–Mo and solid acids. Energ Fuel
2011;25:4675–85.
[12] Kacem SH, Jemni MA, Driss Z, Abid MS. The effect of H2 enrichment on in-cylinder
P. Lu et al.
flow behavior, engine performances and exhaust emissions: case of LPG-hydrogen
engine. Appl Energ 2016;179:961–71.
[13] Kim K, Kim J, Oh S, Kim C, Lee Y. Evaluation of injection and ignition schemes for
the ultra-lean combustion direct-injection LPG engine to control particulate emis-
sions. Appl Energ 2017;194:123–35.
[14] Asami K, Zhang Q, Li X, Asaoka S, Fujimoto K. Semi-indirect synthesis of LPG from
syngas: Conversion of DME into LPG. Catal Today 2005;106:247–51.
[15] Ma X, Ge Q, Ma J, Xu H. Synthesis of LPG via DME from syngas in two-stage re-
action system. Fuel Process Technol 2013;109:1–6.
[16] Ma X, Ge Q, Fang C, Ma J, Xu H. Direct synthesis of LPG from syngas derived from
air-POM. Fuel 2011;90:2051–4.
[17] Ge Q, Lian Y, Yuan X, Li X, Fujimoto K. High performance Cu–ZnO/Pd-β catalysts
for syngas to LPG. Catal Comm 2008;9:256–61.
[18] Ma X, Ge Q, Fang C, Xu H. Effect of Ca promoter on LPG synthesis from syngas over
hybrid catalyst. J Nat Gas Chem 2012;21:615–9.
[19] Zhang Q, Li X, Asami K, Asaoka S, Fujimoto K. Direct synthesis of LPG fuel from
syngas with the hybrid catalyst based on modified Pd/SiO2 and zeolite. Catal Today
2005;104:30–6.
[20] Chen W-H, Lin B-J, Lee H-M, Huang M-H. One-step synthesis of dimethyl ether from
the gas mixture containing CO2 with high space velocity. Appl Energ
2012;98:92–101.
[21] Chen SJ, Fu Y, Huang YX, Tao ZC, Zhu M. Experimental investigation of CO2 se-
paration by adsorption methods in natural gas purification. Appl Energ
2016;179:329–37.
[22] Bao J, He J, Zhang Y, Yoneyama Y, Tsubaki N. A core/shell catalyst produces a
spatially confined effect and shape selectivity in a consecutive reaction. Angew
Chem Int Ed 2008;47:353–6.
Applied Energy 209 (2018) 1–7
[23] Yang G, Tsubaki N, Shamoto J, Yoneyama Y, Zhang Y. Confinement effect and
synergistic function of H-ZSM-5/Cu-ZnO-Al2O3 capsule catalyst for one-step con-
trolled synthesis. J Am Chem Soc 2010;132:8129–36.
[24] Nie R, Lei H, Pan S, Wang L, Fei J, Hou Z. Core–shell structured CuO–ZnO@ H-ZSM-
5 catalysts for CO hydrogenation to dimethyl ether. Fuel 2012;96:419–25.
[25] Yang G, Kawata H, Lin Q, Wang J, Jin Y, Zeng C, et al. Oriented synthesis of target
products in liquid-phase tandem reaction over a tripartite zeolite capsule catalyst.
Chem Sci 2013;4:3958–64.
[26] Li X, He J, Meng M, Yoneyama Y, Tsubaki N. One-step synthesis of H–β zeolite-
enwrapped Co/Al2O3 Fischer-Tropsch catalyst with high spatial selectivity. J Catal
2009;265:26–34.
[27] Sun J, Xing C, Xu H, Meng F, Yoneyama Y, Tsubaki N. Filter and Buffer-pot con-
finement effect of hollow sphere catalyst for promoted activity and enhanced se-
lectivity. J Mater Chem A 2013;1:5670–8.
[28] Shi L, Yang G, Tao K, Yoneyama Y, Tan Y, Tsubaki N. An introduction of CO2
conversion by dry reforming with methane and new route of low-temperature
methanol synthesis. Accounts Chem Res 2013;46:1838–47.
[29] Wang X, Liu G, Wang L, Pan J, Lu GQ, Cheng H-M. TiO2 films with oriented anatase
001 facets and their photoelectrochemical behavior as CdS nanoparticle sensitized
photoanodes. J Mater Chem 2011;21:869–73.
[30] Sun J, Yang G, Ma Q, Ooki I, Taguchi A, Abe T, et al. Fabrication of active Cu–Zn
nanoalloys on H-ZSM5 zeolite for enhanced dimethyl ether synthesis via syngas. J
Mater Chem A 2014;2:8637–43.
[31] Zeng C, Sun J, Yang G, Ooki I, Hayashi K, Yoneyama Y, et al. Highly selective and
multifunctional Cu/ZnO/Zeolite catalyst for one-step dimethyl ether synthesis:
Preparing catalyst by bimetallic physical sputtering. Fuel 2013;112:140–4.